INSTITUT FRANCAIS DU PETROLE PUBLICATIONS

Alain CHAUVEL
Senior Director, Scientific Division, IFP
Deputy Director, Strategy & Corporate Planning Division, IFP
Professor a t the IFP School

Gilles FOURNIER
Engineer, Strategy & Corporate Planning Division, IFP

Claude RAIMBAULT
Senior Engineer, Strategy & Corporate Planning Division, IFP

~n association with Alain PICEYRE

International Petroleum Consultant, BEICIP-Franlab
Head of the Evaluation and Project Inspection Department

Manual of
Process
Economic
Evaluation
New, revised and expanded edition

Translated by BGS
J.V. Guy Bray, M.A., M.Sc, Ph.D., P.Eng., ATA
Santa Barbara, California

2003

t Editions TECHNIP 27 rue Cinoux, 75737 PARIS Cedex IS,FRANCE

Translation of
Manuel devaluation economique des procedes
A. Chauvel, G. Fournier, C. Rairnbault
0 2001, Editions Technip, Paris

All rights reserved.

No art of this ublication may be reproduced or transmitted in any form or by any means, electronic or
mecianical, inckding photocopy, recording, or any information storage and retrieval system, without the
prior written permission of the publisher.
0 Editions Technip, Paris, 2003.
Printed in France
ISBN 2-7108-0836-6
 Foreword 1

Among our most broadly accepted clichks, one that is particularly fixed in the public’s mind
aims to give credence to the idea that “forecasters, and more generally all those who claim to
anticipate the future, are always wrong”. Instead of interpreting this sentence as an inelucta-
ble error, perhaps it would be better to take it as a compliment, like the recognition that is
implicit in the growing role granted to this kind of activity. Indeed, to foresee an event by
endeavoring to deduce it from a set of accessible information and rational hypotheses about
change, will usually produce reactions among the group concerned, whose consequences
automatically lead to the event’s failing to occur. In other words, setting aside chance events,
a forecaster that people pay attention to is unlikely to be right, especially when his reasoning
leads him to envisage crises or tense situations.
In the case of economics the area of forecasts, and more broadly of futurology, is one that
offers favorable conditions for working out the application of its theories and methods. The
particular object of the operation is to present interlocutors with development scenarios that
are designed to identify the most difficult situations, and to prevent their occurrence by sug-
gesting appropriate measures-also to identify the most favorable situations, the ones to be
aimed for, while knowing perfectly well that they represent an unattainable ideal. In other
words, in this kind of activity the economist’s very function, like the forecaster’s, is bound to
lead to a contrary outcome. In practice, his role is not so much to picture a special situation
that he knows a priori will not occur, except under extraordinary circumstances, but rather to
define a space within which the most probable outcomes will be found. The aim is to replace
the irrationality of prediction, which in most cases is doomed to failure, by the deductive logic
of case studies based on various sets of development theories. Here economics comes fully
into play as a decision-making aid, especially when its objective is the regular, systematic
evaluation of Research and Development (R&D) projects, and the quantification of potential
innovations and their consequences.
For this purpose, economics uses some or all of the following procedures:
0 Study of the overall economic situation into which the project under examination will be
inserted, and consideration of the foreseeable evolution of the major indicators.
0 Analysis of the market and the status of competition, so as to learn one of the essential vari-

ables for all subsequent calculations, in other words the unit production capacities that char-
acterize given geographic contexts. Such results are usually derived from an examination of
the price-volume relationships for products or raw materials, and their extrapolation over the
X Foreword

medium or long term. Studies of competitors and industrial projects, both operating and
planned, are also very important for this purpose.
0 Techno-economic comparisons, measured against a criterion represented by a given stan-
dard of profitability, between the results for the technology under consideration-whether
developed in the laboratory, in a pilot plant, or as a prototype-and the results offered by the
existing or potential competing approaches. The procedure utilized in this type of operation
is based on principles similar to those employed by engineering companies or services, by
setting up a construction plan together with assessments of material and energy requirements,
and by sizing and costing equipment. The only major differences are the lack of detailed,
credible information, and also the very short time allowed for completing it. On the other
hand, it is not a matter of producing an absolute opinion, but only of reaching a relative judg-
ment.
0 Sensitivity analysis for technical parameters, so as to identify the essential variables and
thereby to provide a performance framework to guide research towards the achievement of a
successful project, i.e., industrial or commercial production. By establishing an economic
ranking of factors affecting the operation in a given situation, it also enables the necessary
efforts and means to be brought to bear preferentially on improving the most important
parameters4ither to achieve this goal as rapidly as possible, or to abandon the work more
quickly if the levels required for competition prove to be unrealistic.
0 Examining the overall impact, especially the positive or negative consequences of insert-
ing a new technique or equipment into an existing industrial context. In this case it is a matter
of crediting the innovation under consideration with the resulting savings or additional costs
generated by its presence. This type of procedure is widely used in the field of petroleum
refining, and linear programming is an invaluable tool in applying it. Its special interest lies
in enabling the definition of the techno-economic conditions, i.e., the external constraints,
under which the project being studied would turn out to be fully workable.
0 Considering the economic consequences of environmental measures and regulations that
the technology being evaluated must comply with, or that it may give rise to. Generally
speaking, the current trend is to consider all the different kinds of nuisance that may accom-
pany an innovation, including the effects on jobs (creation or loss), on the locality, on trans-
port, etc. In order for an analysis to be as comprehensive as possible and include all the
parameters that may ensure the success or failure of an undertaking (opportunity, acceptabil-
ity, etc.) it is necessary, in the time-honored phrase, to “internalize the externalities” for each
project. The major pitfall that this kind of evaluation runs into is the difficulty of having to
work entirely with forecasts, and to make simultaneous predictions of the evolution of all the
economic elements, both internal and external. Under these circumstances, even if several
scenarios are devised, it is practically impossible on the basis of existing knowledge to
attempt to pick out what may eventually become “strong signals” from among the “weak”
ones. Looked at in this way, evaluation falls back into the error of forecasting.
The present volume represents an updating, combined with a reorganized layout, of the
Manuel d ’dvaluationbconomique des procddds published by Editions Technip in 1976, and
later by McGraw-Hill in 1981. It does not attempt to cover all the various topicsjust mentioned.
 Foreword XI

Firstly, it aims to discuss in somewhat greater depth the subject of market analysis, in
terms of volumes and prices. It thereby enables a better understanding of the modalities of
selecting a unit production capacity, an essential parameter for developing balance sheet cal-
culations, sizing and costing of equipment, and ultimately the definition of investment costs,
operating costs, and the costs required for plant profitability.
Secondly, it also takes up the three-part methodology expounded in the previous editions.
However, their order has been reversed, so as to provide a closer correspondence with the
method that proceeds from market data towards the technical characteristics of the systems
under consideration. Consequently the first section of this methodological triptych concerns
incomes, beginning with profits and the notion of taxation, then the main profitability criteria,
whether empirical or based on the concept of present value, and finally minimum profitable
sales prices. The second section deals with expenses, i.e., cost of price and operating cost,
examining variable and fixed operating costs as well as the cost of capital. The last section
takes a closer look at the structure of these fixed assets and the effects of changing production
capacities, dates, and local circumstances, assuming that the battery limits investments of
production unit or processing system size are known.
Along the same line, the third stage focuses on an examination of procedures for estimat-
ing the amount of these costs, in the reference capital. Next comes a fairly complete synopsis
of the so-called exponential and factorial methods. For this, it is necessary to know at least
the expense involved in acquiring the main items of equipment, which form the core of the
physico-chemical operating units involved in the system being studied.
In these circumstances, when this kind of information is not directly accessible, it can be
useful to have at one’s disposal procedures that enable it to be obtained. As in the earlier edi-
tions, this approach, which is based on the use of graphical representations of CEF (Cost Esti-
mation Formulas), makes up most of the appendices, where it is broken down by major
category of equipment. However, here it differs from preceding editions.
Indeed, to the extent that the considerable progress achieved in computing during recent
years has greatly improved the accessibility and dissemination of equipment size modeling,
the rapid calculation methods previously offered for the more common items of equipment
have not been repeated here. However, other kinds of infomation have been added, notably
for plate exchangers.
It is equally important to point out that the present publication concerns only those indus-
trial plants that operate continuously. Thus, pilot plants and prototypes are excluded, along
with consideration of batch or limited operations. In this last case the procedure for sizing and
costing of equipment is not so very different from the one suggested here. On the other hand,
the way in which modem facilities of this kind are managed is very specific to the conditions
selected: multi-usage or multi-product, employing teams in two or three eight-hour shifts,
five or seven days out of seven, etc.
Finally, contrary to the position taken in previous editions, no examples are given to illus-
trate the suggested methodology.
 Table of Contents 1

Foreword ....................................................................................................................... IX

Abbreviations and Acronyms ......................................................................................... 1

Chapter 1
MARKET ANALYSIS

1.1 The Raw Materials/Products Duality and PriceNolume Relationships ........ 5

1.1.1 Relative Nature of the Raw MaterialsProducts Duality ...................................... 6
1.1.2 The Concept of Elasticity ................................................................................. 6
1.1.2.1 Mathematical ..................................................................................... 7
1.1.2.2 Economic ................... .................................................................. 7
1.1.3 Major Type Situations ..................................................................................... 10
1.1.3.1 Raw Materials and Products are Freely Available on the Market ............. 10
1.1.3.2 Limited Availability of Raw Materials, Very Big Markets for Products ... 11
1.1.3.3 Raw Materials are Available, but the Market for Products is Limited ....... 15

1.2 Forecasting Methods ........................................................................................... 18

1.2.1 Forecasting Methods: Areas of Application and General Character ...................... 19
1.2.2 Frequency of Forecast and Types of Methods ............... 19
1.2.3 Processing and Extrapolation of Historical Data ................................................. 20
1.2.3.1 Internal or Endogenous Methods .... 21
1.2.3.2 External or Exogenous Methods ........................................................... 26
1.2.4 Dominantly Future-Oriented Methods .......................... ............................. 34
1.2.4.1 The Various Interested Parties ............................................................. 34
1.2.4.2 General Procedure for a Typical Study ...................... ................... 36
1.2.5 The Limits of Forecasting Methods ................................................................... 38
XIV Table of Contents

1.3 Specific Application Methods for Forecasting Prices and Volumes ................ 39
1.3.1 Price Applications ........................................................................................... 39
1.3.1.1 Current Currency. Constant Currency. Deflation Index ........................... 39
1.3.1.2 Relationships Between Product Prices and their Raw Materials. Line of
Equivalence ........................................................................................ 41
1.3.1.3 Product Price Structure. Evolution Curves ............................................. 48
1.3.2 Sales Volume Applications ............................................................................... 58
1.3.2.1 Normalization of Evolution Curves ....................................................... 58
1.3.2.2 The Concept of Penetration as a Tool for Determining the Level of Market
Saturation ........................................................................................... 60
1.3.2.3 Real and Average Growth Rate. Application to the Long Term ............... 63

Chapter 2
THE ELEMENTS OF ECONOMIC CALCULATION

2.1 The Study of Project Profitability ...................................................................... 67

2.1.1 The Concept of Taxes ...................................................................................... 68
2.1.2 Profitability Criteria ......................................................................................... 71
2.1.2.1 Typological Classification and Standardization of Expression ................. 71
2.1.2.2 Empirical Criteria ................................................................................ 72
2.1.2.3 Criteria Based on the Concept of Present Value ..................................... 78
2.1.2.4 The Practical Significance of the Cash Flow Concept ............................. 111
2.2 Operating Cost and Cost of Price ....................................................................... 112
2.2.1 Cost Accounting Methods ................................................................................ 112
2.2.1.1 Stream Factor and Load Rates .............................................................. 112
2.2.1.2 Breakdown of the Operating Cost and Cost of Price ............................... 114
2.2.2 Variable or Proportional Charges ...................................................................... 116
2.2.2.1 Raw Materials, Reagents, and Other Consumables ................................. 116
2.2.2.2 Utilities and Miscellaneous Costs ......................................................... 137
2.2.3 Labor ............................................................................................................. 146
2.2.3.1 The Concept of the Shift Worker .......................................................... 146
2.2.3.2 Labor Costs ........................................................................................ 149
2.2.4 Fixed Charges ................................................................................................. 151
2.2.4.1 Fixed Operating Charges ..................................................................... 151
2.2.4.2 Fixed Capital Charges ......................................................................... 154
2.3 Investment Charges ............................................................................................. 164
2.3.1 Definitions and Purposes of Various Investment Charges .................................... 164
2.3.1.1 Type Breakdown ................................................................................. 164
 Table of Contents XV

2.3.1.2 Investment Costs of Production Facilities .............................................. 166

2.3.1.3 Fixed Capital ...................................................................................... 172
2.3.1.4 Depreciable Capital ............................................................................ 178
2.3.1.5 Total Investments on a Manufacturing Site. Working Capital .................. 182
2.3.1.6 Grassroots Investments ....................................................................... 183
2.3.2 Adjustments to Investment Data ....................................................................... 185
2.3.2.1 Adjustment of Capacity ....................................................................... 185
2.3.2.2 Investment Updating ........................................................................... 190
2.3.2.3 The Location Effect ............................................................................ 206

Chapter 3
THE DETERMINATION OF BATTERY
LIMITS INVESTMENTS

3.1 Methods for Applying the Estimation Procedure .............. 223

3.1.1 Background ................................................
............................................ 223
3.1.2 The Principal Stages in the Progress of a Project ................................................ 224
3.1.2.1 The Study or Pre-Study Stages ...................................... 225
3.1.2.2 The Implementation Phases ................................................................. 228
3.1.2.3 The Respective Contributions of Evaluation and Estimation .......... 229
3.1.2.4 The Actors ......................................................................................... 230
3.1.3 Choosing an Estimation Method ........ .......................................................... 231
3.1.3.1 Selection Criteria ................................................................................ 232
3.1.3.2 Classification of Estimation Methods ................................................... 235
3.1.3.3 Selecting an Estimation Method ........................................................... 244

3.2 Methods for Determining Investments .............................................................. 247

3.2.1 Exponential Methods ............................................................. 248
3.2.1.1 The “Extrapolation Factor” Procedure ..................................................248
3.2.1.2 Notional Operating Unit Methods .............................. 248
3.2.1.3 Flow Methods ....................................................................................
253
3.2.1.4 The Significant Steps Methods ......................... ...............................
256
3.2.1.5 Method of Functional Modules ............................................................
263
3.2.2 Factorial Methods ......................................................... 263 .................
3.2.2.1 The Structure of Battery Limits Investments .............................. 263
3.2.2.2 Methods Employing Constant Multiplying Factors ................................ 264
3.2.2.3 Methods Employing Variable Multiplying Factors ................................. 269
3.2.2.4 Methods that Consider the Particular Nature of Each Project .................. 271
XVI Table of Contents

3.3 Methods Developed by the IFP ........................................................................... 291

3.3.1 Background .................................................................................................... 291
3.3.2 “Functional Modules” Method .......................................................................... 292
3.3.3 The PrC-Estime Method ................................................................................... 297
3.3.3.1 Current State of Development .............................................................. 297
3.3.3.2 Development and Application of the Method ......................................... 298
3.3.3.3 Description of the Method .................................................................... 302
3.3.3.4 Validity and Application Limits ............................................................ 305

Appendix 1
FUNCTIONAL MODULES METHOD (FMM)

Al.1 Data Analysis ...................................................................................................... 309

A1.2 Description of the Method ................................................................................. 311
A1.3 Application of the Functional Module Method ............................................... 313

Appendix 2
PRk-ESTIME METHOD

A2.1 Evaluation of Pressure Vessels ......................................................................... 319

A2 . .1 Thickness Calculation ................................................................................. 319
A2.1.1.1 Basic Formula .............................................................................. 319
A2.1.1.2 Values of Constants ...................................................................... 320
A2.1.1.3 Application .................................................................................. 321
A2. .2 Determination of the Prices of Columns, Drums, and Related Reactors ............ 321
A2.1.2.1 Principle ...................................................................................... 321
A2.1.2.2 Weight Determination .................................................................. 322
A2.1.2.3 Determination of Prices of Externals .............................................. 323
A2.1.2.4 Determination of the Price of the Internals ...................................... 330
A2.1.2.5 Determination of Final Price ......................................................... 333
A2.2 Evaluation of Reactors ...................................................................................... 335
A2.3 Evaluation of Heat Exchangers ........................................................................ 337
A2.3.1 Tube-Type Heat Exchangers ........................................................................ 337
A2.3.1.1 Additional Data ............................................................................ 337
A2.3.1.2 Selection of Tube-Type Exchangers ............................................... 341
A2.3.1.3 Prices of Tube-Type Exchangers ................................................... 342
 Table of Contents XVII

A2.3.2 Air-Coolers ................................................................................................ 349

A2.3.2.1 Simplified Method: Principle ........................................................ 349
A2.3.2.2 Preliminary Calculation ................................................................ 350
A2.3.2.3 Sizing Air-Coolers ....................................................................... 354
A2.3.2.4 Determining the Price of Air-Coolers ............................................ 355
A2.3.2.5 Validity of the Method ......................................... 358
A2.3.3 Plate Exchangers .................................... ............................................... 358
A2.3.3.1 General ....................................................................................... 358
A2.3.3.2 Sizing Plate Exchangers ............................................................... 359
A2.3.3.3 Determining the Price of Plate Exchangers ..................................... 360
A2.3.3.4 Comment .................................... ........................................... 362
A2.4 Evaluation of Pumps and Compressors .......................................................... 363
A2.4.1 Pump Prices ............................................................................................... 363
A2.4.1.1 Centrifugal Pumps ....................................................................... 363
A2.4.1.2 Reciprocating Pumps ................................................................... 366
A2.4.1.3 Rotary and Metering Pumps ................... ................................ 368
A2.4.2 Compressor Prices ...................................................................................... 368
A2.4.2.1 Types of Compressors ....................................... 368
A2.4.2.2 Determining Compressor Prices .................................................... 370
A2.5 Evaluation of Furnaces ..................................................................................... 373
A2.5.1 General Observations on Furnaces ............................................................... 373
A2.5.2 Furnace Prices ............................................................................................ 373
A2.5.2.1 Determining the Base Price ........................................................... 373
A2.5.2.2 Correction Factors ....................................................................... 374
A2.5.2.3 Determining the Energy Cost ........................................................ 375
A2.6 Evaluation of Steam Ejectors ............................................................. 377
A2.6.1 Ejector Calculations .................................................................................... 377
A2.6.1.1 Determining the Maximum Suction Flow ....................................... 377
A2.6.1.2 Determining the 20°C Equivalent Dry Air Flow ............................. 378
A2.6.2 Ejector Prices ............................................................................................. 383
A2.1 Evaluation of Special Equipment ..................................................................... 385
A2.7.1 Evaluating Dryers ....................................................................................... 385
A2.7.1.1 Dryer Selection ............................................................................ 385
A2.7.1.2 Dryer Sizing ................................................................................ 385
A2.7.1.3 Dryer Prices .............................................. 391
A2.7.2 Evaluating Crystallizers .............................................................................. 394
A2.7.2.1 Crystallization Efficiency ............ ............................................ 394
A2.7.2.2 Crystallizer Selection .................. ............................................ 395
A2.7.2.3 Crystallizer Prices ........................................................................ 396
A2.7.3 Evaluating Evaporators ............................................................................... 397
A2.7.3.1 Evaporator Selection .................................................................... 397
A2.7.3.2 Sizing Evaporators ................................ .................................. 398
A2.7.3.3 Evaporator Prices ......................................................................... 400
A2.7.4 Evaluating Filters ....................................................................................... 402
A2.7.4.1 Filter Selection ............................................................................ 402
XVIII Table of Contents

A2.7.4.2 Sizing Filters ............................................................................... 402

A2.7.4.3 Filter Prices ................................................................................. 403
A2.7.5 Evaluation of Centrifuges ............................................................................ 405
A2.7.5.1 General ....................................................................................... 405
A2.7.5.2 Sizing Centrihges ........................................................................ 405
A2.7.5.3 Centrifuge Prices .......................................................................... 406
A2.7.6 Evaluation of Crushers and Grinders ............................................................. 408
A2.7.6.1 General ....................................................................................... 408
A2.7.6.2 Price and Performance of Crushers and Grinders ............................. 409
A2.7.7 Evaluation of Gas Cyclones ......................................................................... 411
A2.7.8 Evaluation of Vibrating Screens ................................................................... 413
A2.7.8.1 General ....................................................................................... 413
A2.7.8.2 Prices of Vibrating Screens ........................................................... 413
A2.7.9 Evaluation of Conveyors .............................................................................. 414
A2.7.9.1 General ....................................................................................... 414
A2.7.9.2 Conveyor Sizes and Prices ............................................................ 415
A2.7.10 Evaluation of Instrumentation ...................................................................... 418
A2.8 Evaluation of Units for Providing Utilities ....................................... 419
A2.8.1 Evaluation of Utility-Producing Units ........................................................... 419
A2.8.1.1 Steam Production ......................................................................... 419
A2.8.1.2 Electricity Generation ................................................................... 421
A2.8.1.3 Production of Cooling, Boiler, and Process Waters .......................... 422
A2.8.1.4 Refrigeration ................................................................................ 423
A2.8.2 Distribution of Utilities to the User Units ...................................................... 424
A2.8.2.1 Steam .......................................................................................... 426
A2.8.2.2 Cooling Water ............................................................................. 426
A2.8.2.3 Electricity .................................................................................... 426
A2.8.3 Special Cases .............................................................................................. 427
A2.9 Evaluation of Storage ........................................................................................ 429
A2.9.1 Atmospheric Pressure Tanks ........................................................................ 429
A2.9.1.1 Capacity Less than 100 m3 ............................................................ 429
A2.9.1.2 Capacity Greater than 100 m3......................................................... 432
A2.9.2 Pressurized Storage Tanks ........................................................................... 432
A2.9.2.1 Cylindrical Tanks ......................................................................... 432
A2.9.2.2 Spherical Tanks ........................................................................... 433

Bibliography .................................................................................................................. 435

Index ................................................................................................................................ 449

 1 CHAPTER
Market Analysis I

Before embarking on any analysis intended to provide a technical or economic justification

for a project, it would seem desirable to define the meaning of the term, at least as it applies
to the concerns of an industrial nature that are the subject of the present volume. A project
should be understood as a real or fictitious intention to set up a more or less complex instal-
lation, which may implement a procedure or a conversion; one or more raw materials go in;
one or more products, in the broad sense of the term, come out. The processing applied in the
plant gives the products a certain added value as compared to the raw materials.
The economic evaluation of such a project demands that we have at our disposal a certain
amount of information about the context in which this study will be undertaken: this must
precede any calculation concerning even the kind of equipment to be considered. In fact,
the results obtained are likely to be strongly influenced by the project’s techno-economic
environment, especially by the local characteristics of the site where the plant may be set up,
and by the markets in which the raw materials must be found and where the products must be
sold.

1.1 THE RAW MATERIALS/PRODUCTS DUALITY

A N D PRICE/VOLUME RELATIONSHIPS

When contemplating a new plant, a manufacturer’s first concern must be to confirm that the
raw materials that are needed are in fact available, and under what economic conditions. In
addition, it is essential to estimate the capacity of potential users to purchase the products that
it is intended to manufacture, both in volume and in price terms, and then-by comparing the
situation before and after plant installation-to evaluate to what extent it may be possible not
only to run the plant successfully, but also to make a profit. Consequently, whether they refer
to raw materials or to products, price/volume relationships are a permanent feature of this
type of analysis.
Moreover, the connection between these external constraints, which are a characteristic
feature of the inputs and markets for a given manufacturing process, is what leads the manu-
facturer to:
The decision whether or not to actually construct the corresponding plants.
6 1 . Market Analysis

a The definition of the most suitable unit production or processing capacity. Possession of
this information is in fact essential, because not only the equipment sizing calculations
but also the whole economic procedure of assessing the intrinsic worth of the project
depends upon it.
Before considering the influence of price/volume relations on unit capacity size, it would
seem essential to begin by emphasizing the relative nature of the raw materials/products dual-
ity, and then to define the element that makes it possible to measure the degree of dependence
between these variables: in this case it is the concept of market elasticity.

1.1.1 RELATIVE NATURE OF THE RAW

M A TERIALS/PRODUCTS DUALITY

The distinction between the two makes sense only in the context of a conversion or process-
ing unit that on the one hand utilizes them, and on the other manufactures them: something
that is a raw material in a given situation may become a product in other circumstances, and
vice versa. Looked at in this way, the general study of the evolution of prices or volumes does
not require different procedures depending on whether we are looking at what goes into or
what comes out of a particular plant.
In terms of products, the size of the installations that are to be put to work depends essen-
tially on the quantities that may be accepted or asked for by the users, i.e., on the size of the
market. It follows that if it is not a seller’s market, competitors will have to be displaced; the
volume of sales will be larger in proportion to whether the prices offered, all other things
being equal, are attractive-which means low. If, on the other hand, it is a monopoly situa-
tion, with growing needs, it will be possible to keep prices high and even to raise them all the
more easily as the scarcity of the products increases and demand rises as a result.
For raw materials, the approach is much the same. If availabilities are adequate, the greater
the quantities sought, the better the suppliers’ prices will be, so as to retain the customer, pref-
erably the one who gives them the biggest orders. On the other hand, if there is a shortage
situation, manifesting itself in substitute raw materials or by a need to manufacture the
required inputs, the acquisition costs will obviously rise with the demand.
In any event, a manufacturer who wants to construct a profitable plant must have in mind
a sufficient price difference between the products and the raw materials. If at first certain ele-
ments, such as unit capacity figures, are not available for precise calculations of the minimum
difference, at least there will be some material balance sheets, even energy consumption or
production costs, which will provide reference points or limits that must not be exceeded.

1.1.2 THE CONCEPT OF ELASTICITY

This concept, whose origin may be attributed to the French economist Augustin Coumot
(180 1- 1877), is based on the following observation:
‘<Formanufactured goods, demand should vary more than price, whilefor the most essen-
tial things, as for the most superfluous, demand varies little in comparison to price”.
 1. Market Analysis 7

Generally speaking, elasticity is defined as the ratio between the absolute, or preferably
the relative variations over time of two phenomena, i.e.,

AyIAx or AY Ax
--I-
Y X
Depending on the classification employed, several kinds of elasticity may be distinguished.

1.1.2.1 Mathematical
From a purely mathematical viewpoint, there are:
(a) Empirical or casual elasticity, obtained by calculation for each moment, using two
series of observations x(t) and y(t). This has only very limited interest.
(b) Theoretical elasticity, derived from the parameters of a model, or a combination of
these parameters: y = f(x), and from the corresponding smoothing of the observations (a log-
log model, for example). This type of elasticity is useful for forecasting.
(c) Anticipation elasticity, which includes:
Short-term elasticity, which measures the influence of the rate of change of one vari-
able on the rate of change of another variable, during a certain future period and with
the help of a model that correlates x and y ,
0 Long-term elasticity, which, for a future period, takes into account the cumulative

effects of previous periods. Thus for example: an increase in price leads to a

decrease in consumption at time t, which in turn leads to a rise in price and a reduc-
tion in demand at time t + 1.
Thus we can clearly see that long-term elasticity must always be stronger than the short-
term kind.

1.1.2.2 Economic
Employing an economic point of view, we may introduce the distinction between simple,
two-variable relationships: quantities (q) and price (P), and more complex relationships that
involve a greater number of parameters.
In the first case, we may consider particularly:
(a) Price-demand elasticity for a given product, also called “direct” or “self‘ elasticity,
to strongly emphasize that the price and quantity refer to the same product.
It is expressed by the relation:
3
q=- 4
AP
P
where
q: coefficient of elasticity
q : quantity
P: price.
8 1. Market Analysis

Although the mathematical formulation relating q and P has the form q = f(P), by conven-
tion, or rather out of habit, graphical representations of it always show the independent vari-
able, i.e., the price, on they-axis, and the quantities on thex-axis. Also by convention,elasticity
is negative, since experience shows that price and quantity usually move in opposite directions.
When it is positive, a minus sign therefore has to be inserted. Such a situation may arise in
two ways:
0 When the increase in the price of an essential product results, in spite of everything, in

an increase in consumption; the most disadvantaged social classes transfer their demand
onto this product to the detriment of the consumption of other, less indispensable prod-
ucts (the Giffen effect or paradox),
0 When the rise in a product’s price only heightens its snob or fashionable value, and

thereby causes certain purchasers, usually well-off ones, to buy it for themselves (the
Veblen effect).
Theoretically, the absolute value of price elasticity may vary from 0 to infinity. It is low
for essential and luxury products; in contrast it is higher for everyday or semi-luxury goods.
When it is nil, demand is said to be “inelastic”; when less than 1 , the market is called “tight”;
it becomes “elastic” if it is greater than 1.
If the relation between q and P is linear, the coefficient q decreases in proportion as the
price decreases. In an “isoelastic” graph, q has the same value at all points on the curve; such
a situation is described by the following equation:

k
q = - where k and a are constants
P“

In this case:

where: q = -a
and : q = k P
If q = -a = -1, we obtain the expression for an equilateral hyperbola. In logarithmic
coordinates the elasticity is equal to the slope of the straight line representing the fbnction
log q = f(l0g P).
(b) Cross elasticity of demand for a given product. In this case we are no longer taking
into account only the direct influence of this product’s own price on the demand, but also,
assuming that this latter parameter remains constant, considering the effects of the evolution
of the prices of other goods in the same market. In this case, for two products (1) and (2) for
example, we can measure the elasticity of the consumption of (1) according to the evolution
of the price for (2), when the price for (1) is held constant, by the relation:

dq1 4
%1)/(2) = y
dp2
According to whether these two goods are complementary or substitutable, the cross elas-
ticity will be negative or positive. In fact, when there is complementarity, as a general rule an
 1. Market Analysis 9

increase in P2 also leads to a decrease in q1 (q < 0). For substitutable products, the rise in P2
leads to a fall in 42 but leaves (1) the possibility of increasing its share of the market q1
’
(rl 0).
(c) Substitution price elasticity. This enables the measurement of the relative effects on
their respective markets of a change in the price structure of several given products. If we
limit this analysis to only two of them, (1) and (2), such an elasticity will have the following
form:

In this case complementarity is represented by q, = 0, since the demand structure is not

affected by a change in the price structure. For substitutable products, q, is usually negative.

COMMENT
Generalization of the concept of demand elasticity in relation to price.
0 This concept may first of all be applied by permuting the role of the variables; in

this case we are trying to determine the elasticity of prices as a function of the amount
available, i.e., the supply; it is expressed as:

e is a function of the existing stocks, the possibilities for increasing production, and
the length of time considered, but it is always positive. Indeed, any increase in the
demand increases the pressures on manufacturing and the problems of supply, and
thus leads to a rise in price.
0 Next, one of the two variables may be changed, the price for example, which is

sometimes replaced by purchasing capacity, i.e., income in the case of the consumer,
or the level of activity. In this case, we may define income elasticity by means of the
following relation:

q R is usually positive, but varies significantly according to the products, the services,
or the goods concerned. However, it follows certain empirical natural laws,
expressed in particular by Engel and Schwabe. Such a concept has practically no
application in the field of economic evaluation of projects.
The development of more complex models enables consideration of more numerous
relationships, by increasing the number of variables. The concept of elasticity finds
possible extensions here. Thus, for example, the numerous energy models developed
in the 1970s are based on statistical analyses where, as a first approximation, the
10 1. Market Analysis

overall demand for energy (E) may be deduced from the activity levels (A) or from
the yield and prices (P) of this energy, according to an equation having the form:

E = A ~ p%
A

where qAand qp are the coefficients of “product” and “price” elasticity respectively.

1.1.3 MAJOR TYPE SITUATIONS

With this introduction, and based on these various considerations, how are we in practice to
determine the unit conversion or processing capacity? Several situations may be envisaged.

1.1.3.1 Raw Materials and Products are Freely Available on the Market
In this case, the acquisition costs for feedstocks are the same as those applied elsewhere; this
is also true for products. In contrast, in order to maintain competitiveness one must keep up
with the leading prices, i.e., the prices of the most efficient current manufactures. It is there-
fore necessary to do as well as the most economic processes, and the plants that make the
most of the effects of scale. So the size level to be considered emerges from a study of the
markets, particularly their development and existing capacities.
These leading prices may also come from plants of more modest size, but old enough to
have been paid off, and thus able to take advantage of special economic circumstances. If this
is the case, the unit capacity to be selected should be bigger than the existing ones, by a factor
such that the economy of scale obtained will compensate for the capital costs. It is often dif-
ficult to do this if the size of the operating plants is already large. Again, the answer will come
from a study of the evolution of prices and tonnages, and from the projections to be made
from them.

COMMENT
In this connection, we should also take into account the aging of operating units.
Depending on the particular case, contradictory results may be produced: either addi-
tional outlets and a replacement market if there is a halt in operations, or conversely
a reduction in the share of available tonnages if there is a modernization and debot-
tlenecking, leading to an increase in the unit manufacturing capacity.

Situations of this type are mostly found in intermediate chemical products, both organic
and inorganic. Thus, the production of styrene, a two-stage process:
0 Alkylation of benzene by ethylene to form ethylbenzene,

0 Dehydrogenation of ethylbenzene to form styrene,

operates in a market where the raw materials, benzene and ethylene, are freely available at
prices set by market quotations, that is, by procedures external to the project under consider-
ation. The same applies to the styrene, whose prices and tonnages must be brought into line
with the competition. So if outlets do exist, unless a process can be found that is very different
 1. Market Analysis 11

from the operating ones, the unit capacity to be considered must be in line with those already
in place, particularly the most modern ones, for example, at least 300000 t/yr in Western
Europe.
The concept of elasticity discussed above has nothing to offer here, in the short term.
Although it is indeed helpfbl in terms of predicting the overall market, in practice it is not
used to determine the desired minimum unit production size.

1.1.3.2
r
limited Availabilit of Raw Materials,
Very Big Markets or Products

In situations of this type, the manufacturer operates under two kinds of constraint:
0 The price of the product(s), which must be aligned with the market price, and which may
therefore be considered as fixed, at any given time.
The price of the raw material, which, because of limited availability, may have various
sources, or may even require major processing if it is not directly accessible. In this case,
as scarcity increases, the processing applied becomes more and more complicated and
the cost of the raw material rises in step with the needs.
0 In these circumstances the proposed plant’s unit capacity will depend on a minimal dif-

ference between the costs of the product and the feedstock, corresponding to the operat-
ing expenses and the capital costs for the production envisaged. There will therefore be
a maximum permissible raw material price. An illustration of such a situation is pro-
vided, for example, by the production of MTBE (Methyl Tertio Butyl Ether).
Let us suppose that, in the context of commercial development of lead-free gasoline in
Western Europe, a French, or more generally a southern European refiner wishes to know the
volume of high-octane gasoline conforming to the new specifications that he will be able to
put on the market at an economically acceptable price. For this purpose, considering his refin-
ery’s current structure, he thinks that by 1992 he will be able to fill his needs for additional
octane, half by modifying his refinery flow sheets, and the rest by including 10% of ether as
an additive.
The MTBE may have various origins, and in particular may be produced by the refiner
himself, as follows:
From isobutylene in a C4 steam cracking cut. The relevant unit is already set up in the
neighboring petrochemical area, which includes plants for pyrolysis of naphtha with a
total ethylene production capacity in excess of 500000 t/yr, i.e., sufficient to provide the
isobutylene supply for the manufacture of 100000 t/yr of MTBE, using a standard pro-
cess.
From isobutylene in a C4 catalytic cracking cut. The unit, set up within the refinery
itself, has a 50000 t/yr capacity and employs a reactive distillation technology with a
lower investment cost. In fact, the whole C4 cut is charged to etherification. Now, the
by-products from catalytic cracking have an average isobutylene content much lower
than those from steam cracking, especially after extraction of butadiene. The amount of
diluents directly influences the size of the facilities and consequently the investment
costs: this can lead to a search for less capital-intensive technologies, especially for C4
catalytic cracking cuts.
12 1. Market Analysis

MTBE can also be purchased on the market. In this case, let us suppose that it may have
the following origins:
0 C4 catalytic cracking cut located on another site, supplying an etherification unit in its

neighborhood that employs a standard process, making it an expensive investment.

0 An n-butane-isobutane mixture produced locally in the Middle East, available at low

cost and suitable for supplying, in the same area, a large-scale industrial complex
consisting of a deisobutanizer, an isomerization of n-butane into isobutane, a dehydro-
genation of isobutane into isobutylene, and finally an etherification operating on an isob-
utane-isobutylene cut. Installation of an ether production capacity of 500000 t/yr would
enable delivery of products into the southern European market.
0 Isobutane alone or a mixture of n-butane-isobutane produced locally in Western Europe

(Scotland, for example) which, under conditions similar to or better than those just
described for the Middle East, could supply the same market, as long as the construction
of 500000 t/yr capacity etherification plants appears to be profitable under European
conditions for the sale of the initial C4 and payback of capital.
Each of these MTBE sources has its particular price, as shown graphically in Figs. l.la
and 1.1b. These prices include:
0 The cost of the initial raw material, after deduction of the revenues from by-products, C4

cut residuals for example,

0 The cost of the processing itself, which includes operating expenses and capital costs,

0 The cost of transporting the MTBE from its production site to the plant where it will be

used (ocean freight charges, for example). Based on linear depreciationover ten years, and
remuneration of capital at an internal rate of return of 15% (see Chapter 2, 9 2.1.2.3.D),
the situation in terms of the quantity of MTBE available and the transfer price is summa-
rized in Table 1.1.
Using these numbers, it then becomes possible to define a curve for the average price:

as a function of the quantities available, and to conduct an economic pre-optimization of the

plants that will be using the MTBE, by varying their raw material demands and thus their pro-
duction capacities (Fig. 1.2).
In the past, the price of MTBE was always expressed as a function of the price of high-
octane gasoline: depending on the circumstances, one could thus deduce an overrating in
favor of ether of between 1.2 and 1.5. If we then suppose that the price of crude remains in
the vicinity of $18- 19hbl for a certain time, which would then give a pre-tax price of around
$180/bbl for high-octane gasoline, the price of MTBE should lie between $216 and $270/t.
In fact, since the situation is tending towards a rising demand for octane, and since manufac-
turers always like to maintain some margin of safety, it would seem reasonable to think that
the premium should lie between 1.3 and 1.4. In these circumstances, the price of ether should
be somewhere between $234 and $252/t. The graph in Fig. 1.2 shows the corresponding
quantities available. So if the manufacturer concerned wishes to be prudent, it will base its
production capacity for lead-free gasoline on an MTBE supply of 190000 t/yr, i.e., a market
14 1. Market Analysis

Figure 1.1 b MTBE profitability (influence of feedstock, location, and

MTBElpremium gasoline ratio; crude: $18-19hbl; premium gasoline:
$1Soh). (Source: IFP)

Table 1.1 MTBE production for various raw material sources

I Origin I I
Quantity Price ($/t) I Raw material source
I
1 100000 22 1 ex steam cracking of naphtha
2 50 000 237 ex catalytic cracking (reactive distillation)
3 30 000 262 ex catalytic cracking (conventional etherification)
4 80 000 268 ex n-butanelisobutane (Middle East)
5 50000 273 ex isobutane (Western Europe)
6 90 000 290 ex n-butanelisobutane (Western Europe)
 1. Market Analysis 15

0 100 200 300 400

Cumulative total quantities of MTBE (Wyr)

Figure 1.2 Determination of MTBE production levels versus average

price. (Source: IFP)

5 70075
q=-- z0.1
245.5 13650
In addition, the low value obtained emphasizes the tightness of the supply.

COMMENT
The example selected is a very petroleum-oriented one, but the terms would be much
the same in other areas of activity. Thus, the increased development experienced by
linear low-density polyethylene (LLDPE) leads to a strong demand for ethylene
comonomer, and particularly for n-butene- 1. Now, this product has various possible
sources: extraction (adsorption) from C4 cuts from naphtha steam cracking, dimer-
ization of ethylene, dehydrogenation of n-butane, etc. The average acquisition cost
for comonomer may have a significant influence on the selection of unit production
capacity for LLDPE.

1.1.3.3 Raw Materials are Available,

but the Market for Products is Limited
In a situation of this type the manufacturer has adequate quantities of raw material(s), at mar-
ket prices. However, it finds that sales opportunities are too limited to justify setting up large
scale units, and since it cannot profit from economies of scale, it is obliged to set high prices
16 1. Market Analysis

for the product(s). Under these conditions its chances of success vis-a-vis the competition are
strongly affected. It can only look for new outlets, and in particular consider whether the qual-
ities of its own product(s) might enable it to displace various items from their markets: these
might be less sophisticated items, and thus a priori less expensive ones, but the demand for
them should, by significantly raising the unit capacity, lead to a reduction in the cost of pro-
duction. Of course, this will be possible only above a certain threshold, represented by the
minimum operating expenses for the plant. This usually means the costs that are directly
related to the material and energy balance sheets, and the costs of overhead, labor, and main-
tenance (capital costs, provision for depreciation, and financial costs are not taken into
account). These operating expenses are often confused with the marginal cost l . Such consid-
erations must be addressed as soon as the markets for a known product prove to be too lim-
ited, but they are also important when studying the possibilities of penetrating various
markets with a new product, whose qualities could lead to possible substitutions.
In this connection, the case of polymethyl-4 pentene-1, better known as TPX, is illuminat-
ing. TPX’s monomer, obtained by dimerization of propylene, is currently used in the produc-
tion of linear low-density polyethylene, in competition with other a-olefins. At the time when
the marketing of this polymer was under consideration, its special qualities, and in particular
its transparency, presented an opportunity for use as a substitute for a number of more tradi-
tional materials, such as:
0 Glass for laboratory vessels

0 Plastics for medical syringes

0 Porcelain for dishes

0 Transparent plate glass.

A preliminary market study had moreover shown that for these four markets, and within a
particular geographical area, the annual sales volumes and average prices were as shown in
Table 1.2.

Table 1.2 Potential markets for polymethyl-4 pentene-1

Market
Sales volume (4)
(t/Yr)
20 000
I Average price (P)
(local monetary units)
50
10 000 100
15 000 70
50 000 30
(Source: IFP)

1. Generally speaking, when the quantities produced are proportional to the operating costs, the
short-term marginal cost is effectively equal to these costs. This is not always so, especially at low lev-
els of production. Nor is it true for the long-term marginal cost. These concepts, which deserve a more
detailed analysis, will be considered later. They turn out to be very useful in selecting the optimal unit
capacity for a plant.
 1. Market Analysis 17

These elements, although they were quite scanty, were enough to define the curve q = f(P),
giving the variation in total tonnage of annual sales (q)as a function of price (P). A graph of
this kind (Fig. 1.3) is very useful in making an estimate of the size of a plant to be constructed
and in very quickly deciding whether or not it is worth going ahead. In fact, when this kind
of information is available it is possible, right at the outset and based on a simplified eco-
nomic selection tool, to investigate the configuration that will ensure the most profitable
project. In this case the cashflow (CF, see Chapter 2, 0 2.2.2) is well suited to answer the
question. The object is to maximize the following function:
CF=q(P-D)
an equation in which q represents the potential market, P the product's sales price, and D the
operating expenses of the production plants (see Section 1.1.2).
The maximum value of the function CF is obtained by setting the first derivative to zero,
and confirming that the second derivative is negative, i.e.,

dCF dq
-=q +(P-D)-=O
dP dP

d2CF = 2 -
- dq + ( P - D)-d2q < 0
dp2 dP dP2

20 30 40 50 60 70 80 90 100

Price (arbitrary units)

Figure 1.3 Markets and potential prices for polymethyl-4 pentene-1.

(Source: IFP)
18 1. Market Analysis

if we now introduce elasticity:

the optimum price is given by the following expression:

Now, the relation q = f(P) allows us to calculate the coefficient of elasticity q and thus,
knowing D, to determine the optimum price.
In the selected example (Fig. 1.3) we have:
q = A e-BP (or: q = 2.3.105e4.O3l59

The elasticity increases as the price rises. Then:

or: P = 31.7 + D
At its maximum the price must therefore equal the operating expenses plus 31.7 units.
Such a result, which we obtained without knowing the size of the unit, but only from having
some idea of the operating costs, enables us to reach some conclusions about the market that
can be very interesting. Thus, in the present example, we see straightaway that market num-
ber 4, which is moreover the biggest one, is out of reach because the corresponding permis-
sible price is only 30. So the unit capacity of the plant cannot exceed 45000 tlyr. A hrther
calculation would show that D is greater than 20, which also eliminates market number 1, and
brings the maximum size of the unit down to 25 000 tlyr. We can thus see why it was neces-
sary, at the time when such a study was undertaken, to abandon the plan to market TPX.

1.2 FORECASTING METHODS

One of the essential tasks of an evaluation is to make it possible to judge the economic attrac-
tiveness of projects. To do this we must therefore be able to visualize future situations, and
especially to define their characteristics. For this purpose we have to rely on the available
means, in particular the various known methods of forecasting-to the extent that they
answer our specific needs.
In this connection, it is desirable to emphasize at the outset that evaluations look three,
five, or even ten years ahead, and that they are also expected to make comparisons, on a con-
sistent basis, between several situations that could arise at those periods.
 1. Market Analysis 19

1.2.1 FORECASTING METHODS: AREAS OF

APPLICATION AND GENERAL CHARACTER

Many fields of activity use forecasting techniques; they can handle demographic develop-
ments just as well as energy and raw material availability and demand, or possible variations
in financing capacity, manufacturing prices and costs, sales volumes for more or less sophis-
ticated products, even changes in the behavior of consumers, government policies, societies,
etc.
Market analysis, or more generally technological forecasting, which is all that concerns
evaluation, addresses the narrower field of prices and quantities. However, the methods uti-
lized for forecasting in all these cases are based on the general procedures that have developed
over time, such that their manner of application does not differ from one case to another. The
only thing that changes is the nature of the dependent variable; time remains the common
parameter, whether it appears directly or not.
If they are carelessly applied these methods often give rise to certain disappointments,
when the reality turns out to be different from the anticipated results. It is therefore essential
to define their basic character, and especially their limitations, before going on to examine
them in greater detail.
In the first place, the great number and complexity of the parameters that must be consid-
ered in assessing a situation are such that it proves to be impossible to develop a procedure
capable of yielding completely reliable information, especially as some of these variables
cannot be quantified. It is more a matter of obtaining orders of magnitude, or trends, than
definitive values.
Consequently, it is always desirable to maintain a critical attitude towards forecasting tech-
niques, and to bear in mind the possibility of intervening by including a certain measure of
subjectivity: this can provide elements of judgment, especially about economic matters, that
are hard to quantify or express formally. The level of uncertainty that characterizes forecast-
ing has also had the effect, over time, of multiplying the number of procedures and increasing
their level of sophistication, so that there is now a whole arsenal of methods, each one as
imperfect as the next, but all more or less specific and easy to apply. In this case, as in similar
situations throughout the process, the role of the evaluator consists in selecting the proce-
dure(s) according to the problem being faced, the credibility of the available information, the
degree of accuracy required, and above all the time allowed for providing an answer.

1.2.2 FREQUENCY OF FORECAST AND TYPES OF METHODS

Forecasting methods may be classified according to various criteria, particularly by the time
period envisaged: the intervals are related to the frequency of acquiring or analyzing the data
or parameters concerned. From this standpoint four type frequencies are usually distin-
guished:
0 Daily

0 Monthly
20 1. Market Analysis

0 Quarterly
0 Annual.
The first types are quite obviously better suited to the near term than the latter ones.
Looked at in this way, forecasting methods can theoretically be grouped as follows:
0 Very short-term forecasts, for periods under three months: these rely heavily on descrip-

tive methods such as breakdown into time series, exponential smoothing, stochastic pro-
cesses (especially Box-Jenluns), distribution models (e.g., Gompertz), etc.
0 Short-term forecasts covering periods of from three months to a year: the usual approach

is to extrapolate from the previous forecasts, but simplified regression models are pre-
ferred for use here.
0 Medium-term forecasts, covering one to three years: the preferred procedures in this

case are multiple regressions, econometric models, the Box-Jenkins method with trans-
fer, i.e., with loop or feedback.
0 Long-term forecasts, covering periods of three to five years: the most suitable tech-

niques in this case are econometric models, distribution models, qualitative methods
based on decision theory, for example, scenario methods such as Zwisky’s morpholog-
ical techniques, the Delphi survey, etc.
0 Very long-term forecasts, for periods greater than five years: the procedures concerned

are mostly the methods of what is called futurology.

Under this classification the kind of market analysis used in evaluation would be called
long-term or even very long-term, which in theory would forbid the use of some of the
methods of shorter-term forecasting, but which often have the advantage of being easier to
apply. They should therefore not be completely ignored, because even though he is well aware
that the results will suffer, the evaluator’s objective is only to define a framework for study.
This being the case, it seems preferable to incline towards a different classification, based
in part on the level of complexity of the methods, and even more on the scope and the nature
of the means to be compiled or employed. In this case, we may distinguish:
0 Objective procedures based on analysis of past and present, i.e., on “historical data” and

their extrapolation: these have the advantage of being quantitative and thus making it
easier to apply formulas or modeling. Morphological analysis, whose object is to exam-
ine all the possible solutions for a given problem, one by one, also falls under this
approach.
0 Intuitive procedures that look essentially towards the future: these bring qualitative and

quantitative matters together and thus involve taking subjective factors into consider-
ation. The Delphi survey and scenario theory are prominent examples.

1.2.3 PROCESSING A N D EXTRAPOLATION OF HISTORICAL DATA

This type of procedure finds its justification in the application of the following principles, in
whole or in part:
0 Observed changes occur by means of progressive modification only, without disconti-

nuity: in this case, time is the only independent variable.

 1. Market Analysis 21

0 Similar changes occur when the agents of change are the same: several independentvari-
ables thus have to be considered.
In these circumstances, the resulting methods may be divided into two groups:
0 Those that have only one endogenous variable (time or an equivalent parameter), which
are called “internal”,
0 Those that have this endogenous variable, and also exogenous variables, which are

called “external”.
However, they both have the same three-stage manner of application:
0 Data analysis

0 Modeling to provide forecasts

0 Checking of the forecasts.

Obviously the last step is the trickiest to achieve, especially as the period under consider-
ation is far-off. In practice, this procedure cannot be used except for short-term or very short-
term forecasts. It loses all meaning for the most distant outlooks, since even if there is no dis-
continuity, profound changes may occur over such a period, making it necessary to consider
subjective factors.

1.2.3.1 Internal or Endogenous Methods

These require the availability of chronological (or chronicled) data and observations. For
this reason, in theory they can be applied only to existing products, whether tonnages or
price are concerned. In practice, they also apply, by extension, to substitute products.
Depending on the frequency of the observations, it is possible to distinguish various kinds
of phenomena:
0 A general trend

0 Seasonal variations

0 Cyclical variations

0 Economic fluctuations, which can however be explained and quantified.

Evaluation is in general little concerned with seasonal or isolated effects, as long as they
can easily be recognized as such. More specifically, it is recommended that evaluations
ignore them completely. On the other hand it is essential to identify trends. In this connection,
we may mention the various procedures that follow.

A. Graphical Adjustment and Extrapolation of Data

Individual observations made over a period of time are shown graphically as a series of points
having apparent discontinuities; a quick solution consists in expressing the intuitive or logical
continuity that exists by drawing a straight line or more generally a curve through these
points, virtually freehand, and extending it so as to “project the past into the future”. Anglo-
Saxons sometimes call these exercises simple, even though this term also covers more sophis-
ticated techniques, especially regressions, based on the mathematical processing of infor-
mation without any attempt to identify causes: time, although it is an independent variable,
22 1. Market Analysis

does not qualify as a real cause (moving average methods, least squares, various statistical
smoothings or adjustments, etc.).

B. Moving Average Method

This procedure aims to perform mathematical processing and then to derive a graph that elim-
inates or at least reduces variations that are random, accidental, or due to external causes,
even seasonal variations, so as to portray only the overall trend.
It involves first of all determining the “moving total” at a particular time t, over a period
n + 1, by calculating:

Yi being the observation at time i, and then obtaining the “moving average” from the relation:

in practice, n varies by discrete amounts; in this case, if n is an odd number, we take the value
of the median observation to be the sum of the two observed values that surround time t, and
for n/2 the smaller integer. For example, if n is expressed in months, for n + 1 = 13 the moving
average is obtained by considering the month corresponding to the time t, the six preceding
it and the six following it, and dividing the total by 13. For n + 1 = 12, it is calculated from
the sum of the two consecutive months surrounding t, the five preceding and the five follow-
ing, and dividing the whole by 12.
When the data cover a relatively short period of time, the graphs of moving averages fall
on a “quasi straight trend line”. Over a longer period regular fluctuations appear, defining a
cyclical phenomenon, whose amplitude and period may be calculated and used to determine
cyclic coefficients that can be used for future projections.
It is also possible to smooth out this effect by taking the moving average of several moving
averages and so on, so that at the end of the process we see only a straight trend line, to the
extent that this sort of thing retains any meaning.

C. Linear Regressions: the Method of Least Squares

In contrast to the preceding technique, here we start by looking for a straight “trend” or
“regression” line, or more generally a curve-that can be mathematically straightened.
In so-called regression operations, we are in fact looking for a relation of pre-selected type
that expresses the connection between a variable Y (observation), described as dependent or
explained, and one or more other variables ( X I ,X2, X3, etc.) said to be independent, explan-
 1. Market Analysis 23

atory, or descriptive (time, for example). The graphic expression: j7 = f(xl, x2,x3, ...) defines
a regression 2 curve or surface.
In the case of two variables Xand Y, the dependence gives rise to a set of points M having
the coordinates (x, j ) ,representing the average of they values corresponding to a single value
ofx. The broken line joining these points constitutes the line of regression of Y onX, to which
a regression curve may to be fitted. If this curve is a straight line of the form:
y(x) = ax + b or y(x) - J =a (x - J )
we speak of linear regression, a being the regression coefficient, where:

The operating principle of the least squares method for finding straight lines of regression
is to minimize the sum of the squares of the distances from each point M to this straight line,
measured parallel to the y-axis. Thus for two points Mj(xi, yi) and h$(xi, y(xj)), having the
same x-value, and located on the plane (x, y ) and on the straight line y (x) respectively:

must be set to the minimum.

COMMENTS
1. If y is linked to several variables (xl, x2, x3, etc.), we can use the least squares
method to find a multiple linear regression having the form:y(xl, ~ 2 , ~..3. x,)
, = a1x1
+~ 2 x +2 ~ 3 x+
3 ... + a,x, + b, in which al, a2, a3, ..., a, are partial coefficients of
regression that express the separate influence of each of the variables on y. In an
(n + 1)-dimensional space, this equation corresponds to the regression plane, such
that the sum of the squares of the distances to this plane from the points represented,
measured parallel to the y-axis, is a minimum.
2. If the adjustment is made by using a non-linear function, we speak of a curvilinear
regression.
0 In some cases (logarithmic, exponential, hyperbolic, parabolic, etc.) forms, a

change of variable may allow us to get back to a linear correlation, which will be
much easier to project into the future (anamorphosis).
In other cases, the adjustment may be made on curves representing polynomial func-
tions such as y = a + bx + ex2. The constants a, b, c are defined such that the sum of
the squares of the distances of the experimental points to the curve (measured paral-
lel to the axis of the dependent variable y ) is a minimum.

2. This term, which is very widely employed, especially in statistical data processing, was first
used by Sir F. Galton in 1886. A. Chupov in 1925 preferred the term stochastic relationshipwhich,
however, has been less popular since that time-to better express the random nature of the dependence
being studied.
24 1. Market Analysis

In effect, to minimize:

Ci [vi -y(xi)I2 = Ci bi - a - bxi - cx! )2,

when xi and yi are known, amounts to setting to zero the partial derivatives of this
expression with respect to a, b, c, etc. Solving the corresponding group of linear
equations enables us to calculate these coefficients.
0 The Gompertz distribution form or model, described by the equation:

y = abcf or y = ea--bc‘

where a, b, c are constants and b and c lie between 0 and 1, allows adjustments to be
made using a type of curve that is very often encountered in practice, and which is
particularly useful for describing the variation over time of a given product’s sales.
These are so-called S-shaped curves which, by using the FERT method, bring
out four stages in the life of a product: a first, uncertain or inductive, and called
“funny” (F); a second, exponential development, called “exciting” (E); a third, linear
growth, called “reliable” (R); and finally a fourth, decline, called “trivial” (T).

Figure 1.4 shows various forms of curvilinear regressions (parabolic and cubic), by way
of illustration.

D. Exponential Smoothing
In contrast with the preceding methods, which assign the same weight to all observations
regardless of their age, and give more importance to numbers than to quality, this type of
operation has the advantage of favoring the more recent data over those of more distant date,
by giving them a greater weight. Doing this enables us to better express the continuing effects
of certain parameters, and to limit those belonging to factors that are now inactive-and
thereby to provide a better picture of recent changes in a trend.
The observations are weighted by using an inverse geometrical progression, such that the
further back the observation lies in the past, the weaker its influence becomes. If a represents
the weight of the present as compared with the past, we find:

r, = a 5-1 +(1 -a)r,-l

where
<
Yt-l
: observation or forecast for the current or future period
: last observation
-
YtPl:weighting of the preceding period
a : weighting factor, lying between 0 and 1, calculated approximately from historical
data, and expressed in a manner similar to the preceding equation.
The general expression for Y, is built up little by little as follows:
-
r, = aY,1 + a(1 - a)Yt-z + a(1 - a)2 Yt-3 + ... + a(1 - a)”-’ Y,, + a(1 - a)” Y,-”
If the number of observations is high enough, the last term becomes negligible.
 1. Market Analysis 25

Y Y

9, = a + b x + c x2
Y Y

log ?,= a + b x + c x2
Y Y

9, = a + b log x + c(log x)2

Y

Figure 1.4 Curvilinear regression (example curves of parabolic and cubic

shape). (Source: IFP)

It must also be emphasized that this method can be just as well applied to monthly or quar-
terly data as it can to annual ones. In this way, seasonal or cyclic-type variations may be
brought out. Giving more importance to recent observations does introduce major possibili-
ties for error in making projections. In this case, it is necessary to employ exponential
smoothing only on information from which seasonal effects have been removed.
“Deseasonalization” can be applied to historical data by considering:
0 Either mean monthly differences, for example (Buys-Ballot method), calculated by sub-

tracting from the overall monthly mean a series of observations covering several years,
the averages for this period of time being determined for each of the year’s twelve
months,
26 1. Market Analysis

0 Or coefficients obtained by dividing the various preceding monthly means by the overall
monthly mean.
This is done before processing the data by one of the procedures already described, espe-
cially by exponential smoothing, and then by reversing it after the future projection is made,
to restore the seasonal effects.

COMMENTS
In this type of calculation, it is essential to distinguish cyclic variations from random
fluctuations, which may in particular arise from economic phenomena such as
strikes, breakdowns in supply, temporary incentives or constraints, etc. The corre-
sponding observations have to be eliminated or if possible corrected. The Box-Jen-
kins model, published in the United States in 1970, is a generalized, sophisticated
version of exponential smoothing. It is relatively complex, but on the other hand
more accurate, and like other internal forecasting methods is especially applicable to
the short or very short term. However, the drawback of the complexity is that it finds
few applications in evaluation, which is why we will not consider it at length here.
We will simply take note of its principle: it distinguishes between seasonal and non-
seasonal data, handles a variable number of parameters grouped under three head-
ings (differential or stationary parameters, moving average parameters, auto-regres-
sion parameters) and includes methods for confirmation and diagnosis (in particular,
autocorrelation).

1.2.3.2 External or Exogenous Methods

These are based simultaneously on having time series of data, and on taking into account
external variables, regarded as causal.
In theory, like internal methods, they apply only to known products, since it is necessary
to have observations from the past, but by extension they may also be utilized for new prod-
ucts, when there is direct substitution.
Their credibility level depends on that of the available information. They are also more
complex than their so-called internal equivalents, consequently more difficult to apply; in
addition, identification of the exogenous variables is not always a simple matter.
The main procedures found in this type of technique are as follows.

A. Mathematical Correlation with a Precursor Index

As a rule, this is the preferred tool for making a forecast, and in various forms it is very widely
used in evaluation, although in theory its use for long-term or, especially very long-term pro-
jections is questionable.
The procedures based on it are among the simplest of exogenous methods, both in concept
and in application. In fact they consist of finding an independent variable, referred to as an
“index”, that can be connected to the series of collected information with as high a correlation
level as possible (e.g., above 90%). The mathematical representation of the relationship thus
obtained should allow the observations to be extrapolated into the future as a function of the
 1. Market Analysis 27

already known or predicted evolution of the exogenous parameter, depending on whether or

not there is a shift over time-also called a “log”-between the two types of variable.
The index must be carefully selected: this is the economic analyst’s true job; he finds it
particularly among the macro-economic indicators, whose past and future variations are, inci-
dentally, studied or drawn up by specialized organizations (in France, for example, by
INSEE), with the help of very sophisticated and comprehensive models. Notable ones include
the Gross Domestic Production (GDP) 3, Gross Domestic Product (GDP) 3, Gross National
Product (GNP), the GNP per capita, indices for the variation of industrial production (IPI),
whose variation for Western Europe is shown in Fig. 1.5, wages, prices, energy consumption,
etc. Figures 1.6a and 1.6b provide examples of such a correlation between the energy and pri-
mary organic chemical sectors.

1970 1975 1980 1985 1990 1995 2000

Figure 1.5 Variation of the Industrial Production Index (IPI) for Western
Europe (OECD Europe). (Source: OECD)

3 . As a first approximation, for a given country these indicators correspond to the following defi-
nitions:
Gross National Product (GNP) = Gross Domestic Product (GDP) + Balance of Exchange with the
rest of the world.
Gross Domestic Product (GDP) = Gross Domestic Production (GDP) + Non-commercial Services
(1976 SECN Agreement or 1970 SEC equivalent, i.e., SystPme Elargi de Comptabilite Nationale, or
SystPme EuropCen des Comptes 6conomiques integrb: Expanded European System of National
Accounting, or Integrated European Economic Accounts System).
Gross Domestic Product (GDP) = total of the added values for all goods + total of added values of
commercial services (UN Convention referred to as the SCN =Systtme des Comptes NormalisPs:
Standardized Accounts System).
28 1. Market Analysis

Figure 1.6a Total energy consumption in Western Europe as a function of

the Industrial Production Index. (Source: BP Statistical Review)

80 90 100 110 120 130 140

Industrial Production Index (base 100 in 1985)

Figure 1.6b Ethylene consumption in Western Europe as a function of the

Industrial Production Index (IPI). (Source: SH)
 1. Market Analysis 29

The external variable may also be selected from among various statistical reference series,
particularly demographic ones (marriages, births, education, etc., so as to obtain, for example,
some ideas about consumption of food, clothing, items of equipment, etc.), socio-demo-
graphic ones (distribution of socio-professional categories, etc.), administrative ones (permits
for construction, driving, etc., telephone, broadcasting, and insurance fees, etc.), economic
ones (incomes, purchasing power, etc.), and manufacturing ones (various kinds of packaging,
depending on variations in the demand for the contents, etc.).

COMMENTS
1. In some cases, it may be useful to define relationships with several indices at the
same time, with a view to reducing some of the inertia and thereby, using a single
mathematical formula, to improve the overall correlation, i.e., the accuracy of the
eventual forecasts. The increase in the number of variables considered is accompa-
nied by a much greater complexity in the sophistication and use of the formula, and
in fact makes it necessary to develop true econometric models (see Q 1.2.3.2B).
2. So as to at least partially correct the shortcomings, for evaluation purposes, of
exogenous methods that employ precursor indices, it is often desirable to accompany
this type of procedure by what are called “scenarios”. This is done by assigning a
number of different values to the external variable: values, both higher and lower,
that the variable could have at some future time in various economic situations. The
principal characteristics of these future situations must first be imagined and
described in detail, in order to justify the particular values selected.
The probability of reaching or falling short of a minimum threshold for carrying out
a project, according to the more or less optimistic or pessimistic hypotheses consid-
ered (rosy, neutral, or gloomy scenarios), thus becomes a determining element in
deciding the fate of the project studied.
3. The precursor index method, along with the so-called “analogous” or “historical
comparison” procedures, is particularly applicable to new products, to the extent that
they may be directly substituted for existing materials. It also applies to products
already on sale, but whose development and behavior resemble, with a certain delay,
those of materials that have been available on the market for a longer time. An exam-
ple is offered by the evaluation of markets for household appliances, photographic
equipment (instant cameras, disposables, etc.), televisions, videotape recorders
(whose market is related to the preceding ones, and their replacement, etc.). Both an
example and a counter-example may be found in a comparison of the variation of
demand for polyamides (nylons), acrylics, and polyolefins in the United States
(Fig. 1.7). Until about 1970 the developments observed were essentiallyparallel, with
delays between the first product type and each of the two others, over a very long
period, of around eight years and 14-16 years respectively. Knowing the growth in
demand for nylons, it was thus possible to make very long-term growth forecasts for
the other two fibers. Although this empirical rule later held fairly true for the polyole-
fins, it turned out to be completely wrong for the acrylics, whose decline resulted from
their inabilitj-unlike their competitor products-to adapt to modem application
techniques, which were particularly designed to eliminate certain conversion steps.
30 1. Market Analysis

10 000

1 000

1940 1950 1960 1970 1980 1990 2000

Figure 1.7 Comparative variations in demand for polyamide, acrylic, and

polyolefin fibers in the United States. (Source: SRZ)

B. Econometric Models
These are significantly more complex than simple mathematical correlations with a single
index; but they offer the advantage of taking better account, by quantifying them, of the reac-
tions of the markets when confronted by various economic changes, especially variations in
the prices of raw materials and energy. Although they are better suited to longer term fore-
casting, they are not greatly employed in evaluation, owing to their relative complexity. Con-
sequently, the description given here is deliberately brief. Among the principal models of this
type we may distinguish:
0 Models with several explanatory variables, i.e., correlation and multiple regression

models. These parameters have already been mentioned above: in practice, it is a matter of
extending the so-called precursor index procedure to accommodate the simultaneous use of
several indices. But in this case, the work of the economic analyst is a lot more complicated,
since he first has to identify the exogenous (also termed independent or explanatory) vari-
ables, then compare the time series that he wants to correlate with the ones corresponding to
each of the parameters selected, determine the levels and the coefficients of correlation for
each variable treated separately, and finally define a single regression by adjusting the weight
of each component.
 1. Market Analysis 31

In practice, there are plenty of programs for facilitating statistical data analysis, choosing the
most significant indicators, and finding the best-fitting regression, particularly by seeking to
minimize the sum of the squares of the remainders (squares of the distances, measured paral-
lel to the axis of the dependent variable, between the various observations and their equiva-
lents on the representative curve).
0 Behavioral models. These have a double advantage over the preceding methods.

Firstly, they allow for unusual parameters: ones therefore not included in the list of explana-
tory variables regularly considered, but which may be suggested by an examination of the
context in which the study is being conducted. These are usually phenomena of an economic,
decision-based-even unilateral and arbitrary-nature, but nevertheless capable of operating
over relatively long periods or even of having a decisive effect. This is particularly true of
“heavy” fashion trends, or political or administrative decision-making that could have major
impacts on the demands or prices for various products. By way of example, in some market
analyses conducted for actual evaluations concerning medium-term prospects for refining
and petrochemistry, it is essential to consider the effects of a possible limit on the benzene
content cf gasoline, even though the measure might still be far from being approved, and even
though alternative solutions would immediately be proposed.
Secondly, they make it easier to identify the truly exogenous variables, by integrating the
probable correlations between various parameters that, at first blush, appear to be indepen-
dent. This applies to the volume-price relations examined previously, which, when inserted
into models, remove price as an explanatory factor.
0 Economic analysis models. These deal more specifically with short-term forecasts and are

particularly well suited to defining the prospects for equipment purchases. In fact, in this case
they take into account not only general market trends and corresponding parameters, but also
current economic conditions (inflation, credit terms, household funds, loosening or tightening
of fiscal measures, levels of saving, etc.). However, they may have some uses in evaluation,
particularly in forecasts of the demand for various kinds of equipment (rotating machines,
machine-tools, etc.) or products (catalysts, pharmaceutical products, etc.)--to the extent that
the models distinguish between the initial markets, which depend essentially on techno-eco-
nomic performance, and the repeat markets, which reflect the existing stocks, the probable
lifetimes, and the functional reliabilities of the equipment or the quality of the products.

C. Inter-Industry Exchange Models

These are relatively complex and rigid procedures that apply to a whole industry or an entire
industrial sector, and therefore require the availability of a great number of data. They are not
well suited to the requirements of evaluation, although they have been used for this purpose
in the past. In addition, they have the further drawback of poor sensitivity to variations in
explanatory parameters. Among the best known and most recently developed examples is the
input-output method.
These models are designed to take advantage of both of the common approaches to fore-
casting, namely:
0 Specific market studies

Overall sectoral analyses.

32 1. Market Analysis

The first of these has trouble with accommodating external factors, especially over the
long term; the second yields insufficient detail about products.
Input-output methods derive from the work of Wassily Leontieff, who showed in the early
1930s that a country’s total production could be broken down according to its kinds and allo-
cations. The resulting tables, which are processed mathematically by using matrix calcula-
tions, enable us both to have a clear picture of the economic structure at a given moment, and
to make forecasts. They usually have n + k columns and n + m rows, corresponding to a
nation’s various sectors of activity, and contain zeros or positive numbers representing mon-
etary values.
The elements aii such that 1 I i I n and 1 I j I n express in monetary terms the total goods
supplied by sector i to sectorj.
The elements aii such that 1 I i I n and n + 1 IjI k express correspond to the values added
by sectorj (wages, taxes, profits, etc.).
The elements aii such that n + 1 I i I m and 1 Ij I n express the final demand (private and
public outlays, and commercial transactions).
In summary, the output of a particular sector has two components:
One corresponding to the inputs of the other sectors: this is the intermediate demand or
consumption,
The other brings together all the utilizations as such: this is the final demand or con-
sumption.
As an example, we may suppose that a particular date and for a given country the electric-
ity sector is divided up in the following manner:
Agriculture 238 MF
Fisheries 2
Industries 4 905
Families 1898
Public services 478
Total 7521
In most of these sectors electricity is utilized as such, but in the industry sector it enables
transformations, such that its total production may be expressed as follows:
Total production = Intermediate demand + Final demand
7521 4 905 2616
When such an operation is performed for the sectors considered, vectors are obtained for
the total production and the final demand, and a square matrix for the intermediate demand.
The main problem that is encountered when applying such a procedure to a particular sector
lies in the distribution and grouping of the data (breakdown and aggregation).
When applied to chemistry, for example, an inputloutput matrix for a given date appears
as a double-entry table:
0 The columns correspond to the principal sectors of economic activity for a region, a

country, or a group of countries (automotive sector, building, paper products, food, etc.):
they contain the various products whose markets we wish to study.
The rows correspond to the chemical products selected (polyvinyl chloride, polyethyl-
ene, polyester fibers, etc.).
 1. Market Analysis 33

The figures appearing in this table may be expressed in volumes (tons), and represent the
quantities of each product used in each sector.
For their part, the national accounts provide the results in monetary terms (local curren-
cies) for the various sectors under consideration. They also allow us to define the correspond-
ing vector, so that we can determine a “coefficient of penetration” for each product at a given
point in time in each of the sectors under consideration, by calculating, for example, ratios
such as polyethylene (tons)/automotive sector (dollars).
If we now define a new matrix, comprising:
Columns representing industrial activity sectors, and
0 Rows representing chemical products,

and showing the penetration coefficients, multiplying this table by the vectors of the indus-
trial sectors in monetary terms will return us to the tonnages previously considered. But when
presented in this form the input/output matrix allows us to perform market studies and
medium- to long-term forecasts. In fact, we can use it on two levels:
0 The sectoral analysis level, i.e., using the value vector, whose structure is based:

- Either on accounting or statistical data, when it concerns historical, i.e., existing

information,
- Or on developments we can forecast, particularly by linking them to macro-economic
indicators (GDP, IPI, etc.),
The penetration coefficient level: these can be changed to reflect new possibilities (tech-
nological improvements, different patent situation, new processes, etc.),
and thereby calculate the tonnages for each of the products studied in the various sectors.
lnput/output tables, which were originally intended to portray intersectoral transactions,
also enable price forecasting (Fig. 1.8).

Intermediate Final Overall

demands demand demand
Yi Xi

+
Added
value
vi
IF/tl
-

Figure 1.8 Use of input-output models in quantity/priceforecasting.

(Sources: Battelle & IFP)
34 1. Market Analysis

In fact, each product is manufactured from certain “upstream” components, and contrib-
utes to obtaining various other ones “downstream”. Assuming that losses are negligible, we
can deduce a matrix based on the average outputs of the transformations employed, consisting
of columns and rows for the products considered, and dimensionless (tonnagekonnage) coef-
ficients bij in the table itself.
The demand for each product is the sum of the intermediate demands (transactions) and
the “final” demand for the product itself, utilized as such. It is written:
Xi = CbijXj + Yi
(overall demand) (exchanges) (final demand)

Solving this matrix system enables determination of the vector Xof the demands Xi.
We obtain a similar formula for prices. The price Pi of a given product depends both on
the material cost, Pj (purchase of raw materials), and on the production costs, vi (added
value), such that we may write the resulting relationship:
Pi - C bij Pj + Vi
(product price) (cost of materials) (added value)

The solution of this matrix system enables the determination of the vector P for the prices
Pi. Any forecast based on the matrix of coefficients bij may be expressed in concrete terms
by projections concerning price variations.

1.2.4 DOMINANTLY FUTURE-ORIENTED METHODS

In contrast with methods based on the processing and extrapolation of historical data, which
we may describe as premised on the absence of discontinuity, procedures that place more
importance on direct analysis of the future consider that there can be breaks in development.
These may be caused by sudden changes in conditions, whether economic, political, or tech-
nical. As opposed to the analytical, logical, and somewhat static nature of the first methods,
the latter ones thus offer a more dynamic approach, and the potential to include irrational and
novel factors. They reconcile both qualitative and quantitative aspects, in contrast to methods
that look to the past and consist essentially of modeling, formalizing, and quantifying. Their
very principle makes them especially suitable for forecasting the markets for new products.

1.2.4.1 The Various Interested Parties

Generally speaking, the procedures considered here take the form of investigations, which
can be conducted in various ways and at various levels. They can be divided broadly into
three main, large groups, according to the persons consulted.

A. Representatives
We refer here to vendors. In this case it is a matter of short- or very short-term forecasts,
which are not really of interest for evaluation purposes, because the future they address is too
 1. Market Analysis 35

close, and/or because they are designed for producer companies that have well-developed
marketing departments.

B. Customers

Forecasts based on customer opinions address the short and even the medium term. For this
reason they are closer to the concerns of evaluators and in fact are derived from market stud-
ies, performed either in a purely quantitative manner or semi-qualitatively.
In the first case, a representative group of customers are selected and a formal investigation
is conducted among them; periodic repetitions may provide necessary corrective or moderat-
ing factors, so that by looking at previous results it becomes possible to ensure a better cor-
respondence in the future between the forecast and the reality.
In the second case, more specific for industrial equipment items and new products, several
members of each of the entities included in the sampling are questioned: they are chosen on
the basis of their various functions. The investigations address not only the proposed product
but also its competition. The bulk of the study is devoted to the rationalization, homogeniza-
tion, and interpretation of the results: the latitude of the procedure makes it to some extent
qualitative.
In the case of renewable goods, many models have been developed in an attempt to simu-
late customer behavior, most of them complex and requiring preliminary investigations. In
particular they include the ones proposed by Urban, A.D. Anstutz, and others, and the
“Trade-off’ software based on the analysis of consumer choice criteria.
In addition to these, we must mention the so-called experimental technique, which consists
of conducting full-scale tests, or tests in specific markets, of certain customer reactions to var-
ious changes (price, packaging, advertising effects, etc.).

C. Experts: the Delphi Method

Procedures that rely on expert consultation are specifically technological forecasts, for the
medium and long term. For this reason, they are perfectly suited to the needs of evaluation.
On the other hand, they are usually onerous to apply and require considerable time to yield
results. This inertia reduces their effectiveness and discourages their use.
They were originally developed from the work of 0. Helmer, N.C. Dalkey, and T.J. Gor-
don of the Rand Corporation, published in the 1960s, who perfected the so-called Delphi or
Delphic method. In this technique, the opinions of a group of experts, selected for their qual-
ifications in a particular area of activity or in a given subject, are compared. It is based on the
following two principles:
0 Maintenance of the anonymity of the persons selected, so as to avoid mutual influences

that could distort the results,

0 Repetition of the consultation procedure for several rounds (four, when the method was

first applied), combined with a recirculation of the information to all the experts being
consulted between each round: the objective being to eliminate the extreme opinions at
each repetition, or at least to require a justification, so as to progressively reduce the
spread of results until they become significant for forecasting purposes.
36 1. Market Analysis

In the most recent applications, special emphasis has been placed on improving speed and
efficiency, first by limiting the number of rounds (one or two at most), then by carefully draw-
ing up the questionnaires. These are “quali-quantitative” in nature and allow the experts to
reply very easily and very quickly by simply checking boxes; in addition, this way of doing
things lends itself readily to collecting and processing the information by computer.
By way of example, we may cite the investigation undertaken in the early 1980s by Saga
Petrokjemi AS&Co. (Norway) on the possible potential for using methanol as a fuel for auto-
mobiles in the United States and Western Europe. The goal in this case was to measure the
effects on the exploration and development of gas deposits in the North Sea, which would
then be a preferred source of raw material. The initial questionnaire provided for 35 replies,
some of a quantitative nature (time periods or values considered to be most probable), others
more qualitative in kind (level of agreement with a given statement, for example).

1.2.4.2 General Procedure for a Typical Study

The market where these investigations are considered most effective for defining the devel-
opment potential for a product, a piece of equipment, or a service, is certainly among compa-
nies, and in particular small- and medium-sized companies and industries (makers of
equipment, producers, distributors, service providers, etc.). The procedure to be followed in
this case conforms to the general principles, which it is essential to respect if the operation is
to be successful. However, one should not expect too much from the results, especially the
response rate, which must be considered satisfactory if it exceed 20%. In addition, one has to
bear in mind that this type of method is clumsy to use and involves substantial delays (six
months to one year usually); therefore it is preferable to apply it in successive stages, thereby
reserving the possibility of terminating it fairly quickly and accepting partial information.

A. Internal Audit

The first step consists of assembling in-house as much information as possible on the “prod-
uct” (in the broadest sense of the term). Although it would be nice to have economic or mar-
keting data about it at the outset (markets tested to date, actors, competition, price, volumes
or quantities), the first priority at this stage should be the collection of technical information,
so as to properly identify the performances and especially the preferred area for intervention.
For this purpose, the framework of the main section of the questionnaires that will support
the subsequent steps must be drawn up. This stage may be the time to initiate parallel analyses
of the legal, financial, and other aspects, or of patent rights, whose results sometimes result
in halting any further investigation. This part of the study is in any case no more than an infor-
mal inquiry among the experts of the company that created the product in question.

B. External Audit

The second step consists of undertaking a similar task externally, but in a much more formal
and structured fashion, so as to produce a definitive questionnaire. In this case, it is a matter
of drafting a guide for interviews, whose contents must be validated by a small number of
 1. Market Analysis 37

actors (10- 1 9 , through direct contacts with the persons selected, who will have recognized
skills in the field. The objective of these interviews 4, which should not take longer than an
hour and use advanced draft documents, completedjointly by both parties, which confirm the
seriousness of the operation, is three-fold:
0 To validate the content of the questions already selected and their formulation,

0 To complete the questionnaire, if need be, by adding items addressing the new concerns

of the actors consulted,

0 To trim the document by focusing it on the most promising markets: this particularly

applies to everything that concerns the technical characteristics of the product or its com-
petitors.
The content of the interview guide now remains to be defined. It must already exist in
semi-final form, i.e., it should be as “quali-quantitative” as possible, so as to reduce the time
for obtaining replies to a minimum. In practice, each question should give rise to a matrix (a
table) with a small number of rows (fewer than ten) of pre-selected identification criteria, in
other words the ones considered to be relevant and significant. The quantification is per-
formed in columns with pre-determined variation intervals. The object should be to make the
maximum use of the principle of checking boxes. However, it is also possible to envisage
grading scales (1 to 4 or I to 10 for example). In this case, it is best to choose an even number
of possibilities, so as to avoid a median, neutral grade, and to make the person take a position.
One may also ask for a preferential ranking of the criteria from 1 to n, and, as a last resort, an
evaluation, but it should be quantitative. And the person consulted should always be left some
degree of freedom, by providing a “Miscellaneous” or “Other” heading, while of course ask-
ing for the additions to be specified, i.e., numbered.
At this point, we can see the importance of consulting internal experts, who are the only
ones who can help to define these matrices. We can also see how desirable it may be to offer
a broad range for quantification at the beginning, so that when the interviews are completed
we can be certain of being able to properly re-center the questionnaire on the intervalsjudged
by the external audit panel to be the most promising.
When all is said and done, the compilation of these tables amounts to a genuine multi-cri-
terion analysis which, as far as the interview guides is concerned, actually consists of three
parts:
The first has to do with the company, represented by one or more parameters (labor force,
sales, activities, installations, export figures, level of facilities, preferred suppliers, etc.).
The second is essentially technical and concerns the product studied or, more precisely,
the general characteristics of competitor products on the market. In fact, it is important
not to disclose the level of performance attained, but instead to refer to the ones most
commonly encountered or sought after.
The last concerns marketing. It aims to measure people’s levels of satisfaction, and their
expectations. It also enables an assessment of the person’s view of the market: identifi-
cation and relative weighting of the actors, current price levels, acceptable additional
costs for better performance or to provide supplementary services.

4. In its simplified form, and to gain significant time, the investigation may be halted at this inter-
view stage, which moreover can be conducted partly on the telephone.
 Next Page

38 1. Market Analysis

C. The Investigation Proper

Upon completion of the external audit, the interview guide should lead to a relatively short,
clear, precise questionnaire, concise but nevertheless easy to read, which should not take
more than 10to 15 minutes to fill out. A personalized letter should accompany this document,
stating the purpose of the investigation, and motivating the addressee by assuring him that his
answers are confidential and that he will receive the overall results if he agrees to participate.
A technical brochure may possibly be attached. The number of addressees should preferably
be limited to 100 (maximum 200).
In this type of procedure, it is essential to avoid vertical filing, i.e., direct transfer to the
wastepaper basket. To do this, it therefore becomes important to carefully identify beforehand
the persons and not the companies to which it is mailed. It is better to have a limited list of
this type than to have a bigger but anonymous sampling. An effective solution consists of
making an introductory phone call before the mailing, and if no reply is received within a cer-
tain time (one to two months), to send it out again. But the bottom line is that the operation
may prove to be both arduous and costly. Thus, while waiting for a better solution, another
attractive method of communication to be considered is e-mail, or fax, which conceals the
banality of the questionnaire by including it among a shower of announcements and miscel-
laneous advertising. Here again the procedure may prove onerous and difficult, particularly
when the geographic coverage is a large one: finding the right addresses and numbers is not
always easy.

1.2.5 THE LIMITS OF FORECASTING METHODS

In theory, forecasting procedures should be applied in three stages:

0 Collection and analysis of information, whether it consists of historical data or the

results of an investigation,
Construction of a model for forecasting purposes,
0 Checking the forecasts.

The reality is very different, especially in evaluation. In fact, experience shows that the
third stage is almost never reached. It is rare for any follow-up to be scheduled for a market
study. Because they are generally one-time events, such studies are only very seldom
repeated at regular intervals or sufficiently close together in time for it to be considered
worthwhile-all other things being equal-to confirm the level of agreement between the
initial forecasts and the reality. The checks performed are always too partial or too general.
In the evaluator’s defense, it must be remembered that he wields an extraordinarily change-
able weapon, both in its design and as concerns its possibilities and its credibility, and that
what is asked of him is first of all to define a frame for further work. In that same defense, we
should mention the general disillusion displayed in recent years by all forecasters and all
kinds of user vis-a-vis models, particularly towards the most complex and sophisticated ones,
most of which have failed miserably to live up to their claims. Finally the information is not
always available, at least when it is needed: the compilation and upkeep of data-banks, or
even the mere subscription required to have access to one, let alone the conduct of meaningful
Previous Page

1. Market Analysis 39

studies and the correct interpretation of their results, often appear too burdensome and too
lengthy to undertake.
In conclusion, in the area of market analysis, as well as in the areas that come within its
field of application, the only course for evaluation is to make the most of some of the panoply
of forecasting methods available, while continuously adapting to the reliability of the infor-
mation available and to the time constraints imposed. From this point of view, methods based
on the extrapolation of historical data, combined with the use of several scenarios, are an
evaluator’s most suitable choice and, relatively speaking, the most efficient one. This is in
spite of their mediocre record for the medium and long term, and their even worse one in the
very long term.

1.3 SPECIFIC APPLICATION METHODS

FOR FORECASTING PRICES A N D VOLUMES

Although the same procedure may in theory be applied to processing both price and volume
data, experience shows that the specificities of each of these two fields should be considered
before making a choice among the many methods available. In fact, in practice there is some
segregation, and a preferential use of certain ways of performing regressions and extrapola-
tions rather than others, depending on the kind of information concerned.
Here, we will not be dealing with the various kinds of price usually encountered in the area
of marketing, nor with the mechanisms that govern their setting. This matter is discussed in
detail in Chapter 2, 5 2.2.1.2. The main emphasis here is on forecasting and methodological
aspects.

1.3.1 PRICE A PPLICA TIONS

Concerning price, the attitude taken and thus the choice of a forecasting method are essen-
tially a function of the type of calculation envisaged. Particularly in the case of evaluation,
the method is characterized by a priori reasoning from a real industrial situation, both in eco-
nomic and technical terms, and trying thereby to limit additional uncertainties by reducing
the number of exogenous variables and their associated lack of precision. Consequently, it is
often useful, if not indispensable, to begin by omitting the effects of inflation. One way of
doing this is to operate in “constant currency.”

1.3.1.1 Current Currency, Constant Currency, Deflation Index

Operating in current currency or prices consists in analyzing and extrapolating the set of
observations available, as such, over time. This means considering the nominal values of this
information at each moment in time.
Working in constant currency or prices consists first of all in selecting a reference date
(usually a year), and then expressing the nominal data available for other dates in equivalent
40 1. Market Analysis

values for this reference. To do this, it is necessary to have a correction factor, or preferably
a ratio of indices, for a given country, that takes into account the depreciation of the national
economy. This may be a variation in the purchasing power, the gross national product, the
average price, etc.
This so-called “deflation index” is premised on the fact that the “standard of value” usually
undergoes a deterioration that it is useful to correct, so as to make realistic comparisons at
different dates.
For a manufacturer, this depreciation roughly corresponds to the general rise in product
prices and shows up in its sales figures, i.e., in the price-volume relationships.
In the simplistic theoretical case where only one product is being considered, the sales fig-
ure Vat a given moment is such that:
V = q P where q = quantity produced
P = unit price
0 for t = to, we have VO = qo Po
0 fort=t vt = 4, pt
V, - Vo represents the variation of production in nominal values, when q and P vary simulta-
neously.
To compensate for the price variations and find the real changes in production, we must
determine:

vto = 4t Po
Vp - Vo then represents the variation in real terms.
Now:

PdPt constitutes a correction or deflation factor, which, of course, is of no interest in the

case of a single product. On the other hand, when projected over a whole range of produc-
tions, or a nation, the concept has a lot more justification.
Doing this consists, in fact, of correcting V, by the ratio p / p of the weighted average
mo . mt
product prices, or more generally of the production of goods at time 0 and t. Accordingly:

vp=v‘ mP o lPmt

or:

or: Po = pt Pmo / Pmt

In this form, such an expression shows that for each nominal price value, we can find a
corresponding value deflated to a given date. Every country has many deflation indices that
can take account of the deterioration of the standard of value, either at the national level itself,
or at the level of the various industrial or non-industrial sectors. In the United States, for
 1. Market Analysis 41

example, the GNP Deflator is used; in Western Europe, and more specifically in France, the
index of consumer price development is utilized, or the implicit index of the GDP.
By way of indication, Table 1.3 presents the information concerning this latter method of
calculation since 1970. Figure 1.9 shows an application of deflation to the price of crude
petroleum since 1970, under this same procedure. Graph 1.9a shows US$ per barrel and
Graph 1.9b French francs per ton. Both of them clearly portray the effects of the two energy
crises of 1974 and 1980. However, the first one brings out the decrease in real value of a bar-
rel of petroleum in US$ after this date, while the second, because of the increase in the
exchange rate of the dollar against the French franc between 1980 and 1985, shows the con-
stant value of the cost of a ton of petroleum in local currency during this period.
Among the indices that are best suited to an industrial sector or a particular kind of product,
we may mention, for example:
0 For France: the monthly and annual indices of wholesale energy prices, motor vehicles,
pharmaceutical specialties, processed and non-processed foods, and indices related to pro-
duction sales prices (before tax) for refractories, glasses, etc., inorganic chemicals (excluding
fertilizers), fertilizers, organic chemicals 6 , etc.
For the United States: The Nelson indices published in The Oil and Gas Journal, partic-
ularly for inorganic (code 613) and organic (code 614) products, and for various items of
equipment, the Chemical Engineering and Chemical Week indices, etc.
Their specific application is discussed in more detail in the section on methods of calcu-
lating operating and cost prices (see Chapter 2, 6 2.2). Generally speaking, any index of the
variation in product price, equipment cost, value of a particular good, etc. may be used to cal-
culate deflation.

1.3.1.2 Relationships Between Product Prices and their Raw Materials

Line of Equivalence
Many products often have their prices strongly affected by the price of the raw materials from
which they come. This relationship is approximately linear and has the form:

The coefficients a l , a2,a3,etc., represent the weights (or volumes) of outputs from the pro-
cess that converts the starting materials into the product studied; a priori, they are constant,
at least for a certain level of unit production capacity; bo is a term that includes all the other
manufacturing costs, and as a first approximation it too may be taken as constant.

5. This is a matter of a weighting among 295 categories of expense, performed by working house-
holds or employees residing in towns of all sizes; the analysis is based on about 160000monthly state-
ments, conducted at 30000 points of sale in a hundred urban areas.
6. These indices are calculated and published by the INSEE, particularly in the Annuaire Statis-
tique de la France (French Statistical Yearbook).
42 1. Market Analysis

Table 1.3 Examples of deflation indices 1-2

France USA

GDP implicit GNP/GDP Implicit price

Year
price index in constant US$ index
(1985 = 100) in current US$ (1985 = 100)
(1985)
1970 25.2 1015.5 - -
1971 26.8 1102.7 - -

1972 28.7 1212.8 2 894.5 41.9

1973 31.1 1359.3 3 040.9 44.7
1974 34.8 1472.8 3 030.5 48.6
1975 39.3 1598.4 2 987.7 53.5
1976 43.7 1782.8 3 133.2 56.9
1977 47.7 1990.5 3 284.7 60.6
1978 52.5 2249.7 3 439.9 65.4
1979 57.9 2 508.2 3 542.7 70.8
1980 64.4 2 732.0 3 534.3 77.3
1981 71.8 3 052.6 3 604.0 84.7
1982 80.2 3 166.0 3513.9 90.1
1983 88.0 3405.7 3 638.6 93.6
1984 94.6 3 772.2 3 884.9 97.1
1985 100.0 4014.9 4014.9 100.0
1986 105.1 4231.7 4 124.5 102.6
1987 108.3 4524.3 4268.2 106.0
1988 111.5 4880.6 4432.9 110.1
1989 114.8 5250.8 4565.9 115.0
1990 118.2 5 546.1 4621.8 120.0
1991 22.1 5 724.8 4594.5 124.6
1992 24.7 6244.4 4867.0 128.3
1993 27.8 6558.1 4983.4 131.6
1994 29.7 6947.0 5 153.6 134.8
1995 31.8 7265.5 5 385.8 137.9
1996 33.3 7636.0 5438.7 140.4
1997 34.5 8 110.9 5 672.0 143.0
1998 135.5 8400.0 5813.1 144.5
1999 136.8 (p) - - 146.2 (p)
2000 138.6 (p) - - 148.8 (p)
@): provisional.
1. Until 1968 the information provided by the OECD was based on the GNP; since then, the figures have come
from the GDP
2. The Gross National Product or GNP corresponds to the total of the gross added values of domestic producers,
plus total taxes (including VAT), minus subsidies not included in the product values.
The Gross Domestic Product (GDP) is equal to the total gross added values of domestic and non-resident produ-
cers, plus total taxes (including VAT), minus subsidies not included in the product values.
The implicit price deflator for the GNP or the GDP measures the average annual growth in prices for the whole eco-
nomy (consumption, production, etc.) over a given period. It is calculated by applying the least squares method.
(Source: OECD).
 1. Market Analysis 43

80 -
70.
cc
60.

50 .
-2
i 40-

30 . h

20 .

10 ..

01
1970 1975 1980 1985 1990 1995 2000

Figure 1.9a Variation in the crude petroleum price in US$/bbl. (Source:IFP)

400

350

300

250
-g!
$ 200
LL

150

100

50

0
1970 1975 1980 1985 1990 1995 2000

Figure 1.9b Variation in the crude petroleum price in FFIt. (Source: IFP)
44 1. Market Analysis

If one of the raw materials involved has a dominant economic weight, and if the prices of
the others are held constant, the preceding expression is simplified:
Pproduct = a P m +b where b = Cte
If, in addition, b is small, we find:

---
‘produc t
= Cte
R
‘M

Situations of this kind are found in many areas of industrial activity, particularly in sophis-
ticated chemical products (as long as we exclude the costs of packaging, advertising, invest-
ment in research, etc., i.e., marketing costs), and also in refining, where as shown in Fig. 1.10
we have relationships of the form:

1.2
‘naphtha
-- = 1.2 or
‘diesel
-- ~

crude crude

In practice, these are average values, around which we find fluctuations of ilO% ?. Corre-
lations of this type prove to be less evident in other areas, such as petrochemistry for example,
so that b may no longer be ignored. However, certain trends that result from smoothing, and
consequently certain extrapolations, may be obtained from a study of the variation of such
ratios over time. Figure 1.11 is an illustration of this for the prices of the major intermediate
petrochemical products ethylene, propylene, and benzene, which are mainly obtained from
naphtha, at least in Western Europe. In spite of some major fluctuations, mostly due to eco-
nomic phenomena linked to production incidents that led to appreciable imbalances between
supply and demand, there is an obvious relationship between the prices of these products and
that of the raw material from which they come, and this can be used to make short- or even
medium-term forecasts.
Relations linking the prices of products and their raw materials may help to make techno-
logical forecasting much easier. Two examples based on the concept of the equivalence rela-
tion and the division of areas will provide illustrations.

7. Figure 1.10 is presented as an example to show the roughly parallel variations of diesel fuel and
naphtha, except for the period covering the first half of 1986. This apparently erratic variation is
explained by the fact that at that time the British coal producers were on strike, and the fuel then pref-
erentially consumed in the United Kingdom for residential or community heating was coal. There was
a transfer of the corresponding demand onto domestic fuel oil (which is an almost identical product of
the diesel fuel production process). The pressure on supplies led to a very sharp rise in the price of
middle distillates.
In conclusion, it should be remembered that before defining a relationship it is important to elimi-
nate aberrant situations (remove accidents, remove seasonal effects, etc.) by endeavoring to identify
the events that produced them.
 1. Market Analysis 45

1.8

1.ti
0
U
5
c
0
$
I

5-
V
1.4

1
I
1985 1986 1987 1988

Figure 1.10 Variation in the ratios of the Rotterdam spot prices of naphtha
and diesel to the price of Brent crude. (Sources: ECN, Platt 's,etc.)

V .

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Figure 1.1 1 Variations in the spot prices of ethylene,propylene, and benzene

against the price of naphtha (Northwest Europe). (Sources: ECN, Platt 's)
46 1. Market Analysis

First example
This deals with two ways of synthesizing methanol, using either coal or natural gas as the
starting material. We thus obtain, depending on the case:

Prnethanol = Pcoal +

or Pmethanol = Pnatural gas + b'

The price equivalence according to one or the other of these two possibilities also leads to
a linear relation:

b'-b
Coal = a
U'
f'natural gas + a
represented graphically (Fig. 1.12) by a straight line whose slope is unchanged over time as
long as there is no improvement in the yields, but whose intercept on the y-axis is valid only
for a given date, because it depends heavily on the investment costs of the plants. This straight
line delimits two areas, one favoring the use of coal (on the side of the natural gas price axis),
the other favoring the use of natural gas (on the side of the axis graduated in coal prices).
Depending on the relative variation of prices of the two raw materials (expressed in constant

200

150

s
e loo
al
0
.-
-hm
0

50

0
0 2 4 6 8 10
CIF price of natural gas ($rrmMBTU)

Figure 1.1 2 Comparative economics of producing methanol from coal and

from natural gas. (Source: IFP)
 1. Market Analysis 47

*
currency at the time of the study so as to be free of the constraint imposed by changes in the
y-intercept of the equivalence line) we can immediately decide, for a particular location
(Western Europe), not only on the economic potential of one technology versus the other, but
also on the probability of a change in conditions occurring in the long run. In the example
given, it would be difficult to imagine that coal, delivered in Northwest Europe, could ever
compete with methane in the synthesis of methanol, at least without the occurrence of major
upheavals.

Second example
The second illustration deals with a single technology capable of being applied, as before,
to two different raw materials (benzene and toluene), but leading to two products that are also
different: one established, styrene, the other a new one, p-methylstyrene. Insofar as the cor-
responding polymers have properties that are at least comparable, if not better for the second
one, here again we may look for the equivalence line for installations, either new, i.e., not yet
depreciated ones, or revamped ones (Fig. 1.13). In this case, it is a matter of adapting the sty-

5 000

4000

st
.-0
h 3000
Q)

w
:
2 000

1000
1000 2000 3000 4000 5000
Toluene price (FF/t)

Figure 1.1 3 Comparative economics of producing p-methylstyrene and

styrene (constant 1988 currency). (Source: ZFP)

8. Deflation performed using either the GNP Deflator Index (United States) or the consumer price
index (France).
48 1. Market Analysis

rene-producing units, which operate either in the liquid phase (oldest units) or in the vapor
phase (newest units), to obtain p-methylstyrene. The investment costs, which vary from one
case to the other, appear as three different straight lines of equivalence, each of which divide
the diagram into two areas, one favoring the production of styrene (on the side of the toluene
price axis), the other favoring p-methylstyrene (on the side of the axis graduated in benzene
prices). In addition, Fig. 1.13 shows the simultaneous variation in the prices of benzene and
toluene in Western Europe and the United States, in constant currency for the end of the
period. The result provides little encouragement for the construction of p-methylstyrene
plants, unless the resulting polymer offers distinctly better performances than polystyrene,
which would justify a price increase. It is not favorable to the revamping of liquid-phase sty-
rene units either; on the other hand, however, it does favor Mobil-type plants that operate in
the vapor phase. Unfortunately platforms of this kind are still uncommon and of recent date,
and thus far from having been completely paid off; in addition, the demand for toluene fuels
(lead-free gasolines) is not trending in the direction of a decrease in its price relative to that
of benzene.

1.3.1.3 Product Price Structure. Evolution Curves

Generally speaking, the price of a product at any given moment may be regarded as consisting

Time

Figure 1.14 Variation of the margin and floor price over time.
(Source: IFP)
 1. Market Analysis 49

0 The floor price (or minimum profitable sales price) whose method of calculation is given
in Chapter 2, represents the lowest price that the market’s best-established producer would
ask. In a particular geographic area, this leader would thus be the one that currently had the
most efficient plants, both in terms of unit production capacity, giving it the greatest econo-
mies of scale, and of applied technology. These installations, which would necessarily be
recent, would therefore not have been amortized, requiring capital charges to be included in
the cost of manufacture.
The floor price also corresponds to the price that this same producer would ask for the con-
struction of a new plant having the same characteristics. So it is a reference price, which the
competition must try to meet if it wants to hold onto its share of the market.
The margin represents the difference between a product’s actual price at a given moment,
and its floor price. It depends on the demandsupply situation and may be very high, espe-
cially where new products are concerned. But, because they are attractive these high margins
are unstable: in fact, they encourage other producers to invest, which results in temporary pro-
duction overcapacities compared with needs, and thus leads to a general fall in sales prices.
In this regard, in fact, we must distinguish strong trends from short-term movements, over
the life of a product.

A. Strong Trends
Chronologically, there are three periods to consider.

a. Disappearing Margins

A reduction in the initial price first affects the margin, although its variation is only very grad-
ual: in order to maintain or increase its influence, each producer will seek to sell a little more
cheaply than its competitors, while still making maximum profits; usually, it is in no-one’s
interest to “break” the market. The fall in sales price is also partly due to a reduction in the
floor price: the opening of markets results in the construction of unit capacities of greater size,
and a better optimization of the techniques employed. Experience shows, however, that the
decline in the margin is by far the more significant: for new products it operates at an expo-
nential rate of 15 to 20% during the early years. In the past, this situation was even set up as
a principle by certain economists who extrapolated this downward trend, maintaining for
example: “If the sales volume grows two- or three-fold, the price will be approximately equal
to two thirds of the starting price”. Moreover, until the 197Os, i.e., before the first global
energy crisis, the reality did conform to this principle, as particularly shown in Fig. 1.15, for
six intermediate and 16 final products of the United States organic chemistry industry. This
graph, which normalizes the observed phenomena by making the origin of the time axis the
year in which the price began to fall, and assigning an index of 100 to the initial value, imme-
diately brings out the existence of a correlation. It also shows a regular, exponential decrease
that leads to an average reduction of about 50% in current currency after ten years.
The diminution in the margin continues until its complete disappearance, i.e., until the
moment when the sales price reaches the floor price. After that, we see fluctuations around
this value (Fig. 1.14). However, a manufacturer cannot maintain a price level below the the-
50 1. Market Analysis

0 2 4 6 8 10 12 14 16
Years since prices began to fall

Figure 1.15 Relative fall in a product’s sales price when

production rises: a U.S. example.
(Source: Chem. Egineering Progr.)

oretical minimum for very long, without shortly having to shut down its plant. The recent
history of the steel, coal, refinery, and petrochemical industries shows what happens: the
disappearance of the least efficient plants is found to be necessary for the re-establishment
of a price scale that will not only allow the remaining units to function in an economically
satisfactory manner, but also enable a level of profitability that will encourage new
investments. This first takes the form of improvements in the utilization rate at existing facil-
ities.

b. Increase in Floor Price

At this stage in the life of a product, many authors have proposed that its sales price should
now fall to its floor price, based on the fact that it appears impossible to re-establish the mar-
gins. In order to justify this attitude, we may quickly describe the main components of the
floor price, without examining its structure in detail (see Chapter 2, 9 2.2.1.2). It has the fol-
lowing constituent items:
Fixed operating costs (maintenance, taxes, insurance, overheads, etc.), linked to the
costs of the production units alone,
Capital costs (provision for depreciation, financing costs, return for investments), a
function of the overall investment,
 1. Market Analysis 51

0 Variable or proportional costs (cost of materials, expenses for catalyst, solvent, chemi-
cal products, energy and other utilities, etc.), resulting directly from the operation’s
material and energy balance sheets,
0 Marketing costs (packaging, advertising, distribution, etc.) and the costs of research.
As Fig. 1.16 shows, fixed operating and capital costs have a tendency, if not actually to
decrease as a result of the economies of scale, at least to stabilize because of the effects of
inflation on the investments; marketing costs tend to remain approximately constant. On the
other hand the variable costs, which are directly proportional to the cost of raw material, are
very sensitive to any change in its price. This explains why the rise in the price of petroleum,
and more generally in the cost of energy, from 1974to 1985,was directly reflected in the floor
prices of products, pulling them higher. During this same period, the high rates of inflation
and of return on capital that were encountered and pursued in many countries only served to
accentuate the phenomenon, by erasing what economies of scale were still possible. Ulti-
mately, the variation of the floor price for most products over time presented a U-shaped

Marketing, administrative,

\ and R&D costs

\ Variable costs

Fixed costs - Capital charges

-5 0 5 10 15 20
.
25
Years

Figure 1.1 6 Variation of the floor price over time.

(Source: A.A.C.E.)
52 1. Market Analysis

curve. For a while, this observation too was set up as a principle, just for the purpose of mak-
ing forecasting easier. In fact, such a representation could be expressed mathematically by an
equation of the form:
P=ao+al t + a 2 t 2 + a 3 t 3
P being the floor price and t the time.
The coefficients ag, al, a2, a3, are derived by solving the set of equations obtained by con-
sidering four values of P at different periods in the past, i.e., by assuming that a large part of
the U-shaped curve is known, in particular the coordinates of four points on it, and that they
correspond to the prices calculated for the most efficient plants at each of the dates consid-
ered.

c. Reconstructing the Margins

This eventuality, which appeared impossible to forecasters in the 1975-1985 period, has
become a reality in recent years in various industrial sectors, particularly in organic chemis-
try. Several things have helped to create a favorable situation: some have enabled the floor
price to be lowered significantly and then held steady, others have made it possible to follow
this by raising product sales prices.
For the first group the determining factor has been the fall in the real value of crude petro-
leum, which began to appear in the United States after 1980, but was in fill swing from 1986
onwards and accelerated in 1998-1999, reaching levels similar to those seen after the first oil
crash. In spite of some subsequent fluctuations and a marked recovery in producer country
prices (more than US$30/bbl in the third quarter of 2000), it appears that we may expect a
measure of stabilization at around US$17-20/bbl over the medium term. Another major factor
for most of the industrialized countries has been control of inflation, accompanied by a low-
ering, albeit moderate, in the interest rates set by the central banks. An exception is Germany,
where this option has been delayed as long as possible. However, the current economic recov-
ery should eventually result in some revival of inflation. Of course, the crises that have in
recent years successively affected Eastern Europe, then Southeast Asia, and finally Latin
America are evidence to the contrary. For all that, since they are essentially financial in ori-
gin, they can probably do little more than slow down a steady trend towards a growth in
demand signs of reflation are already visible in certain countries.
For the second group, we must especially bear in mind the recovery in demand and then
the scarcity of supply: by restoring the utilization rate of the production facilities these factors
first enabled the plants to benefit from near-maximum economies of scale, and then to bring
pressure to bear on availability, i.e., to drive up prices.
The resulting reconstruction of margins is also facilitated by certain governmental mea-
sures aimed at reducing taxes on company profits. To the extent that manufacturers are
increasingly encouraged to reinvest, in spite of the caution that past events must induce, the
main question concerns the duration of the phenomenon. The construction of new units that
could relieve pressure on available supplies, and thus on prices, cannot have its fill impact in
less than two or even three years from the date of the decision. From this viewpoint, for exam-
ple, projects announced in 1988-1989 implied that there would be little to fear before 1991-
1992: until that time both the sales prices and the margins should therefore remain high.
 1. Market Analysis 53

Afterwards, the possible proliferation of new facilities could only reduce the average utiliza-
tion rate for all facilities, pushing the floor price up and leading to a diminution of the mar-
gins. We must nevertheless point out that other factors also contributed to this situation
(maintenance of a high level of price for raw material, erratic behavior of intermediates, etc.).
When all is said and done, although fluctuations may appear in short- and medium-term
forecasts of the variation of sales prices over time for most products, if we look at the longer
term and smooth out the historical data, it appears that the U-shape described above may still
apply. A polynomial equation is therefore sufficient to describe the phenomenon mathemat-
ically.
Such an approach is supported by the illustration presented in Fig. 1.17, which shows the
variations in the posted prices of various major intermediate petrochemical products in the
United States. Some of these, relatively new ones like acrylonitrile, show a more or less steep
fall at the start of the period examined; other older ones such as ethylene have an almost flat
profile. The results for major mineral or energy products would be similar. Obviously it
would be necessary in this case to go back to an earlier time period to observe the phenome-
non. It can also be brought out by applying a deflation index so as to work in constant cur-
rency for the end of the period (constant US dollars), instead of in current currency (current
US dollars). Figure 1.18 presents the result of doing this, for the major intermediate petro-
chemical products already mentioned. In addition, eliminating the effects of inflation has the
advantage of moderating the fluctuations and consequently making modeling and extrapola-
tion easier.

Figure 1.17 U.S. price variation for various chemical products, in current
currency. (Source: Chem. Marketing Reporter)
54 1. Market Analysis

Figure 1.18 U.S. price variation for chemical products, in constant currency
(1986 US$). (Source: ZFP)

B. Short-Term Variations

Fluctuations in the posted prices of high-tonnage products, which may be considerable even
in a period of relative economic stability, usually seem to be unpredictable. In fact, they
depend on too many parameters to permit any general interpretation that could lead to a suf-
ficiently reliable forecasting method. However, some of them are more significant than oth-
ers, and all by themselves may bring out subtler, i.e., shorter-term trends than the ones
mentioned previously. In this way, without undertaking a detailed analysis of the economic
events that produce these periodic saw-tooth variations, which are described as “volatility”,
it is nevertheless possible to use them to explain the overall past course of events, and conse-
quently to derive succinct rules for extrapolation.
Thus, for products that are already mature, and have modest profit margins, we find that
slight variations in demand sometimes lead to significant changes in price. It can be the same,
too, when unexpected plant stoppages follow incidents or accidents, and then if more of these
chance events occur there is a veritable surge in prices. In other words, variations of a few
percent (no more than &5%) in needs or availabilities at the world level, or within a given
geographic region, often have significant effects on price variations.
These temporary phenomena, whose magnitude we try to predict in spite of everything,
must be separated from longer-term ones that are the reflection and the consequence of so-
called strong trends, such as the opening of markets due to demographic expansion or
 1. Market Analysis 55

improvement in living standards, and inversely by major economic recessions. These large-
scale variations result in the construction of new plants or conversely the permanent closure
of less efficient ones. The fluctuations, which are a priori more erratic in the very short term,
indicate a more or less complete matching of the market’s immediate supply potential to the
demand. In a way, they serve to recognize the existing mechanism’s lack of operational flex-
ibility at a given moment, and its inability to respond or adapt to additional disturbances.
Such a shortcoming is felt even more strongly as the production required approaches its
theoretical limit. In other words, as the utilization rate of the facilities increases, their sensi-
tivity to manufacturing incidents, to interruptions in output, etc., and generally to “marginal”
effects, grows exponentially. This is why various authors have come to define a “tightness
index”, expressed as follows:

capacity
tightness index = loglo
capacity - production

When extended to continental or even global scale, or even just to the scale of a single
country, depending on its level of industrialization, this concept requires some adjustments,
especially when it deals with widely sold products. In fact, a unit in continuous production is
generally calculated as actually working for a period of less than a full year, so as to take
account of scheduled stoppages for essential maintenance; it is sized according to a particular
work factor (see Chapter 2, Section 2.2. I. 1). Experience shows that when the number of facil-
ities considered is large enough, it is impossible to reach the maximum cumulative manufac-
turing potential: there are always one or more sites that are idle because of servicing.
Furthermore, owing to previous major accidents, some workplaces undergo prolonged shut-
downs. When it is possible to draw up a balance sheet for the facilities that are really in oper-
ating condition at a given moment, it is preferable to use the term “capability” instead of
“capacity”.
Except for this nuance, the tightness index proves to be a relatively efficient tool for short-
term forecasting. In fact, historical analysis shows that the level of correlation between this
index and some economic criterion, such as the ROI, the cumulative discounted cash flow, or
the IRR (see Chapter 2, Section 2. I), designed to measure the profitability of a manufacturing
process, is usually high. To illustrate this we have only to look at Figs. 1.19a and l.l9b, for
example, which show the United States market for p-xylene. The first displays the variation
over time of the gross profit in % together with the tightness index: the two curves are essen-
tially parallel. The second shows that a linear correlation may be obtained from them:
profitability = a x tightness + b
a and b being constants, in the present case respectively 39 and -28, so that: gross profit (“!)
= 39 TI - 28.
We will show later that for any value of a profitability criterion, a corresponding sales price
PV may be calculated, as long as certain economic data are available (investments Z,produc-
tion expenses D, etc.). Thus, for example, for the ROI we have the following equation:

ROIxZ+D
PV =
capacity
56 1. Market Analysis

Figure 1.19a U.S. market for p-xylene. Comparative variation of gross

profits and tightness index. (Sources: IFP, Chem-Systems, etc.)

0.6 0.8 1 1.2 1.4 1.6 1.8

Tightness

Figure 1.19b U.S. market for p-xylene. Correlation of gross profits and
tightness index. (Sources: IFP, Chem-Systems, etc.)
 1. Market Analysis 57

In conclusion, any prediction about how flexible the functioning of a group of installations
will be, must allow us to make predictions about prices 9.

C. Floor Price and Ceiling Price

In contrast to the floor price (magnetic in nature and thus constituting a leading or leader
price, because it reflects the operating conditions at the newest installations, which are a priori
the most efficient and have the largest unit production capacities), we may define a ceiling
price. In this case, we refer to a level of manufacturing cost that constitutes a disincentive, a
“laggard” price that has fallen behind the prices of competitors or that other plants will allow:

900

800

700

-
*
600
Other
kinds of
Naphtha
with limited
Naphtha
and/or liquid
500 eedstock flexibility & integration feedstock
.-s

L1
ij 400
a
‘0
& 300

200

100

0
0 2 4 6 8 10 12 14 16 18
Ethylene capacity (Mt/yr)

Figure 1.20 Variation ofthe minimum profitable sales price with cumulative
production capacity (for ethylene in Western Europe, 1991).N.B.: in that year,
the economics of steam cracking using ethane feedstock were better than for
naphtha. (Sources: IFP, SM, etc.)

9. This method must however be applied cautiously, since the concept of tightness is much more
complex than indicated here, and is linked not only to the balance between supply and demand, but
also, in particular, to the composition of the “pool” of suppliers and/or customers. As an illustration,
simply imagine a global market (demand) of 1000000 tlyr, fed (supply) either by 20 installations of
50000 t/yr each, or by 2 units each one by itself producing 500000 t/yr. The risks of an imbalance
occurring are much higher in the second case than in the first.
58 1. Market Analysis

a price, when all is said and done, that leads to plant shut-downs. It is usually a reflection in
economic terms of market conditions affecting those production units that may be character-
ized, whether wholly or in part, as having too small an output, insufficiently modern technol-
ogy, or a physical plant that is too old. However, in this connection we should note one
positive feature: the possible lack of capital costs, resulting from full amortization of equip-
ment after a certain period (generally 10 years); these may compensate for competitors’ tech-
nological advances or economies of scale.
If one constructs a graph whose x-axis shows the cumulative production capacities of the
various production units for a given product within a selected geographic region, and whose
y-axis shows, in ascending order, the corresponding minimum profitable sales prices for each
production unit, given its specific features, one generally finds a very flattened, S-shaped
curve, displaying a very wide “plateau” (Fig. 1.20). The first decile of production capacities
serves to identify the most favored plants, and the last decile the most threatened ones. In this
way it is possible, for example, to assess the size of a potential replacement market, or alter-
natively the limits of fluctuation of a product’s price with respect to an average level, identi-
fied by the y-value of the curve’s inflection point.

1.3.2 SALES VOLUME APPLICATIONS

Just as for price variation, variations over time in the size of the demand, i.e., the quantities
sold, have certain specific characteristics. In addition, they are accompanied by some partic-
ular concepts, notably the penetration rate and the growth rate.

1.3.2.1 Normalization of Evolution Curves

Generally speaking, as already stated in paragraph 1.2.3.1,the curve of fluctuations in a prod-

uct’s sales over time is found to have a characteristic “S’ shape. It shows four stages of mar-
keting: inductive, exponential and linear growth, and decline. Figure l .2 l illustrates this for
the evolution of acetylene consumption in the United States since 1935: a more detailed anal-
ysis of the situation shows that the observed decline is in large part due to a change in the raw
material used for the synthesis of vinylchloride, where for economic reasons the use of eth-
ylene was preferred.
Every product, in the broad sense of the term, has a corresponding, specific curve of this
type, which may be more or less complete, depending on its marketing status, i.e., its degree
of “maturity”. At a given moment, such representations for various products reveal temporal
and quantitative discrepancies that are often quite large, and prevent any direct comparison,
even for products in the same field of activity. In some cases the differences are big enough
that the compilation and smoothing of historical data lead to a result in apparent contra-
diction with the shape of curve described above. Thus, the correlation is often essentially
linear; in these cases, we are dealing with widely used, already mature products, for which
the period under review is too short: the start-up and exponential growth stages are not
shown.
 1. Market Analysis 59

1935 1940 1945 1950 1955 1960 1965 1970 1975

Figure 1.21 Variation in U.S. acetylene consumption, 1935-1975.

(Sources: IFP, SRI,etc.)

Time

Figure 1.22 Evolution of the petrochemical industry in Western Europe.

(Sources: ECMRA, IFP)
60 1. Market Analysis

Now, in considering variation in demand generally, it is essential even in short-term fore-

casting to identify carefully the nature of the development phase in which the extrapolation
is to be made, so as to limit the sources of error in selecting the most appropriate techniques.
A handy means of doing this is to reduce the sales curves of several products (from the same
family or the same industrial sector) to a single curve. For example, major intermediate pet-
rochemical products and polymers in current use, products from the steel, mineral, and phar-
maceutical industries, etc., and specialty polymers.
This normalization operation requires working in “reduced coordinates”, obtained by
dividing the momentary values of these parameters by those for the curve’s maximum, which
by definition is taken to be 1 or 100. By analogy with thermodynamics these are the charac-
teristics of the critical point, in the present case the point where needs reach a ceiling,
enabling calculations to be made. In practice, these data are available historically for only a
small number of products: in chemistry, for example, acetylene, styrene-butadiene rubber,
DDT, etc. It is however possible to obtain them for other products so as to arrive, for example,
at the result shown in Fig. 1.22, for the West European petrochemical sector. On this normal-
ized, S-shaped graph, generally referred to as a “maturity curve”, we can see that the ABS
resins are still in the exponential growth phase, while polypropylene and high and low density
polyethylenes have entered the linear stage, isopropyl alcohol and ethylene oxide have
reached a ceiling, and finally polyvinyl chloride and styrene-butadiene rubber are in decline.
In practice, to produce such a representation, i.e., to be in a position first of all to calculate
the reduced coordinates:

QR =-
Qt and TR =- r,
Q 00 Tl 00
where Q = volume or quantity and T = time, and then to establish the degree of progress or
maturity of the various products, we must find a way to determine the characteristics, even
approximates ones, of the maximum for each one, especially for the ones that have not yet
reached or passed that stage. For this purpose, it is of the greatest importance to correctly
identify the most appropriate parameters, preferring the ones that can be of assistance in mak-
ing extrapolations. It would therefore appear that the concept of penetration is well suited to
normalization as regards quantity (y-axis), while time (x-axis) should no longer be expressed
as dates but as relative durations, by adopting as the origin (reference date) the moment when
the product began to be industrially manufactured and sold.

1.3.2.2 The Concept of Penetration as a Tool for

Determining the Level of Market Saturation

By definition, the meaning of this expression is “the level of presence of a product or a brand
in a market for goods or services”, and the rate of penetration is “the ratio between the number
of customers using a product from a given manufacturer, and the total number of potential
customers in the market for this product”. By extension, it is also defined as being the ratio
between the quantities actually sold and those potentially saleable; in this case, it represents
the market share taken up. According to different writers, the concept has been more or less
 1. Market Analysis 61

broadened and misused. From this viewpoint, the mathematical correlation of the demand
with a precursor index such as the number of inhabitants or the industrial production index
(IPI) permits a kind of generalization. Thus, the consumption per capita or per index point
constitutes an indication of the market penetration and may be taken as a measure of the sat-
isfaction of the potential needs of the customer base or of an industrial sector:If the latter are
expressed in the same units the ratio between the two is equivalent to the penetration rate as
originally defined, because the element of correlation is then eliminated.
In its general form, however, this parameter offers the advantage of immediately introduc-
ing the concept of reduced coordinates. It is based on the maximum possible market, and
thereby justifies its selection for the construction of normalized graphs.
The additional interest of utilizing it in a derived form, especially when correlated with a
precursor index, arises from the possibilities offered for extrapolation and the rapid identifi-
cation of “critical” conditions.
In this connection, various writers, including D. Stevenson, have demonstrated, first, the
effectiveness of using the industrial production index (see 0 1.2.3.2.A) as a forecasting tool,
by testing it on a large number of chemical intermediates. They showed that although the rela-
tionships might be linear for considerable periods, overall they displayed an S-shape, like the
demand curves over time, differing only in that the slope of the tangent at each point provides
a measure of the speed of market penetration, in the broad sense of the term. Figure 1.23,por-
traying the consumption of styrene-butadiene rubber in Western Europe since 1960,provides
an illustration of this. It should moreover be noted that the graph of the variation of consump-
tion against index points has an S-shape too.

0.8

3
Y
I
0

E 0.6
8
a
m

.
v)
0

0.4
0

0.2
40 50 60 70 80 90 100
Industrial Production index (base 100 In 1980)

Figure 1.23 Demand for SBR in Western Europe, 1960-1983.

(Sources: ECMRA, IFP, SM, etc.)
62 1. Market Analysis

D. Stevenson also suggested a simplified method for determining the limits beyond which
extrapolation should not be pushed, i.e., the approximate position of the demand ceiling. The
approach consists of using logarithmic coordinates so as to obtain a linear representation of
the graph illustrating the relationship between the penetration for a particular year, and that
for the previous year. In these circumstances the first bisector is the place where the penetra-
tion values are the same from one year to the next, indicating zero growth. The intersection
of the representative straight line with this first bisector will therefore correspond to a station-
ary demand, i.e., to the desired ceiling; its coordinates will enable the identification of both
the timing and the size of the saturation threshold.
This method is illustrated in Figs. 1.24 and 1.25, showing the variation in demand for sty-
rene-butadiene rubber in Western Europe. The first shows the linear relationship obtained
using data from the period 1961- 1973; the intersection with the first bisector indicates a the-
oretical maximum penetration of 11 (thousands of tons per IPI point). The second enables
comparison of this result with the reality, by looking at all the information available for the
1960-1983 period. The difference for penetration is about 10%; as for the date of 1975 pre-
dicted by the theory, it actually occurred in 1971-1972. This procedure is of interest not so
much for its accuracy, which is all the more uncertain because the slope of the representative
straight line is near 45”, but rather for the possibility that it offers for placing limits on the
extrapolation and for estimating the characteristics of the “critical point”, whose identifica-
tion is essential for normalizing the variation curves for needs, and for determining levels of
maturity.

5 10
Log (penetration rate for the previous year)

Figure 1.24 Determination ofthe market saturation plateau: SBR in Western

Europe. (Source:ECMRA)
 1. Market Analysis 63

Figure 1.25 Variation of the penetration of SBR into Western Europe,

1960-1983. (Sources: IFP, SM, etc.)

Experience shows that application of this method to products that are already in decline
leads to estimates that exceed the reality by around 10 to 15%, which is explained by the
impossibility of actually saturating the potential markets. In these circumstances, so as not to
cause disparities with products in the expansion phase, normalized graphs should be con-
structed from the calculated critical parameters, and reduced by 10% even when actual fig-
ures are available.

1.3.2.3 Real and Average Growth Rate. Application to the Long Term
For products that are known, for example by using the above procedure, to be developing in
an exponential or linear manner, either growing or shnking, it may be of interest for extrap-
olation purposes to employ a parameter that is often mentioned in economic publications: the
growth rate. This concept is based on the following approach.
Let Do be the initial demand for a particular product. After one year, it will have increased
or decreased in a certain proportion rlDO, such that it is equal to:
D1 = Do(1 + Y1)
where rl represents the rate of growth between 0 and 1.
At the end of the second year, it becomes:
0 2 = Dl(1 + ~ 2 =) Do( 1 + ~ 1 ) (1 + 72)
64 1. Market Analysis

and after n years:

D,=D0(1 +rl)(l + r 2 ) . . . ( 1 +r,,)
where rl, r2 ... r,, are the growth rates for the successive years under consideration. Indeed,
they are the average rates for each year. In fact, we may increase the frequency of this kind
of information: monthly, daily, etc., and theoretically ad infinitum, as long as the correspond-
ing statistical data are available.
Generally speaking, if r represents the “real” or “instantaneous” growth rate, i.e., in prac-
tice the rate for the periodicity considered, the average growth rate r, over n periods is
defined by the equation:
t=n
rm = !. I r dt
t
t =o

A. Constant Real Growth Rate

In the particular case where the periodicity is annual and where:
y1 F r2 = y3 = ... = r, = r = Cte

we find: D,,=DO(l +r),,

Moreover, Eq. (1.1) gives:
rm=r
or: D,,= Do (1 + r,)‘

COMMENT
This expression is mathematically similar to the one for the continuous compound
interest rate i; in fact, i is such that, if a sum So is borrowed at this rate in year 0, a
sum S,,:
s,,= so (1 + i)”
must be repaid in year n.

The Eq. (1.2) may also be written in the following manner:

log-Dfl - n log(1 + r,)

DO
The graph showing the variation of log D,,lDo = n log (1 + r,) as a hnction of n produces
a straight line of slope log (1 + r,). As long as r, is much less than 1, which is usually the
case for most major products:
log(1 +r,)Er,
As a first approximation, the average annual growth rate corresponds to the slope of the
straight line obtained.
 1. Market Analysis 65

B. Variable Real Growth Rate

An exponential growth in demand over a given period t may be expressed as follows:

D,,= D, em
'' (1.4)
where rm= average growth rate, linked to t by Eq. (1.1).
Note that the expression (1.3) may be written:

4 -- rmt
log -
DO

The corresponding curve as a function o f t is a straight line of slope rm,which agrees with
the above conclusions.
The extrapolation of needs by the exponential method expressed in Eq. (1.3) does not con-
sider the effect of market saturation, and very quickly leads to conclusions that are badly
wrong.
To accommodate this slowing down phenomenon, W.W. Taddle and J.B. Malloy, for
example, suggested a modification of the above model, first of all by introducing an element
that would provide a progressive correction, and then by setting a threshold below which the
growth in demand is highly unlikely to fall, during the extrapolation period considered. This
limiting growth rate r , in fact corresponds to a balance between the consumption of the prod-
uct under consideration, and a value characteristic of the economy, for example the GNP, the
IPI, etc.
So in the worst case, we may suppose-as before, of course, during the expansion phase-
that after an infinite time the growth of the sales volume will coincide with the value that the
institutional forecasters envisage for such an indicator. Under these circumstances, we find
that the growth rate at a given date is composed of two terms, one representing the limiting
growth rate r,, and the other decreasing exponentially over time from the initial value ro to
r,, i.e., with a time constant k such that:

The resulting expression for r then has the following form:

Putting this new definition of r into the Eqs. (1.2) and (1.3) for the exponential variation
of demand, we obtain:
t f

0 0

or, by integration:
66 1. Market Analysis

and: D,, = D, erm' = D, e

[ r,t + (I -.-if)]

In the particular case when r, = 0, we have:

r = YO eckt

'0 (I - ,+)
and: D=Doek

The use of this model assumes that the available data are sufficient to calibrate it to begin
with, i.e., to enable the value of k to be determined. To do this we must first divide the historic
data into two groups: older ones leading up to a definition of ro in the not too distant past, and
more recent ones, for example covering the last two or three years, providing the value of Y
for that period. We then calculate k using Eq. (1S),and retain the result for use in the same
equation when it is used for forecasting, r then becoming the unknown. However, applying
such a procedure demands certain precautions: it is desirable to avoid the use of statistical
information resulting from exogenous events (war, economic crisis, sudden inflation, deval-
uation, etc.). The curve in Fig. 1.19a illustrates this, with the 1942-1944period corresponding
to the Second World War, during which the demand for acetylene in the United States
reflected not only the traditional markets but also temporary ones related to the massive man-
ufacture of explosives.
 CHAPTER 2
The Elements of
Economic Calculation

2.1 THE STUDY OF PROJECT PROFITABILITY

Whatever the kind, the purpose, or the degree of sophistication of a project, an industrialist’s
primary concern is to know whether it can be done profitably, or in other words, whether it
will generate profits once it has been carried out. This type of question first presents itself in
absolute terms. It can also be looked at in a relative way, in the sense that, on the one hand,
several projects may be competing with each other within a single company, and it is a matter
of pushing the ones that will generate the most income. The need for comparison also arises
when it becomes necessary to decide whether it is worthwhile to enter a given market, in com-
petition with other actual or potential alternatives, or to make a ranking of the various possi-
bilities, including that of not undertaking the project.
In practice, there are many ways of obtaining an answer, i.e., of assessing the intrinsic or
relative economics of a project. This statement alone shows that in this matter the recom-
mended procedures include neither a universal solution, nor interchangeable ones. In fact,
their applicability mainly depends on the kind, the volume, and the reliability of the informa-
tion available. Thus, some of them may require prior collection of a large volume of accurate
data: for example, carrying out a medium- or even long-term market study involving raw
materials and products, from three points of view-quantitative, qualitative, and price. In
contrast, other, shorter procedures are used when information about market situations, sup-
plies, geographic location, or operating conditions is lacking. These approaches find applica-
tions that relate their level of sophistication to the stage reached by the project: the more basic
ones are best suited to the preliminary study phase, the more complex ones to phases whose
outcome is the decision to begin construction or not. Beyond that point, there is no further
need to assess the operation’s economic attractiveness, unless particular events fundamen-
tally change the calculation’s original conditions.
For these reasons, profitability analyses are in practice most often applied in the area of
evaluation than in the studies carried out in a related discipline, estimating, which is more the
specialty of engineering companies. In their more sophisticated applications, on the other
hand, they are closer to the methods developed and utilized by economists and accountants.
But at the outset we must note an essential difference: some of them operate in an a priori
68 2. The Elements of Economic Calculation

manner, on the basis of prospective data alone, the others can base their calculations on actual
balances, and thus are able to perform analyses a posteriori. In addition, for evaluators, who
must always look towards a distant future, the important thing is always to choose the simple
approach. Consequently, they usually resort to “light-weight’’economic or accounting proce-
dures and to simplified working hypotheses that allow these to be applied, even though they
may often be incorrect in principle-as long as they remain valid in terms of their credibility
and their objective.

2.1.1 THE CONCEPT OF TAXES

This concept is applied to companies on every added value. It particularly applies to so-called
taxable profits, which represent the difference (assumed to be positive) between the products
or the revenues, and the charges that the tax authorities allow as deductions. It has signifi-
cance only where the existence of a private business sector is acknowledged by the state. Its
thus reaches its full importance in market economy countries, where the financing of projects
and the management of operations are left to the citizens’ own initiatives. Countries belong-
ing to planned economy (statist or socialist) systems are able to ignore it. In this case, indeed,
a given company cannot operate exclusively on public funds and at the same time give back
part of the money to these very financing institutions, calling it a tax: on the contrary, the prof-
its are entirely turned over to the community, which acts simultaneously as shareholder,
investor, and administrator.
In market economy countries, the taxation of companies, or more precisely the tax rate,
varies widely from one state to another. It also varies over time, depending on economic con-
ditions. By its nature, its purpose is to furnish part of the national budget, and thereby to con-
tribute, along with other sources of taxes, to the maintenance and improvement of public
services. In addition, depending on its level, it constitutes a direct intervention by the State in
the management of companies, to the extent that high tax deductions reduce by that much the
distribution of dividends or discourage reinvestment; conversely, low ones may encourage
reinvestment, innovation, and job creation.
Generally speaking, taxation is applied to companies as it is to individuals, at different lev-
els. Although we are discussing its effect on profits here, it must not be forgotten that com-
panies must also pay other kinds of tax: local taxes, training levies, sales taxes, customs
duties, ... value added tax (VAT). In this case, the company does the Government’s work,
collecting the tax for it, i.e., returning the difference between the deductions it takes from its
customers and those made by its suppliers, which are “recoverable”. When the difference is
negative (revenues less than investment expenses during the first years of a plant’s operation,
for example), the fiscal legislation usually grants a tax credit.

COMMENT
The usual unit of time for accounting purposes is a year: balance sheets and their results
are compiled only at the end of the period. This leads to discrepancies, particularly in
the tax area, that are hard to take into account when considering the future, which is
what evaluation does. In addition, the fiscal year does not necessarily coincide with the
calendar year, depending on the country (e.g., Japan and the United Kingdom).
 2. The Elements o f Economic Calculation 69

In theory, there is a profit or loss as soon as there is an added value or a minus value, i.e.,
a difference between a good’s sale or resale price and its net book value. Here the tax situation
sometimes differs depending on the delay between the good’s acquisition and its transfer,
whether it is altered or not. In fact, the longer this delay becomes, the greater is the gap
between its apparent and real values, owing to the general rise in prices, which justifies a
reduction in the tax rate. Thus, a distinction is often drawn between:
0 Short-term added values-these are the ones that apply when a good is sold again within
two years of its acquisition, or upon the transfer of depreciable assets for no more than
the amount of depreciation already taken. They are like regular profits and are conse-
quently taxed in the same way.
0 Long-term added values-these correspond to transfers of goods, whether depreciable
or not, that have been held for more than two years, as long as (for depreciable ones)
there is a positive difference between the sale and purchase prices. This kind of added
value may be taxed at lower rates.
By way of illustration, Table 2.1 provides information on the 1997 rates of taxation
applied to company profits by various states.

Table 2.1 Tax rates on company profits (P) in various countries during 1997

Country Rate (YO) Comments

Germany 30 P - distributed
45 P - undistributed
+ Special surtax of 7.5% on the amount of the tax
Austria 34 General
22 Financing companies for SMI
Belgium 39 P > 13 million FB (on entire profit)
28 to 39 Progressive (overall, not in steps) P < 13 million FB
+ “Economic” surtax equal to 3% of the tax
Denmark 34 General
Spain 35 General
30 SMI with sales I 250 million pesetas and P I15 million pesetas
25 Mutual insurance companies
20 Cooperatives
USA 15 Progressive: P < US$50000
25 Progressive: US$50 000 < P < US$75 000
34 Progressive: US$75 000 < P < US$lO million
35 Progressive: P > US$lO million
+ Surtaxes:
34 to 39% for US$lOOOOO < P < US$335000
35 to 38% for US$15 million < P < US$18.3 million
Finland 28 General
70 2. The Elements of Economic Calculation

Table 2.1 (continued) Tax rates on company profits (P) in various countries during 1997

Country Rate (YO) Comments

France 33.33 General
+ Special tax: 10%
19 SMI profits reinvested (actual tax 20.9% including the previous tax)
Greece 45 General
40 Companies with bearer shares not listed on the Athens exchange,
and banks
Ireland 28 P I50 000 pounds
(1/4/97-31/3/98)
36 P > 50 000 pounds
10 Industrial, computer services, and financial companies, and
companies exporting Irish goods
Italy 53.2 Includes 37% State tax (remainder local)
36.2 Includes 20% State tax (remainder local) for SMI listed on stock
exchange with assets < 500 billion lire
Japan 37.5 General
Companies with capital < 100 million yen:
28 Portion of P < 8 million yen
37.5 Portion of P > 8 million yen
Luxembourg 32 P > 1.312 million FL (on entire profit)
20 to 32 Progressive. P < 1.312 million FL (overall, not in steps)
+ Surtax: 4% of the tax for unemployment fund
Netherlands 36 Graduated. P < 100000 guilders
35 Graduated. P > 100000 guilders
Portugal 39.6 General (State + local)
27.5 Real estate investment companies and certain agricultural
companies
United Kingdom 31 P > 1500 000 pounds
21 P < 300 000 pounds
21 to 31 Progressive, 300000 pounds < B < 1500000 pounds
Sweden 28 General
(Source: Minisdre de 1 'Economieet des Finances - Ministry of Economicsand Finance - September 1997)

For some of these countries, the variations recorded during recent years indicate an overall
reduction in taxation, intended to foster reinvestment and job creation. This policy is very
markedly displayed by some countries such as France, where profits distributed to sharehold-
ers are more heavily taxed; in Germany, the opposite is the case, no doubt to encourage the
acquisition of goods from beyond its frontiers.
 2. The Elements of Economic Calculation 71

2.1.2 PROFlTABlLlTY CRITERIA

There are many economic criteria that can be employed to assess the attractiveness of a
project. Most of them are based on knowing the revenues or the sales prices, and take into
account the concepts of gross or real profits, i.e., that they apply before or after tax, depending
on the case. However, some of them, such as cost of price or operating cost, do not have this
alternative meaning, since their very definition excludes the revenue from sales.

2.1.2.1 Typological Classification and Standardization of Expression

Generally speaking, they fall into two main categories:
Empirical criteria-these are a matter of applying “cookbook recipes”, designed by field
personnel, particularly those responsible for manufacturing products, for marketing them,
and so on: they are intended to be simple, and operate outside the realm of economic theory,
which they simply disregard.
In spite of their lack of rigor, they are widely used because of their simplicity, easy to under-
stand and apply, and above all capable of giving an idea of the profitability very quickly and
with a minimum of information. Among the best known, we must mention: the Payout-Time
(POT), the Return on Investment (ROI), the Turnover Ratio, and the Break-even Point,
among others. Their English names indicate both their origin and the geographical area where
they are most utilized. Moreover, so as to avoid confusion with certain other, more rational
criteria based on a similar approach, it is recommended that they be employed under their
Anglo-Saxon names.
Criteria based on the concept of discounting or present value-these are built on a detailed
or simplified application of the rules of economic calculation, sometimes with a few approx-
imations that may introduce some errors when compared with a strictly a posteriori, account-
ing approach. Indeed, it should not be forgotten that a priori reasoning, which is necessary in
the preliminary investigation of projects, has to be satisfied with data that are incomplete, or
assumed, or of varying credibility, whose processing neither requires nor allows any great
rigor. The most widely used criteria of this kind are: the capitalized discounted cash flow, the
internal rate of return, the discounted payback or recovery period, the economic cost of price,
and the relative capital enrichment.
The very fact that we have available a large number of alternatives and methods for assess-
ing the economic worth of a project shows how relative and unpredictable are the verdicts
obtained by using each of them in isolation. Not one is able to express accurately the full
potential of a project. Some are acknowledged to be more reliable, but it is usually essential
to apply several of them at the same time to provide a better coverage of the realities and to
confirm the results.
Under these circumstances, so as to be in a position to make a uniform presentation of the
information required by the use of these various criteria, and thus to bring out their similari-
ties and differences more easily, it is desirable to adopt the same nomenclature, i.e., to stan-
dardize their expression.
72 2. The Elements of Economic Calculation

With this in mind, we intend to utilize the following formal system:

V : annualrevenues
C : operating cost
D : annual operating costs
A : provision for annual depreciation
F : annual financial costs
B : annual gross profits
a : tax rate
CF: cashflow
Z : investment, i.e., in fact, depreciable capital
f : working capital
Z, : residual value
i : discount or interest rate.
Most of the concepts that such a nomenclature is based on will be defined and explained
in more detail later on. However, we can define the following relationships right away:
B=V-C (2-1)
The gross profits represent the difference between annual income and expenditures.
C=D+A+F (2.2)
The annual outflows, i.e., the operating cost described above, consists of the following
items:
0 Annual operating expenses, tied to the actual functioning of the facilities,

The annual allocation to provide for depreciation. This concept deserves further explana-
tion: however, roughly speaking, it is a notional expense intended to compensate for the
economic disadvantages due to the ageing of pieces of equipment, and the eventual need
to replace them or to substitute an equivalent or better system (see Section 2.2.4.2.A),
0 The financial costs from possible loans obtained for the construction of the production

units.
CF=B(l -a)+A (2.3)

The cashflow or annual income, also called cash from operations, represents the effective
earnings. It thus covers:
0 The real profits, i.e., profits after taxes, which are obtained from the gross profits by sub-

tracting the amount of the fiscal deductions, at a rate a,

0 That part of the income, arbitrarily deducted as an expense from the operating cost, that

constitutes the allocation to provide for depreciation, and that under this procedure-
which, moreover, is allowed under the fiscal legislation of most market economy coun-
tries-is deducted from the tax.

2.1.2.2 Empirical Criteria

As shown above, these criteria do not arise from the rational application of economic theories,
but are the result of a more pragmatic use of information.
 2. The Elements of Economic Calculation 73

A. The "Payout Time" or Simplified Payback Period Method

Generally speaking, the payback or recovery period for an invested sum is defined as being
the length of time at the end of which the total earnings produced by the operation of the cor-
responding facilities, i.e., the amount of the overall revenues less expenditures, including
taxes, are equal to the amount of the initial investment.
From a strictly economic viewpoint this measure should normally be based on the concept
of present value; this aspect will be discussed in more detail in Section 2.1.2.3.E.
Here we will describe a simplified version, to be applied as such, in spite of the approxi-
mations, and even the criticisms or contradictions concerning management or economic prin-
ciples that it entails.
By definition (and this must be particularly stressed), the Payout Time is given by the fol-
lowing expression:
Z
POT = - (2.4)
CF
This is then the time required for the cash flow to become equal to the initial investment,
in this case the Depreciable Capital, whose content will be examinedbelow, in Section 2.3.1.1.
Using Eqs. (2.1)and (2.2)this becomes:
Z - I
POT =
B(1- a ) + A (V - C)(1- a ) + A

The Payout Time method thus applies under the following assumptions:
0 Revenues and expenses are constant from one year to another: prices and quantities of

products, raw materials, reagents, by-products, catalysts, various energy or other require-
ments, and labor needs are consequently supposed to remain fixed over time.
0 The provision for depreciation, which is itself regarded as constant, is derived from a nec-

essarily straight-line calculation procedure, and assumes a complete recovery of the original
investment, i.e., that the equipment has no salvage value (see Section 2.1.2.3.B.a(5)).
0 By using operating cost and its mathematical expression (2.2),the POT accepts the exist-

ence of financial or equivalent costs on all of the initial capital. It thus assumes that this cap-
ital outlay comes from a real or notional loan subject to repayment at a real or pseudo rate of
interest.
Taxation is taken into account; consequently the POT constitutes a assessment criterion
essentially confined to market economy countries.
The POT is measured in years. Actually, the investment Z is expressed in local or reference
currency, and the cash flow CF in annual value. The ratio between the two is thus denomi-
nated in years.
The shorter the recovery period thus expressed, the higher the earnings-represented by
the cash flow-must be, since the initial investment is fixed; and so, a priori, the better the
economics of the project being considered must be. When comparing two projects or alterna-
tives, the one with the smaller POT is regarded as having the better economics.
74 2. The Elements of Economic Calculation

The expression (2.5) shows that the POT is less than the depreciation period n = I/A-
unless the operation loses money and thus has negative profits, a situation that would
condemn the project at once, since it would not even recover the initial investment. By way
of indication, for a large proportion of industrial units n is of the order of ten years (see Sec-
tion 2.2.4.2.A.b). Experience shows that an investment should not be considered in such cases
unless the POT falls below seven years, and preferably below five l. This limit is of course a
function of the economic situation: it is higher when economic conditions are favorable. In
difficult times, such as those experienced during the recent energy crises, the threshold is
around two years. This observation merely indicates the gap between revenues and expendi-
tures, i.e., between a product's sales price and its cost of price; the size of this difference
reflects the extent to which demand exceeds supply, which requires the economy to be
expanding.
In this regard, it should be noted that the expression for the POT can be written using unit
costs, i.e., per ton produced or that could be produced, or more generally per unit or per piece.
To do this, we have only to divide the elements used in its formulation either by the actual
production or by the production capacity. We thus obtain, in the latter case:

I'
POT =
(VI- c')(l- a ) + A'
where
I' : Ilcapacity = unit investment
V': Vlcapacity = sales price
C': Clcapacity = cost of price
A' : Alcapacity = unit provision for depreciation.
Such a presentation has the advantage of showing that, for a given POT, it is possible to
deduce a sales price, assuming that the other parameters are known.

V'= (I'/POT) - A' + c'

(2.7)
I-a
If the value assigned to the POT is assumed to be a maximum (five years, for example) the
value of I" given by Eq. (2.7) is, for the same conditions, a "minimum profitable sales price".
It expresses the lowest product value that must be attained in order to recover the invested

1. Comment. In many cases, the life of the facilities is relatively short, or assumed to be for calcu-
lation purposes: usually less than or equal to ten years. This leads to even shorter payback periods, and
a need for rapid recovery of the initial investments. In practice, it is too difficult or even impossible to
obtain major loans from specialized institutions for such short periods. The financing for a given
project is usually obtained from the company's cash flow, which can take advantage of broader and
more favorable situations. In this case, the source of capital for a project under examination is
unknown, and from a strictly economic viewpoint we should employ the concept of present value. This
leads to an apparent contradiction, which may be brought up, inasmuch as the company treasury acts
as a lending organization at a notional interest rate that actually represents a present value calculation.
Consequently, we must take care to see that the interest rates used in the POT calculation are fully con-
sistent with the discount rates that are normally applied (see Section 2.1.2.3.A.d).
 2. The Elements of Economic Calculation 75

capital within a certain time. It should therefore be near the actual market prices, particularly
when substitutable products are concerned.
The simplified Payback Period method represented by the POT has the advantage of
enabling a quick analysis of a project’s economics: for this reason, it is well suited to the aims
of a preliminary evaluation. On the other hand, it displays all the drawbacks that generally
result just from utilizing this kind of criterion (see Section 2.1.2.3.E). In particular, the least
capital-intensive projects, the ones that require the smallest initial contribution, are the high-
est rated, and there is no consideration of the period after the recovery of the invested capital.
The POT constitutes a measure of the risk incurred by the investors, not by the producers.

B. The Return on Investment (ROI) Method

In its original definition, the ROI is calculated before taxes. For this reason, it is suited to the
analysis of projects in planned-economy countries. It is expressed in percent, and is described
by the following equation:
V-D
ROI = -. 100
I
In practice, it consists in relating gross annual earnings, i.e., the difference between reve-
nues and operating expenses, to the capital initially invested. The pre-tax calculation leads us
to exclude any parameter that is used in the procedure for determining the amount of these
taxes, such as the depreciation provision, and to some extent the real or notional financial
costs. Considering the comment made on this subject for the POT, this latter item should be
included, since the states that provide project financing generally apply discount rates, usu-
ally low ones (in the vicinity of 3%) at the national level. But the definition of the ROI must
be retained, and applied as is.
Experience shows that the return on investment falls in the vicinity of 30 to 35%, which is
equivalent to a POT of about five years. The higher the ROI, the better the project’s econom-
ics, and when two of them are compared, the one with the higher return is considered to be
the more attractive.
In terms of unit costs, the ROI is expressed as follows:

V‘-D’
ROI=--. 100 (2.9)
I’
where
V’: Vlproduction capacity = sales price
D‘: Dlcapacity = unit operating cost
I ’ : Ilcapacity = unit investment.
For a given ROI, expression (2.9) allows a sales price to be calculated. If we consider the
value thus selected to be a minimum, V’ then constitutes a “minimum profitable sales price”:

(2.10)
76 2. The Elements of Economic Calculation

By extension, some users have modified Eq. (2.8), which applies to planned systems, in
order to make it easier to compare projects from countries with different structures, on a pre-
tax basis and using the same kind of measures. For application in a market economy, this rela-
tion then becomes:
v-c
ROI = -.lo0
I
Again by extension, the concept of after-tax return on investment has been introduced. In
this case. the formulation is as follows:

(V - C)(1- a ) + A
after tax ROI = .loo (2.1 1)
I
where: C=D+A+F
Combining Eqs. (2.5) and (2.1 1) gives:

after tax ROI = -100

POT
The after-tax return corresponds to the inverse of the POT. It is 20% when the POT is five
years.
By analogy with the terms employed for profits before and after taxes, these two general-
ized forms may be described as the “gross” and “net” returns on investment.
With a view to simplification, the word Return is commonly employed without specifying
whether it is being applied before or after tax. This leads to confusion, which is all the greater
because the Internal rate of Return criterion, to be considered in Section 2.1.2.3.D, is itself
often simply called the “return” by many who use it. The uncertainty can be removed by bear-
ing in mind the average orders of magnitude that are equivalent to a POT of five years:
30% before tax ROI
20% after tax ROI
15% internal rate of return

C. The Other Empirical Criteria

There are several other more or less empirical criteria of profitability, designed to correct cer-
tain defects in earlier methods, or to emphasize certain particular aspects. As a rough guide,
we will mention:

Turnover ratio
This expresses the ratio between annual sales (V)and invested capital (Z) and is well suited
to the ranking of projects of completely different kinds, involving construction plans that are
far apart in their design, and consequently have markedly distinct initial investments. It con-
stitutes a measurement of the rapidity of the renewal or reconstitution of the amount of the
sums tied up.
 2. The Elements of Economic Calculation 77

Break-even point
In practice, this concept appears in various forms; for this reason its significance and its
content are usually explained differently by different authors and users. In its original defini-
tion, the break-even point is the effective production threshold below which the operations of
a plant designed to provide a certain level of manufacture, i.e., having a certain production
capacity, should be halted.
To determine the minimum utilization rate of the equipment (the ratio between the actual
and potential quantities coming from the units), a graph (Fig. 2.1) is constructed to show the
simultaneous variations of sales and operating cost in terms of this parameter. By sales we
mean the total turnover, after deducting possible related expenses (packaging, sales and
advertising costs, repayment of research and development costs, etc.). The break-even point
corresponds to the intersection of the two curves (usually straight lines). Its x-value represents
the rate of utilization for which the revenues and the expenses directly related to the units
under consideration balance each other. Experience shows that for facilities that consist for
the most part of pressure tanks, whose costs vary roughly according to the power 0.6 of the
quantities processed in the equipment (see Section 2.3.2.l), the threshold of economic utili-
zation corresponding to the break-even point is itself around 60%, for widely used products.
Generally speaking, the lower the value obtained in this way, the better are the project eco-
nomics, since they are less sensitive to fluctuations in demand. The advantage of this criterion
is that it introduces an upper limit to the production capacity derived from a market analysis:
in particular when, in seeking to improve the situation by economies of scale, a proposed
development is brought forward in time.

Total sales
Income after

deduction of taxes

m
e
W
s

Fixed charges

J
0 10 20 30 40 50 60 70 60 90 H)
Utlltzatlon or load rate (YO)

Figure 2.1 Determination of the break-even point or equilibrium threshold

from the utilization rate. (Source: ZFP)
78 2. The Elements of Economic Calculation

The EMIP (Equivalent Maximum Investment Period) method, and

its complement the IRP (Interest Recovery Period) method
This procedure is based on the Payout Time procedure and attempts to correct certain
shortcomings, in particular by suggesting an approach beyond the payback date (in this case
called the break-even point) that enables the further evolution of the project to be considered.
This economic analysis model was introduced in the early 1970s but has not been developed
significantly. In these circumstances, we will simply point out that it consists of constructing
a graph that shows the variations over time of the annual income (which may be expressed as
present values), from the time 0 when the decision to construct was taken. The EMIP is the
ratio of the area defined by the x-axis and the curve up to the break-even point, to the maxi-
mum cumulative expenses. The IRP is the period obtained by demarcating the equivalent area
beyond this point. Both are expressed in years. The lower they are, the better are the project
economics. Standardizing the curves enables comparison among several alternatives.

2.1.2.3 Criteria Based on the Concept of Present Value

This type of criterion reflects the principles of economic calculation, in particular the idea of
present value, which will therefore be considered first.

A. The Concept of Present Value

a. Background

Project financing involves locking up sums whose final total is equal to the overall invest-
ment. The disbursements are of course made progressively, throughout the planning and con-
struction stages, but they have to be provided without any expectation of short-term benefit,
since no income is likely to be generated by the operation of the facilities. Without special
incentives or other compensations, the act of investing thus leads to a definite loss. This sim-
ple observation doubtless explains why some return is expected from any sum that is locked
in, whatever its nature or origin.
Theoretically, all or part of the required project investments may be obtained via loans
contracted with banking organizations, or from the managing company’s equity funds. Now,
although we all know that loans require the payment of interest, there is no reason why other
sources of financing should not make similar arrangements. In fact, they have a right to
expect at least the same level of return, to make up for the losses or damages involved in pos-
sibly having to make their own loans to third parties, or to make acquisitions for more imme-
diate future profits or benefits. The concept of present value has no other purpose than to
provide economic guarantees of proper returns from all the sums invested.
In practice, distinguishing between capital sums on the basis of their source can be a tricky
business in the case of larger industrial companies: several projects may be under way at
once, their financing handled by a single central treasury. The treasury obtains its funds both
from the company’s own varied operations, and from various shorter- and longer-term loans.
In such a situation, it is impossible to define the real financing terms of a particular project.
Effectively, it is as if all the sums required for the project were borrowed from the central trea-
sury and repaid at a pseudo interest rate.
 2. The Elements of Economic Calculation 79

In some cases, however, for example with companies that are smaller in scope or specially
created for the purpose, it is sometimes possible to differentiate the various capital sums by
origin, and to treat them accordingly.
Generally speaking, we may distinguish three types of financing:
0 Regular loans obtained directly from everyday lending organizations,

Equity funds, or funds regarded as such,

Contributions from a central treasury, or fictive loans arranged by the treasury.
In the first type, capital remuneration comes from the imposition of an interest rate; in the
other two a related mechanism is required: the discount rate. However, we still have to spec-
ify the conditions for equating these returns. From the a priori viewpoint implicit in the eval-
uation, the real terms of the financing operations are usually unknown or provisional: in such
circumstances, it therefore seems more realistic to employ constant rates throughout the
period of project planning.

b. Borrowed Capital
The payment of interest in return for sums that are borrowed or loaned may be calculated in
various ways.

Simple Interest
In this case, the interest rate, defined as the amount that the unit of capital costs or returns
in the given unit of time (year), is held constant. At the end of k years, the sum Sk that must
be paid back or recovered from the loan or investment of the initial capital So, at simple inter-
est rate i, is:
Sk = SO (1 + ki)
where So x ki represents the total interest.
When the operating period is less than one year, the interest due or recoverable is prorated
to the actual number of days that the loan or investment was in effect, as a fraction of either
360 days (ordinary interest) or 365 days (exact simple interest).

Compound Interest
Simple interest, whether in the form of partial payments at the end of each period of time,
or as a lump sum at the end of the operation, yields exactly the same amount of money. In the
second of these cases the lending organization experiences a loss of earnings, because it could
have made other profitable investments during that time, using the money from the partial
payments. To compensate for this disadvantage, compound interest is utilized: here, any
interest left unpaid at the end of each basic time period (for example, a year) is itself treated
as an integral part of the loaned or borrowed capital during succeeding years.
Under this arrangement, investing or obtaining a sum So in year 0 leads to a return, or the
obligation to repay:
in year 1: SI = So (1 + i )
in year 2: S, = S1 (1 + i) = So (1 + i)2
80 2. The Elements of Economic Calculation

s k represents the capitalized value at year k of the initial amount SO,and i represents the
interest rate applicable to the fractional unit of time considered.
One year is usually selected as the unit of time for calculations, but in the case of short-
term loans or borrowings it is often necessary to employ shorter periods-a month, or even a
day-and to set interest rates in consequence.
Because it is essential to make comparisons on a uniform basis, even though the interest
rates may have been determined under differing conditions, it is customary to reduce them all
to a common period of time: the year.
This method then leads to the definition of:
0 An effective interest rate (ir): the value of the interest rate i applied to the period of time

considered, prorated to a period of one year,

A nominal annual rate of interest (i,): the product of the numberp of partial time periods
in one year, and the value of the rate i applied to that time period, or:
i,=pXi
In fact, the solution is usually derived from the inverse procedure:

i = ”1
P
Generally speaking: i, 2 i, where the equality represents a period of one year. In practice,
the effective rate of interest i, is used for calculation.
To determine it, since i, is known we employ the expression:

applied to a one-year period, that is to say, to k = p , or:

(2.12)

p being the number of unit periods in a year.

Since we also have:
s, = S o ( l + i , ) (2.13)

by combining Eqs. (2.12) and (2.13) we obtain:

Continuous Compound Interest

The industry generally uses annual compound interest: sometimes it resorts to monthly
interest, but a daily rate is rare. In this regard, one could theoretically consider even shorter
periods, even infinitesimal ones, and thereby define a continuous compound interest.
 2. The Elements of Economic Calculation 81

This may be calculated by taking the limiting case of the following expression, where p
tends to infinity:

i,=
(1+“
a)” -1

which gives, when p is infinite: i = e’a - 1

k
and consequently: eiak = (1 + ir)

Under these conditions the expression:

becomes: s k = soeiak

or:

Which returns us to the general expression for the capitalized annual value of an initial sum
SO.

c. Present Value. Discount Rate.

When a company invests capital to finance a project, it must be compensated for the lost earn-
ings or the “psychological” constraints and losses that ensue. In particular at least a partial, if
not a full equivalent should be provided for disadvantages such as:
0 Depreciation of the currency,

The impossibility of making investments with banking or related organizations instead,

0 The inability to make more immediately profitable or more diversified investments,

Or to increase its working capital, distribute higher dividends, engage in stock exchange
operations, etc.
The bottom line is that having a certain sum now and having the same amount at the end
of k years is not the same thing. Let C
A be a coefficient representing all the inconveniences
mentioned above, some of which cannot be directly expressed in figures, such that it shall be
equivalent to have, on an annual basis:
0 F1 in a year’s time

0 Or C; F1 today.
82 2. The Elements of Economic Calculation

In this case, C‘; F1represents the current or at least the discounted value of the sum F1 Ci
then constitutes a present value coefficient, also called a “psychological discount coefficient”
for the next year versus the current year.
k
In the same way, we may accept that it is possible to define coefficients Cf , C; , ... ck-1,
etc., such that it is the same thing to have a sum Fk in the year k, or an amount Fk in the year
k-1.
This, then is the same as having Fk in the year k, or in year 0 having A , available, such that:

A , being the current value of Fk

If: c,k = c:,c: c; ... c;-,

we obtain: h=C,kFk
c,k being the present value coefficient for year k with respect to year 0.
Placing reality above possibility, i.e., favoring the present over the future, usually leads us
to choose values of C f - , that are less than one. It may be different in certain particular situ-
ations: it is otherwise when we include past events. In this case, the result is comparable to
the one obtained when we use deflation indices (see Chapter 1).
In order to better convey the tendency of present value coefficients to be less than one, it
is convenient to express them in the following form:
1
c;-, =- k
1 f ak-1
in which which is itself less than one, has the properties of a rate and is therefore called
the discount rate or the psychological discount rate for year k with respect to year k-1 .
In this case, we have:

Fk
A,=

In addition, when:

we obtain:

which can also be written:

and so: A,=- Fk

(1 +u p
 2. The Elements of Economic Calculation 83

d. Properties of Interest and Discount Rates

The process of evaluation may employ the concepts and methods of economics, but at least
it has to apply them in simplified form, or with certain simplifying assumptions, in view of
the credibility of the information available and the conditions under which it operates.
With this in mind, it may be useful to consider a perfect financial market, whose principal
characteristics are the following:
The existence of a single rate of interest for all transaction,
The possibility of investing or borrowing any sum at this rate, i.e., a sum of any size,
The rate remains constant over time.
In this case, it amounts to the same thing:
Either to have the sum SO in year 0, or sk = SO(1 + i)k in year k,

Or to have Fk in year k or: A, = ~ Fk in year 0.

(1 + i ) k
In the first alternative, we have the expression for taking interest into account; in the sec-
ond, we recognize the equation for the present value concept. Consequently, in a perfect cap-
ital market, the discount rate is equal to the market's interest rate. From a purely formal point
of view, it thus becomes possible to express both concepts in similar terms, with the discount
rate acting as a pseudo interest rate.
It may also be desirable to free ourselves from certain constraints that, although they are
external to the study being conducted, introduce additional uncertainties. Depreciation of pur-
chasing power is such a case: as shown in Section l .3. l . l during the consideration of product
price variation, we can rid ourselves of this problem by operating in constant money, instead
of current money. To do this, we must generalize the concept of the deflation coefficient or
index developed for that purpose.
The approach consists in fact of seeking homogeneity by expressing the entire flow of
funds, not just the prices, either for example in current dollars or euros at the market values
for the year under consideration, or in constant dollars or euros relative to a reference year
and consequently having a fixed purchasing power over time.
To maintain the consistency of this approach, we must operate in terms of a discount rate,
and distinguish between:
i' = discount rate for current money
and: i = discount rate for constant money.
For the practical justification of this distinction, we go back to the meaning of the coeffi-
cients CA, C: , ... C i - , ,and separate within each of them the respective annual depreciation
of purchasing power: DA, Q2, ... Dkk_l.This gives equations of the form:

k - k E k
ck-l - Dk-l k-1
84 2. The Elements of Economic Calculation

Now: ct-I
1.e.:

or:
If moreover we assume, as is the case in evaluation, that the coefficients are constant from
one year to another, the equation becomes:

C,k = ( C ) k = ( D ) k ( E ) k

If we now postulate, for reasons analogous to those given above, that:

c=-1 1 1
, D=- and E = -
1 +i' l+d l+i
i and i' being defined as above and d being the rate of depreciation of purchasing power, we
obtain:

--
1 - 1
1+ i' (1 + d ) ( l + i)

or 1 + i' = (I + d)( 1 + i) as long as none of these factors is zero.

Thus, as a first approximation:
i'=i+d
i', i, and d being less than 1, and i x d being negligible. In other words, the discount rate for
current money is effectively equal to the discount rate for constant money, increased by the
rate of purchasing power depreciation.
In spite of the subjective character that its name implies, the psychological discount rate,
which is the discount rate, can at least partially be explained in more concrete terms. Thus,
within a company, its value is linked to a number of considerations, and more particularly to
two of them:
The terms under which the company treasury is financed,
How the available funds are managed.
At any given time, a company's financial resources are made up of cash flows of several
kinds.
They may be short-, medium-, or long-term loans, the company's equity funds from vari-
ous sources, earnings and subsidies of different kinds, etc. Each of these sources carries an
effective interest rate if it relates to a loan or an investment, or a pseudo interest rate if it does
not. With occasional exceptions, the financing of projects from a central treasury disregards
the nature of the capital involved. The remuneration of the sums laid out must at least reflect
the treasury's average condition, i.e., it must express a certain balancing between the various
methods by which the treasury is funded, depending on the size of their respective contibu-
tions and their specific financial terms.
 2. The Elements of Economic Calculation 85

Looked at in this way, the minimum discount rate employed should therefore be equal to
the arithmetic mean of the various funds involved, each one associated with its own particular
interest or discount rate.
The result of this calculation actually represents only a threshold, or a point below which
it is unreasonable to fall, on pain of early bankruptcy. It is also necessary to consider how the
overall management of the funds available to the company is to be handled. As already noted
in the discussion of empirical criteria, and as we shall also see when we turn to those based
on present value, the purpose of these economic indicators is to enable projects to be ranked,
particularly in decreasing order of attractiveness, independent of any measure of their intrin-
sic profitability. Obviously, a company will only make new investments that fall within its
budgetary envelope. Preferably, it will get involved in the operations that appear most attrac-
tive, specifically the least risky and most lucrative ones. The economic criteria are there to
make its choice easier. This does not prevent a company with available cash from deciding,
in spite of everything, not to invest, or if it does nevertheless propose to go ahead, from decid-
ing to set a minimum profitability level for doing so. Below this threshold it will prefer to use
its funds for other purposes, in particular to reward its shareholders by giving them dividends.
Studies of the internal rate of return (see Section 2.1.2.3.D) illustrate how to put such an
approach into practice.
In fact, investing comes down to taking on risks, to accepting low discount rates: in par-
ticular, ones that are close to the values that emerge from analyses of financing modalities. In
short, the lower this rate, the more projects will appear worthy of interest. To generalize, a
conservative company will make few new commitments, and will favor payments to its share-
holders: consequently it will use high discount rates. Conversely, a dynamic company will
want to invest and to embark on new projects: it will tend to lower its discount rates. Actually,
it will do so only so far as it is able, depending especially on its size and its financial guaran-
tors. Thus, the most solid establishments use the lowest rates, but they cannot be lower than
the ones used by the most financially credible organizations of all, i.e., the ones belonging to
the State where the industrial facilities are located. Indeed, state institutions are usually the
most credit-worthy, inasmuch as they always have the resources to give back to the taxpayers,
to erase their debts. Of course, this is not necessarily the case, since some countries, particu-
larly developing ones, may be forced to suspend payments. Except for this kind of situation,
where international organizations must make up the deficiency, when States-particularly
the most heavily industrialized ones-need to finance major public-sector projects, they are
obliged to invest and recover their money over the long term. To ensure a minimal return on
the capital tied up in this way, they set a discount rate by reference to the one used for long-
term loans (= 5%), to which they add management charges and a little danger money (= 2%).
In the past, certain countries periodically published outlines of their major national economic
trends, specifying the value of the discount rate at which they intended to repay their invest-
ments in future years. The worldwide liberalizing trend that is currently under way, with the
privatizing of major State corporations and the growth in personal shareholdings, is down-
grading public financing and emphasizing financing by international banking organizations
and private institutions. Consequently, even if the concept of planning sometimes persists, it
now represents no more than a guideline. Table 2.2 provides examples of the constant money
levels advocated by various countries. They show a variation over time that reflects the eco-
nomic conditions and, possibly, the sector concerned (e.g., USA). These variations have no
86 2. The Elements of Economic Calculation

Table 2.2 Examples of the discount rate in various countries

Country Applicable year I Rate (YO)

Belgium 1969 7.0
1976 4.0
1982 4.6
USA
Water 1974 40.0
1978 3.5
Agriculture 1978 5.0
Defense 1978 10.0
Development assistance 1978 8.0 to 12.0
Fast-breeder reactor 1978 5.0
Oil and gas 1980 8.0
France 1960 7.0
1965 7.0
1970 10.0
1975 9.0
21985 8.0
United Kingdom 1961 2.0
1969 10.0
1976 7.0
1979 5.0
USSR 1969 8.0
(Source: IFP)

other purpose than to encourage manufacturers to get involved to a greater or lesser extent in
new operations or the modernization of existing facilities.
The discount rates recommended by states are an indication of their political will to pro-
mote a more or less aggressive industrialization.

B. Calculation of the Discounted Cash Flow and the Net Present Value
The application of methods based on the concept of present value first of all requires us to
determine the annual income, then to reduce it to its present value, and finally to examine its
variation over time, i.e., in the absence of any simplifying assumptions, to draw up a schedule.
From an accounting viewpoint, we can look at the revenues and expenses for each year
that are actually credited or obligated.
In the a priori, or evaluation approach, we can estimate these data only by looking at the
future, so as to define a production program and to predict the probable variation of the prices
of various products, power sources, etc. involved in the project under consideration. Conse-
quently, although the construction of a schedule may be justified in one case, it may be unnec-
essary in another, particularly when the available information is insufficient or unreliable.
 2. The Elements of Economic Calculation 87

a. Calculation Methods

Nevertheless, before undertaking an approach of this kind certain details must first be speci-
fied, e.g.:
(1) The type of procedure selected: current money or constant money. So as not to intro-
duce too many assumptions and uncertainties, constant money is generally preferred for eval-
uations; this requires working with a constant discount rate.
(2) The date, actually the year, for which the work is being done, since under this approach
there has to be a reference. Usually the choice for the base year 0, in whose money all the cash
flows are expressed, is the one when the facilities are actually put into operation.
(3) The duration of the study: n (in years). For the sake of convenience, or a desire to keep
it simple, n is generally made equal to the depreciation period. Sometimes a different number
is selected, particularly when the profitabilities of two competing projects are to be compared.
This possibility can best be illustrated by an example: the comparison of two alternative
versions of the same facility for producing ammonia. The first proposes the construction of a
single, high-capacity unit (300000 t/yr), with start-up in year 0 and a ten-year depreciation
period. The second proposes a two-stage construction, to match the growth in demand: a first
facility of smaller size (1 50000 t/yr), starting up in year 0 and depreciated over ten years, and
a second one, identical in its operation, starting up in year 5 and also depreciated over ten
years. The comparison will be valid only if it has a uniform basis; consequently there are two
possible solutions for the duration of the study:
0 Ten years, in which case only the 300000 t/yr and the first 150000 t/yr units will be

depreciated by the end of the period. The unrecovered portion of the second smaller unit
still has to be accounted for, for example by applying the concept of salvage value.
0 Fifteen years, where from the tenth year onwards no depreciation provisions for the

300000 dyr and the first 150000 t/yr units need be deducted from profits.
In both cases, the schedule must include, at year 5, an intermediate investment correspond-
ing to the second 150000 t/yr facility. In spite of the additional uncertainties that may result
from it, the preferred solution should be the one with the longest duration, so as to take better
account of effects related to the prospects for market growth.
(4) The terms of taxation, i.e., the possibility of making calculations before or after their
deduction from profits, which amounts to deciding whether or not to consider fiscal impacts.
(5) The salvage value of the facilities. At any time, the equipment at an industrial site may
be employed for other purposes or sold, even if only at scrap metal prices.
The price of these materials is essentially tied to their proposed usage. At that time, it rep-
resents their “salvage value”.
This approach can be applied during or after the depreciation period, and may have a real
or notional character.
From an accounting viewpoint, which reflects the actual cash flows on the actual dates
they took place, the calculation of the salvage value of a facility is possible only when it is
sold, if a sale takes place, or if not, when the units are dismantled. In this case, it is desirable
to consider the possible added values and their related taxes.
88 2. The Elements of Economic Calculation

For evaluation purposes, the economic tool can be much more rudimentary, since it is
employed to analyze facilities that are usually notional. In addition, it is intended to make the
choice between several alternatives easier, and constitutes a handy means of comparison. So
the principal concern is to define identical study conditions for the various alternatives con-
sidered, and, particularly, to adopt the same reference date and duration. Although it is a priori
a tricky matter, even impossible, to foresee the end of the project, and thus to know the value
of the equipment at the end of its life or of its utilization in a given facility, it is possible that
at the end of the selected period it may not yet have been completely depreciated. To express
these possibilities the salvage value may, unless otherwise indicated, be taken to be zero, or
equal to the fraction of the invested capital that is unrecovered at the end of the study period.

b. General Expression
Although the chosen formulation permits a broader generalization, it seems preferable to
limit it to the more common circumstances encountered in project evaluation. Thus, calcula-
tions are usually conducted in constant money, and the reference year 0 is the year when the
units are put into industrial production. Finally, the facility is assumed to be in a market econ-
omy country, where profits are subject to taxation, and certain charges, in particular the
depreciation provision and financial costs for contracted loans, are deductible from the tax.
The recommended approach for determining the annual income then consists of calculat-
ing the result of the revenues and expenses that are produced by the operation of the facilities,
i.e., the corresponding cash flows (the differences between the annual revenues and
expenses). This exercise is made easier by setting up a schedule showing the variations from
one year to another of the various constituents of the flows of funds. Thus, for each year of
operationp, the cash flow (CF), is expressed as follows:
(CF), = V, - Dp - F, - ‘p - (Vp - D, - F, - A,) u
or: (2.14)
in which, as noted in Section 2.1.2.1, V, represents the operating revenues for the yearp, D,
the corresponding expenses, A, the provision for depreciation, F, the financial costs for pos-
sible loans, a the tax rate, assumed constant for the period of the study, and 1, the total invest-
ments made in year p .
Introducing the operating cost:
C, = D, i-A, + F,
allows us to simplify Eq. (2.14) above, which may then be written:
(CF), = (V, - C,) (1 - a ) + Ap -,Z (2.15)
Following the procedure mentioned in Section 2.1.2.3.A.q determining the current value
of the cash flow for yearp amounts to dividing by (1 + i)p, i being the discount rate for con-
stant money under the assumptions of this calculation, i.e.:
 2. The Elements of Economic Calculation 89

c. Principal Type Applications

When the terms of the calculation selected are specified or changed, the general expression
(2.16) may take on different forms. Among the principal type applications, some particular
ones may accordingly be highlighted.

Financing Provided from the General Treasury

In project evaluation, financing is in practice assumed to come from a treasury that is itself
supplied on terms that are usually unknown. In fact, the evaluation study is still too prelimi-
nary for the financial setup to be known and included, or too vague for the possible variation
of discount rates, tax rates, etc. to be envisaged. In these circumstances, it is simply not pos-
sible to determine the amount of the financial costs Fp, because the debt percentage has not
been explicitly established. In addition, it would seem out of the question to be able to deter-
mine a specific discount rate i for the project(s) under consideration, because it is designed
precisely to eliminate the effect of loans, if there should be any. So we must take care in using
Eq. (2.16), that on the one hand Fp = 0, and that on the other hand i does indeed correspond
to the usual discount rate for the company concerned, or to the one for the country where the
facilities are located.
The present values of the annual incomes can then be written:

Depending on the value p , lying between the reference year 0 and the final year of the
study, n, they have varying significances, as follows:
0 In year p = 0, revenues and operating expenses are nil, depreciation has not yet taken
effect, and ZO is equal to the initial investment, i.e., the sum of the investments in depreciable
capital (Z) and working capital v), i.e.:
Z,=Z+f
which leads to:

0 At some year p , lying between the years 1 and n - 1 (1 5 p 5 n - l), there are usually no

new investments, hence I, = 0. This is not always the case. Sometimes, we need to consider
additions of capital andor working capital for process improvements, clearing bottlenecks,
etc.: in the previous example concerning the staging over time of the construction of two iden-
tical ammonia-producing facilities, so as to match the growth in demand, it would be neces-
sary at the year p = 5 to take account of the depreciable and working capital for the second
facility, put into operation at that time. The revenues and operating expenses vary according
to the quantities produced and the average prices charged from one year to another. The allo-
cation for depreciation A, is generally assumed to be constant over time, unless a declining-
balance system is used, or of course if new investments are made, which amounts to propor-
90 2. The Elements of Economic Calculation

tionally increasing the allocation from the moment when the units become operational. The
expression for the present value of the incomes then becomes:

0 At yearp = n, i.e., at the end of the study period or the presumed life of the facilities, -I,,

represents an income that is the value of the recoverable capital. It covers the working capital
J; if it has not changed in the interim, and the salvage value I,, corresponding to the total of
the undepreciated portion of the facilities, if there is any at that time, and the revenue from
their sale or re-use. During this operation, there would be added values that must be consid-
ered. In evaluation, where the calculation concerns the comparison of projects that are usually
imaginary, and is conducted without any consideration for accounting principles,f is assumed
to be constant and I, either zero or equal to the portion of the capital not yet recovered. This
procedure avoids the problem raised in accounting by the tax on added values. Note that over
time the working capital becomes progressively smaller in real value by an amount equal to
the discount: this explains why it is legal to increase it proportionally, without it necessarily
constituting a concealment of profits. The total I, corresponds to the income -(I, +A; the rev-
enues and expenses are the ones for the year n: V,, and C,. The allocation for depreciation A,,
corresponds to the last annual payment. In these circumstances, the discounted income
becomes:

Financing on Known Terms

Opportunity or Shadow Cost of Equity Funds
In some circumstances, it is possible to define, for a particular project, the portion of the
investment cost that is financed from loans, and the portion that comes from the company's
equity funds. Two kinds of question then arise, concerning either the best utilization of these
funds respectively, or the use of the total sum required. The general expression (2.16) is still
valid, but its application requires caution.

First Possibility
The first possibility is that the financial costs Fp are not zero, they represent the interest at
yearp on the sums initially borrowed, or, if need be, when additional investments are made.
Consequently the value of the discount rate employed, which reflects the average state of the
company's treasury based on its sources of capital, the amounts of the respective contribu-
tions, and their repayment terms, must exclude everything that relates to the loans and involve
only the company's equity funds. In this calculation:
At year p = 0, I0 must include the sums actually laid out by the company responsible for
the project. If the initial depreciable capital I consists partly of a loan E, and partly of equity
funds P, i.e., if
I=E+P
 2. The Elements of Economic Calculation 91

we obtain:
I o = I +f - E = P + f

E in fact represents an income received at time t = 0. For the purpose of simplification, the
working capital is assumed to come exclusively from equity capital. In practice this might not
be the case: it might be partly financed using loans. If so, the financial costs Fp must be mod-
ified appropriately. The total discounted revenues may be written as follows:

At any yearp, lying betweenp = 1 andp = n - 1 ( 1 Sp S n - l ) , In,regardless of any addi-

tional investment, is no longer zero. It is equal to the repayment Rp of the Principal for year
p, calculated such that:
p=n
CR,=E
p=l

given that in year 0, RO= 0.

The expression for discounted income becomes:

The allocation to provide for depreciation A p still refers to all the capital initially required for
the construction of the facilities. If there is strict proportionality, i.e., use of the straight line
depreciation method, Ap and R p are constant, so that:

with A y ) = allocation to provide for depreciation of equity funds.

Consequently:
(Vp - Cp)( 1 - U ) + A p - Rp
- (VP-DP-AP-FP)(l -a)+Ap-Rp

with: D p = D, +R,
We recognize the initial expression, but with the advantage of considering only the equity
funds, and thus being, in theory, more consistent.
At yearp = n, -In represents on one hand the value of the recoverable sums, or (I, +A,
and
on the other the last repayment of principal R,. Then:
-In = (Z, +f)- R ,
92 2. The Elements of Economic Calculation

and:

Second possibility
In the second possibility, which attempts to consider the best utilization of all the capital
laid out, even though its origins as loans or equity funds may be known we must return to the
initial concept in which Fp = 0 and the discounting is the result of a representative weighting
of the complete average basket of funds. To illustrate this, it is best to work with an example.
If the initial depreciable capital then consists of 20% of equity funds, for which the discount
rate is 15%, and 80%of loans that pay an average of 5%, the approach is as follows:
0 Either we look only at the equity funds, in which case for I = 100, E = 80 and P = 20,

Fp = 0.05 x 80, i.e. 4, and i = 0.15,

0 Or we consider all the invested sums together. For I = 100, we then have:

0.05 x 80 + 0.15 x20

Fp = 0 and i= = 0.07
100
Here, of course, Rp = 0.

No taxation
When calculations are made for countries with planned economies, the concepts of fiscal
debits and deductions, particularly provisions for depreciation, do not exist. In expression
(2.16), Fp and Ap become zero.
In the case where financing is provided by funds fiom the company treasury or the state
concerned, the discounted income is expressed as follows:

since, as we do not know where the funds come from, Fp is also zero.
In the case where the financing structure is known, since Fp # 0, the income becomes:

where Ip includes the repayments Rp. This can happen when a State contracts loans from
international banking institutions.

C. Present Value Criteria of Profitability

Using expression (2.16) to determine the discounted annual income makes it easier to define
and to apply various criteria of profitability, notably the following:
Capitalized discounted cash flow
 2. The Elements of Economic Calculation 93

0 Internal rate of return

0 Payback period with discount.
Other criteria are based on the same procedure: some of them will only be outlined here.

a. Capitalized discounted cash flow, or Cumulative net annual discounted cash flow,
or Cumulative net present value 2
By definition a project’s capitalized discounted cash flow, or its total discounted income, is
equal to the sum of the present values of the annual cash flows from its operations:

p=n
(Vp - C,)(l- a ) + A, - zp
or: B = C (2.17)
p=o (1 +i)P

This criterion allows us to look at a project over the entire life of the facilities that it puts
to work, or at least over the depreciation or study period, i.e., for a sufficiently representative
length of time for the principal parameters to be considered. When it is a question of whether
or not to go ahead, the method first determines whether the operation is economically attrac-
tive, or if the capitalized discounted cash flow obtained is positive. In other words, the dis-
counted cumulative value of the revenues should exceed that of the overall expenses.
Consequently, when this is so, then the higher this criterion--expressed as a value-
becomes, the better the profitability of the project concerned is said to be. When two or more
projects have to be compared, the ones with the highest capitalized discounted cash flows are
considered to be the most profitable.

General case where financing comes from company treasury

The origin of this capital is unknown. Hence, the financial costs cannot be defined, and
Fp = 0. The discount rate i is used for all of the company’s treasury funds.
Depending on the year being considered, ZP has the following values:
p=o Zo=I+J
llpln-1 zp = 0
p=n In = -(Zr +f)
which gives the following expression, derived from Eq. (2.17):

2. In these expressions, the word “cumulative” or “capitalized” is often omitted, which may lead
to confusion with the simple annual income (see Section 2.1.2.3.B).
94 2. The Elements of Economic Calculation

or:

when Fp = 0.
In practice, in order to calculate a capitalized discounted cash flow we must proceed in
stages, since most of the parameters vary from one year to another: income, expenses, possi-
ble provision for depreciation, even discount and tax rates, working capital, etc. In these cir-
cumstances, it is best to set up a schedule like the one in Table 2.3, for determining the annual
incomes. To obtain the present values, we must also have the value of the following coeffi-
cient, which depends on the values of i and p :

a=- 1
(1+i)P

Table 2.4 shows how to do this. The accumulated value of the discounted cash flows, on
the bottom line of the schedule, is the desired capitalized discounted cash flow.
These calculations may at first appear tedious, especially when they become repetitive:
commercial software is now widely available to help out.

Special cases

When thefinancing method is known

Expression (2.17) becomes more complicated when only the profitability of the equity
funds matters. The financial costs are identifiable, so Fp # 0. The discount rate i does not
apply to the loans. Depending on the year, Zp has the following values:
p=o Z,=P+f
llpln-1 Ip = Rp
p=n I,,= +I,. +f)+ R,,
The capitalized discounted cash flow can then be written:

(vP- cP)(1- a ) + A, - R,

+E p=l (1 + i)'

where Fp f 0.

When there is no taxation

The concept of tax vanishes: a = 0. The expression (2.17) is simplified to:
 Table 2.3 Capitalized Discounted Cash Flow method. General case: financing provided from company treasury

Calculation
Comments
Years 0 1 2

.
i
Investments In general Ip = 0 -11 -12
for 0 < p < n and I, = 0
Sales volume V1 v2
Operating DP Variable charges, labor, DI D2
expenses(') maintenance, tax
and insurance, overhead,
and head office
Depreciation(') AP A1 A2
Financial costs(') FP These costs are zero: F, =0 0 0
Operating cost CP Cp = Dp +Ap + Fp Cl c
2
with F, = 0
Net profits
(for given a)
Depreciation AP
Cash Flow (CF)p

Discount factor 1/(1 + i)P 1/(1 + i)

(for given i)
Discounted income (CF)J( 1 + i)P (CF),/( 1 + i) (CFh/( 1 + i)2
Capitalized
Discounted
Cash Flow I D B =

(1) Optional calculation if operating cost is known.

x-
p=n

p=o
(CF)p
( 1 +i)P

Very important note: Be sure that the financial costs are zero in this operating cost calculation.
(Source: IFP)
96 2. The Elements of Economic Calculation

1
Table 2.4 Determination of the coefficient a:a = -
(1 +i)"

n 0.04 0.05 0.06 0.08 0.10 0.12 0.15 0.20 0.25 n

1 0.96 15 0.9524 0.9434 0.9259 0.9091 0.8929 0.8696 0.8333 0.8000 1
2 0.9246 0.9070 0.8900 0.8573 0.8264 0.7972 0.7561 0.6944 0.6400 2
3 0.8890 0.8638 0.8396 0.7938 0.7513 0.7118 0.6575 0.5787 0.5120 3
4 0.8548 0.8227 0.792 1 0.7350 0.6830 0.6355 0.5718 0.4823 0.4096 4
5 0.8219 0.7835 0.7473 0.6806 0.6209 0.5674 0.4972 0.4019 0.3277 5
6 0.7903 0.7462 0.7050 0.6302 0.5645 0.5066 0.4323 0.3349 0.2621 6
7 0.7599 0.7107 0.665 1 0.5835 0.5132 0.4523 0.3759 0.2701 0.2097 7
8 0.7307 0.6768 0.6274 0.5403 0.4665 0.4039 0.3269 0.2326 0.1678 8
9 0.7026 0.6446 0.5919 0.5002 0.4241 0.3606 0.2843 0.1938 0.1342 9
10 0.6756 0.6139 0.5584 0.4632 0.3855 0.3220 0.2472 0.1615 0.1074 10
11 0.6496 0.5847 0.5268 0.4289 0.3505 0.2875 0.2 149 0.1346 0.0859 11
12 0.6246 0.5568 0.4970 0.3971 0.3 186 0.2567 0.1869 0.1122 0.0687 12
13 0.6006 0.5303 0.4688 0.3677 0.2897 0.2292 0.1625 0.0935 0.0550 13
14 0.5775 0.5051 0.4423 0.3405 0.2633 0.2046 0.1413 0.0779 0.0440 14
15 0.5553 0.48 10 0.4173 0.3152 0.2394 0.1827 0.1229 0.0649 0.0352 15
16 0.5339 0.4581 0.3936 0.2919 0.2176 0.1631 0.1069 0.0541 0.0281 16
17 0.5134 0.4363 0.3714 0.2703 0.1978 0.1456 0.0929 0.045 1 0.0225 17
18 0.4936 0.4155 0.3503 0.2502 0.1799 0.1300 0.0808 0.0376 0.0180 18
19 0.4746 0.3957 0.3305 0.23 17 0.1635 0.1 161 0.0703 0.03 13 0.0144 19
20 0.4564 0.3769 0.3118 0.2145 0.1486 0.1037 0.061 1 0.0261 0.01 15 20
21 0.4388 0.3589 0.2942 0.1987 0.1351 0.0926 0.053 1 0.02 17 0.0092 21
22 0.4220 0.3419 0.2775 0.1839 0.1228 0.0826 0.0462 0.0181 0.0074 22
23 0.4057 0.3256 0.2618 0.1703 0.1 117 0.0738 0.0402 0.0151 0.0059 23
24 0.3901 0.3101 0.2470 0.1577 0.1015 0.0658 0.0349 0.0126 0.0047 24
25 0.375 1 0.2953 0.2330 0.1460 0.0923 0.0588 0.0304 0.0 105 0.0038 25
26 0.3607 0.2812 0.2198 0.1352 0.0839 0.0525 0.0264 0.0087 0.0030 26
27 0.3468 0.2678 0.2074 0.1252 0.0763 0.0469 0.0230 0.0073 0.0024 27
28 0.3335 0.255 1 0.1956 0.1 159 0.0693 0.04 19 0.0200 0.0061 0.0019 28
29 0.3207 0.2429 0.1846 0.1073 0.0630 0.0374 0.0174 0.005 1 0.00 15 29
30 0.3083 0.23 14 0.1741 0.0994 0.0573 0.0334 0.0151 0.0042 0.0012 30
31 0.2965 0.2204 0.1643 0.0920 0.0521 0.0298 0.0131 0.0035 0.0010 31
32 0.285 1 0.2099 0.1550 0.0852 0.0474 0.0266 0.01 14 0.0029 0.0008 32
33 0.2741 0.1999 0.1462 0.0789 0.043 1 0.0238 0.0099 0.0024 0.0006 33
34 0.2636 0.1904 0.1379 0.0730 0.0391 0.0212 0.0086 0.0020 0.0005 34
35 0.2534 0.1813 0.1301 0.0676 0.0356 0.0189 0.0075 0.00 17 0.0004 35
36 0.2437 0.1727 0.1227 0.0626 0.0323 0.0169 0.0065 0.0014 0.0003 36
37 0.2343 0.1644 0.1 158 0.0580 0.0294 0.0151 0.0057 0.0012 0.0003 37
38 0.2253 0.1566 0.1092 0.0537 0.0267 0.0135 0.0049 0.00 10 0.0002 38
39 0.2 166 0.1491 0.1031 0.0497 0.0243 0.0120 0.0043 0.0008 0.0002 39
40 0.2083 0.1420 0.0972 0.0460 0.0221 0.0107 0.0037 0.0007 0.0001 40
41 0.2003 0.1353 0.0917 0.0426 0.0201 0.0096 0.0032 0.0006 0.0001 41
42 0.1926 0.1288 0.0865 0.0395 0.0 183 0.0086 0.0028 0.0005 0.0001 42
43 0.1852 0.1227 0.0816 0.0365 0.0166 0.0076 0.0025 0.0004 0.0001 43
44 0.1780 0.1 169 0.0770 0.0338 0.0151 0.0068 0.0021 0.0003 0.0001 44
45 0.1712 0.1 113 0.0727 0.03 13 0.0137 0.006 1 0.00 19 0.0003 0.0000 45
46 0.1646 0.1060 0.0685 0.0290 0.0125 0.0054 0.0016 0.0002 0.0000 46
47 0.1583 0.1009 0.0647 0.0269 0.01 13 0.0049 0.0014 0.0002 0.0000 47
48 0.1522 0.0961 0.0610 0.0249 0.0103 0.0043 3.0012 0.0002 0.0000 48
49 0.1463 0.0916 0.0575 0.0230 0.0094 0.0039 3.001 1 0.0001 0.0000 49
50 0.1407 0.0872 0.0543 0.0213 0.0085 0.0035 3.0009 0.0001 0.0000 50
(Source: IFP)
 2. The Elements of Economic Calculation 97

or:

depending on whether the source of capital is not known or is taken into account.

When income and expenses are assumed to be constantfrom one year to another,
along with the allocationfor depreciation
This artificial situation, which is rarely encountered in the real world of economics or
accounting, has a special advantage in the evaluation of projects, where it is sometimes diffi-
cult, even absurd, to introduce arbitrary market fluctuations, given the lack of precision in the
available information. Its attraction is therefore to allow use of a simplifying procedure,
which replaces the schedule with a single equation. For example, in expression (2.18):

when Fp = 0.
Vp, Cp, and Ap are constant, and respectively equal to V, C, and A , such that:

p=n
+[(V-C)(l-a)+A] ~ 1 (2.19)

p=l ( I +i)'
when Fp = 0.

p=n 1 (1 + i)" - 1
The factor ~ can be calculated; it is equal to: p=
p=l (1 +i)' i( 1 + i)"
Depending on the values of i and n, it can be obtained by reference to Table 2.5.
In the case where the method of financing is known, this simplification can be applied only
if the repayments Rp and the financial costs Fp are also identical every year. It then becomes
necessary to resort to a calculating trick by defining an average interest rate, as shown in
Section 2.2.4.2.B.
Since, in addition to these simplifying assumptions, the salvage value of the facilities is
often assumed to be zero in an evaluation, at the end of the depreciation period the expression
(2.19) reduces to the following equation:

(2.20)
i( 1 + i)"

when Fp = 0.
98 2. The Elements of Economic Calculation

Table 2.5 Determination of the coefficient p: p = (1 +i)" - 1

i( 1 + i)"
I
I
~

n 0.04 0.05 0.06 0.08 0.10 0.12 1.15 0.20 0.25 n

-
1 0.962 0.952 0.943 0.926 0.909 0.893 0.870 0.833 0.800 1
2 1.886 1.859 1.833 1.783 1.736 1.690 1.626 1.528 1.440 2
3 2.775 2.723 2.673 2.577 2.487 2.402 2.283 2.106 1.952 3
4 3.630 3.546 3.465 3.312 3.170 3.037 2.855 2.589 2.362 4
5 4.452 4.329 4.212 3.993 3.791 3.605 3.352 2.991 2.689 5
6 5.242 5.076 4.917 4.623 4.355 4.111 3.784 3.326 2.95 1 6
7 6.002 5.786 5.582 5.206 4.868 4.564 4.160 3.605 3.161 7
8 6.733 6.463 6.210 5.747 5.335 4.968 4.487 3.837 3.329 8
9 7.435 7.108 6.802 6.247 5.759 5.328 4.772 4.03 1 3.463 9
10 8.1 11 7.722 7.360 6.710 6.145 5.650 5.019 4.192 3.571 10
11 8.760 8.306 7.887 7.139 6.495 5.938 5.234 4.327 3.656 11
12 9.385 8.863 8.384 7.536 6.814 6.194 5.421 4.439 3.725 12
13 9.986 9.394 8.853 7.904 7.103 6.424 5.583 4.533 3.780 13
14 10.563 9.899 9.295 8.244 7.367 6.628 5.724 4.61 1 3.824 14
15 11.118 10.380 9.712 8.559 7.606 6.81 1 5.847 4.675 3.859 15
16 11.652 10.838 10.106 8.851 7.824 6.974 5.954 4.730 3.887 16
17 12.166 1 1.274 10.477 9.122 8.022 7.120 6.047 4.775 3.910 17
18 12.659 11.690 10.828 9.372 8.201 7.250 6.128 4.812 3.928 18
19 13.134 12.085 11.158 9.604 8.365 7.366 6.198 4.843 3.942 19
20 13.590 12.462 11.470 9.818 8.514 7.469 6.259 4.870 3.954 20
21 14.029 12.821 11.764 10.017 8.649 7.562 6.312 4.891 3.963 21
22 14.451 13.163 12.042 10.201 8.772 7.645 6.359 4.909 3.970 22
23 14.857 13.489 12.303 10.371 8.883 7.7 18 6.399 4.925 3.976 23
24 15.247 13.799 12.550 10.529 8.985 7.784 6.434 4.937 3.981 24
25 15.622 14.094 12.783 10.675 9.077 7.843 6.464 4.948 3.985 25
26 15.983 14.375 13.003 10.810 9.161 7.896 6.491 4.956 3.988 26
27 16.330 14.643 13.211 10.935 9.237 7.943 6.514 4.964 3.990 27
28 16.663 14.898 13.406 11.051 9.307 7.984 6.534 4.970 3.992 28
29 16.984 15.141 13.591 11.158 9.370 8.022 6.55 1 4.975 3.994 29
30 17.292 15.372 13.765 11.258 9.427 8.055 6.566 4.979 3.995 30
31 17.588 15.593 13.929 11.350 9.479 8.085 6.579 4.982 3.996 31
32 17.874 15.803 14.084 11.435 9.526 8.112 6.591 4.985 3.997 32
33 18.148 16.003 14.230 11.514 9.569 8.135 6.600 4.988 3.997 33
34 18.411 16.193 14.368 11.587 9.609 8.157 6.609 4.990 3.998 34
35 18.665 16.374 14.498 11.655 9.644 8.176 6.617 4.992 3.998 35
36 18.908 16.547 14.621 11.717 9.677 8.192 6.623 4.993 3.999 36
37 19.143 16.711 14.737 11.775 9.706 8.208 6.629 4.994 3.999 37
38 19.368 16.868 14.846 11.829 9.733 8.221 6.634 4.995 3.999 38
39 19.584 17.017 14.949 11.879 9.757 8.233 6.638 4.996 3.999 39
40 19.793 17.159 15.046 1 1.925 9.779 8.244 6.642 4.997 4.000 40
41 19.993 17.294 15.138 11.967 9.799 8.253 6.645 4.997 4.000 41
42 20.186 17.423 15.225 12.007 9.817 8.262 6.648 4.998 4.000 42
43 20.371 17.546 15.306 12.043 9.834 8.270 6.650 4.998 4.000 43
44 20.549 17.663 15.383 12.077 9.849 8.276 6.652 4.998 4.000 44
45 20.720 17.774 15.456 12.108 9.863 8.283 6.654 4.999 4.000 45
46 20.885 17.880 15.524 12.137 9.875 8.288 6.656 4.999 4.000 46
47 21.043 17.981 15.589 12.164 9.887 8.293 6.657 4.999 4.000 47
48 21.195 18.077 15.650 12.189 9.897 8.297 6.659 4.999 4.000 48
49 21.341 18.169 15.708 12.212 9.906 8.301 6.660 4.999 4.000 49
50 21.482 18.256 15.762 12.233 9.915 8.304 6.66 1 4.999 4.000 50
(Source: IFP)
 2. The Elements of Economic Calculation 99

Although the search for a less complex and more easily applied formulation may be justi-
fied by its ability to provide a quick approximation to an answer that is in any case based on
unreliable data, it nevertheless represents an invalid approach in terms of accounting and eco-
nomic theory. In fact, it is not so much using it that is the problem, as not fully understanding
the level of error being committed. Indeed, not only revenues and expenses, but also provi-
sions for depreciation, behave in different ways, depending on whether the calculations are
made in constant money or current money, i.e., in discounted values or not.
A straight-line depreciation gives the same amount each year only in current money. So to
assume that in constant money, the depreciation provision is the same from one year to another
implies that there has been a revaluation (which is legal, as long as it is also applied to the invest-
ments, and the possible added values are considered). Conversely, to use similar revenues and
expenses over time in current money actually amounts to accepting that the productivity of the
facilities diminishes over the assumed life of the project. Thus, strictly speaking, we cannot log-
ically assume that all the parameters used in the calculation remain constant at the same time.

b. Internal Rate of Return (IRR), or Internal Economic Return,

or Discounted Cash Flow Rate of Return

This method arises directly from the preceding one, since by definition a project’s internal
rate of return i, is equal to the value of the discount rate i, which cancels out the capitalized
discounted cash flow calculated over n years.
We must therefore find i, such that:

i.e., such that the expression (2.17), or the expressions derived from it for particular cases,
becomes zero.
To do this, we need to draw the graph of the capitalized discounted cash flows against the
discount rate (Fig. 2.2a). The required value of i (i,) is the one for which the curve intersects
the x-axis. This is then the maximum discount that will allow repayment of the capital used
to finance the project, without making a loss.
When using capitalized discounted cash flow, for an operation to be considered profitable
by a company the outcome resulting from the application of the discount rate must be posi-
tive; similarly, for the internal rate of return, the value has to be greater than the discount rate.
If it is not, the project must be rejected.
In theory, a project is said to be more attractive, in economic terms, in proportion as its
internal rate of return, expressed in percent, is high; when the comparison concerns several
of them, the one with the highest IRR is considered to be the most attractive. This criterion
has the advantage of being an assessment method that is independent of the more psycholog-
ical and arbitrary concept represented by the discount rate; moreover, it is just as concrete and
evocative, since it also has the character of an interest rate.
In practice, certain restrictions have to be placed on its application, and it is sometimes
necessary to confirm the results obtained by resorting to other decision criteria. In particular,
when several projects are being compared, it is useful to examine:
100 2. The Elements of Economic Calculation

Figure 2.2a Determination of the Internal Rate of Return: illustration of

method. (Source: IFP)

0 The relative slopes of the variation curves of the capitalized discounted cash flow
against the discount rate. In fact, the most attractive project is often the one with the lowest
slope. Because of this, it is most sensitive to possible fluctuations in income, and most depen-
dent on the reliability of the information on which it is based, i.e., on the accuracy of the data
initially available or on the validity of the assumptions used in the calculation, which often
come from market analyses (Fig. 2.2b).
The relative position of the curves. When they intersect, the comparative profitability of the
projects can be measured only in terms of the value of the discount rate corresponding to the
intersection (Fig. 2.2~).In fact, depending on the case-that is, on the i concerned-the project
with the highest internal rate of return is not necessarily the one that also has the highest capital-
ized discounted cash flow. This assumes of course that the DCF is positive, which automatically
implies IRRs higher than the chosen value of i. Thus project (2), whose internal rate of return is
larger than the rate of project (l), does not yield a higher capitalized discounted cash flow, and
with it an economically equivalent conclusion, unless the company’s discount rate i is greater
than iA, the x-value of A: the intersection point of the two curves (1) and (2). In other terms:
iA<i<i <i
1‘ ‘2

Alternative (2) is more attractive than (1) whatever criterion is applied. The capitalized dis-
counted cash flow obtained is lower if i < iA; in this case the conclusion reached is the oppo-
site of the one obtained by comparing the IRRs, since i < i .
‘I ‘2
 2. The Elements of Economic Calculation 101

Figure 2.2b Internal Rate of Return: comparison of two projects. General

case. (Source: IFP)

R
Discount rate ( I )

'r1

Figure 2 . 2 ~Internal Rate of Return: comparison of two projects.

Special case. (Source: IFP)
102 2. The Elements of Economic Calculation

However, there is a case for preferring the results given by the capitalized discounted cash
flow. In fact, applying expression (2.17) and the others derived from it shows that i, increases
as the depreciable capital I decreases. Consequently, projects that are less capital-intensive,
and thus require lower initial investments, show to greater advantage. This type of analysis is
therefore disastrous for major construction projects, whose profitability may be among the
best over the long term. This situation is often encountered when projects with very different
investment costs are compared, or the study period is too short.
As pointed out in Section 2.1.2.3.A, the concept of the internal rate of return allows us to
get a practical grasp on one of the essential factors in determining the value of a company's
discount rate: how its capital is managed. The lower the value of i, the greater the number of
projects with i, values that are higher, and vice versa. A ranking can be set up on the basis of
decreasing internal rates of return, so as to identify the most favorable projects, and in parallel
the cumulative total of the corresponding initial investments can be shown. Setting a value
for i then amounts to establishing a minimum threshold of interest, but it also defines the lim-
its of the budgetary envelope required for projects considered to be profitable. In practice the
opposite procedure is generally employed: the amount of available capital is known or deter-
mined, and the value of i is deduced. Obviously, the larger are the sums for investment, the
lower the discount rate hurdle can be set, and the more numerous and more ambitious the
projects are likely to be.
The internal rates of return that are acceptable to the major industries, while they may not
be as low as those of State organizations, which a r e - o r are said to be-the most securely
based, nevertheless have minimum values around 12 to 15%. These figures are roughly
equivalentto a POT of 5 years and an ROI of 30 to 35% (before tax). In addition, if we employ
an approach similar to that described during the discussion of these two criteria, we can set a
minimum IRR and calculate a corresponding sales price, which we will again call the mini-
mum profitable sales price. Following the notation for unit costs used above, i.e., per ton of
production, the simplified Eq. (2.20) gives us:

I' + f'(1 -a)- A j L + C'

.I=[
L
P J
1-a

a=- 1
with:
(1 + ir)"

(l+ir)n-l
P=
ir(l + ir)"

and F' = 0.

COMMENTS
0 The internal rate of return can be confused with the ROI. The minimal values
assigned to these two measures may perhaps enable them to be distinguished.
 2. The Elements of Economic Calculation 103

0 If an existing facility is expanded or restructured it may be desirable to determine

the “relative” internal rate of return for this operation alone. This then concerns only
the additional income generated by the extra production, and is obtained by reducing
an expression comparable to the ones above, so that only the additional investment
required for the changes or additions is considered.

c. Payback Period with Discount (Discounted POT)

In practice, it is not so much a matter of setting out to determine a duration, but of first setting
a date. Thus, by definition, the recovery or repayment date of a project is the date k when the
total discounted cash flows become positive. Therefore, it is such that:

It corresponds to the moment when the incomes generated by the construction and opera-
tion of the facilities are sufficient to balance the initial investments, these investments being
repaid on the basis of a discount rate set by the manufacturer who sponsors and operates the
plant.
By extension, the payback period is defined as the period separating this date k from the
one selected for the start of the economic calculation, generally the time 0 when production
starts. This discounted equivalent of the POT is also expressed in years. For a given project,
the shorter it is, the better the profitability is said to be. When several competing alternatives
are compared, the one with the shortest discounted recovery period is considered the most
attractive.
From a mathematical viewpoint, applying this measure consists of finding the lowest value
of k, between 0 and n, such that:

p=k
(vp- cp)(1 - u ) + A, - z,
c
p=o (1 + i)’
20 with F p = 0 (2.21)

General case
Financing is provided from the company treasury if k < n, which is desirable and the most
common; from expression (2.21) we obtain:

VP- CP)(1- a ) + A, 20 with Fp = 0

p=l (1 + i ) P

The salvage value Zrand the working capital are not recovered, in terms of their present val-
ues, until year n, and thus cannot be considered during the study period.
On a graph, the curve representing the growth in cumulative discounted cash flows over time
(Fig. 2.3a) cuts the x-axis at a point that gives the value of k for the date of full recovery.
 2. The Elements of Economic Calculation

Figure 2.3a Payback time with discount: illustration of method.

(Source: IFP)

Special cases
0 The financing method is known and only the repayment terms for the company’s equity
funds P matter. In this case, the expression (2.21) becomes:

-( P + f)+ gp=l
(VP - CP)(1- a ) + A,
(l+i)P
- R,
20 with Fp#O

0 Taxation is ignored; in other words, a = 0. The expression (2.21) can then be written:

20
p=l
(1 + i ) P

0 Revenues, expenses, and the depreciation provision are assumed to be constant from one

year to another, with financing provided from the company’s treasury. In this case, Eq. (2.21 )
takes the following form:
k
(1 + i ) - 1
-(I + f)+[(v - c)(1- a ) + A ] 20 with F = O
i ( l + ilk
 2. The Elements of Economic Calculation 105

Like the POT method, the minimum discounted payback period method leads to neglect-
ing the economic elements for the period later than a certain date, in this case time k.
For this reason, it may produce different results from those obtained, for example, by rank-
ing projects in order of decreasing capitalized discounted cash flows, since that criteriontakes
into account the income and expenses from the use of the facilities over their entire lifetime,
or at least their depreciation period.
Hence, as shown in Fig. 2.3b, for two projects (1) and (2) requiring different initial invest-
ments Zland Z2 (Zl> Z2 for example), the growth curves for the cumulative discounted cash
flows over time show that the payback dates are such that:
(POT)1 > ( P W 2
Consequently, alternative (2) appears more profitable than (1). In fact, in the example
selected, the slope of curve (2) (which actually consists of successive straight line segments) is
on average less steep than the slope of curve (1). After n years the incomes for project (1) become
greater than those for the competing alternative (2): the previous conclusions are then reversed.
Such a situation is rare when two projects from the same industrial sector have to be com-
pared. However it is much more frequent when it is a matter of deciding between proposals
of widely differing types, for example concerning activities as sharply contrasted as refining
and nuclear energy. For this reason, the payback period criterion is most valuable for inves-
tigating variants of a single project, whose competitive situations are essentially equivalent,
the technologies compared often related, and the investments of the same general size.

Figure 2.3b Payback time with discount: comparison of two projects.

Special case. (Source: IFP)
106 2. The Elements of Economic Calculation

Generally speaking, if market forecasts were perfect (the “assured future” hypothesis) the
use of the payback period as an evaluation criterion would probably give incorrect results. On
the other hand this way of working can be very useful for shedding light on the investment
burden and its impact on incomes in a project’s later years. This favors projects whose pay-
back date is compatible with the risks that they may run. In other words, the payback period
is an effective method for assessing the hazards incurred, especially when competition is
intense; in this case there is every reason to evaluate the profitability by using this criterion,
particularly in its simplified form: the payout time.

D.Other Criteria Based on the Present Value Concept

As with the empirical methods, there are many other ways of calculating a project’s profit-
ability by using the concept of discounted income. They rely on more or less complex, even
artificial economic or accounting data; some of them, by emphasizing certain particular
aspects, are designed to correct the defects of the procedures described above, or to empha-
size certain characteristics. Their diversity shows that none of them is a universal solution.
However, they are all based on the same information, and its lack of precision sometimes
destroys the credibility of the very results that these approaches provide.
By way of example, two other methods will be outlined: relative capital enrichment
(RCE), and the economic production cost.

a. Relative Capital Enrichment (RCE)

This criterion is defined as the ratio of a project’s overall capitalized discounted cash flow to
its investment cost. If n is the presumed lifetime of the facilities in years, then using the above
notation we have:

(CF)P
/1 P
p=o (1 +i)
RCE = . ~ O(in
O %)
Z

The higher this rate, the more profitable is the situation. When two are compared, the one
with the higher RCE is considered to be the most attractive from an economic viewpoint.
Like the preceding ones, this method has its limits. It may prove to be inadequate, or even
faulty, when the objective is to determine the optimum dimensions of a facility. For example,
this is the case when a curve illustrating the variation of capitalized discounted cash flow
against investment has a sigmoid form (S-shaped curve, Fig. 2.4).The RCE is represented
graphically by the slope of the straight line joining the origin to the point on the curve that
represents the selected capital, i.e., the defined production capacity. It reaches a maximum
when this straight line coincides with the tangent drawn from the origin. However, this invest-
ment and consequently the size of the unit that it defines are smaller than the ones that would
be obtained from the optimum capitalized discounted cash flow, which corresponds to the top
of the curve.
 2. The Elements of Economic Calculation 107

investment
corresponding
to maximum
capitalized
discounted
cash flow
investments

Figure 2.4 Relative Capital Enrichment (RCE). (Source: ZFP)

b. The Economic Production Cost, or Economic Cost of Price

This concept arises from the desire to reconcile the attitude of accounting, which considers
past expenses and incomes on an annual basis, and the attitude of economics, which looks to
the future, usually a distant one, but which adopts the year as its unit of time.
In the most common case, where a project’s financing is provided from the company trea-
sury, i.e., when the annual financial costs are nil, the operating cost is written, using the
adopted nomenclature, as follows:
Cp = Dp + Ap (2.22)

With its two terms, one describing the operating expenses, the other capital costs, this is
simply an “accounting production cost” for the yearp just elapsed.
We can imagine an analogous calculation being made for a long period in the future, but
based on a reference year, i.e., employing the concept of present value. In these circumstances
we would speak of a “discounted” or “economic” annual production cost, to distinguish it
from its accounting equivalent. But its nature must be such that whether the total is obtained
by cumulatively entering the figure for each year of operation, over the period of the study,
108 2. The Elements of Economic Calculation

or by doing the same thing for its components, i.e., the actual stream of similar expenses
(operating costs and capital costs in particular), the result is the same.
The criteria considered so far involve the concept of discounted incomes, which assumes
the existence of revenues linked to the sale of products at market prices. Conversely, for each
particular value that may be assigned to them, there are annual returns that, when divided by
the actual quantity produced during the period, define an average unit sales price. We can call
this the “minimum profitable” price if the level of return selected is itself considered to be
minimal. In this case, calculating on a zero profitability basis signifies that the revenues are
equivalent to the operating expenses. Considered in terms of production units, these expenses
define an “economic unit production cost”, which is just the minimum sales price, as shown
previously, above which the discounted cash flow is positive, and below which it is negative.
As such it amounts to a new economic criterion or a special application of the use of cash
flows, which may prove very effective for a rough analysis.
Indeed, a simple comparison of prices quickly shows whether or not .projectis making a loss.
However, in this case it is often convenient to make a constant money comparison with a
sales price that is itself stable: this assumes, independent of any technological improvements,
that the performance or the productivity of the facilities, i.e., the annual or unit economic pro-
duction cost, is held constant. Here, two situations may arise.

The Stream of Operating Expenses is Assumed to be Invariant

Looking for a constant annual economic production cost amounts, under expression (2.22),
to defining a discounted depreciation (called “economic” to distinguish it from the corre-
sponding accounting concept) that is also fixed. If A is the value sought, it must be such that:
A A A

i.e.:

or:

Note that the annual decreasing depreciation provision thus applied is higher than the lin-
ear depreciation of the accounting equivalent (see Section 2.2.4.2.A), since it includes not
only the depreciation itself but a repayment of capital at a rate equal to the discount rate i.
In contrast to accounting depreciation whose duration follows the legislative provisions,
economic depreciation is normally calculated over the effective or presumed duration of the
facilities’ operations.
This way of working answers an earlier concern due to an apparent logical inconsistency
in assuming that the incomes, expenses, and depreciation provisions could all be held con-
stant at once.
 2. The Elements of Economic Calculation 109

Operating Expenses are Allowed to Vary Over Time

The annual economic production cost can be fixed only if the depreciation varies, because
it is ultimately calculated for each year by difference. The procedure is thus as follows:
First, the stream of expenses incurred over n years must be reduced to their equivalent in
year 0, i.e.:

p=l p=l p=l

Now:

so:

Next we have to find a constant annual equivalent for Co, i.e., a fixed annual economic cost
price P:
p = c o = c o i(l+i)”
P +
(1 i)” - 1

The economic depreciation for any yearp is found by calculating the difference: P - Dp.
As with the other criteria considered so far, for each annual concept there is a correspond-
ing way of working in production units, for example: per ton. This leads to the definition of
a “constant unit economic production cost”, also called the “average discounted cost”. Its
value is such that, if the entire production is sold at this price level, there is neither profit nor
loss. In other words, it is equivalent to a constant sales price for which the discounted income
is zero. This condition obtains if the total discounted revenues are equal to the total dis-
counted expenses. If Q represents the quantities produced in yearp, and P the constant price
at which they are sold, then:
p= n p=n
VPepJ cP
p=l
(1 + i ) p

or: (2.23)

The unit economic production cost is therefore equal to the ratio of the total discounted
expenses to the total discounted quantities produced.
 Next Page

I10 2. The Elements of Economic Calculation

COMMENTS
By making Cp = Dp + Ap, we obtain:
p=n

p=l
P= (2.24)
p=n

0 If Qp = Cte = Q, and if Cp = Cte = C , the expression (2.23) becomes: P = C/Q.

P is simply the annual cost of price per manufactured unit (see Section 2.2.1.2).
In the form (2.24) it can be written:
I
-+D
p = -P (2.25)
Q
(1 + i)" - 1
where: P=
i(l+i)"

0 In this simplified form (2.25) the economic production cost is equivalent to a min-

imum pre-tax profitable sales price, like that defined earlier when discussing the
internal rate of return. The equation obtained then:

V'=[
I' + f ' ( 1-a)- A'J 1+ c'
1-a

can in fact be written (as long asf' is assumed to be zero or included in I t , and a = 0):

I' - A' + c'

V' = -
P
knowing that: C' = A ' + D '

I' + D'
V' = -
or:
P
I
-+D
so: p = -P
Q
since, by definition, V', I' and D' represent unit quantities.
Previous Page

2. The Elements of Economic Calculation 111

2.1.2.4 The Practical Significance of the Cash Flow Concept

As illustrated in the Fig. 2.5 flowchart, the sales figures generated by marketing, in other
words the revenues V, enable the financing of raw material and various other purchases, and
also the continued operation of the facilities, via other operating costs: wages, social benefits,
management costs (maintenance, taxes, insurance, overhead and head office costs, etc.). The
difference between the revenues V and the sum of all the outlays D constitutes: the gross
profit cashflow.

T I ,
Growth
Gross profit Net available
- cash flow -c
cash flow cash flow
(undistributedprofits, etc.)
A A
Direct operating Capital repayment
Financial
+ costs - costs
- + cash flow
(wages, costs of benefits. etc.) (distributed profits. etc.)

Employment Cash flow

available
Sales of funds
for financing

Figure 2.5 The various cash flows. (Source: IFP)

The additional withdrawals for capital expenses such as financial costs and taxes on profits
lead to: the net available cashflow.
These flows can then be divided into three main categories:
Maintenance cashflow: this is for possible renewal of existing plant, so as to maintain
productivity, and to make the corresponding provisions (allocation for depreciation,A).
Growth cashflow: part of the net profits, distinct from the allocation for depreciation, is
invested in new projects.
Capital remuneration cashflow: another portion of the net profits is distributed to share-
holders as dividends.
Only the first two are likely to bring about investments: together they represent thefinanc-
ing capacity cash flow, corresponding to the deployment of concrete means of production,
both to hold on to existing market shares and to win new ones by the marketing of innovative
112 2. The Elements of Economic Calculation

products, and also by creating new jobs. After all is said and done they are the source of
potential sales growth and, consequently, of possible profits.
Conversely, the dividends paid out to shareholders appear unproductive when looked at in
this way. This may enable a better understanding of the “psychological” character of the discount
rate selected by a company or a state. In fact, it appears to be the reflection of the desire either to
repay the shareholders(high rate), or to enable investment and risk-taking (low rate). So, it is an
indirect measure of the distribution of the net profits and the dynamism of the company.

2.2 OPERATING COST AND COST OF PRICE

In the previous section, these two concepts were frequently evoked in the nomenclature
employed. On the one hand, they are linked by the unit capacity, on the other the operating
cost ( C ) appeared as the resultant of two principal terms: the operating expenses (0) and the
capital charges (A and F).
Here, we need to explain them, to provide more detailed definitions, and to see what they
cover in practice.
Actually, evaluating a project’s profitability consists of drawing up a balance sheet of rev-
enue versus expenses for a production facility. The revenue side, arising from the sale of man-
ufactured products or from services provided, can be estimated from forecasts based on
market studies, which give both the probable volume of transactions and their expected price
levels. However, the expense side still has to be grasped, and its overall scope expressed in
terms of operating cost or cost of price.

2.2.1 COST ACCOUNTING METHODS

2.2.1.1 Stream Factor and Load Rates

Depending on the person for whom the evaluation is intended, or the circumstances of its
preparation, the calculations may be conducted:
0 In annual costs: the economists’ and accountants’ approach; for their purposes they draw

on information from earlier studies, supposedly available,

0 In costs per ton of product(s) or feedstock, or production unit: the sellers’ or, more

broadly, the marketing department’s point of view,

0 In hourly costs: the point of view of the manufacturing personnel, but also of consulting

firms and, more generally, of the engineering companies responsible for sizing the facil-
ities.
We will retain the first two of these three approaches, which respectively address the con-
cepts of “operating cost” and “cost of price”. The question then becomes one of how best to
manage the passage from one to the other: they are roughly subsumed in the unit production
capacity, but in reality are more complex.
In this context, in fact, there are several factors to consider.
 2. The Elements of Economic Calculation 113

A. TheoreticalStream Factor
This corresponds to the duration of operation of the production facilities over the year, with
allowance for anticipated routine maintenance shutdowns. In this regard, we may distinguish
between sites that run continuously, 24 hours a day, including Sundays and holidays, and
ones that operate semi-continuouslyor discontinuously, whose schedules can be very diverse
(five or seven days per week, one, two, or three eight-hour shifts, etc.). For the sake of con-
venience in this discussion, we have limited it to plants whose size is sufficient to justify con-
tinuous operation. In this case, the stream factor may be expressed in various ways: for
example, 8 000 hours or 340 daydyear, instead of 8 760 hours or 365 days maximum for a
regular year. This is also why it is sometimes formulated in terms of a ratio or a percentage,
such as 0.9 or 90% (ratio of the operating time or stream days to the length of the year or cal-
endar days). The figure of 8 000 dayslyear is commonly employed: although it has the disad-
vantage of not yielding a whole number of days per year (333.33 dayslyear), it is very
convenient for consulting firms. For them, in fact, the basic unit of time is the hour. It is used
as much for material and energy balances as for the sizing parameters representing equip-
ment, and thus their costs. This approach simply aims to base the conclusions of market stud-
ies on a theoretical nominal capacity, allowing for a reasonable number of hours or days of
downtime over the year to permit maintenance and periodic inspections of equipment, possi-
ble overhauls, replacement of catalysts, adsorbents, solvents, etc. It reflects optimal usage of
the equipment, at its best possible production capacity.

8. Effective Stream Factor

This allows for unprogrammed stoppages, due to incidents or accidents, strikes, breakdowns
in supplies, stock shortages, etc. On the other hand, it ignores operation at reduced rates and
thus merely expresses the actual annual operating time, for example: 6000 hourslyear,
3 10 dayslyear.
Like the previous one, this indicator still has a slightly theoretical or artificial character, in
that it says nothing about the facilities’ actual production and its possible fluctuations during
the period.

C. Utilization and Load Rates

The ratio between the effective and theoretical stream factors defines a technical rate of uti-
lization, i.e., for example 6000 hourslyear: 8000 hourslyear = 0.75 or 75%, which expresses
the real level of activity at the site, and thus directly concerns the manufacturing personnel.
From an economic or accounting viewpoint, it is not so much this aspect that counts, as the
potential to produce, and thus to sell, during this same period. Now, in practice, a given facil-
ity is very unlikely to operate at its full potential over long periods. First of all, at startup it
does not immediately begin to function at full capacity: it only very gradually attains its opti-
mal rate. In addition, it has to adapt to market hazards and often the outlets are smaller than
the potential production.
The monitoring of economic changes thus leads to the definition of an instantaneous load
rate (zi),which, for short periods, is simply the ratio between the observed hourly production
114 2. The Elements of Economic Calculation

(qi)and the corresponding sized hourly capacity. Expressed in annual terms, such an indicator
can be used in an equation like the following one:
Q (real annual production) =

C (theoretical annual production capacity) x (utilization rate) x

c
t=lyear

t=O
CiTi

c
t =I year

t=O
ci

t=lyear
where: C = effective stream factor
t=O
t=lyear

and:
ct=O
t=lyear
CiTi

= average load rate for the year (2)

t=O
This approach assumes a knowledge of the results from operation of the facility under con-
sideration for a period of a year, and thus reflects an accounting viewpoint. In evaluation
work, where the studies naturally have a very pronounced predictive character, it is not pos-
sible to have access to such detailed information. Consequently, in these cases it is customary
to adopt an effective stream factor whose value is equal to the theoretical stream factor (pref-
erably 8 000 hourdyear), and a load rate that is equally arbitrary-although based on the fore-
cast of market growth-but assumed to be constant for relatively long periods: a year at least.
In this case, the concept of utilization rate vanishes and is merged with that of the load rate,
giving only one coefficient, comparable to an average value. It represents the ratio between
the forecast production and the calculated capacity, and may equally well be called the load
rate or the utilization rate.

2.2.1.2 Breakdown of the Operating Cost and Cost of Price

Although companies may differ in the ways in which they allocate the expenses of operating
and managing industrial facilities, in practice it is always possible to make the necessary
adjustments to place them in the usual three major categories, as follows (Fig. 2.6):
0 Variable charges or variable costs

Labor
Fixed charges or fixed costs.
The “variable charges” are so called because they are directly proportional to the actual
production of the existing facilities, not to their nominal manufacturing capacities. These
expenses reflect the results of the material and energy balance sheets that go with the opera-
tions concerned, i.e., with the yields.
 ~ ~~ ~~

2. The Elements of Economic Calculation 115

Figure 2.6 Composition of the operating cost. (Source: IFP)

Expressed in annual sums (euros per year), as recommended when utilizing operating cost
and accounting rules, they obviously reflect the demand, and track its variation.
In the commonest application, however, they are related to the unit quantity physically
manufactured, especially to the ton of raw material processed, or of product. In this case, as
long as the effect of scale does not lead to changes in the balance sheet, there is no difference
in energy or raw material content from one ton of product to another. In other words, the so-
called variable costs, when reckoned by the year, i.e., in terms of operating cost, become con-
stant costs if they are expressed as cost of price, In this case, the French often use the expres-
sion “mise au mille” (per thousand) to indicate that the costs, or the elements used in
determining them, are in per ton figures, i.e., per thousand kilograms of product.
The “fixed charges” are the expression of the operating costs that follow directly from the
costs of investment (see Section 2.3). They are so called because they are drectly tied to the
116 2. The Elements of Economic Calculation

processing capacity of the facilities, and are consequently defined the moment this parameter
is known. In particular, they do not depend on the quantity actually produced. For example,
if there is a fortuitous but lengthy shutdown of the facilities, the fixed charges do not stop, but
continue to appear on the financial balance sheet. In these circumstances:
0 In terms of operating cost, i.e., annual expenses, they appear quite constant,

0 In terms of cost of price, i.e., per unit quantity manufactured, they become dependent on

the actual annual production, and therefore variable. They rise as the load rate falls, and
vice versa.
“Labor”, particularly the operating personnel, gives rise to costs that are difficult to assign
to either of the two previous categories of expense. The maintenance, general service, admin-
istrative, and marketing personnel, who do not do shift work, are accounted for under various
entries covered by the fixed charges. The others, however, who are obliged to work in shifts
to ensure the continuous operation and supervision of the facilities, must be accounted for
separately. Their numbers increase with both the capacity and the actual production. They do
not grow proportionally, but in discrete amounts; they generate costs that are roughly constant
over the year, continuing during short shutdowns, but they can fall off quickly if the situation
lasts a little longer, since the onsite personnel cannot remain idle too long: they will be trans-
ferred temporarily to other activities, and charged to other projects.

2.2.2 VARIABLE O R PROPORTIONAL CHARGES

These consist essentially of two items:

0 Raw materials and other consumables

0 Utilities.

To these must be added related costs, often grouped under the label “miscellaneous”: they
are generally of secondary importance, even negligible. To determine the amount of the costs
generated by the first two items we must return to the concept of the quantity-price relation-
ship, discussed in Chapter 1. Both of the constituent parameters must be defined and ana-
lyzed, because the ultimate objective is to calculate their product.

2.2.2.1 Raw Materials, Reagents, and Other Consumables

At this level, the question takes on a three-fold aspect: not just tonnage and price, but also
definition.

A. Nature and Characteristics of Consumable Products

There is a need to distinguish among various categories of consumable substances employed
in a conversion process whose objective is to obtain one or more products that are specifically
wanted, and thus directly marketable, but which may also lead to unwanted products that nev-
ertheless have to be made use of, if possible. Because of this, the economic inventory con-
cerns consumption as much as it concerns credits tied to such uses.
 2. The Elements of Economic Calculation 117

With this in mind, the main headings to consider are the following:
0 Raw materials

Co-products
0 By-products

0 Reagents or chemical products

0 Heterogeneous phase catalysts, molecular sieves, alumina, contact masses, etc., and

more generally any kind of solid or liquid packing material

0 Homogeneous phase catalysts, various additives, solvents, dehydration and discoloring

agents, etc., and overall, any liquid additive in the processing stream
Cartridges, filters, electrodes, lamps, etc, and other items of equipment that are fairly
quickly worn out and frequently replaced, and which are directly related to the operation
of the facilities.

a. Raw Materials
The expression “raw materials” describes the essential starting items that go towards defining
the operation’s weight material balance. They are an integral part of the products and co-
products that result, and consequently they have the same nature, i.e., organic or inorganic
depending on whether they belong to one or the other of these two classes. In hybrid or com-
posite materials the two categories are represented in proportion to their contributions. Their
amounts provide a measure of the order of magnitude of the substances to be produced, and
they must conform to strict commercial specifications.

b. Co-Products
The term “co-product” covers all the products that are jointly produced from the same man-
ufacturing line, as long as they come from the same raw materials and can be directly sold in
their respective markets, i.e., as long as they have properties or specifications at least equiv-
alent to their commercial counterparts. They must also be able to take advantage of similar
economic conditions, in particular the same effects of scale, which means that each of the
substances thus manufactured must be available in volumes of similar size.
From this standpoint, steam cracking of naphtha, which simultaneouslyproduces ethylene,
propylene, isobutene, butadiene, and benzene, is a type example in which tonnage is the unit
for accounting purposes. In such a case, it may be difficult to draw up an economic balance
sheet in terms of cost price, i.e., per ton of products. There are two possible approaches:
Choose one of the substances produced as a reference (e.g., ethylene). It constitutes a
“key product”. The other co-products, expressed in quantities per ton of key product, are
entered as credits.
0 Define a notional ton of product, whose composition reflects the respective proportions

of the various co-products manufactured. The calculations are then conducted by refer-
ence to an “average ton” of this notional product.
The production of composite molded objects of various shapes (bottles, toys, etc.) provides
another example, in which the accounting unit is the piece. It requires the simultaneous use
of several thermoplastic materials. In this case, a balance sheet based on cost of price can be
drawn up only for a particular type of object, with the others appearing as credits.
118 2. The Elements of Economic Calculation

c. By-Products

“By-products” have two different origins. They arise from secondary, undesirable transfor-
mations, linked either to the conversion of the raw materials itself, or to the use of reagents
or accompanying chemical products, even solvents, additives, or other additions. For this rea-
son they may or may not appear in the operation’s weight material balance, depending on
whether they come wholly or partly from the initial feedstock, or, on the other hand, entirely
from the use of reagents. As with co-products, their nature is strongly dependent on that of
their parent materials, but generally speaking they are much more numerous, and so their
compositions are even more varied. Their two principal characteristics are, first, that they are
produced in small amounts, the smaller the better, and, second, that most of the time they do
not meet commercial specifications. In the economic balance sheet, they appear as credits or
debits according to whether they can actually be sold, or whether, in order to get rid of them,
specific additional transformations must be carried out or a provision made for disposal costs.
However, even if it is possible to sell them, they have two considerable drawbacks: on the
one hand their presence in regular products is heavily penalized; on the other, with few excep-
tions, bringing them up to specifications demands additional processing, which is charged
directly to them. The bottom line is that even in the best circumstances the potential market
value of by-products has to be marked down by an amount that represents their cost as prod-
ucts in their own right. Unless otherwise indicated, this minus-value is roughly of the order
of 50%.

d. Reagents

In theory, “reagents and chemical products” do not appear as major items in the material bal-
ance. They have an accessory role, to the extent that they are designed to facilitate the oper-
ations of certain units, or to supplement them by eliminating by-products or surplus reagents
and catalysts, etc. They are often different in kind from the raw materials and co-products,
e.g., inorganic when the conversion involves organic substances. They must meet strict com-
mercial specifications, more exacting than the ones required for manufactured products or
their direct precursors, so as not to contaminate the main transformation. However, they are
generally used in smaller quantities. For this reason, since they do not benefit from economies
of scale, they also do not become more affordable in better economic times.

e. Heterogeneous Phase Catalysts and Various Packing Materials

Their essential characteristic is that they are involved right from the start, in significant
amounts, and that the full complement must be in the facility from the moment it begins pro-
duction. This is true for all the solid contact masses such as catalysts, molecular sieves, alu-
mina, activated charcoal, diatomaceous earths, etc., and also for the quantities of liquids,
especially solvents used in extraction or absorption, various oils, coolants, mineral salts, and
liquid metals (e.g., mercury amalgams). The corresponding expenses, called first load or ini-
tial load expenses, which must be incurred even before the equipment starts operating, are
treated as investments (see Section 2.3). However, as time goes by the reserves of these prod-
ucts are used up, so that on one hand they have to be regenerated, whether continuously or
not, and on the other partly or wholly replaced, so as to make up the losses that are an inevi-
 2. The Elements of Economic Calculation 119

table consequence of processing. The bottom line is that there is a consumption that must be
treated as a variable charge, since it rises in proportion to the production.
However, each situation constitutes a special case deserving detailed analysis, so as to
make clear what has actually been spent. In this regard, it would seem more useful to study
some specific examples. One of the most complete is for catalysts operating in the heteroge-
neous phase, and is therefore worth spending a little time on.
There are various kinds of solid catalysts: in particular, they are described as mass or sup-
ported, depending on whether the active phases are mixed with binding agents before being
formed, or are deposited on a pre-formed carrier. In both cases, the result is a porous structure
onto which compounds are adsorbed, under controlled pressure and temperature conditions,
before being activated, converted, and desorbed. Certain parasitic transformations lead to
products that are hard to strip, and which, when heated, eventually form gums or carbon
deposits. This leads to a partial blocking of the pores and a masking of the active components.
It is accompanied by a similar phenomenon that is due to attrition, i.e., mechanical wear from
liquid and gaseous reagents circulating through the catalytic beds. The resulting fines seal off
the pores little by little, and lead to an overall reduction in activity. Another cause of deacti-
vation, especially for mass catalysts, may arise from the volatilization of certain active
phases: the temperature is far from uniform throughout the reactive mass, and local hot spots
may affect the stability of the existing systems.
This progressive aging of the catalysts over time can be monitored in various ways, in par-
ticular by measuring the loss of charge from the bed, which increases steadily, or the rise in
temperature that is needed to maintain their performance. This latter parameter has moreover
proven to be a particularly useful tool for control. In fact, every catalytic reaction has a par-
ticular initiation temperature, i.e., a heat level sufficient to obtain significant production.
Deactivation leads to progressively raising this temperature, until it reaches a level above
which the danger of destroying the active phases becomes too great. At this point, it is con-
sidered necessary to halt operations and restore the catalyst’s properties: we then say that we
have completed a run. Regeneration is usually performed in situ by careful combustion of the
carbon and gum deposits, either by using steam, or by controlled burning in air with reduced
oxygen. In some cases the catalyst may also be sieved, to eliminate fines, and made up to vol-
ume.

COMMENT
This operation is sometimes done offsite, by specialized companies with their own,
specially equipped facilities. To avoid a prolonged shutdown the system is immedi-
ately replaced by a preactivated one. This approach, which consists of keeping inter-
ruptions to a minimum by using two identical loads, is being more and more widely
adopted: it also frees operators from the technical and economic constraints involved
in catalyst management, by using service companies who thus act as rental agents.

In fact, after such reactivation treatments it is very uncommon to get back to the original
operating conditions: the initiation temperature is a little higher, such that the second run is
shorter than the first. Since the length of each succeeding run diminishes with each regener-
ation, there comes a time when it is preferable to stop the procedure and completely replace
the catalyst, which is said to have reached the end of its life. We then have to purchase a new
120 2. The Elements of Economic Calculation

load, and it is as if the first one had been completely consumed during the preceding period,
in two years for example. The corresponding expense, expressed in terms of cost of price,
consists of dividing the cost of this load of catalyst by the actual production of the facility
during this lifetime.
When the active phases consist of precious metals such as platinum and iridium, or semi-
precious ones such as cobalt and nickel, the calculation becomes a little more complicated. In
fact, it is possible to re-process used catalysts to recover these elements, and then recycle
them.
The resulting savings are far from negligible, considering the prices of these metals and
the recovery rates, which may be as high as 99% in the case of platinum. Under these circum-
stances, when considering production expenses during the catalyst’s period of utilization it is
appropriate to charge only for the cost of the support, the impregnation, and lastly the supple-
ment of active phase needed to return it to its original operating condition.
The other kinds of solid contact masses, such as molecular sieves, alumina, activated char-
coal, etc., behave like catalysts from an economic viewpoint, at least in a general way. How-
ever, they typically create expenses that are low enough to be ignored, or included with a
larger group of miscellaneous consumables, depending on the level of accuracy required and
the information available.

f. Homogeneous Phase Catalysts and Various liquid Additives

Generally speaking, their economic behavior is very similar to that of reagents. However,
they may give rise to additional costs to the extent that their utilization leads to pollution
related to the disposal of wastes. Therefore, it is very important to identify these possible
drawbacks early on, and then to determine whether the wastewater treatment facilities that
are traditionally found on industrial sites, or that could be set up for the project being consid-
ered, are capable of handling the required purifications. If not, the price of the additive will
have to be raised by the corresponding cost, to the extent that this can be estimated-either
directly, by companies specializing in this area, or by using a procedure analogous to the one
described in this book, with a few adjustmentsrelated, in particular, to the size of the tonnages
involved.
As for the regeneration of liquids, it is done on a continuous basis or not at all. In either
case, the make-up material needed to compensate for the operating losses is usually expressed
in terms of the unit quantity of manufactured product. In other words, the resulting expenses
resemble those described for reagents.

g. Consumable Equipment Items

These are items that wear out fairly quickly, i.e., their average life is short when compared to
the anticipated life of most of the basic components of the facility. In these circumstances,
depending on the rate of replacement for these items and their unit prices, it is possible to find
an annual cost, which may be calculated per ton of manufactured product or per unit quantity
produced in the plant. Here again, certain specific problems may arise, whose consequence
is to generate additional costs that must be taken into account in the purchase price adopted,
or included in the broader category of general services. They particularly involve questions
 2. The Elements of Economic Calculation 121

of availability, shipping delays, ease of application, etc., which may for example impact on
various investment items such as the stock of spare parts, working capital, etc., or lead to spe-
cific labor or maintenance costs, etc.

B. Unit Prices of Consumable Products

a. General Situation
The mechanisms of price formation have already been specifically examined in Chapter I.
However, the matter was mainly considered from a theoretical standpoint and with a view to
the future, since the objective was to use information and observations from the past (histor-
ical data) to predict variations over the next three, five, or even ten years-where every
project must play out. This kind of analysis is risky by its very nature, even though it may be
possible to limit the uncertainty to some extent, as much by using deflation or updating, as by
discounting, which definitively enables the calculation to be performed in constant money,
particularly money of the current year: generally speaking, this comes down to assigning pri-
ority to the present.
This theoretical study showed that a product’s price may be broken down into two compo-
nents: a floor price and a margin. The first emerges from a profitability calculation, and cor-
responds, with a few modifications, to the concept of a minimum profitable sales price,
discussed on several occasions in Section 2.1. The second component, which depends mainly
on external factors, especially the economic situation, is mainly regarded as being the result
of fluctuations in supply and demand.
The reality is actually much more complex than this rather schematic version, and many
parameters can be invoked to explain (more or less) the very delicate mechanisms that govern
price formation. The procedure suggested here presents some difficulties of its own when it
comes to application. For example, to calculate the floor price we need to introduce the costs
of raw materials, chemical reagents, solvents, and so on. In the interests of uniformity and
consistency it might appear reasonable to adopt prices for these various inputs that are also
determined according to the same principle. But such an attitude is not only unrealistic, as
regards the scale of the calculations that it would involve or the information that it would
require, but also erroneous, since the announced objective is to determine the thresholds cor-
responding to average economic conditions of the day. We cannot expect to have every single
factor on our side at the same time. Moreover, to add a margin that depends on the balancing
effect of the markets alone, no matter what the input or its weight in the final result, appears
just as arbitrary.
After all, although the procedures described in Chapter 1 may be fully appropriate for man-
ufactured products and their raw materials, particularly for identifying leading prices and
their underlying growth trends, they cannot be applied to everything.
Nevertheless, when all is said and done the determination of a product’s price does require
consideration of the prices of its ingredients, and of the products that are essential for its man-
ufacture. In order to do this, we can only base ourselves on the current or previous prices for
such items.
Unfortunately, the diversity of the situations that are encountered, and the desire either to
offset monopolistic attitudes, or to avoid amplifying erratic fluctuations, have led to the
122 2. The Elements of Economic Calculation

development of regulatory methods for each case, and these very actions have created a mul-
tiplicity of reference price systems.

b. The Main Kinds of Reference Price

The arrangements and the methods of application covered by these reference systems, whose
number is itself a reflection of their shortcomings, are not applicable to every product. How-
ever, as a general rule, as we go back up the manufacturing chain towards materials of an
energy-related or strategic nature, such as petroleum, or primary products such as the major
chemical intermediates, the more complicated the situation becomes and the more numerous
are the references utilized. In addition, although the wording or the formulation of some of
them may seem to indicate a very broad field of application, in practice each sector of activity
employs somewhat different ones. Roughly speaking, the main types of reference price in use
are the following.

Posted Prices or List Prices

This concept was introduced in the United States at the end of the 19th century, among
petroleum producers, of whom there were so many that Standard Oil, the principal refiner and
transporter of the day, decided to “post” the prices at which the company was prepared to buy
crude at the wellhead. These prices were published regularly, and served as a reference for
transactions. The idea was later picked up and extended to all producers. Generally speaking,
these are predetermined prices, whose main purpose is to define a framework for discussion,
but also to enable the assessment of one product’s qualities by comparison with another’s:
this product, here called a “marker”, should be widely available, and have well-known char-
acteristics, so that a relative premium or discount could be deduced. They are never used as
contract prices.

Official Selling Prices or Producer Prices

This concept has progressively replaced the preceding one, and expresses the reaction of
supply to demand. It accompanied the attainment of independence and industrialization of
developing countries that produced the raw materials for energy, minerals, agriculture, etc.
However, the initial concern is unchanged: to define overall rules for commercial transactions
that would assist in the negotiation of contracts.

Spot Prices
These are prices connected to day-by-day transactions, as they occur, usually concerning
limited quantities and referring to specific locations where the shipment is assumed to be
available. For example, for petroleum products the most important spot markets are Rotter-
dam, also called ARA (Amsterdam, Rotterdam, Antwerp), for Northwest Europe, Genoa for
the Mediterranean, Caribbean or Gulf Coast for America, Singapore for Southeast Asia, and
so on. The best-known locations for grains are particularly Chicago, London, New York,
Paris, etc.; for metals, New York, London, Paris, etc.
These are also called ‘‘free” or “cash” markets: they effectively constitute exchanges
whose quotations are regularly published in the more specialized periodicals: dailies such as
Les kchos, the BIP, etc., weeklies such as the European Chemical News, Chemical Marketing
 2. The Elements of Economic Calculation 123

Reporter, L ’Usine nouvelle, Chimie Actualitis, etc., monthlies such as Informations chimie,
Petroleum Press Service, MatiBres plastiques, etc.
Certain organizations or associations such as Reuters, ICIS-LOR, Platt’s, Petroleum
Argus, Telesate, London Oil Report, Petroflash, etc. make a specialty of this kind of informa-
tion, which they collect, process, and publish, either on a daily basis or on request, via banks,
servers, and data transmission networks.
Spot prices are the product of transactions conducted either by independent traders 3, or by
the trading subsidiaries of producer or user companies. These private companies, which are
independent of government, are merchants responsible for buying and selling on the spot
markets real or notional shipments that they neither utilize nor distribute. They have no indus-
trial restrictions, but face strong constraints on the financial level. Traders must be distin-
guished from brokers, who are paid on commission and who have no negotiating power,
being charged simply with bringing their co-contractor and a customer together.
A single shipment may change hands several times (two or three times on average) before
being used, which gives this type of exchange a manifestly speculative character, even
though, when all is said and done, this aspect appears to be the necessary counterpart of cov-
ering price-risk.
Generally speaking, the spot price is a good indicator of the supply and demand situation.
However, in theory it only takes into account exchanges representing a very small fraction of
actual business or industrial activity: it does not reflect either mutually agreed swaps or cap-
tive uses. In practice, its influence is far greater, particularly during periods of over-produc-
tion, and to the extent that the liberalization of trade, combined with growing industrialization
in the developing countries, is drawing more and more actors into the free markets and aug-
menting the number and volume of transactions.

Contract Price
Mutual agreement transactions for the great volume of raw materials and products take
place in the context of contracts, arranged to cover all or part of the needs for a given period.
These agreements are likely to cover several years, or to be “evergreen”, i.e., renewable for
an indefinite duration, but liable to termination without notice by one or other of the co-con-
tractors. The aim of companies, depending on their purchasing policy, their level of integra-
tion, and their market access logistics, etc., is to assure their regular or principal supplies or
shipments as far as possible; the remainder to be covered by day-to-day trades on the spot
market. Such an approach is made even more necessary by the increasing trend among man-
ufacturers to limit the size of their stocks for economic and safety reasons, applying just-in-
time operation.

3. As a rough guide, some traders are multi-product (petroleum, metals, grains, etc.) and have
major financial scope and worldwide influence, such as Phibro (short for Philips Brothers), Marc Rich,
Tradax, etc. Others prefer to specialize in certain markets and rely on their long history or their reputa-
tion for reliability or aggressiveness, etc. Examples include, for petroleum products: Coastal States,
Intemorth, Mobanoft, Scantrade, Stinnes, Tampinex, Vitol, etc.; for the major petrochemical inteme-
diates: BMS (Belgium Marketing Services), Helm, Lucky Gold Star, MG Metal (Metallgesellschaft),
Tolson, Trammochem, etc.; for agricultural products: ADM, Bunge and Bom, Cargill, Dreyfus,
Fermzzi, etc.
124 2. The Elements of Economic Calculation

The contracts involve sometimes-lengthy negotiations between the parties, leading to a

document that defines quantities, duration, pricing terms and revisions, methods of delivery,
terms of payment, etc., and includes general and specific legal language. “Local quotations”
(for different places) are also provided for. A summary document is often drafted (to be faxed,
telexed, etc.) if the matter is urgent.
Contract prices, when they are negotiated on a monthly or quarterly basis, usually refer to
spot prices, so as not to penalize one of the contracting parties with respect to the other, as a
result of the frequent and substantial fluctuations caused by temporary imbalances between
supply and demand on the free market. For example, a monthly basis might be expressed as
follows:
Price for month M = average of the Platt’s prices for the month A4 or the month M- 1
The reference for certain transactions may also be given as the average of the n daily quo-
tations preceding the loading or delivery date. The prices of products that are sold by the bag
or in bulk in a very competitive market (plastics for example) are often re-negotiated, some-
times on each occasion.
In addition to the spot prices published in the specialty press, contracts also take account
of previous agreements and the basis adopted for transactions, which are sometimes pub-
lished. They also rely on consultants’ studies on the national and international competition,
or customs statistics that provide a picture of the traffic (source and destination) and prices.

COMMENTS ON SPOT PRICE-CONTRACT PRICE RELATIONSHIPS

0 Comparisons between contract and spot prices sometimes show quite large dis-

crepancies, whose size is actually a reflection of the magnitude of the momentary

imbalance seen to exist between supply and demand, which, moreover, is often exac-
erbated by exchange rate fluctuations.
0 These differences may partly be explained by the small quantities exchanged or by

the limited number of transactions on the spot market, when the unit quantities for
contracts are much more important. In this case, the two reference prices are not
directly comparable, since the exchanges do not concern related tonnages that are
likely to have similar coefficients for economy of scale.
In addition, spot prices are very “volatile”. An accidental stoppage in a major pro-
duction unit, in a just-in-time market, can lead to an explosion in prices that will be
reflected in the contract price, as long as it lasts for a certain period.
However, generally speaking, contract prices ultimately have greater stability,
usually being based on the spot prices for the preceding period, with the extreme
variations suppressed. By way of example, Fig. 2.7 illustrates this feature in the case
of ethylene. Another evident effect is a certain time-lag (of the order of three months)
between the maxima and minima, which clearly brings out the indicator or precursor
role of the spot market.
0 The US.dollar is very often utilized in inter-continental exchanges, for both spot

and contract prices. Within Western Europe they are generally drawn up in the cur-
rency of the price-setting country, i.e., the one that is the net importer of the product,
for example Germany for propylene, France for butadiene, etc.
 2. The Elements of Economic Calculation 125

0 For captive products such as ethylbenzene, almost all of which is locally converted
into styrene, or for cyclohexanone, transformed successively into oxime, caprolac-
tam, and finally a nylon salt before being marketed, the spot prices, once published,
are generally disproportionate to what one might expect from the corresponding or
equivalent contract prices (average transaction or internal transfer price). This dis-
parity actually expresses a one-time, exceptional customer need: to fill it, the pro-
ducer had to make temporary or special arrangements that interfered with the normal
operation of its facilities and generated additional costs, which then had to be
charged entirely to the marginal quantities extracted. Such quotations are in no way
representative of the state of a product’s market or its regular commercial value.

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001

Figure 2.7 Comparative variations of the spot and contract prices for
ethylene ($/t: Northwest Europe). (Sources: IFe ECN, Plattk, etc.)

Forward Pricing
Forward markets are above all financial markets, also called “paper” or notional markets,
where contracts are exchanged: these are undertakings to buy or sell a certain quantity of a
product, having a particular quality and deliverable at a particular place. To simplify and
rationalize matters, these various criteria are standardized, given that in such a context phys-
ical deliveries are always numerically negligible in comparison with the volume of transac-
tions that they create.
The appearance of forward markets is a consequence of the very great price disruptions
that characterized the last three decades of the 20th century. Indeed, in stable times there is
no need to guarantee oneself against price fluctuations; when instability does appear, major
companies are relatively undisturbed, in that they are able to spread the risks over the various
126 2. The Elements of Economic Calculation

items in their product range and still maintain their financial cover. It is different for smaller
companies, which are financially more vulnerable and whose activities are more concen-
trated.
The principle behind forward markets consists of limiting losses (and possible gains) via
a systematic cover provided by notional commercial exchanges, coming from a succession of
sales followed by purchases-of paper, not product. It involves making contractual commit-
ments that ultimately result in the payment (or recovery) of the simple difference between the
anticipated price at a given date (the forward price) and the actual spot price on that date (the
cash price), when the physical transaction actually takes place, or on the assumption that it
takes place at that time. In the terminology used by brokers, if the difference between the
actual cash price and the forward price is positive (difference > 0), we have a backwardation;
if on the other hand the difference is negative we have a contango (from continue and go).
Differences between two “forward” positions (three and six months for example) are handled
in the same way. Note that, although cash prices may vary dramatically, forward prices do
not: they fluctuate in a much more gradual manner because they are, in a way, the resultant
of the underlying market trends perceived by the professionals, with possible allowances for
economic events (Fig. 2.8).
The principal existing forward markets are, in the petroleum area: the Nymex (New York
Mercantile Exchange), the IPE (International Petroleum Exchange) in London, and also the
exchanges in Singapore, Chicago, Rotterdam, etc.; for grains and oleaginates: Chicago, etc.;
for sugar: London, New York, Paris, etc.; for cocoa and coffee: London, New York, etc.; for
potatoes: Lille, etc.; for rubber: Kuala Lumpur, etc.

11/98 12/98 01/99 02/99 03/99 04/99 05/99 06/99 07/99 08/99 09/99 10/99 11/99 12/99
27
25
23
21
19
17
15
13
11
9

-1.5.-’
-2.5
Contango

11/98 12/98 01/99 02/99 03/99 04/99 05/99 06/99 07/99 08/99 09/99 10/99 11/99 12/99

Figure 2.8 Variation in Brent prices over twelve months ($/bbl).

(Source: Platt 5)
 2. The Elements of Economic Calculation 127

Calculated Prices
In a market exposed to economic hazards and major fluctuations, both buyers and sellers
have felt the need for more stable and realistic reference systems, capable of taking into
account the expenses involved in manufacturing and marketing a product.
In the petroleum field, the best-known mechanism of this type is the "netback". This arises
from a theoretical calculation that allows a refiner to determine the maximum price for which
it can buy a barrel of crude on the spot market, at the shipping port (FOB). The procedure
relies on knowing four components:
0 The real prices of products sold on the market, in the various geographic regions,

0 The yields for these various products, resulting from the refinery equipment available

and the quality of the crude being processed,

0 The marginal cost of refining,

The spot price for ocean freight.

Combining the first two allows us to obtain the cost of an equivalent ton of petroleum,
from which the latter two are subtracted to give the corresponding value for the raw material:
the price sufficient to enable the marketing of products under these actual conditions.
In increasingly volatile markets, the netback has become one of the most widely used pro-
cedures for calculating the price of crude.
Equivalent systems exist in other sectors of activity. For example, we have:
0 Net raw material cost (NRMC),

0 Marginal cost, in the case of linked production,

0 Internal transfer price, particularly for captive products, with or without inclusion of a

margin,
Average transaction prices, comparable to actual minimum profitable sales prices, like
those obtained from calculation procedures similar to those discussed in this book.
The purpose of doing this is to have lists of prices that are uniform or consistent among
themselves, readily updated, and that can be used as techno-economic references in market-
ing discussions. As a rough guide, Table 2.6 gives some idea of the results produced by such
methods.
Threshold Price
In the context of national policies designed to protect certain sectors of activity from for-
eign competition that is considered unfair, in that it receives special advantages, other price
systems are often employed. This happens, for example, with agricultural products, which,
both in the USA and in Europe under the Economic Community, are directly or indirectly
subsidized. Examples include:
Intervention price or guaranteed price, at which a State or group of States undertakes to
buy a product from its own farmers. The value chosen is generally distinctly higher than
the world price; sometimes limits (quotas) are imposed, specifying the maximum quan-
tities to be sold in this way ( e g , milk, beetroots, etc.),
0 Minimum import price, providing a basis for determining the amount of deductions lev-

ied on foreign products,

0 Export refunds, corresponding to the difference between the guaranteed price and the

world price, etc.

128 2. The Elements of Economic Calculation

Table 2.6 Transaction prices for selected chemical products in the USA
($ per metric ton), January 1999

Products Prices Products Prices

Vinyl acetate 990 Hydrogen peroxide 35% 540
Acetone 880 Ethanol (fuel) 410
Acetic acid (technical) 840 Ethanol amines 1260
Nitric acid 40°B 240 Formaldehyde 37% 290
Nitric acid 94-98% 400 Propanol 750
Sulfuric acid (pure) 85 MEK 1010
Terephthalic acid 640 Ethylene oxide 1080
Acrylonitrile 1170 Propylene oxide 1410
Anhydrous ammonia 190 Phenol 860
Acetic anhydride 1100 Polyester (resin) 1720
Maleic anhydride 1170 Polyethylene HD (resin) 750
Phthalic anhydride 840 Polyethylene HD (molding) 880
Butadiene 330 Polyethylene LD (resin) 880
Butene-1 620 Polyethylene LLD 770
n-Butanol 1120 Polypropylene 640-860
Secondary butanol 1500 Polystyrene 840- 1650
Tertiary butanol 1500 Liquid caustic soda (50%) 350
Caprolactam 2 000 Caustic soda pellets 680
Chlorine 280 Styrene 480
Diethyleneglycol 770 Urea 46% N 220
(Source: Chemical Marketing Reporter)

c. Sharing of Transport Costs and Risks Between Seller and Buyer

Standardized language

Regardless of the reference price selected, in order to obtain a proper understanding of a prod-
uct’s price it is also necessary to state just how the seller will make it available to the buyer,
so that the latter may take into account the additional costs of transferring it to the place where
it will be used. To avoid any ambiguity concerning the allocation of costs and risks assigned
to each of the parties, rules have been enacted which are expressed in the standardized lan-
guage drawn up by the International Chamber of Commerce. The vocabulary, called “Inco-
terms” describes particular situations, illustrated by diagrams in French and English. Their
manner of application is shown in Fig. 2.9a.
Between the factory that manufactures the product and the one that transforms it, several
steps must be distinguished:
Transport by truck, railroad, waterway, etc. to the production site
Customs control (customs clearance for export)
 2. The Elements of Economic Calculation 129

0 Unloading and possible loading onto a ship, a plane, a train, another vehicle, or contin-
uation by the same mode of transport
0 Transfer from one frontier to another
0 Unloading and possible loading onto a train, a truck, a barge, etc., or continuation in the
same vehicle
0 Customs control (customs clearance for import)
0 Transfer by truck, railroad, etc. to the place where it is used.
Each of these stages involves specific expenses: costs for transport, insurance, taxes, and
handling fees, whose payment by one or other of the parties must be clearly stipulated.
Among the numerous situations considered, we can only look at three principal ones:
0 “Factory” or “ex works”, when the price of the goods is identified as coming from the

manufacturing plant. In this case the buyer must take delivery at that point: all the costs
and risks connected with transfer to the utilization site are its responsibility.
0 FOB (Free On Board), when the goods are delivered under customs seal for export by

the selected mode of transport (ship, for example) to travel from one country to another.
The costs and risks to the buyer must cover the shipment to the unloading point, the bor-
der crossing, the customs clearance for import, and delivery to the factory where it is
used.
0 CIF (Cost-Insurance-Freight),when the goods are delivered under import customs seal

by the seller to the arrival point of the selected mode of transport (ship, for example).
The seller then bears all the corresponding costs and risks, except for those for insuring
the cargo, which are the buyer’s responsibility. The buyer also pays for loading onto
another mode of transport for delivery, customs clearance for import, and shipping to the
place of utilization.
To simplify matters, we often speak of prices that refer to the port of embarkation (FOB
Gulf Coast, for example), or of unloading (for example, CIF A M ) . Note that all these oper-
ations are accompanied by formalities (obtaining export licenses, buying insurance policies,
various invoicing, legal waivers, issuing certificates of origin, etc.) for which each party is
responsible, according to the stages falling under its control.
Figs. 2.9b and 2 . 9 ~
are diagrams of the FOB and CIF price situations, which themselves
give rise to several variants, in particular when it is a question of who is responsible for the
handling and border-crossing costs.

d. Pricing for Evaluation Purposes

The diversity of reference price systems does not make it easy to set up simple or general rules
for project evaluation. However, with certain exceptions, we may adopt the following princi-
pal approaches:
Where raw materials, products, and co-products are concerned, the most practical solution
is to refer to average transaction prices, i.e., to the minimum profitable sales prices as deter-
mined by a procedure like that suggested in this book. It may be recalled that these resemble
floor prices that are very close to contract prices, as often confirmed by actual cases, when
this kind of information is available. Considering the tonnages involved, the principal deter-
minant is the economy of scale.
130 2. The Elements of Economic Calculation
2. The Elements of Economic Calculation 131
*
v)
v)
u
0
rc
0
2a
.-M
Y
 d

Wd

FOB
FOB means "Free On Board. The merchandise h)
W
is placed on board a ship by the seller, at the

FOB
FOB means "Free On Board. The merchandise h)
port
is of embarkation
placed on board named
a shipinbythe
thesales contract
seller, at the
Theport
riskofofembarkation
loss or damagenamedto the merchandise
in the is
sales contract
transferred
The risk from the
of loss orseller
damageto the buyer
to the when the is
merchandise
FREE ON BOARD merchandise crosses the shio's rail.
transferred from the seller to the buyer when the
FREE ON BOARD merchandise crosses the shio's rail.

Seller's primary duties Difference from FAS Buyer's primary duties

Seller's
- Deliver primary
the goods duties
on board FAS is considered
Difference fromtoFAS be a sale in the domestic Buyer'scarrier
- Nominate primary duties
- Provide
- Deliverexport
the goods
licenseonand
board pay
market
FAS is
arrange
Therefore,
considered
anyTherefore,
thetoseller
export license.
be a does
sale in
FOBdoes
not have
the
IS annot
to
domestic
export
- Contract
- Nominate
for carrier
the carriage and
- Provide
export taxesexport license
and fees, and pay
if required
market
sale, so the seller has
the seller
to provide the
arrange any export license. FOB IS an export
have to
export
- Contract
pay for the carriage and
the freight
export taxes
- Provide a cleanand onfees,
boardif required
receipt license
sale, at
soits
theown expense
seller has to and risks,the
provide and pay
export pay theexport
- Obtain freightlicense and pay
- Provide
- Pay loading a clean
costs on board receipt
according to the
all license
necessary export
at its fees and and
own expense taxes.risks, and pay - Obtain
export taxesexport license
and fees, and pay
if required
all necessary export fees and taxes.
- Pay of
custom loading costs
the port according
to the to the
extent that Required documents export taxes and fees, if required
custom
they of the
are not port tointhe
included theextent
freightthat - commercial invoice
Required documents
- customary
- commercial clean receipt. usually mate's receipt
invoice
they are not included in the freight - export licenseclean receipt. usually mate's receipt
- customary
- export license
Optional documents
- export license
Optional documents
- certificate of origin
- export license
- other documents
- certificate needed for export and import
of origin
clearance
- other documents needed for export and import
clearance

Figure 2.9b Free on Board price. (Source: Incoterms)

Figure 2.9b Free on Board price. (Source: Incoterms)
 CIF means “Cost. Insurance, and Freighr. This is the same as CBF, but in addition

CIF
---
COST, INSURANCE, FREIGHT
the seller must provide maritime insurance against the risk of loss or damage during
transport. The seller contracts with the insurer and pays the insurance premium.
The- buver
--,- will note that in this case. in contrast to the exDression ‘FreiohVCarriaae
& Insurance paid to ”, the seller is solely responsible for providing the minimum
insurance coverage (called FPA terms)

Seller’s primary duties Difference from C&F Buyer’s responsibilities

- Conclude the shipping contract and The only difference between C&F and CIF - Accept delivery of the goods
is that under the latter term the seller upon dispatch, on receipt by
pay the freight to the designated assumes the additionalobligation of
destination providing insurance. the buyer of the invoice, the
- Deliver the goods on board insurance policy (certificate),
N
- Provide export license and pay export and the bill of lading; the bill of
taxes and fees, if required lading is supposed to “represent 3
- Insure the goods during transport and Required documents the merchandise”
- commercial invoice
pay the premium - bill of lading
- Pay unloading costs if not
- Provide the buyer with an invoice, a net - export license included in the freight
-insurance policy
bill of lading, and a maritime insurance
policy or certificate Optional documents
- Pay loading costs - certificate of origin, consular invoice
-other documents needed for import
- Pay unloading costs if included in the clearance
freight

Figure 2 . 9 ~Cost, Insurance, and Freight price. (Source: Incoterms)

134 2. The Elements of Economic Calculation

If the starting raw material is energy-related or strategic in nature, such as natural gas,
petroleum or derived petroleum cuts, coal, etc., or biomass, it may be difficult to obtain more
than a price range. It is best either to select several scenarios, with clear definitions and
descriptions, or to conduct a price sensitivity study. This approach may be extended, if need
be, to first generation chemical intermediates, to minerals, precious and semi-precious met-
als, basic agricultural products, etc., whose prices are subject to very large fluctuations
depending on the economic situation.
As for by-products, which are usually produced in small tonnages, their price reference is
the spot market. However, the costs of purification to meet commercial specifications must
be deducted, or failing this, a discount of around 50% should be adopted.
In some situations, to be studied on a case-by-case basis, by-products may have to bear the
costs of removal or disposal, or worse still, the expense of eliminating them. Sometimes, it
involves accessory production from the regeneration of startingmaterials (the major petrochem-
ical intermediates, for example); the reference price is then the one used initially for this type of
product, with a deduction for the possible costs of returning it to the original specifications.
Reagents, which are also used in relatively limited quantities, are like by-products, best rep-
resented in terms of price by the spot markets than by floor prices. However, in certain cases
it may be difficult to find the quotation in the usual market price lists. They may instead be
found in producer catalogs, designed primarily for chemical laboratory supplies, or industries
such as pharmacies, photography, etc., where purity is essential. As a result, the prices shown
are very largely over-valued for the applications intended for reagents. In this situation, two
approaches are possible: either apply a discount of at least 30%, or calculate by taking the unit
quantities to the power 0.8, or determine a floor price and increase it by about 30%, as long as
it is possible to find the way back through the manufacturing chain to a raw material with a
published spot price. In such an approach it is not necessary to know all the elements, partic-
ularly the energy costs or the investment costs. Simple rules show that the more sophisticated
the reagent, the smaller is the contributionof these two items to the floor price. Roughly speak-
ing, particularly for solvents, the cost of materials accounts for 70 to 80%. In other words, the
important thing is to carehlly identify the successive reactions and the corresponding yields,
so as to make a quick calculation. As a guide, in the absence of more specific information, an
approximate solution is the following: on the basis of an assumed standard unit production
capacity of 100000 Uyr, assess the average share of the cost of raw materials in the product
price, as a function of the output and the degree of complexity of the transformation to be per-
formed, starting with major organic or inorganic chemical intermediates (Table 2.7). The
capacity adjustment can then be made by applying the following expression:
Y

with
x : % of raw material cost in the floor price for unit capacity C1
y : % of raw material cost in the floor price for unit capacity C2
f:extrapolation exponent.
When x is obtained from Table 2.7, C1 = 100000 andf= 0.7.
 2. The Elements of Economic Calculation 135

Table 2.7 Amount of raw material cost in a product’s floor price. Unit capacity: 100 000 t/yr

Yields (%)
Complexity
>90 70-80 50-60
Low (hydrogenation, alkylation, oligomerization, straight fertilizer, etc.) 75-80 80-85 85-90
Moderately low (polymerization, simple oxidation, nitration, chlorination, 45-60 50-65 60-75
phosphoric acid etc.)
Moderately high (complex oxidation, sulfuric acid, etc.) 30-40 35-50 40-55
High (steam cracking, co-production, chlorine, electrolytic soda, etc.) <20 10-30 15-35
(Source: IFP)

As concerns price, the behavior of homogeneous phase catalysts is comparable to that of

chemical reagents: the spot market serves as a reference, as long---obviously-as the main
inputs and their respective proportions have first been identified. With heterogeneous cataly-
sis the problem is often trickier, in that one is dealing with a product that is itself finished,
where it is hard to determine not only its composition but also how it was made. In the
absence of precise information, the best way to estimate its price is still, as with the more
sophisticated reagents, to make out the starting inorganic materials. In this regard, examining
patents can provide useful information for estimates, particularly concerning the kind of sup-
ports and metals, and their precursors (oxides, chlorides, etc., for example) that were utilized.
The amounts of these constituents are the first thing to look for, all the more necessary when
they are precious or semi-precious elements. The approach to follow may then be as follows:
For mass catalysts: refer to the spot prices of the metallic oxides and corresponding met-
als; weight them according to their proportions in the final formula. Add 30% to take
simultaneous account of yields (95% min. wt.), energy costs, labor, and fixed operating
and capital costs.
For supported catalysts: separate the support and the active phases; treat the support in the
same way as the mass systems; add the share of costs for the metals, allowing for losses of
around 1 to 2% in weight, which are even more important from the economic viewpoint when
they are precious metals, and consequently expensive. The costs for impregnation may be
taken as included in the additional 30% applied to the “support” portion (alumina, silica, etc.).
Every general rule has its exceptions, so it is always wise to take a hard look before apply-
ing the ones expounded here. Examples include enzymatic catalysis, generally employed in
homogeneous situations, but also likely to be supported; chlorinated alumina; supported zeo-
lites, etc. A useful alternative is to proceed by analogy or similarity. This approach can more-
over be generally applied, as long as equivalent information is available from applications
other than the ones being considered (quantities, productivity, date, location).
Adjustments for performance (particularly productivity, i.e., LHSV (volume), LHSV
(weight), residence time, etc.) are made by direct proportion; those for date are of the same
kind as those likely to be made for investment matters, which will be taken up later (see
Section 2.3.2). In this case, we rely on indices, where the ratio of values for two separate dates
enables us to take a known cost from the past and update it. Differences in location, whose
136 2. The Elements of Economic Calculation

Table 2.8 Nelson-Farrar indices for chemicals

Organic products Inorganic products

Year
(Code 614 EX 612) (Code 613 EX 611)
1965 90.2 107.6
1966 90.4 109.1
1967 90.8 113.2
1968 90.0 117.8
1969 87.9 119.5
1970 89.0 119.1
1971 89.2 121.9
1972 87.4 123.1
1973 88.1 127.8
1974 149.4 157.7
1975 203 .O 221.3
1976 213.2 238.2
1977 216.5 246.6
1978 215.4 254.1
1979 264.2 267.9
1980 33 1.8 310.1
1981 362.6 375.5
1982 338.7 405.1
1983 332.1 381.6
1984 330.6 380.4
1985 320.2 389.7
1986 302.0 383.7
1987 319.9 380.6
1988 363.3 406.3
1989 395.5 438.9
1990 382.1 451.5
1991 376.3 452.4
1992 366.1 442.5
1993 374.9 437.4
1994 386.0 445.9
1995 445.0 473.5
February 437.3 472.9
May 457.5 474.1
August 454.2 474.5
November 43 1.2 472.5
1996 432.9 481.6
February 419.0 487.0
May 439.6 486.6
August 434.6 474.9
November 438.3 478.1
1997 437.1 471.5
February 443.4 471.3
May 433.9 463.2
August 435.9 471.7
November 435.2 479.7
1998 411.9 477.6
February 423.4 483.8
May 410.9 482.2
August 408.2 478.9
November 405.2 465.6
1999 - -
Februarv 402.8 445.9
May 401.1 428.2
iource: The Oil & Gas .I )
 2. The Elements of Economic Calculation 137

influence on investment matters is very strong (see Section 2.3.2.3), have only a negligible
impact here. Consequently, the situation is made simpler, since the kinds of index that we can
call on are international in nature. As a rough guide, among the regularly published ones we
may mention Nelson-Farrar’s: Table 2.8 shows the main values it provides in the area of
organic and inorganic compounds, in general terms.
COMMENT
The approach suggested for catalysts can be extended to reagents, when the compo-
sition of the corresponding item is unknown, but only if we can identify the overall
or dominant nature of the inputs (organic or inorganic).

2.2.2.2 Utilities and Miscellaneous Costs

A. Definition and Treatment of UtXty Costs

In practice, the word “utility” expresses the provision to a facility of its requirements both for
energy (electricity, fuel, steam, cooling) and for related items (particularly the various cate-
gories of water: cooling, process, boiler, treated, etc.; inert or pressurizing gases such as nitro-
gen, carbon dioxide, rare gases, etc., compressed air perhaps, etc.). It also covers products or
by-products of the same type that may result from a conversion process (steam, condensates,
combustible gases, etc.).
Following a procedure similar to those used for raw materials and reagents, the costs of
utilities are obtained by knowing both the quantities to be employed (annual or per unit,
according to how the calculations are performed, in terms of operating cost or cost of price)
and their respective prices per unit, or their usual values.
Generally speaking, in project evaluation studies the utility costs that must be considered
have to cover the following items:
0 Manufacturing costs proper, in specific plants, but usually without an exact location,

0 Distribution costs, via the appropriate networks for each kind of utility,

0 If necessary, expenses for recycling effluents that are still valuable after being used, such

as condensed steam reinserted into the boiler circuit, and cooling water that, after warm-
ing, is returned to its initial condition by being passed through air-cooled towers,
0 Make-up amounts of starting materials, for replacing losses (water, liquid coolants, con-

densates, etc.).
These prices thus include:
0 Variable charges, which primarily cover the costs of “raw materials”. Actually, these are usu-

ally the utilities used as precursors to the one concerned, which in this case has a role analo-
gous to the products that are wholly manufactured elsewhere among the various units that
coexist in a complex industrial installation. For example, it could be the natural gas required
for the cogenerationof steam and electricity.This item also coversthe costs of chemicalprod-
ucts (particularlyfor obtaining the various categories of water and for their possible treatment
before recycling), and, of course, utilities (electricity for refrigeration, for example),
Labor costs,
0 Fixed charges covering the costs of investment for the construction and operation of the

facilities (see Section 2.3).

138 2. The Elements of Economic Calculation

0 Some provision for making a profit on the invested capital, via a minimum value for an
economic criterion, or following accounting rules used for a given site or manufacturing
company.
All in all, they are comparable to actual internal transaction prices, or minimum profitable
sales prices as defined in Section 2.1. This approach offers the advantage of affording a mea-
sure of autonomy to the utility production units, whose role in the management plan is in
every way similar to that of the facilities providing products that must meet precise specifi-
cations. It also enables the centralization and rationalization of the energy and other needs
arising from the whole assemblage of units that make up the industrial platform. It is just as
if each one of them were to charge its own utility costs at their real prices, like its raw material
and chemical product requirements.
However, since each entity has a specific function, which leads to operating conditions that
vary from one unit to another, it may prove necessary to make certain adjustments to the cen-
tralized distribution system. For example, there may be a request to regulate the saturated
steam supply by partial flashing, or it may have to be superheated or its pressure raised. The
resulting investments costs must then be charged entirely to the facility that uses the steam,
without changing the initial access costs for these utilities.
In the particular case of the provision of refrigerants at a certain thermal level, the corre-
sponding investments are usually charged to the consuming unit alone, since such facilities
are not very versatile or flexible, and are generally compelled to operate under very exacting
conditions. In this case, only the energy necessary for their operation is included in the eco-
nomic calculation, also possible expenses for replacements and additions of coolant liquid.
The same procedure is applied when a given unit has such large utility requirements that
they cannot be handled by a central entity that services other plants.

B. U f X f yPrices

These are very variable from one site to another, and in particular very dependent on the ease
of access to the basic energy materials required to produce them, or to natural water resources
(rivers, groundwater, seas, etc.). They are also closely linked to the existing regional or local
infrastructures where the particular manufacturing center is set up, and even to national deci-
sions in the area of energy policy.
Thus the Scandinaviancountries and mountainousregions are known for their low electricity
prices, and in Asian and African countries the scarcity of water is often compensated for by the
low cost of fuel. Governmentchoices, in terms of energy, also have the consequenceof favoring
cheaper sources: petroleum (Saudi Arabia, the Emiraks, etc.) natural gas (Great Britain,New
Zealand, etc.); coal (South Africa, Eastern Europe, etc.); nuclear electricity (France); and so on.
In sum, each project has its own specific utility prices. However, within the same geo-
graphic zone, we see that the fluctuations that affect it remain within certain limits. In the
absence of contrary indications reflecting local conditions, this observation enables the use
of average values, or reference to internal transaction price calculations, similar to those dis-
cussed for raw materials and chemical products; alternatively, prices that are periodically
published may provide the reference.
As a rough guide, the following information may be kept in mind.
 2. The Elements of Economic Calculation 139

a. Fuels

These are usually directly available on the market. For this reason, they follow the same rules
and procedures that were discussed for raw materials. Sometimes they may be liquid or gas-
eous by-products from the various manufacturing operations at the consumer’s site. In this
case, if they can be used as is, close to their point of production, they are inserted into the
existing systems and charged in proportion to their calorific power. In such circumstances, if
they substitute for commercial products they may at best be credited for their useful value. If
it proves necessary to process them to required specifications before they can be used, their
price-as with any by-product that is marketed-must be reduced by the amount of the cost
of purification.
For late 1998-early 1999, the prevailing prices on the fuel markets were as shown in
Table 2.9a-d.

Table 2.9a Average fuel oil prices in various countries (late 1998-early 1999)

Uses Exchange
*ate&)for local
Industrial ElectricitJ Domestic currency
Country
1SC Heay ,SC Heavy Light Heavy Light
1997 1998
(US$lt) (US$/t) ~ ~ $ / m 3 )(US$/t) (us$/m3)
Australia nd nd nd nd nd 1.30 1.31
Austria 91.3 170.1 201.3 C 276.7 13.51 13.50
Belgium 116.7 111.0 161.5 116.7 195.5 36.66 36.80
Canada 105.8 nd 150.9 nd nd 1.17 1.16
Denmark nd 124.3 220.0 nd 503.7 8.32 8.38
Finland nd 146.3 207.8 96.9 253.6 5.91 6.04
France 95.7 109.7 207.8 nd 303.7 6.52 6.5 1
Germany nd 98.0 174.1 100.5 201.5 2.00 2.01
Ireland 159.4 nd 235.6 119.1 ( l ) 352.6 0.68 0.69
Italy 160.6 146.0 693.8 nd 832.6 1595 1621
Japan 123.6 186.1 175.4 154.2 (l) 274.5 163.52 163.00
Netherlands nd 156.2 nd nd 360.9 2.05 2.07
Norway nd 253.1 299.7 - 387.8 9.20 8.99
Portugal 207.1 225.0 770.3 99.4 808.8 122.60 125.00
Spain 150.5 178.5 290.1 178.2 336.5 123.89 126.00
Sweden - 162.1 ( l ) 176.4 nd 418.8 9.62 9.63
United Kingdom 122.1 nd 164.5 105.6 190.1 0.65 0.66
USA 84.9 nd 11 1.6 86.7 241.5 1.oo 1.oo
OECD Europe 137.1 nd 272.4 55.5 296.5 - -

OECD 177.5 nd 175.3 54.6 283.9 - -

(1)
. , 1997. nd: not available
(2) versus the dollar. C : confidential
- : not applicable
(Sources: AIE, Energy Prices and Taxes, 2nd Qtr. 1999)
140 2. The Elements of Economic Calculation

Table 2.9a (continued) Average fuel oil prices in various countries (late 1998-early 1999)

Uses Exchange rated2)

Industrial Electricity Domestic for local currency
Country
Heavy
1997 1998
(US$/t)
Argentina 152.75 410.10 152.75 nd 1.o 1.o
Brazil 173.12 353.30 173.12 nd - -

China nd nd nd nd 8.29 8.28

India 129.24 223.72 ( l ) 129.24 59.86 36.3 42.6
Indonesia 33.26 24.45 nd 26.15 2909.00 10715.00
Russia nd nd nd nd 5.785 6.125
South Africa 172.79") nd nd nd 4.6 5.4
Thailand 155.34 nd nd 283.09 31.4 41.3
Venezuela 40.73 87.43 40.73 nd 488.6 540.0
( I ) 1997. n d not availabl
(2) versus the dollar
(Sources:AIE, OLADE)

Table 2.9b Average natural gas price (early 1999). France

1 Source I Internal I Gaz de France (GdF)

Supplied Industry Heating
for Fuel Major
consumer >5 lo6 <5 lo6 >5 lo6 <5 lo6
gas
kWyr kWyr kWyr kWyr
Price 0.78 1.07 1.22 1.83 1.22 1.91
 2. The Elements of Economic Calculation 141

Table 2 . 9 ~Average natural gas prices in various countries (late 1998-early 1999; US$/106 kcal)(I)
~~

Uses Exchange r a t e ~ ( ~ ) flocal

o r currency
Country
Industrial Electricity Domestic 1997 1998
Australia 14.06 C2) nd 34.41 (2) 1.30 1.31
Austria 14.18 nd 38.07 13.51 13.50
Belgium 12.81 (2) 12.92 (2) 40.43 36.66 36.80
Canada 9.02 nd 21.71 1.17 1.16
Denmark C C 5 1.55 8.32 8.38
Finland 12.20 10.47 14.88 5.91 6.04
France 13.60 (2) nd 38.18 (I) 6.52 6.5 1
Germany 16.44 (2) 13.68 (2) 36.07 (2) 2.00 2.01
Ireland 17.38 10.50 42.58 0.68 0.69
Italy 18.36 nd 74.55 (2) 1595 1621
Japan 34.28 (2) 15.67 (2) 95.29 (2) 163.52 163.00
Netherlands 11.80 11.68 34.25 2.05 2.07
Norway - nd - 9.20 8.99
Portugal nd nd - 122.60 125.00
Spain 16.33 15.21 63.22 123.89 126.00
Sweden nd nd nd 9.62 9.63
United Kingdom 10.01 11.63 30.44 0.65 0.66
USA 11.78 9.28 26.26 1.oo 1.oo
OECD Europe 17.16 (2) nd 43.29 (2)
OECD 15.38 (2) nd 35.77 (2)
(1) high heating value. n d not available
(2) 1997. C : confidential
(3) versus the dollar. - : not applicable

(Sources: AIE, Energy Prices and Taxes, 2nd Qtr. 1999)

Table 2 . 9 ~(continued) Average natural gas prices in various countries

(late 1998-early 1999; US$/106 kcal)(')

Uses Exchange rates(3)for local currency

Country
Industrial Electricity Domestic 1997 1998
Argentina 13.315 nd 22.220 1.o 1.o
Brazil 12.315 nd 9.920 - -
China nd nd nd 8.29 8.28
India nd nd nd 36.3 42.6
Indonesia 1.815 nd 0.934 2 909 10715
Russia nd nd nd 5.785 6.125
South Africa 13.641(2) - - 4.6 5.4
Thailand nd nd - 31.4 41.3
Venezuela 1.982 1.982 9.469 488.6 540.0
I) high heating val n d not availab
1) 1997. - : not applicable
(3) versus the dollar.
(Sources:AIE, OLADE)
142 2. The Elements of Economic Calculation

Table 2.9d Average coal prices in various countries (late 1998-early 1999; US$/t)

Uses Exchange rates(2)for local currency

Country
Industrial Electricity Coke 1997 1998
Australia nd nd nd 1.30 1.31
Austria 58.37 C nd 13.51 13.50
Belgium nd 33.72 54.73 36.66 36.80
Canada nd nd nd 1.17 1.16
Denmark nd nd - 8.32 8.38
Finland 74.06 38.05 93.77 5.91 6.04
France 91.55 35.01 50.40 6.52 6.5 1
Germany nd 41.29 51.64 ( I ) 2.00 2.01
Ireland nd 44.74 (1) - 0.68 0.69
Italy 43.49 54.00 60.33 1595 1621
Japan 37.23 46.46 44.13 163.52 163.00
Netherlands nd nd nd 2.05 2.07
Norway nd nd C 9.20 8.99
Portugal nd 51.18 52.57 122.60 125.00
Spain nd nd C 123.89 126.00
Sweden nd nd nd 9.62 9.63
United Kingdom 53.60 45.99 nd 0.65 0.66
USA 35.76 28.62 50.60 1.oo 1.oo
OECD Europe 48.12 36.92 50.66
OECD 35.68 27.29 41.33
(1) 1997. nd: not availab
(2) versus the dollar. C : confidential
- : not applicable

(Sources: AIE, Energy Prices and Taxes, 2nd Qtr. 1999)

Table 2.9d (continued) Average coal prices in various countries (late 1998-early 1999; US$/t)

Uses Exchange rates(2)for local currency

Country
Industrial Electricity Coke 1997 1998
Argentina 8 1.68 81.68 8 1.68 1.o 1.o
Brazil 3 1.79 3 1.79 nd - -
China 37.71 nd 42.55 ( I ) 8.29 8.28
India nd nd nd 36.3 42.6
Indonesia 8.68 nd 47.96 ( I ) 2 909 10715
Russia nd nd nd 5.785 6.125
South Africa nd nd - 4.6 5.4
Thailand nd nd - 31.4 41.3
Venezuela nd nd 1.93 488.6 540.0
(1) 1997. nd: not available
(2) versus the dollar. - : not applicable
(Sources: AIE, OLADE)
 2. The Elements of Economic Calculation 143

b. Steam
This can be generated specifically to meet exact needs, in special boilers. It is also often co-
generated during electricity production.
In the first case, every expense connected to its production is chargeable. Its internal trans-
action price is then directly dependent on the price of the fuel employed; it is also a function
of the size of the plant, the pressure at which it must be delivered, and the scale of the distri-
bution system (see Appendix A.2.8).
As an example, Table 2.10 shows representative values for different fuel prices.

Table 2.1 0 Examples of internal transaction prices for steam

(€ mid-2000h)

Fuel price (€/lo6 kcal) 7.5 10 14

Size of facility (a) >250 <25 >250 <25 >250
Pressure (bar) 10 3.8 6.1 5.2 7.5 6.7
40 5.0 7.6 6.9 9.4 8.5
100 5.5 8.2 7.3 10.2 9.1
(Source: IFP)

When steam and electricity are co-generated, the costs, whatever their source (variable or
fixed charges), must be charged in proportion to the amounts of the utilities that are jointly
produced.

c. Electricity

A distinction must be made between the average price tariffs offered by the major national
distributors such as EDF in France, ENEL in Italy, etc., and those found at heavily industri-
alized sites, for example refineries or chemical complexes, which provide for their own elec-
trical energy needs internally, usually, moreover, by using combustible by-products.
For the major distributors, the average prices offered to customers are not only a function
of the nature and structure of all the power stations in operation, i.e., in fact, a function of the
country concerned, but also of the sectors of activity being serviced and the scale of consump-
tion. In this regard, Table 2.1 1a gives a comparative idea of the price levels in operation in
1998-early 1999, among various members and non-members of the OECD.
For an average load rate of 8 000 hourslyear, for example, Table 2.1 1b shows the prices in
France for various power requirements.
At the producing sites, the calculations can only be made in term of costs, or of internal
transaction prices, which are very dependent on the kind of fuel used and the technology
employed. They also depend greatly on the unit capacity of the plant.
144 2 . The Elements of Economic Calculation

Table 2.1 l a Average electricity prices in various countries (late 1998-early 1999; US$kWh)

Uses Exchange rates(*)for local currency

Country
Industrial Domestic 1997 1998
Australia 0.0584 (l) 0.0829 (I) 1.30 1.31
Austria 0.0719 0.1528 ( l ) 13.51 13.50
Belgium 0.0538 ( I ) 0.1635 ( l ) 36.66 36.80
Canada nd nd 1.17 1.16
Denmark 0.0545 0.1703 8.32 8.38
Finland 0.0444 0.0866 5.91 6.04
France 0.0423 0.1168 6.52 6.5 1
Germany 0.0589 0.1391 2.00 2.01
Ireland 0.0602 0.1248 0.68 0.69
Italy 0.1015 0.1707 1595 1621
Japan 0.1078 (l) 0.1534 (l) 163.52 163.00
Netherlands 0.0598 0.1227 2.05 2.07
Norway nd 0.0562 9.20 8.99
Portugal 0.1350 0.2221 122.60 125.00
Spain 0.0697 0.1826 123.89 126.00
Sweden 0.0271 ( I ) 0.0804 ( l ) 9.62 9.63
United Kingdom 0.0598 0.1111 0.65 0.66
USA 0.0403 0.0822 1.oo 1.oo
OECD Europe 0.0622 0.1304
OECD 0.0424 0.1010
(1) 1997. nd: not available
(2) versus the dollar. - : not applicable
(Sources: AIE, Energy Prices and Tares, 2nd Qtr 1999)

Table 2.1 1a (continued) Average electricityprices in various countries (late 1998early 1999; US$kWh)

Country
Uses I Exchange rates(2)for local currency
Industrial 1 Domestic 1997 1998
Argentina 0.079 0.139 1.o 1.o
Brazil 0.058 0.133 - -

China nd nd 8.29 8.28

India 0.080 (I) 0.034 36.3 42.6
Indonesia 0.050 nd 2 909 10715
Russia nd nd 5.785 6.125
South Africa nd nd 4.6 5.4
Thailand 0.061 ( I ) 0.074 (I) 31.4 41.3
Venezuela 0.029 0.01 1 488.6 540.0
(1) 1997. nd: not available
(2) versus the dollar. - : not applicable
(Sources: AIE, OLADE)
 Next Page

2. The Elements of Economic Calculation 145

Table 2.1 1b Price of electricity (c€/kWh, France, mid-2000)

Required power Electricity price

(Mw) (CukWh)
7.6
5.0
4.1
3.3
(Source: IFP)

d. Refrigeration
The internal transaction price for cooling depends on three main parameters:
0 The thermal level

0 The unit capacity required

The price of electricity.

For France in mid-2000, it can vary from 60 to 200 elO6 kcal, over a range of tem-
peratures between 0 and -100°C and for consumptions from 5 to 15 MkcaVyear (see Appen-
dix A.2.8).

e. Cooling, Process, and Boiler Water

The average internal transaction price for cooling water, for France in mid-2000, was
between 0.03 and 0.5 a m 3 , depending on the quantities to be supplied. The calculation
assumes an addition of river or ground water corresponding to 5% of the volume circulating
in the distribution system; it also allows for an average heating of 10°C, from 25 to 35"C,in
the facilities where it is used (see Appendix A.2.8).
As for process water, it has the same characteristics as high-pressure (100 bar) boiler
water. Its internal transaction price is very dependent on the plant's treatment capacity, as
shown in Table 2.12.

Table 2.1 2 Internal transaction prices of process water. France, mid-2000

Unit capacity Production cost

(m3W (am3)
4.5 3.0
45 1.2
100 0.6
Previous Page

146 2. The Elements of Economic Calculation

C. Miscellaneous Costs
Among the expenses that may be included in the variable charges, and which have not been
considered under the headings listed so far, the royalties described as “running” are particu-
larly conspicuous. Their essential characteristic is to be proportional to the actual plant pro-
duction, i.e., to continue over several years, in contrast to the overall initial “paid up”
royalties, which are linked to the unit size capacity and are charged as investment costs, to be
examined in detail in Section 2.3.
Generally speaking, “royalties” are intended to at least partially repay the R&D work that
leads to the industrial implementation of innovations in equipment, processing, products, cat-
alysts, etc. Their main effect is to remunerate the research organizations, which do not market
their discoveries themselves, but grant licenses to third parties to exploit them commercially.
In return, the owner of the innovation is obliged either to give the operator certain exclusive
rights, or to undertake to protect the original aspects through patents.
Running royalties, according to the provisions of the agreement signed between the two
parties, are payable by annuities over the plant’s depreciation period, or according to the
period (or semi-period, sometimes) of the validity of the patents (18-20 years). These pay-
ments may be adjusted over time, to take account of the level of competition, but they remain
linked to the actual production, not the installed capacity, so that they are subjected to the
fluctuations of market demand, and all the hazards of operating the units. Roughly speaking,
we may say that for calculation purposes ten years of running royalty payments are equivalent
to a single initial payment. Intermediate solutions are sometimes applied, including both a
cash payment, and subsequent “running” payments over a pre-determined period and accord-
ing to a schedule established by contract.

2.2.3 LABOR
As mentioned earlier, this item covers only the expenses for the manufacturing personnel,
working in shifts on the facilities, which they operate continuously, semi-continuously, or
intermittently. It concerns only the labor responsible for the actual operations, as well as the
permanent and current maintenance of the equipment. It particularly does not include the
costs of general and special maintenance personnel, workshops, warehouses, control and ana-
lytical laboratories, or services of an administrative nature (accounting, management, market-
ing, food services, on-the-job medical, works council, etc.).

2.2.3.1 The Concept of the Shift Worker

Some economic calculation schemes, in particular when they apply to complete manufactur-
ing complexes, usually not especially complicated ones, include a breakdown of wages by
activity (manufacturing, maintenance, supervision, analysis, etc.). In the present work, the
chosen procedure is more general, since it is usually impossible to obtain the information that
would allow such a distribution, and also because it concerns the desirability of inserting new
products (in the broad sense of the term) into an existing manufacturing situation. In these
circumstances, as suggested in Section 2.2.4, the maintenance costs are rolled into the fixed
 2. The Elements of Economic Calculation 147

charges, as a given percentage of the investment for the production units. The other kinds of
expense are included in the overhead and head office costs, which are themselves calculated
as a certain fraction of the costs of the plant.
The maximum length for an effective period of work without a break is generally eight
hours (sometimes twelve hours). In theory then, continuous 24-hour operation demands no
more than working with three identical shift teams. In reality, a number of factors have led to
a gradual increase in the number of teams necessary to ensure continuous operation through-
out a normal year (8000 hours on average): legal measures enacted over time, such as those
regulating weekly and yearly holidays; temporary unavailability due to illness or authorized
absences; reduction in the length of the work week (39 hours or less per week, for example);
part-time employment, and so on.
Originally four, then five, the figure is now 6.5, at least for large-scale units such as refining,
metalworking, and basic or major intermediate chemistry. It is less for smaller continuous-
operation facilities, and obviously for those where the operation is semi-continuous or inter-
mittent: three eight-hour shifts, two eight-hour shifts, seven days a week, or five out of seven.
Technical publications quote numerous examples of labor requirements, usually expressed
in numbers of shift operators, according to the type of manufacturing unit concerned
(Table 2.13). This approach amounts to giving the overall number of workers in the team. The
broad growth in unit manufacturing capacities, combined with increasingly strict safety con-
trols, as well as ever more advanced methods of automation, is leading to greater centralization
of plant operation in control rooms. In these constructions, the visual display of the various
operations being supervised has introduced the concept of the control panel and made it a stan-
dard. Today, a single system operator can manage the operation of not just several sections of
the same unit, but also of several units, thereby reducing the overall need for personnel.
On the other hand, such a development is only possible when there is an improvement in
the operators’ qualification standards, and consequently their pay. However, their numbers
can grow by only a few at a time, as a function of the scale and also the complexity of the
facilities. There are no exact rules here, but certain quick rules of thumb have been suggested,
enabling approximate solutions for a new manufacturing facility. As a rough guide, we
present the following expression, which allows the number of operator hours to be calculated:

Number of operator hours number of process stages

=tx
tons of product (capacity in tons/~iay)~.~*

t is a coefficient which has the value:

23 for intermittent operations
17 for continuous plant operations with average instrumentation
10 for continuous plant operations with good instrumentation
7 for continuous plant operations with remote control lines.
This type of calculation is all the more tricky and controversial in that, during the actual
operation of the units, it happens that night shifts usually require supervision only, whereas
activities aimed at modifying operating parameters are generally performed during the day.
The teams also sometimes need to be reinforced on these occasions, particularly by non-shift
personnel: depending on changing needs, these personnel may work on various facilities.
148 2. The Elements of Economic Calculation

Also worth noting is the increasingly widespread trend towards using sub-contractors, partic-
ularly for maintenance or more generally for non-shift operations.
We should also note that, roughly speaking, the number of shift operators varies on aver-
age in proportion to the power 0.2-0.25 of the unit production capacity.

Table 2.1 3 Typical examples of numbers of operators per team, by industrial sector concerned

Industria1 Number of Industrial Number of

Process Process
sector operators sector operators
Refining Atmospheric 3 Organic Acetic acid from 3
distillation chemical methanol
intermediates
Vacuum distillation 1 Isopropyl alcoho; 2 to 3
Catalytic reforming 3 Maleic anhydride 4
from benzene
Catalytic cracking 4 to 6 Phthalic 3
anhydride
Alkylation 4 Terephthalic acid 9
Hydrocracking 4 Caprolactam ttom 25 to 30
cyclohexane
Viscosity breaking 2 Ethanol from 5
ethylene
Sasic Steam cracking of 7 Ethylene oxide 5
organic ethane
chemistry Steam cracking of 10 Propylene oxide
naphtha
Methanol from 6 - chlorhydrin 6
natural gas
Aromatics by 4 - from isobutane 9
extraction
Benzene by 2 Phenol acetone 9
hydrodealkylation via cumene
p-xylene (adsorptior 3 Vinyl chloride 5
and isomerization) from ethylene
?olymer- HD and LD 7 to 9 [norganic Ammonia 5 to 6
ization polyethylenes chemistry
Polypropylene 8 Nitric acid 2 to 3
Polystyrene 4 to 6 Sulfuric acid 6 to 10
PVC 5 Phosphoric acid 7 to 8
Chlorinelsodaby 8
electrolysis
(Source: SrU)
 2. The Elements of Economic Calculation 149

2.2.3.2 labor Costs

The general information about the number of personnel needed for a given site must be used
with caution, considering on one hand the ever more advanced automation and centralization
of tasks, which tends to significantly reduce the theoretical requirements for shift personnel,
and on the other hand the labor policies of companies and regional and national authorities,
which can have the opposite effect. However, even more caution is required when dealing
with payroll costs and wage scales.
In this regard, various methods of calculation may be considered. The most accurate con-
sist of looking at the structure of the team, the qualification levels of the workers in it, and the
corresponding base wages. As an example (Table 2.14), reference to the French collective
agreement for the petroleum industry allows us to use the coefficients for minimum wages in
its hierarchic wage scale. To obtain the resulting annual costs, we must not only allow for the
bonuses (seniority, vacations, productivity, shift, wage basket, etc.), but also add employer
costs for employee social benefits (insurance, pension, etc.).
The foremen and engineers who make up the supervisory staff do not really carry out the shift
work themselves. Under normal conditions, they are on the manufacturing site during the day
only. On the other hand, apart from that time period, they are tied to a permanent operating sched-
ule (nights, weekends, holidays) and must respond to incidents.In adhtion, they are often respon-
sible for supervising operations at several facilities at the same time. Thus, from an accounting
viewpoint, it isjust as if only a certain fraction of their time was devoted to a particularunit. Con-
sequently, in making a detailed calculation of the overall cost of shift labor, they must be sepa-
rated from operators who really have to work as a team. A practical way of accounting for them
is to representthe supervisionthey perform by a fixed percentage of the total amount of the wages
of the actual shift operators. As a general rule, this rate is reckoned at between 20 and 25%.
In the absence of exact information about the structure of the teams at the site, an approx-
imate method for determining the operating cost of labor consists of looking at it as a whole.
Such an approach amounts to selecting a reference skill or hierarchical position and, based on
assumptions about the average levels of bonuses, employer costs, number of teams needed
for continuous operation of the plants, etc., to calculate the annual cost of a work shift. When
this is done, the overall amount of the labor expense is easily obtained by simply knowing the

Table 2.14 Typical examples ofpayments and wage costs (early 1999)
for shift personnel in the French petroleum industry

Average Annual wage costs

Monthly gross wage
Type of position hierarchical including bonuses
before bonuses (9
coefficient and benefits (9
Outside worker 185 1390 40400-43500
System operator 215 1590 45 700-49 500
Shift boss 250 1820 52600-56400
Foreman 310-340 2230-2430 56400-61 700
Manufacturing engineer 450-550 3 120-3830 65 500-80800
(Source: IFP)
150 2. The Elements of Economic Calculation

number of operators per team and the number of work shifts, often incorrectly expressed as
“operators per shift”. As a rough guide, such a cost, which amounts to 0.35 lo6 € per shift on
average (mid-2000, EU conditions), results from a consideration of the following data:
average monthly gross reference wage (€) 1750.0
seniority bonus (%) 10.0
vacation and productivity bonuses (“YO) 13.5
wage basket and shift bonuses (“YO) 30.0
employer costs (“A) 60.0
supervision (%) 20.0
number of teams 6.5
average annual cost of a work shift (lo6 €) 0.35
In order to be able to make a similar calculation for countries other than France, it may be
useful to work either by using indices, which allow local specificities to be expressed by
means of a ratio, or by reference to the average hourly costs used (employer costs included)
in the different countries. By assuming that the number of hours actually worked is equal to
the theoretical load factor for the most widely used plants in continuous operation, i.e.,
8 000 hourdyear, we can then find the annual labor cost per work shift, which incorporates
local, and in this case national, disparities. For a certain number of countries, Table 2.15 pro-
vides information that will assist in this kind of calculation.

Table 2.1 5 Comparison of hourly labor costs(’) in various countries (November 1999)

Local currency
Country Index US%
Name Hourly cost
Austria 104.7 26.9 Schilling 348.69
Belgium 111.1 28.5 Franc 1085.10
Denmark 106.9 27.4 Krone 192.38
Finland 92.2 23.6 Markka 132.65
France 100.0 25.7 Franc 158.79
Germany 121.8 31.3 Mark 57.73
Greece 40.8 10.6 Drachme 3 245.00
Ireland 66.5 17.1 Pound 12.68
Italy 80.3 20.6 Lire 37630.00
Netherlands 99.8 25.6 Guilder 53.25
Norway 109.3 27.8 Krone 2 16.79
Portugal 30.1 7.7 Escudo 1460.00
Spain 67.9 17.4 Peseta 2737.00
Sweden 91.4 23.1 Krona 191.04
Switzerland 104.7 27.0 Franc 40.68
United Kingdom 86.7 21.5 Pound 13.37

I
I
Japan
USA
I
114.7
96.0
I
29.4
24.6
Yen
Dollar
3 338.00
24.63
(1) Including employer costs.
(Source: Rexecode)
 2. The Elements of Economic Calculation 151

COMMENT
Some treatments have a special line for expenses relating to laboratory and manufac-
turing control personnel, showing them alongside the expense for shift labor, and
reckoning them at about 20% of the latter cost.

2.2.4 FIXED CHARGES

This item covers two major expense categories, whose treatment depends on the nature of the
fiscal regime of the country in which the plants are located:
0 Fixed operating charges such as maintenance, taxes, and insurance, and overhead and

head office costs, whose essential characteristic is that they always form part of the oper-
ating cost or cost of price, regardless of any fiscal consideration.
0 Fixed capital charges, such as the allocation for the depreciation allowance and financial

costs, related to any loans that have been made, or more generally to the locking up of
the sums required for the investments. They are recorded as expenses or pseudo
expenses only in market-economy countries, where the concept of taxation of profits
exists, and where special provisions aimed at encouraging manufacturers to invest grant
them tax exemptions.

2.2.4.1 Fixed Operating Charges

As their name suggests, these are connected to the operation of the plants, i.e., to the utiliza-
tion of the equipment that constitutes the unit proper, and to the partial use of the storage and
general services that support the related administrative operations. They are therefore usually
calculated as a hnction of the investments that these plants represent, and not in comparison
to the total investment burden represented by the depreciable capital and the working capital
(see Section 2.3). These concepts take into account all the essential expenses that must be
incurred before manufacturing can begin, but they do not all necessarily generate operating
costs: their main interest is to meet fiscal concerns rather than manufacturing ones.

A. Maintenance

Three possible kinds of maintenance work must be distinguished:

Routine inspections: periodic checkups and inspections of the various components of a
plant,
0 Specific scheduled services, performed during regular, planned shutdowns of the units,

0 Urgent and occasional repairs made necessary by unpredictable operating incidents.

In manufacturing operations, each one of these is generally undertaken by a different

group of personnel. The first is usually taken care of by the shift operators, particularly by
outside employees, so that the resulting expenses are treated as part of the operating cost or
cost of price, on the line for labor costs.
152 2. The Elements of Economic Calculation

The other two, which require adjustments, even replacements of all or part of one or more
pieces of equipment, generate both personnel costs and additional investment costs. As far as
the labor requirements are concerned, the skills required are in no way comparable to those
of the operators, since they generally call for the use of equipment manufacturers, construc-
tion firms, or workshops. For this reason, specialized crews, different from the ones that oper-
ate the units, perform such work. Usually, the scheduled services are assigned to outside
companies or to contract personnel. However, although urgent repairs are also performed by
specialists, in this case they work for the manufacturing company responsible for the produc-
tion. For all that, here too there is a trend towards increased use of sub-contractors, such that
only the most urgent or safety-related work is done in-house, and that outsiders are very
quickly brought in to get the apparatus itself running again.
All things considered, the correspondingexpenses, whose practical effect is the revamping
or alteration of various components of a facility, are comparable to the costs allowed for the
original construction, whether they are the costs of materials or of construction labor. Conse-
quently, it is just as if, in the course of operations, a portion of the investment was renewed.
It would therefore seem reasonable, in this case, to roughly value the maintenance costs as
being a fixed percentage of the initial cost of both the production units themselves, and the
units that directly support their operations (storage and general services).
In the so-called heavy industry sectors (refining, petrochemical, major intermediates, inor-
ganic chemistry, metalworking, etc.), it is customary to reckon maintenance expenses at an
average Of 4% per year of the cost of the plant, i.e., of the battery limits investments, as well
as for general services and storage (see Section 2.3). In fact, this percentage is very dependent
on the kind of products that are processed, being higher when they are considered corrosive
or likely to foul the equipment; it also depends on the type of equipment installed, being
higher here too when the moving parts cause wear on the static ones. Slight differences may
also be produced by differences in the characteristics of the materials employed: the higher
their quality, the better suited they are to their purpose, and thus the more resistant they are.
Ultimately, concrete constructions, which are both static and corrosion-resistant,require only
minimal maintenance. This is why some people recommend using only 3% for general ser-
vices and storage, while still using 4% for the production units. When the products are so
aggressive that special equipment has to be used (internal coatings, traps, etc.) maintenance
costs may reach as much as 10% per year of the battery limits investments. The same applies
to the processing of solid products, particularly minerals, which mainly affect the operation
of rotating machinery, which can be very sensitive to wear and to the fatigue caused espe-
cially by vibration.
However, we must make some minor modifications to this approach to the problem of
maintenance costs:
0 Firstly, technological advances have a tendency to progressively reduce differences in

performance, both of materials and of equipment, by ensuring a minimal working life-

time.
0 In addition, as will be described in detail in Section 2.3, an indirect way of handling the

maintenance of units consists of, on the one hand, providing at the outset for emergency
equipment (particularly pumps) in case of unexpected shutdowns, and including them in
the facilities part of the investments, and on the other hand of adding an item for spare
parts in the calculation of depreciable capital. This is mainly to cover rotating equip-
 2. The Elements of Economic Calculation 153

ment, and its importance appears only when considering anticipated delays in the deliv-
ery of replacement parts.
0 Finally, from a strictly logical point of view, treating maintenance costs as a fixed per-
centage of investments is also not very satisfying. For example, if the units are shut down
for a lengthy period, these expenses may diminish substantially, even if they do not
entirely disappear.

B. Taxes and Insurance

These terms cover the following items:
Various taxes, particularly local and regional taxes, property taxes, license and other
payments made for improvements, environmental protection, etc.,
0 Insurance costs, not only for coverage against damage by fire, flood, weather, etc., of the

equipment itself, i.e., essentially the production units, general services, storage, etc., but
also for the loads of catalysts and other consumable goods, products, and raw materials
tied up and stored in this equipment, and also for damages caused to third parties, to the
environment, and so on. These costs are periodic in nature, and have to be paid at about
the same amount every year, unless significant changes have been made to the manufac-
turing complex being insured.
It is hardly possible to put a number on the respective contributions of each one of these
items to the total amount of these expenses without consulting specialists, looking at a
detailed list of the goods to be used, and considering the particular site. In fact, both taxes and
insurance premiums are extremely variable. They depend not only on the nature and scale of
the facilities, but also on how they are set up and on local tax procedures, which, moreover,
tend to evolve in the course of time. Thus, local, regional, and even national authorities or
governments may grant companies certain tax advantages, and various exemptions and
incentives, with a view to encouraging job-creating industrial developments in areas devel-
oped for this purpose. In other circumstances, however, deterrent measures may be taken, to
thwart developments that are inappropriate, polluting, that produce an increase in nuisances,
or that threaten a site.
Roughly speaking, and in the absence of exact information, we may estimate that the cost
of taxes and insurance is proportional to the industrial importance. In other words, the cost
will rise as the surface area taken up increases, or as the unit manufacturing capacity, which
also controls the scope of the related installations, grows. For this reason, it is customary to
use the investment cost of the facilities (production units, general services, and storage) as a
reference when calculating the annual expense for taxes and insurance, using a fixed percent-
age of around 2% per year.
Sometimes a distinction is made between:
Taxes and general insurance (1 to 2% of the investment cost of the production units)
Obligations of an environmental nature, which at the local level means specific contri-
butions to the existing or proposed and extended processing and purification system
(0.5% of the investment cost of the production units).
154 2. The Elements of Economic Calculation

C. Factory and Head Ofice Overheads

The costs assigned to a factory’s non-productive components or its ancillary services (various
workshops, stores, food services, security, reception, community health, administration, etc.)
or to those of a company (senior and junior management, contracts, licenses, patents,
accounting, etc.) have to be carried by all the activities that lead directly to marketable prod-
ucts, considered to be productive.
Different companies may allocate these expenses in different ways. Consequently, since it
is impossible to look into the accounting procedures of each one, and since we cannot reason-
ably hope to guess what they are, a convenient solution for estimating such costs is to take an
overall view. Here again, we can expect the costs to rise with the scale of the manufacturing
facilities. In this case the easiest reference to use is still the investment cost of the plant (pro-
duction units, general services, and storage). Here also, it is customary to take a fixed percent-
age, of the order of 1% per year.
In some presentations a distinction is made between:
0 The general expenses that are described as direct, expressed as a percentage of the cost

of shift labor (40 to 50%) and often repeated under the labor heading,
0 Related management expenses, calculated in proportion to both:

- Maintenance costs, and

- Shift personnel.
The total (about 60%) thus includes a link to the investment costs for the production units
themselves.

COMMENTS
0 Maintenance, taxes and insurance, overheads, and head office costs represent a

total of about 7% annually of the amount of the investment for equipment. This cal-
culation does not include the cost of the production facilities for the utilities. This
approach makes sense, because these utilities are included in the operating cost and
cost of price, being considered as actual products and billed at their internal transac-
tion prices, i.e., they are assumed to include the fixed operating costs, determined
according to the above procedure. In this case, the reference is composed of the
investment cost of the production unit for the utility concerned, and the cost of the
share of general services allocated to it.
0 The total of the variable costs, labor, and fixed operating costs constitutes the oper-

ating cost as defined in Section 2.2.1.2.

2.2.4.2 Fixed Capital Charges

In contrast to the fixed operating charges, which are directly linked to the actual operation of
the plant, and which reflect real expenses, this type of outlay has a somewhat artificial char-
acter, and arises more from an accounting procedure than from an industrial reality. This
explains why they not only concern the capital invested for the construction of the production
units proper, but also apply to the funds locked up even before the equipment is ready to pro-
duce (see Section 2.3).
 2. The Elements of Economic Calculation 155

A. Allocation to Provide for Depreciation

a. Background

Depreciation measures an investment’s loss of value. By registering the annual charges,

whose total corresponds to the capital invested, excluding working capital, it both enables the
recovery of the original outlay, and spreads the cost over a certain period, called the “depre-
ciation period”.
In fact, the costs for construction and startup of a plant or a piece of equipment constitute
a necessary contribution, but they are preliminary to any actual operation. During the opera-
tion, and with the objective of offsetting the capital initially tied up, we should therefore not
be considering the actual expenses, but rather making provisions, by deducting them from
profits, and allocating them to a particular account. This approach explains the expression:
“allocation to provide for depreciation”. It is based on the following considerations:
0 Unlike a piece of property or a building, whose resale value is usually higher than its pur-

chase price, the values of plants and equipment of an industrial nature diminish considerably
in the course of time.
The deterioration and the gradual wearing out of equipment during its operation leads to a
reduction in its efficiency, which is expressed as an increase in manufacturing costs, by
higher expenses for maintenance, and by unexpected technical stoppages that are more and
more frequent. In this way, its value as a production tool diminishes.
The most recent plants, based on more efficient or more economical processes and tech-
nologies, create difficulties for older ones in terms of their profitability and product quality.
New constructions also benefit from having higher unit manufacturing capacities, the result
both of technological improvements and of the progressive opening of markets. This further
penalizes the existing plants, which are generally of more modest size, through economies of
scale.
0 The products themselves, manufactured under older procedures, are liable to see their mar-

ket shares dwindling away or disappearing, to the benefit of new products that are better
suited to the demand; thus they may face certain problems of distribution and sale. This kind
of situation often appears at the finished product stage: rubber, plastics, synthetic fibers,
detergents, etc.
When we look at these considerations, only the one concerning the wear and deterioration
of equipment seems susceptible of being assessed with any precision. The others, particularly
technological aging and product substitution, appear very hard to quantify. Nevertheless,
when all is said and done, when facing these problems a manufacturer will not be inclined to
invest in building new facilities unless he is encouraged to do so, in particular by compensa-
tory measures. As a response to this concern, in market economy countries the provisions of
the tax code offer the possibility of exempting a portion of profits from taxation, by treating
as an expense a provision for depreciation that is intended to restore the initial capital invest-
ment, over the operating life of the units.
156 2. The Elements of Economic Calculation

However, it should be emphasized that the fiscal legislation imposes certain limitations:
not all investments are necessarily eligible for depreciation, only the ones that fall under the
definition of “depreciable capital” (see Section 2.3). In particular, the following are excluded:
0 Working capital, which has a much shorter rate of repayment than the other investment

costs,
0 The cost of the property on which the plant is constructed,

0 And, generally speaking, all investments that can be recovered or sold at a value close

to the original figure.

b. Operating Period and Depreciation Period

The period of application of the regulatory dispositions affecting depreciation, as well as the
way in which they are applied, also have a greater or lesser incentive effect on manufacturers.
The administrative authorities can in fact make use of this, particularly to allow for the kind
and magnitude of risks incurred: these, moreover, turn out to be different from one sector of
activity to another, and give rise to specific treatments.
Although the depreciation of a unit or a piece of equipment may be known and perfectly
planned ahead of time, the duration of the depreciation should, strictly speaking, cover the
operating life of these facilities. In practice, since it is evidently impossible to measure the
depreciation of goods during each of the operational accounting periods, depreciation can be
estimated only according to pre-defined conventions or provisions, and regulated in the same
way.
Accordingly, two type situations may be encountered:
0 As long as a manufacturing unit is economically viable, considering the costs for mainte-

nance and repair that it requires, and the market restrictions, it may continue to operate. In
this case, we understand that the effective life of this facility may be longer than the legal
period of depreciation. Meanwhile, moreover, improvements may or may not be made
(changes in the equipment, re-sizing of the apparatus, additional pieces of equipment, etc.),
which may demand additional investments and, in consequence, the initiation of correspond-
ing depreciation procedures, which start when the new outlays take effect.
0 On the other hand, it may happen that the unit ceases to be operational even before it is

completely depreciated. In this case, at the end of the facility’s life a fraction of the initial
investment cost remains unrecovered: in economic calculations it is accounted for by using
the concept of salvage value (see Section 2.2.4.2.A.c).
In real-life industry, the period considered for an investment project depends on the antic-
ipated operating life of the equipment, not on the depreciation period used for accounting. For
evaluation purposes, it may be convenient to make the two coincide, and if need be to deter-
mine the salvage value of the facilities just at the end of this time. For all that, we must rec-
ognize a situation that lies between the ones described above, and which may appear
somewhat artificial, but much easier to handle in terms of calculations. In this one, the oper-
ating or project period selected corresponds to the depreciation practices in force in the coun-
try where the equipment is located. Of course, there are significant differences from one
country to another, but there are also variations in legal provisions, depending on the purpose
of the investments. Ultimately, certain variations occur over time, such that in this area it is
 2. The Elements of Economic Calculation 157

difficult to define rules that are applicable in every situation. However, roughly speaking, at
least for the heavily industrialized countries, we may refer to the information in Table 2.16.

Table 2.1 6 Typical depreciation periods for various facilities

Type of facility 1 Depreciation period (years)

Organic chemical unit 10
Refining unit 10
Utility production unit 12 to 15
General facilities 12 to 15
Inorganic chemical unit 15
Manufacturing unit with major civil engineering 20
Industrial buildings 20
Commercial and living accommodation 20 to 50
Furniture 10
Static equipment 10
Rotating equipment 5
Automobiles and rolling equipment 5
(Source: IFP)

COMMENT
The very reason for evaluating projects is to enable the comparison of several of
them, or of several variants of one of them. In some cases, it may happen that the
depreciation period is not the same as the project period. To illustrate this kind of
situation, it is best to choose an example similar to the one already discussed in
Section 2.1.2.3.B, namely, the production of methanol by two different procedures:
1. Startup in 1995 of one large-scale unit with a unit manufacturing capacity of
500 000 tlyr.
2. Startup in 1995 with a first facility of 250000 t/yr unit capacity, and five years
later, namely in the year 2000,startup of a second unit exactly the same as the pre-
vious one.
In the first case, it is possible to take advantage of a better economy of scale than in
the second; however, the second alternative permits a better response to market
growth, which takes place in a relatively gradual fashion. The economic comparison
of these two variants of a single project demands that the project period selected be
the same for both, so that we can work on a uniform basis. However, in this case there
are also two ways of looking at the matter:
Choose a period of ten years for the examination of the two possibilities under con-
sideration. In this case, in the year 2005,the 500000 tJyr unit will be completely
depreciated, and so will the first 250000 tJyr one. However, the second one, with the
158 2. The Elements of Economic Calculation

same capacity, will be only partially depreciated, such that the unrecovered portion
of the corresponding investment must appear as a salvage value.
0 Choose a common project period long enough for the invested capital to be com-

pletely depreciated whichever alternative is considered, i.e., 15 years. In this case,

starting in 2005 the allocation for depreciation will vanish from the calculations of
operating cost or cost of price for the 500 000 t/yr unit; only the allocation for the sec-
ond 250000 t/yr facility will remain to be considered in the context of a staggered
investment schedule.

c. Salvage Value of Facilities

At any moment in the life of a unit, the equipment it comprises may be adapted or sold. The
price of this equipment, which at that moment depends on the use to be made of it (use value),
represents the “salvage value” of the facility concerned. This operation may take place during
the depreciation period or later, so that the calculations must not only include the sales values
related to their potential use, but also the fact, if necessary, that a certain fraction of the
invested capital has not yet been depreciated.
From an accounting viewpoint, where it is a matter of duplicating the actual movements
of funds on the real dates when they took place, and drawing up financial balance sheets at
the end of each year, that is, after the event, the salvage value can be determined only in the
year of sale, if it is sold, or if not, at the end of the project period considered. In the first case,
it is necessary to account for the possible capital gains arising from the marketing or placing
on sale of the equipment recovered (market value), and the taxes or levies that they generate.
In the second, it is desirable to draw up a schedule detailed enough to show the replacement
value of the existing pieces of equipment at each moment, and thus to be able to estimate their
level of aging, or better still, of depreciation (book value or unamortized cost).
In evaluation, where the economic tool is employed in a much more basic fashion, and pri-
marily as a means for comparing competing techniques or projects that utilize notional facil-
ities, the main concern is to perform comparative calculations on uniform bases, particularly
by selecting identical periods or durations for study. It may therefore be useful, on the date
when the procedure is applied, to take into account the probable salvage values of the facili-
ties corresponding to another date, fixed beforehand and in a more or less arbitrary manner.
In practice, unless otherwise indicated, and with a view to simplification, the salvage value
(ZJ at the end of the depreciation period is generally taken to be nil. During the depreciation
period it is taken to be equal to the unrecovered portion of the capital originally invested.
However, it may happen, depending on the calculation method adopted (for example, declin-
ing balance depreciation: see Section 2.2.4.2.A.d), that Z, must be determined. In this case, a
small percentage of the depreciable capital is selected, not more than 2 to 3%, or the total
weight of metallic elements in the unit is estimated, and valued at the price of scrap metal on
the date of the study.

d. Calculating Depreciation
Various methods are in use for determining the depreciation provision. The most widely
employed are the ones called straight-line, and declining balance. But some fiscal regulations,
 2. The Elements of Economic Calculation 159

particularly in the USA, permit the use of other methods, such as depreciation by linearly
decreasing annuities (Sum of the years’ digits method), or by depreciation funds or constant
economic depreciation (Sinking fund method).

Straight-Line Depreciation
This method, which was the one most commonly used before 1959, consists of recording,
each year, the same percentage of the initial “depreciable” capital, a term to be defined later (see
Section 2.3). Thus, if I represents the original outlay, and n (years) the duration of the depre-
ciation and assumed operation of the facilities, the annual depreciation cost is constant, and
equal to:

A = -I
n

Under the simplifying assumptions made, the salvage value of the equipment (I,) is
assumed to be zero at the end of the period concerned.
The depreciation rate in this case is equal to:

1
u1 =n
It has the same value each year, and enables the complete recovery of the nominal value
of the starting investment in the n-th year.
Since n is usually ten years, ul will be 10%. For accounting purposes, the depreciation
begins on the day the equipment is first used. To provide for the exact time elapsed between
the start of production, or of effective utilization of the facilities, and the end of the project,
the first and last annual amounts are calculated on aprorutu temporis basis.

Declining Balance Depreciation, or Constant Rate Method

This kind of procedure expresses the need for a depreciation that varies over time, but is
greater during the early years of a facility’s operation, thereby conforming more closely to
economic reality. This approach has a double advantage:
0 It reduces the risks connected to rapid obsolescence of the technology or the equipment

employed, when facing increasingly severe competition and exponential progress in

technology, by allowing quicker recovery of the sums invested during the period when
the equipment is, on the face of it, more efficient, and the markets most favorable.
0 It diminishes the share of gross profits-the difference between revenues and expenses-
shown during the early years, and thus correspondingly reduces the income tax burden.
This method of calculation tends to encourage the reinvestment of recovered funds more
rapidly than does straight-line depreciation. It also makes it easier to create jobs and to
put innovative techniques into operation. In doing so it constitutes a legal way of partially
avoiding taxation; consequently, its application lends itself to specific measures that may
differ over time and from State to State, depending on the economic situation and its
imperatives.
160 2. The Elements of Economic Calculation

The principle of this method assumes that the annual depreciation is a fixed percentage of
the value of the facilities at the start of the year. If ad is this constant factor, the fraction depre-
ciated during the first year is: ad (Z- Z,.),and the value of the equipment is:
at the end of the first year : (I- Zr)(1 - ad)
at the end of the second year : (Z- Zr)( 1 -
__________--_____________
at the end of the n-th year : (Z- Zr)( 1 - ad)n

At this point, the value of the unit corresponds to its salvage value Z,.(the capital not yet
depreciated, or the resale value, or the re-use value).

COMMENT
In practice, I,. is not cancelled out until the end of a time period that is infinite in com-
parison to the duration n of the project study or of the depreciation. In fact, if the
equipment stops operating at the end of this time, the items that are still functional may
be re-installed at other facilities, and so on until they can only be used for scrap metal.

From the expression: (I-Zr)(1 - ad)n= Z., we can deduce the declining balance depreciation
rate:

a,=1- [ -
Z q n

In this case, in any year p, such that 0 I p I n, the annual provision for depreciation may
be written:
A, = (I - I, 1)- ad)P-*
( ad

As shown in Fig. 2.10, the annual provision is higher under declining-balance than under
straight-line depreciation, during the first years of a comparison; on the other hand it is lower
later on. The recovery of the original outlays is incomplete, since the curve representing its
variation approaches the time axis asymptotically. In addition, this approach may lead man-
ufacturers to employ or call for a higher rate of decline, depending on their specific situations.
The usual trend is of course to “hold” rather than “run”, i.e., to depreciate as quickly as pos-
sible, as long as revenues permit it.
This shows that, here again, the regulatory provisionswere designedto restrict tax deductions.
Thus until 1954, in the United States, it was not possible to have a declining-balancerate
of more than 150% of the straight-line depreciation rate. After that date ad was allowed to
reach a value of double the straight-line rate (“double declining balance”).
In France, article 37 of the law of December 28 1959, and the accompanying regulations
of May 9 1960,provided for a declining balance system based on the lifetime of the facilities:
less than 3 years a d = 1 x a1
from 3 to 4 years ad= 1.5 X
from 5 to 6 years ad= 2 x
more than 6 years ad=2.5 X U ]
 2. The Elements of Economic Calculation 161

0 1 2 3 4 5 6 7 8 9 10
Years

Figure 2.1 0 Straight-line and declining-balance depreciation.

(Source: IFP)

In other words, the largest part of the equipment of an industrial nature is likely to receive
the highest values. Moreover, in 1989 these provisions were strengthened by the addition of
accelerated declining-balance depreciation, which was authorized for investments for saving
energy and economizing raw materials. Beyond the three-year mark, all the figures shown
above were increased by 0.5: for industrial chemical and refining facilities, for example, this
method allowed a declining-balance depreciation rate of three times the straight-line one.

COMMENT
If ad is greater than the straight-line rate, at the end of the depreciation period I, is
theoretically not zero. In some cases, particularly in evaluation, it might be desirable
to apply a constant depreciation rate, which implies a declining-balancemethod, and
nevertheless recover all of the funds originally invested, even if additional income or
capital gains appear should the equipment be re-used or re-sold at the end of the
period. To answer this type of request, the government allows the use of combined
methods: declining-balance for the early years, and straight-line for the later ones.
Actually, from a legal viewpoint, when the annual amount of a declining-balance
depreciation falls below the straight-line amount for the remaining residual value,
162 2. The Elements of Economic Calculation

i.e., when the slope of the tangent to the declining-balance curve is parallel to the line
representing the straight-line depreciation, it is possible to finish up by using the lat-
ter method. In practice, the delays involved in setting up the accounting for this
mixed system make it just as if the straight line that is followed were to join the curve
tangentially at yearn on the time axis.

Linearly Decreasing Annuities (Sum of the Years’ Digits) Method

Because of a calculating trick, this resembles the constant rate method. If n is the duration
of the depreciation, the first annual amount is proportional to n, the second to n - 1, and the
p-th to n -p + 1.
In other words, the sum of the first n whole numbers being equal to n(n + 1)/2, the annual
amounts are as follows:

2(n- p + 1)
A, = (z-zr)
n(n+1)

which amounts to calculating the annual amount for a given year by subtracting a constant
amount corresponding to A , from the previous year’s figure.

Constant economic depreciation (Sinking fund) method

This procedure consists of working with constant annual payments, by using the economi-
cally discounted value of the transfers of funds, particularly the repayments. By definition, at
the end of the depreciation period the total of the sums recovered and discounted must be equal
to the value, in the initial reference year (year 0), of all the investment costs, including possible
re-sales of equipment, i.e., their salvage value. The resulting equation is therefore the following:

where A represents the desired economic depreciation constant for the period n, i being the
discount rate used by the company responsible for the project concerned, so that:
I-I,
A=
 2. The Elements of Economic Calculation 163

i( 1+ i)"
or: A = ( I - I,)
(1 + i)" - 1

A is the constant annual amount by means of which the initial investment could be repaid
over n years, if the investment had been borrowed at an interest rate equal to the discount
rate i.
The economic depreciation annuity is higher than that for an accounting depreciation
spread over the same number of years, since it contains both a portion covering the depreci-
ation of the facilities proper, and also another portion for the repayment of capital via the dis-
count rate i. In addition, since the annual payment A is constant throughout, and the share for
the financial costs necessarily decreases over time, this type of method leads to a slower
recovery of initial investments than does the straight-line method, during the early years. The
risks incurred by the investors are therefore greater, which to some extent justifies their reluc-
tance concerning such a procedure. On the other hand, it is just the reverse for the lending
organizations.

B. Financial Costs (Interests)

Financing a project means locking up sums equal in amount to the overall initial outlay, i.e.,
the so-called depreciable capital and the working capital (see Section 2.3).
As discussed in Section 2.1.2.3, in market economy countries these investment costs give
rise to repayments, via:
0 Either a real interest rate when they are obtained by loans from banks or lending agen-

cies,
0 Or a pseudo interest rate, also called a psychological interest rate or a discount rate, when

they concern the internal funds of the company running the project, or its shareholders,
or when the source of capital is unknown, or even assumed to be.
For empirical, after tax methods, such as the POT or the ROI, (see Section 2.1.2.2), it is
just as if the full amount of the initial sums required had been borrowed from the company's
treasury, which would then charge the project for real or notional financial costs calculated
on the basis of an interest rate peculiar to itself, and equivalent to its discount rate.
The procedure used for this is that of continuous compound interest, as discussed in
Section 2.1.2.3.A.a, which involves the concept of effective or declining interest i,, which
should apply only to the sums remaining to be repaid or recovered, i.e., in any given year par-
ticularly the portion of the investments not yet depreciated. Also, when a straight-line depre-
ciation is adopted for the purpose of simplifying the evaluation calculations, it leads to a
similar rule for the financial costs. But just as it is handy to have a constant depreciation rate
for the depreciable capital concerned, it may be useful to take a similar approach concerning
the effective or notional interests that must be considered, both for this same capital and for
the working capital. For this reason, rather than applying a constant interest rate to sums that
are declining in a straight-line manner, it is often preferable to apply, by means of a calculat-
ing device, a constant average interest rate to the clearly identified initial outlays. This aver-
age value ,i is derived from the effective value i, by means of the following equation:
164 2. The Elements of Economic Calculation

.
1,
. n+l
= lr -
2n
n being the assumed length of time during which the sums concerned are locked in.
Thus, in the case of depreciable capital for which in general n = 10, if i, is taken to be 12%
per year, we obtain:
im = 6.6% i.e., = 7%
For working capital, for which the renewal period is usually of the order of two years:
im= 9%.
It must be remembered that this calculating trick has no accounting or economic justifica-
tion, and is accordingly rejected by the corresponding authorities. On the other hand, it is very
valuable when there is a need to obtain an order of magnitude by incorporating the component
that expresses the cost of tying up capital.

2.3 INVESTMENT CHARGES

Examining the profitability of projects by applying economic criteria, as well as the determina-
tion of the amount of the fixed charges included in the operating cost or the cost of price,
requires that one first possess figures for the depreciable capital and also for the working capital.
In practice, moreover, this overall approach to the investments authorized for the construction
and operation of new plants or equipment is inadequate in itself. It appears necessary, in fact, to
look at the individual items, i.e., the various investment charges that addup to these investments.
The purpose of the present section is therefore to analyze them in detail, more particularly
by emphasizing the basic element of such a breakdown: the battery limits investments.

2.3.1 DEFINITIONS AND PURPOSES OF

VARIOUS INVESTMENT CHARGES

Depending on the company concerned, on whether it belongs to a productive sector or the ser-
vice and consulting side, and also according to the current fiscal provisions applicable to a cer-
tain project, the composition of the depreciable capital and working capital may be defined
differently. So it is not surprising that there may be differences both in the items presented and
in the partial groupings that are made of them. However, as a general rule, it seems that a con-
sensus can be reached about the major topics that are currently likely to be considered here.

2.3.1.1 Type Breakdown

As diagrammed in Fig. 2.1 1, the principal investments, or investment costs that a manufacturer
has to accept, even before the corresponding facilities or equipment are in a position to operate
and generate profits, may be grouped under four main headings, each one nested inside the next:
 2. The Elements of Economic Calculation 165

The investment cost for the facilities proper (unit investments), comprising:
- The battery limits investments (on-site or on-plot investments)
- General services and storage (off sites or off plot),

Fixed capital, comprising:

- Investments related to the units
- Consulting and engineering fees

- Supply of spare parts

- Contractors’ costs or licensor’s fees,

0 Depreciable capital, whose main components are:

- Fixed capital
- Initial loads

- Interest on construction loans

- Startup costs,

Figure 2.1 1 The various investment charges. (Source: ZFP)

166 2. The Elements of Economic Calculation

Total investments, including:

- Depreciable capital
- Working capital.
This presentation relates to plants constructed on industrial sites that are assumed to be
available and ready for use. They particularly concern the integration of a new unit into an
existing complex, or more generally of a piece of equipment into an operating production sys-
tem. It may happen that the operation is designed for a remote area: in this case a fifth heading
must be considered:
0 Total grass-roots investments, comprising:

- Total investments on an industrial site

- Landcosts
- Yard improvements.
This type of analysis does not attempt to make an exhaustive list of the various headings to
consider when making a comprehensive calculation of investment charges, nor even strictly
to cover all their allocations. By way of illustration, we may mention as an example that cer-
tain companies sometimes include the costs of the first facility, when the technology or the
equipment employed has not yet been validated or tested industrially. Another example con-
cerns the initial loads of catalyst, which, depending on their application, may or may not be
considered depreciable. While not wishing to add to the list of exceptions, it is nevertheless
useful to draw attention to the need to maintain a critical attitude towards the procedures sug-
gested, which must be adjusted to suit local conditions or changing times, and whose principal
merit is to provide a framework for applying the principles or for thinking about the problem.

2.3.1.2 Investment Costs of Production Facilities

On an industrial site of typical size, four major types of facilities are generally to be found:
0 Production units proper

0 Facilities for providing and distributing utilities

0 Storage facilities

0 Services and buildings whose use is shared, called general services.

At workshops and smaller industrial platforms, these same entities may still be identified,
but they are usually well integrated into the manufacturing process itself. In such situations
we can justify limiting the investment cost of a plant to the outlays for the battery limits
investments, particularly if it produces only one product. In practice, this is another case
where it pays to look critically both at the information being processed and the application of
the suggested procedure to the particular situation.

A. Battery Limits investments

In theory, these concern only the money invested in the truly productive components of a
manufacturing complex, thereby excluding ancillary equipment and systems, which may of
course be essential to the units’ operation, but are not specifically assigned to a particular pro-
duction line. In fact, the battery limits investments correspond to the erected costs of the
equipment, i.e., they cover the primary and secondary equipment items, civil engineering,
 2. The Elements of Economic Calculation 167

erection, connection, and indirect workplace costs, whose definition and contents are given
in Chapter 3. Provisions for accidental or unforeseen expenses are often included.
Sometimes the costs of planning work are also included, particularly engineering studies.
To make calculation and presentation easier, it is preferable to separate them and, depending
on the approach suggested, to show them under fixed capital costs.
As shown in Fig. 2.12, at a given industrial site of a certain size there will generally be sev-
eral sub-units for manufacturing or conversion, arranged in series or in parallel, whose activ-
ities are complementary. This complementarity is technological in nature when each of them
constitutes an essential link in a chain of processing or a production line; it becomes eco-
nomic when their purpose is to supply several markets at once, with different natures and
characteristics, from the same raw materials. By identifying the battery limits investments,
we thus mean to isolate the sums that concern each one of these steps only. Looked at in this
way, their geographic limits or boundaries become the locus of:
“Imports”, shown at their respective access costs, i.e., as discussed in Section 2.2.2, at
their market or internal transaction prices, and meeting precise specifications, either
commercial or as required by the operation concerned:
- Raw material(s),
- Common utilities: electricity, steam, various waters (cooling, process, and boiler
water, etc.), fuels (gaseous, liquid, or solid), refrigerants, industrial gases (com-
pressed air, inert gases, etc.),
- Chemical products, catalysts, solvents, absorbents, sieves, etc., depending on require-
ments,

raw materials + products (co-products)

STORAGE 4 I ,STORAGE

STORAGE
chemicals
4 i + by-products
I
,STORAGE

+ used chemical products

I
,WASTEWATER
TREATMENT

utilities -+ exhausted utilities

RODUCTION OF UTILITIES4
f PRODUCTION OF UTILITIES
~

GENERAL SERVICES (administration, benefits, cafeterias, stores, workshops, etc.)

Figure 2.1 2 Schematic diagram of a manufacturing complex.

(Source: IFP)
168 2. The Elements of Economic Calculation

0 “Exports”:
- Products manufactured or processed in the facilities, meeting particular marketing or
internal specifications,
- By-products, organic or inorganic, generated by the main transformation or ancillary
processes, possibly including combustible gases, tars, various residues and wastes,
residual waters, etc., which may possibly be sold or disposed of, but which first
require some additional treatment,
- Utilities described as “exhausted”, because their exit characteristics differ from their
entry ones (for example, cooling water, refrigerants, etc.); but here we must consider
the corresponding by-products, particularly condensates, fumes, and various resi-
dues. Each of them must be either recovered or rejected, but in both cases they require
additional treatment, which for reasons of economy of scale are generally performed
together on all the effluents of the same kind emanating from the various manufac-
turing or conversion stages.
The battery limits investments specifically do not cover the stocks of raw materials or fin-
ished products. However, it may happen that specific modifications have to be made to a
facility (particularly cyclic systems or semi-continuous operations) and that these lead to
intermediate or buffer stocks being established. In these circumstances it is customary to con-
sider such items as integral parts of the manufacturing or processing procedure, and to treat
the investments accordingly.

B. Facilities for the Production and Distribution of Utilities

The different sub-units that make up the various conversion stages at a manufacturing site
cannot be viable as such, whether they are considered separately or even as a whole. To
become operational, they must be provided with accompanying “services”, whose primary
characteristic is to be as general as possible, i.e., to meet the requirements of the majority of
the units and to avoid being too specific. The most conspicuous examples of this kind of facil-
ity, because of their very applications, are the facilities for generating utilities and the corre-
sponding distribution circuits. In spite of their service functions, these entities have a special
role, which is expressed in a specific manner of treating them for economic purposes, partic-
ularly in terms of investment.
In fact, there are two ways of looking at them, which actually correspond to the three most
common situations:
0 The first, which is also the one usually encountered, concerns existing industrial sites,

already provided with services, where it is a matter of examining the effects of inserting
new sub-units or replacing such facilities with more efficient equivalents.
0 The second concerns industrial complexes that are totally new, where we must not only

consider the production facilities themselves, but also provide everything required for
their operation. This particularly happens in industrializing countries.
0 A third situation, which in its investment consequences is comparable to the preceding

one, and requires the same economic treatment, is the one where the insertion of new
components creates disturbances because of their size, so that it becomes essential to
construct service facilities that are wholly devoted to meeting their needs.
 2. The Elements of Economic Calculation 169

To address the first situation as simply as possible as far as the utilities are concerned, it is
best to treat them like raw materials or chemical products. In this case, in fact, we regard them
as inputs from outside the system studied, available under particular conditions, both of a
technical (specifications) and an economic (market or internal transaction price) nature.
These prices, which are derived from a calculation method similar to the one presented in this
chapter, or determined by the laws of supply and demand discussed in Chapter 1, particularly
incorporate the investment charges that are required for the manufacture of the desired “prod-
ucts’’ to the required specifications. In other words, as long as the utilities are accounted for
at their market or transaction value, under the operating cost or cost of price, there is no need
to add in the investment cost for obtaining them. Otherwise, it would amount to counting the
same capital costs twice, by entering them at two different places in the operating cost:
0 As fixed costs, determined by the overall amount of the investment for the units (includ-

ing those required to produce the utilities),

0 And as variable costs, which include the costs of utilities, here charged at a price that

already includes the investment charges corresponding to their manufacture.

As for the other two situations, the opposite approach is the most realistic. It consists of
drawing up a list of the requirements for utilities, and operating as if they were “products”
entirely obtained in situ. In this way, the overall battery limits investments are obtained by
adding together the investments for the processing facilities themselves, and their equivalents
for producing utilities. But, in line with the observations made for the first situation, when
determining the operating cost or cost of price one must be careful to include only the con-
sumption of utilities that relate to the energy and basic materials used in their production. In
other words, although the strict definition of the manufacturing facilities may be extended to
cover the production of utilities, it is best to maintain these same boundaries for capital and
operating cost.

C. Storage

The essential purpose of storage is:

To maintain sufficient reserves of raw materials, reagents, catalysts, solvents, etc., and
generally speaking any input delivered in bulk, so as to avoid interruptions in supply to
the unit concerned. These flows, which are imported into the facilities, come either from
manufacturing platforms on the same industrial site, but earlier in the processing chain,
or from outside suppliers. During a given period, storage provides remedies: firstly for
short-term shutdowns of the units providing these inputs, whether scheduled or not; and
secondly for delays or incidents of a technological, social, climatic, or other nature, that
are likely to affect deliveries,
0 To temporarily retain finished products, by-products, even certain wastes, before they

are shipped, either to other facilities in the manufacturing complex concerned, or to cus-
tomers, or are possibly removed to the appropriate treatment centers, which may func-
tion semi-continuously or in batch mode. Here again, the stocks act as buffers in the
event that the downstream units and the distribution system are temporarily unable to
function properly.
170 2. The Elements of Economic Calculation

As a general rule, companies try to limit the size of their storage, for economic reasons.
Moreover, the trend is more and more to work on a just-in-time basis, and this includes high-
production facilities. The fact remains that, without pushing matters to such limits, manufac-
turers do attempt to reduce the number of drums and tanks allocated for such purposes, by
allowing them some flexibility in utilization. Thus, these items are not used for a single facil-
ity only, but, as long as the kind of content remains the same or changes very little, they are
made to serve several potential users in parallel. This approach explains why their investment
cost is usually separated from the costs of the various production units proper, and why in
practice only a portion of it is allocated to a particular facility.
The share that must thus be charged actually depends on the internal structure of each oper-
ating company. If it concerns the construction of a new industrial complex, it is easy to cal-
culate the amount of storage required, and to distribute it among the production platforms
according to a rule fixed ahead of time, for example by making the various products, whether
finished or not, carry the correspondingexpenses, either in proportion to their respective sales
or to the battery limits investments.
If it concerns the installation of a new facility on an existing industrial site, we must take
account of the equipment in place, their capacities, and the possibility of enlarging them or
the need to construct new ones.
For a specific industrial project, an exact calculation can be made, but in evaluation such
details are not specified: in particular, the place where the unit is to be set up is often
unknown. So we must fall back on empirical rules. Thus, roughly speaking, it is accepted that
the cost of the investment for storage is of the order of 25% of the cost of the battery limits
investments. Such a figure amounts to having a margin of eight days of raw materials or major
products.
Obviously, this figure must be raised if special legislations require longer durations: this
is the case for strategic products, such as those derived from petroleum.

D. General Services and Common Facilities

If storage facilities are essential for the proper functioning of the production units found on
an industrial site, in the same way a complex cannot be operational in the absence of a mini-
mal infrastructure. These shared improvements and facilities, called “general services or
common facilities”, whose investment cost is far from negligible, cover the following main
headings:
0 Servicing the site:

- Site preparation
- Roads and access
- Railroad tracks
- Loading and unloading docks, ancillary handling equipment
- Fences and security
- Lighting
- Parking lots
- Landscaping, etc.
Buildings to house materials, when they are not included in the battery limits investment
envelope, sheds, warehouses, etc.
 2. The Elements of Economic Calculation 171

0 General services proper (buildings and materials or furniture): workshops, stores, labo-
ratories, food services, community health services, safety, administration, management,
etc.
0 Centers for purification or waste treatment, particularly fumes, wastewaters, solid
wastes, etc.
0 Systems for sewage, blow-down, flaring, etc.
0 Holding tanks, retaining walls, etc., and other systems for safety, environmental protec-
tion, observance of prevailing regulations, etc.
This list, which is certainly far from complete, shows-in case there was any need to do
so-the inevitable nature of the additional investments that a manufacturer is obliged to con-
sider, over and above the cost of the actual production facilities themselves, and even before
this equipment becomes operational. Since they produce no revenues themselves, for finan-
cial purposes these ancillary investments have to be charged to the manufacturing units 4.
But, as with storage, the rules for allocating these costs are in practice very dependent on the
internal structure of each company.
However, here again it is possible to separate the case of a new facility being inserted into
an existing industrial complex, from that of the construction of a complete plant. Attention
must also be paid to whether or not it is necessary to modify or to add to what is already there,
when the disturbances produced by the introduction of production platforms completely dis-
organize the existing services. This type of situation is relatively rare, in fact, since from a
size point of view, general services facilities have much less influence than, for example, stor-
age buildings.
Nevertheless, the empirical approach to the investment cost of general services consists of
calculating it in proportion to that of the production units, at a rate of about 15%.
This approximate rule is based on the fact that the more costly a conversion or processing
facility is, the bigger it is, the more site space it takes up, the more employees it has, and so
on, and the greater the burden it places on outside services.

COMMENTS
0 General services and storage taken together represent about 40% of the battery

limits investments. Such an average figure reflects current practices in sectors such
as refining (not including strategic stockpiles), inorganic and organic chemistry, spe-
cialty chemistry (drugs, paint, ink, photography, etc.), food processing, metalwork-
ing, etc.
It is sometimes different in other areas of activity such as nuclear technology, auto-
mobile manufacturing,etc. Accordingly, arbitrary generalizations should be avoided.
0 General services and storage are often grouped together as “offsite” items, as

opposed to “onsite” investments, corresponding to the battery limits costs. In engi-

neering companies the terms “off plot” and “on plot” are also employed to denote the
same things.

4. Note that the costs of operating these common services are charged against operating cost under
the item “factory overhead”.
172 2. The Elements of Economic Calculation

2.3.1.3 Fixed Capital

Apart from the investments for the production facilities themselves and the infrastructurethey
require, fixed capital covers other expenses that a manufacturer must incur as a preliminary
to any commercial operation, and consequently before there can be any possibility of repay-
ment of the sums locked up. These particularly include the following main outlays.

A. Project Study and Engineering Fees

Before undertaking the actual implementation of a project, a company is first obliged to con-
firm the technical and commercial validity of its plans. In addition, it must estimate the finan-
cial risks incurred, and for this purpose define the budget envelope to devote to the operation,
even if it is only to envisage the financial setup for acquiring the necessary funds, in the most
common case where it does not have enough to cover the total itself. This means preliminary
studies, called feasibility studies, which require consultation with the owners of processes or
technologies. It may conduct these studies itself; however, for want of the necessary skills or
personnel, companies generally assign analyses of this type to specialists, particularly con-
sulting companies or similar departments in engineering companies.
When the preliminary work has been completed and the construction decision has been
confirmed, the company must go ahead with the actual implementation of the plant. More
detailed studies of sizing and costing become essential, leading to accurate budgets (see
Chapter 3). The next stages are invitations for tenders and the placing of orders with equip-
ment manufacturers, then onsite activities involving contractors from a variety of occupa-
tions. Procedures have to be set up both to schedule the activities of each of these, and to
provide for the monitoring and control of their services, so as to ensure proper coordination
and management of the planned work. At the end of the process, the company also has to be
able to take over the facilities and to approve their performance during startup tests.
All these operations have heavy impacts on the financial level: they require qualified
people and for this reason are very difficult to carry out internally. For this type of activity,
even more than for feasibility studies, manufacturers are obliged to turn to outside organiza-
tions.
Engineering companies, whose job it is to answer such needs, act like project managers
when they are awarded a project construction contract, playing a role that is comparable to
that of an architectural firm. Moreover, their coordinating activities often lead them to sub-
contract and thus to supervise some of the tasks that are given to them. This is particularly the
case when some of the services go beyond their own area of expertise, and even more so when
the work plans and the schedule become too tight.
The engineering fees represent repayment for all the services provided for the manufactur-
ers. They are responsible for them because when all is said and done these fees replace the
costs of work that would normally have been done internally-less efficiently, and conse-
quently at much higher cost.
In theory, their amount depends on the magnitude and the quality of the services supplied,
and on the nature and capacity of the facilities provided. In practice, it is very difficult for an
outside observer to have a clear picture of them. Indeed, considerations of a commercial
nature often interfere with an exact reckoning, which is itself tied to the preliminary budget
 2. The Elements of Economic Calculation 173

estimate that may have been made by the engineering companies during the initial feasibility
studies (see Chapter 3). Thus, when compared to a theoretical estimate, the figures proposed
to potential clients may either be significantly lowered, or on the other hand raised. The
explanation for such fluctuations may be attributed to the very nature of service activities,
which depend on the vagaries of the economic situation, on very intense competition, and
often on pressures of a political nature. In such circumstances, various situations may arise,
among which two extreme cases are found:
The first applies to an engineering company whose order books are half-empty and whose
personnel are under-employed. Rather than lay off trained staff who may be desperately
needed later on, when better times come along, the company prefers to keep a minimum level
of operations going. For this purpose it will make every effort to secure the contract by cutting
its prices, i.e., by keeping its operating budget as low as possible, even by under pricing, and
of course by reducing its own fees correspondingly.
0 The second case concerns an engineering company whose order books are full, and whose

personnel are over-worked. In preliminary tenders, it will tend to quote the real price, even to
over-price. In fact, if its reputation acquired in previous dealings leads to its being selected in
spite of its rates and its conservative budget forecast, it will have to hire additional personnel
and quickly train them, or sub-contract, while still looking after the supervision of the work
and guaranteeing it to its client, using highly qualified experts. In both cases, there will be
additional costs that explain the higher fees that are demanded.
Regardless of these practical considerations, which show how dangerous it is to treat the
prices in contracts as if they represented the truth, as certain publications sometimes sug-
gest-and a similar caution must apply to figures obtained through personal contactsder-
tain empirical rules do nevertheless allow a more accurate estimate of engineering fees.
The first alternative consists of relating them to the costs of the equipment that is being
analyzed or monitored, i.e., to the pieces of equipment that require drawings, size calcula-
tions, quotes, etc., orders, receiving, assembly, and so on. In this case it is a matter of taking
a given percentage of the total of the battery limits investments and the investments for ancil-
lary infrastructure (general services and storage, perhaps production of utilities, depending on
the procedure adopted). An average figure, generally accepted and verified, is 12%.
In fact, the engineering fees are not directly proportional either to the unit manufacturing
capacity, or to its complexity. They become smaller as the scale increases and the facilities
become more complex, and vice versa. In practice, it may be best to choose a compromise
between values lying between 7 and 30%, depending on whether, for example, steam crack-
ing or selective hydrogenation is being considered, a rolling mill or a specific refining pro-
cess, an automobile assembly line or the production of replacement catalytic converters, etc.
A second alternative, which is more accurate and more realistic, requires that we first be
in possession of a schematic diagram of the production facilities, showing all the main items
of equipment (including pumps). Storage and general services are not involved in this
method, which consists firstly of counting the items in the manufacturing units only, and then
applying the product of an average number of hours worked (calculation, drafting, monitor-
ing, checking, supervision, etc.) and an average hourly labor cost for the consulting firm,for
each company concerned. As a first approximation the following values may be accepted:
174 2. The Elements of Economic Calculation

0 1500 consulting company hours per item

0 55 to 75 €/hour (mid-2000, including benefits and related costs),
i.e., for a facility of average size comprising 100 items:
8 to 14 lo6 € (mid-2000).

B. Spare Parts Stock

Any new plant must have a stock of spare parts to deal with the mechanical incidents that are
likely to occur. In some cases, this stock can represent a considerable investment, for example
when the equipment is subjected to severe abrasion or repeated strains that produce rapid
wear, or when the unexpected shutdown of a particular piece of equipment, whose replace-
ment involves a fairly long delivery time, could paralyze a whole production line and thereby
even threaten the profitability of the project.
In heavily industrialized countries, the problem is not so critical, and the investment for
this item is generally quite small, even nil. In fact, the most sensitive of the essential pieces
of equipment are manufactured on neighboring industrial sites or in surrounding countries, so
they can be shipped quickly. In addition, the most needed equipment happens to be machines
with rotating parts, and especially pumps, whose unit costs are generally modest when com-
pared with the rest of the apparatus. For this reason engineers provide for some duplication,
even on the diagrams for the manufacturing or processing facility. In practice, since reliability
has significantly improved over time, we do not systematically duplicate every pump in a
facility: the current trend is rather to provide one spare for every three or four operating items,
as long as their characteristics are fairly similar. In the case of compressors, rotating filters,
centrifuges, etc. and other more costly rotating machinery, sometimes two identical pieces of
apparatus may operate in parallel, each one, for example, at 60% of its nominal capacity,
rather than having a single, larger one running at full load. This gives the manufacturer some
additional flexibility in case market demand requires greater production than the original siz-
ing provided for, and as long as it does not run into any additional bottlenecks. There is a
reduced risk if one of the two pieces of equipment breaks down, since the other can continue
to supply most of the customers. This way of operating, applied both to pumps and possibly
to other equipment, implies that a stock of spare parts will automatically be included in the
amount of the battery limits investment.
For industrializing countries, which import most of their equipment, it is sometimes best
to provide a significant reserve of components and complete pieces of equipment from the
outset. There are several ways of determining how large it should be. One consists of sepa-
rating the part of the battery limits investment that represents rotating equipment, excluding
pumps, from the part representing static equipment, and then allowing for an outlay for the
spare parts stock equivalent to the money spent on the first type of equipment. Another alter-
native resembles the one for industrialized countries, by favoring an increase in the battery
limits investment to allow for two identical units with a nominal size greater than 50% of the
required output, instead of a single unit covering the whole requirement. A third possibility
is to take an arbitrary percentage (for example, 30%) of the battery limits investment. This
idea is based on the fact that certain items of static equipment (tubular reactors, cyclones,
exchangers, boilers, etc.) may also be affected by problems of wear or premature failure.
 2. The Elements of Economic Calculation 175

In practice, at the end of the day it is up to the evaluator to make the best possible judgment
of the situation and to decide on the measures to be applied. This is where an evaluator’s
expert knowledge and professional skills come into play.

C. Contractor Fees
Generally speaking, the engineering company selected by the manufacturer to construct the
project under consideration is neither the designer nor, more broadly, the inventor, of the pro-
cesses, technologies, and equipment, etc. that its calculations are based on. In other words, the
R&D work that leads to the innovations, and the work of putting them into commercial practice
as industrial facilities,usually come from separate organizations.This leads to interface problems:
0 Firstly, economic ones, concerning the proper remuneration for research work,

0 And then technical ones, involving the adequacy of the formal expression of the labora-

tory or pilot development results, so that they can be conveyed to and properly applied
by the engineering company.
COMMENT
Until recently, many companies operated their own in-house R&D departments.
Even where this type of arrangement still exists, the implementation stage is taken
care of by an outside engineering firm, which in no way reduces the problem of trans-
mitting and processing the basic information. In addition, the laboratory and the
operation itself, even though they may belong to the same organization, represent
two different worlds, each one seeking recognition for its own contributions. To
answer such demands, it would seem essential, even in-house, to include a notional
payment for R&D in the accounting.

a. Ways of Paying for R&D

Generally speaking, and leaving aside each company’s own particular methods for remuner-
ating its own basic work and promoting creativity, the company that does the research, devel-
opment, and perfecting of a process, a technology, or a product for industrial use in a project,
will try to protect the results of its work by obtaining and maintaining patents. It holds the
right to exploit them, but it may transfer this right, either exclusively or not, and for a partic-
ular geographical region, against payments. This repayment is effected by the collection of
royalties, which can be of two kinds:

Paid-up Royalties
These are lump sums paid to the owner of the technologies, by the manufacturer wanting
to apply them, in order to acquire the operating license. They are usually linked to the size of
the facilities envisaged, and proportioned to the tons of unit capacity of production or pro-
cessing. Depending on the kind of technique concerned, they vary from a few Euros per ton
of installed capacity for mass-market products and for less complex applications (for exam-
ple, refining processes), to several hundreds of Euros for the most sophisticated manufactures
and for the most specialized markets (pharmaceutical products, for example).
Moreover, a declining tariff may be applied, with continuous or stepwise reductions,
which takes into account the effects both of scale (the rates diminishing as the nominal unit
176 2. The Elements of Economic Calculation

capacity grows) and of the progressive aging of the invention (catching up by the competi-
tion). In some cases, especially when the first industrial use or a demonstration plant is con-
cerned, special terms may be granted, which may go as far as royalty-free operation.
The payment terms are set by contract. The general rule is not to make an overall payment,
but to proceed by installments. For example, a portion (1/4) is paid upon signature of the
agreement assigning the license, another (1/4) during construction of the facilities (at the
beginning or the end), and finally the remainder (1/2) is received after handover testing of the
unit, during which the attainment of the advertised performance is verified.
In this connection, it should be mentioned that the owner of the process or technology is
liable to penalties, deductible from the amounts owed, if the advertised results are not obtained.
However, the risks are reduced by having the contract stipulate guaranteed performances that
are generally less than those actually obtained. In practice, both the user and the licensor make
a joint effort to operate the facilities as far as possible in the manner originally intended, iden-
tifying possible causes of malfunctioning and making the necessary changes when the circum-
stances dictate. The resolution of financial disputes takes place only afterwards.

Running Royalties
Sometimes, by agreement with the holder of the process, or, more broadly, with the owner
of the invention, the payments are made in the form of annuities over a period related, for
example, to the depreciation period of the operating plant, or the life of the patents that protect
the original aspects of the innovation. These sums obviously incorporate financial, insurance,
and administrative costs, etc., connected to the way these arrangements are currently handled
by banking institutions.
Running royalties are so called because they relate to the actual production of the facilities
and not to their unit capacities, as in the payment method discussed previously. For this rea-
son, they are subject both to fluctuations in demand, and to accidental or other incidents that
may affect the normal operation of the manufacturing units. To some extent, this approach
amounts to making the inventor assume a share of the commercial risks. For the same reason,
the amount of the payment is, by way of compensation, usually calculated as a percentage of
the sales price of the final product, even if only to allow for depreciation of the currency.
Since the originality of the innovation fades over time, this payment rate is also periodically
revised downwards, unless additional improvements are made, justifying its remaining in
force for a longer period (for example, Ziegler catalysis). The details of how the royalty is
calculated and applied, or revised, are expressed in agreements between the parties.
The two ways of paying for R&D work discussed above are not completely independent.
Roughly speaking, they conform to the following empirical rule: a paid-up royalty is equiv-
alent to a ten-year running royalty. Intermediate or combined alternatives are sometimes con-
sidered: these include an initial cash payment, followed by a payment proportional to the
actual production over a particular period, and according to a scale fixed by contract.

b. Provision of Information and Payment for Corresponding Services

In the course of the various contacts that a manufacturer makes, either directly or via a con-
sulting firm, and later on with an engineering company, and during which all or part of the
preliminary and construction plans for its project are disclosed, many kinds of technical and
 2. The Elements of Economic Calculation 177

economic data are provided to the end user by the inventor or holder of the process, technol-
ogy, etc. This information becomes more and more specific and detailed as the negotiations
proceed. The early information, provided at the outset of discussions, is usually for promo-
tional purposes, intended to attract the manufacturer or its representative and to get through
the pre-selection stage. Accordingly, its nature is sufficiently general that it does not lead to
specific studies, and can be presented free of charge.
For example, the interested party may set conditions of unit production capacity, and pos-
sibly of date and regional location, and request specific information on the following points:
The operating principle of the facilities, supported by a diagram showing the basic phys-
ico-chemical operations involved in the conversion or the processing, the volumes pro-
cessed, and the operating parameters (especially average pressures and temperatures),
The overall material balance sheet(s) for the operation, with specifications for the feed-
stocks and principal products, also in what form they are produced within the battery
limits,
General consumptions of utilities and chemical products, by the hour or by the ton (of
feedstock or product),
0 The quantity (the quality, even) of the labor needed to operate the units, often expressed

as the number of shift workers,

0 The investment costs, particularly the battery limits investments, the amounts for the ini-

tial loads of catalysts, solvents, etc.; sometimes the royalty and startup costs, etc.
These data constitute the typical contents of a document that can take the form of a con-
tract, supplied free of charge, and often called a preliminary proposal. The subsequent stages
are rather different.
Thus in the final step of the discussions, when a choice has been made among the various
bidders, the owner of the process or technology selected must provide the client with suffi-
cient information to enable the engineering company contracted to do the actual construction
of the project to conduct its own calculations and studies. This information, which is far more
specific and detailed than what was originally provided, is brought together in a document
called a detailed proposal, or a Process Data Book. It normally carries a price fixed by con-
tract, to be paid by the manufacturer requesting it, in that its production represents a major
and specific contribution from the owner of the technology. In fact, it should provide:
0 More specific diagrams of the facility and of each of the main pieces of equipment that

comprise the outline being described, with their regular operating parameters, and the
transfer lines between these elements,
Material balance sheets, no longer generalized but line-by-line, allowing the composi-
tions and the kinds of constituents in the intermediate flows to be identified, as well as
their circulation parameters (average pressures and temperatures),
0 Utility requirements for each of the main pieces of equipment that consume them-and

possibly produce them,

0 A more detailed investment breakdown, especially for the unit itself, giving individual fig-

ures for primary and secondary equipment, erection and connection, indirect construction
costs, etc., perhaps even accompanied by a list of the primary equipment broken down by
type of apparatus: reactors, columns, tanks,furnaces, exchangers, compressors, pumps,
etc., and sometimes for the secondary equipment and the two other major headings.
178 2. The Elements of Economic Calculation

The owner of the process also creates another document: it too carries a price because it is
used in training the personnel to work on the facilities. This is the “Operating Manual”, which
sets out the equipment startup and shutdown procedures, with special attention to the precau-
tions for avoiding major incidents (explosion, breaks in circulation, etc., premature deterio-
ration of the catalytic or solvent systems, etc.).
The cost of the process data book and the operating manual, which usually go together,
obviously depends on the complexity of the technology concerned and the accounting meth-
ods employed by each company. However, it does not vary with the unit capacity, so as a
rough guide it will fall between 0.2 and 0.5 million € (mid-2000).

COMMENT
Among the owners of technology, it has become more and more common not to sep-
arate the costs of providing the contract documents (process data book and operating
manual) from the royalty cost, and to group them all under the heading of contractor
fees.
In this case, an approximate method for estimating their total cost is to calculate them
as a proportion of the cost of the production units plus general services and storage,
i.e., roughly 5%.

2.3.1.4 Depreciable Capital

Besides what are called fixed investments, which a manufacturer must make before the cor-
responding facilities become operational, there are other expenses that must be incurred prior
to startup. These additional outlays are not only unavoidable but also have two special fea-
tures:
0 First, unlike the previous undertakings whose certainty is only too evident, but which

often take the form of a “turnkey” contract that includes an agreement to a specific price,
these supplementary expenses arise from specific negotiations with various parties and
may vary depending on the economic situation or current events.
0 Next, the location of these expenses in the statement depends on the government tax pro-

visions, which determine whether they are considered to be permanent or recoverable,

i.e., depreciable or not.
The main headings arising from such an analysis are the following.

A. Initial Loads
When the various pieces of equipment are installed, the manufacturing facilities of which
they are a part are not necessarily ready to produce. Some of them, indeed, require that there
be a minimum amount of solid or liquid content inside the vessel: this represents an invest-
ment that must be made ahead of any startup, admittedly at the last minute, but an essential
one. This additional outlay constitutes the initial loads: depending on the particular case, after
a period that is generally shorter than the supposed life of the equipment it will either be used
up and therefore require replacement, or be recovered if the necessary regeneration proce-
dures are performed.
 2. The Elements of Economic Calculation 179

The additions referred to are usually solid catalysts, adsorbents, molecular sieves, desic-
cating agents, etc., solvents, liquid coolants or refrigerants, etc., electrodes, lamps, and other
elements. To some extent, by analogy with the construction industry, they may be said to rep-
resent moveable property, as opposed to real estate.
Since, these initial loads must either be regularly replaced, or periodically reconditioned,
they may be compared to “consumables”. The corresponding expense should therefore
appear under “operations”, i.e., show up in the operating cost or the cost of price. They must
be provided for via the working capital, a temporary investment not subject to depreciation
(see Section 2.3.1.6). This is the procedure applied when the products concerned must be
replaced within a period of time less than or equal to the theoretical renewal period of the
working capital. If this is not the case, the tax regulations usually define it as an outlay that
may be deemed a regular investment.
The problem becomes more complicated when, as a safety precaution, the equivalent of
two initial loads of catalysts, adsorbents etc. are purchased at the outset. One is actually put
into the equipment, the other stored away, to be available for quick replacement in case of a
major operating incident or a sudden loss of activity. In such cases the relevant expenses may
be divided into a portion that is permanent, and thus depreciable, and a remainder that is
renewable, i.e., part of the working capital.
In the same way, it may also sometimes be useful to analyze the costs of the initial loads
in more detail. Thus, when a portion of the investments made for this purpose can be recov-
ered, in one way or another, after a period of time that is significantly shorter than the facili-
ties’ theoretical depreciation period, that sum may be considered to be similar to those spent
on the products in storage, which are treated as working capital. This particularly applies to
catalysts containing precious metals and deposited on a support (see Section 2.2.2.1.A.e). In
fact, at the end of a catalysts life, when the regeneration cycles are so short that it has become
uneconomic to keep repeating them, 99% of the precious metal fraction can be recovered,
while getting rid of the support. To take these economically distinct behaviors into account,
the costs of the metals may be treated as working capital, and only those for the support and
its impregnation as initial load costs. The same approach may also be considered for particu-
lar adsorbents or special, costly solvents that can be regenerated in batches.
The amount of the initial loads is generally determined on the basis of three specific
parameters:
0 Two are size-related:

- The theoretical unit production

- The productivity, which can be approximated by using measures such as the LHSV
(volume) or the LHSV (weight), the space velocity or the residence time, the solvent
rate, etc.
0 The third is economic: the unit price (see Section 2.2.2.1 .B) of the products concerned.

B. Interest on Construction Loans

A certain time, which is different for each project, elapses between the date when the decision
to construct a facility is taken, and the moment when it starts up. On an existing site, and for
relatively simple unit diagrams, this period may be no longer than twelve to fifteen months.
However, when much more complex technologies have to be implemented-steam cracking,
180 2. The Elements of Economic Calculation

catalytic cracking, blast fiunaces,rolling mills, power stations, etc.-and especially when grass-
roots sites have to be developed, it can be well in excess of two years, even as much as four.
During this period, which precedes the active production phase, the manufacturer has to
gradually release or borrow a significantportion of the investment, sometimes almost all of it.
The manufacturer must, in fact, successively finance the construction plans, the development
work and civil engineering, order the equipment, pay for it, set it up, assemble it, and so on.
These outlays may be partially or wholly covered either by direct loans obtained from the
appropriate organizations, or by the treasury of the company undertaking the project. As soon
as these capital sums are lent, there is an obligation to pay effective interest on them; when
equity funds are employed, or, generally speaking, when the company treasury acts as a lend-
ing organization, it is also desirable to earn a return on the sums invested by means of a
notional or psychological interest rate that reflects the company’s financial policy and its
management methods. This matter has already been extensively discussed in Section
2.1.2.3.A,which dealt with the concept of present value.
The amount of these effective or notional interests, called construction interests, obviously
depends on the current cost of money. A rough way to assess them consists of first assuming
that the outlays are made in a linear fashion throughout the construction period, and then fol-
lowing the procedure in Section 2.2.4.2.B.The final investment cost, in this case the fixed
capital, serves as a reference amount, on which an average interest rate (i,) is levied, as
defined by the equation:
.
1,
. n+l
= z, -
2n
where n is the assumed duration of construction and i, the effective interest rate applied by
the lending organizations (banks or government agencies).
Thus, if n = 2, and i, = 12%, we obtain im= 9%.
The highly empirical nature of this method of calculation must be emphasized: from a
strictly financial or economic point of view it has no validity.
This construction loan interest item is often ignored in preliminary evaluations. Indeed, the
amount it represents is sometimes smaller than the margin of error of the known or calculated
investments for the production or processing facilities proper.

C. Startup Expenses
The startup of a manufacturing facility is generally performed under the joint technical
responsibility of the representatives of the engineering company, which undertook the
planning and construction work, and of the owner(s) of the license(s) involved. This opera-
tion takes place in the presence of the employees of the operating company who have been
assigned to the regular running of the equipment. The corresponding costs cover two main
aspects: the services performed on this occasion, and the related expenses that they create.

a. Services

The first objective during startup is to get the unit up to operating speed, so as to confirm that
the equipment is capable of doing what it is supposed to, that the products obtained conform
 2. The Elements of Economic Calculation 181

properly to the required specifications, that the theoretical nominal capacity can be reached
without difficulty, that the performances advertised and guaranteed for the various systems
concerned (catalysts, solvents, adsorbents, electrolytes, etc.) meet the contract conditions,
that the shutdown and startup procedures are effective and agree with the descriptions in the
operating manual, and so on.
At first sight, it may appear surprising that the services provided by the engineering com-
pany and the license holders in carrying out or participating in these acceptance tests, which
are intended to confirm the proper functioning of the facilities, should be paid for by the cli-
ent. In fact, this operation is not only meant to produce an actual warranty certificate, but also
to obtain an idea of the possibilities actually offered, which often go beyond the undertakings.
In the contrary case, it also presents a means for identifying the causes of malfunctions and
of course for correcting them, the corresponding expenses then being charged to one or other
of the parties present, depending on the source of the problems, or shared.
In addition, the second objective may also serve to explain why these startup costs are
charged to the manufacturer. In fact, during the initial operating period there is a genuine
transfer of knowledge to the personnel responsible for managing the units, and this training
is performed as much by the engineering company’s representatives as by those from the
license holders. Moreover, a veritable after-sales service is often established, including peri-
odic visits to the facilities, assistance during certain shutdowns or restarts (for example,
change of catalyst), minor improvements and modifications in the equipment or its operating
parameters, etc.

b. Related Expenses
Although they may be significant, the cost of the above services is on an altogether different
scale from the related costs produced by starting up the facilities. During this period, in fact,
the equipment has to be operated, if only to tune it or make minor adjustments. The products
that emerge from the units are not necessarily up to marketable specifications, or acceptable
to downstream operations. There are consumptions of raw materials, chemicals and other
products, utilities, etc., personnel activities-so many kinds of expenses that are not balanced
by income from the sale of finished products.
The duration of this veritable breaking-in of the equipment, and its takeover by the opera-
tors, depends on how complex the facilities are. It may be relatively short when it concerns a
facility on a pre-existing industrial site, with a well-controlled environment and experienced
crews, and when the technique being applied is simple. In such cases a fortnight is generally
enough. On the other hand, it can extend over several months for more complex installations
(steam cracking and related operations, for example), when it concerns plants on grassroots
sites, or when the personnel have not had suitable training (particularly in industrializing
countries).
The loss of earnings represented by the starting up of these units, from their operation
alone, is substantial. It can be calculated in equivalent months of operating cost, excluding
charges for capital, i.e., in months of operating expenses, as long as this term does cover the
consumption of raw materials, various related products, utilities, and shift labor. In fact, there
are two ways of accounting for it, depending on whether the manufactures obtained during
the startup period are marketable or not.
182 2. The Elements of Economic Calculation

0 If they are not, and cannot be sold later, the raw materials that produced them must be
treated as definitively lost. In this case, the startup costs are often taken as being equal
to 15 days of operating costs (cost of materials included).
0 #en the products obtained meet the required specifications, or can after all be sold
when a few matters have been taken care of or some minor adjustments made, these
starting expenses are in practice evaluated at three months of operating costs (cost of
materials excluded).
The purity standard provides a specific example of a situation in which a product, although
it does not immediately meet the required performance levels, is capable of being sold. If the
market requires a level above 99.5% by weight in this regard, and the average value for the
product obtained during startup operations is only 99.2%, it often happens that once the facil-
ities are operating properly they are capable of reaching 99.7% or more. The product that is
unfit for direct use only has to be stored long enough to use it up by gradual mixing with the
batches whose quality is greater than the required minimum.

2.3.1.5 Total Investments on a Manufacturing Site

Working Capital
To complete the list of investments that a manufacturer has to undertake before being in a
position to start making profits, we must also add to the depreciable capital the portion rep-
resented by working capital.
This outlay is not in itself an investment of the same kind as the ones discussed above.
It really represents a temporary investment, linked to the actual operation of the facilities,
and-unlike the others-it is repaid in an amount close to its initial value, at least if we ignore
the currency depreciation or changes in prices and exchange rates that may occur during the
repayment period.
In economic calculations, the working capital f is treated as a non-depreciable expense
made in year 0. As a pseudo investment, it can be repaid at any time, particularly at the end
of the real or theoretical lifetime (n years) of the equipment. As an investment, it represents
a loss of earnings that can be balanced either by financial costs and a corresponding notional
or effective interest rate, or by employing the concept of present value. In this case, it is
shown in the accounts for each yearp (0 < p < n) in its discounted form:

f
(l+i)P
where i is the discount rate, generally falling between 0 and 1. This shows that the value of
the working capital falls year by year. This explains why government tax provisions permit
periodic re-evaluations, as long as they do not lead to capital gains, which would then be tax-
able.
Working capital essentially constitutes a current allowance that includes the following
main headings:
0 The value of the regular reserve stock of raw materials: loads, various chemical prod-

ucts, homogeneous phase catalysts, solvents, other consumable goods such as elec-
trodes, electrolytes, lamps, etc.,
 2. The Elements of Economic Calculation 183

0 The value of the fluids and solids in circulation, or present at a given moment, in the pro-
cessing units themselves, in the various stages of conversion,
0 The value of the normal stock of finished products and by-products,
0 The amount of the costs to be incurred by unexpected or other shutdowns of the facili-
ties, to allow for continued product shipments, and receiving and paying for raw mate-
rials, according to the terms of the contracts with suppliers and buyers, and also to pay
the regular equipment operating and maintenance personnel,
0 If necessary, the value of the current stocks of spare parts, wrapping, packaging, etc.,
0 Coverage for the delays experienced between the issuance of invoices and receipt of
payment. These delays result both from the regulatory provisions of each particular gov-
ernment, and from commercial practices that may vary over time, depending on the eco-
nomic situation. Thus, in France, payment of invoices is made 90 days after month end;
however, during the crises of the last twenty years this period has sometimes fallen to
30 days after month end. In the USA payment is generally made within two weeks, leav-
ing aside postal delays.
In practice, the current credit may be evaluated in various ways, the most important items
being those concerning supplies and the reserves of product; a handy method consists of
expressing them as the equivalent of months of operating cost, excluding capital costs, i.e., in
the habitual terminology, as months of operating expenses. Custom dictates that this shall be
roughly equal to double the amount of the startup costs for the facilities. In other words, the
current credit thus corresponds to:
0 Either one month of operating expenses, cost of materials included,

0 Or six months of operating expenses, cost of materials excluded.

Depending on whether one is inclined to take a conservative view or not, the higher or
lower value is selected.
The working capital also sometimes includes investments for non-consumable goods that
can be compared to a real temporary stock of products (initial loads of precious and semi-pre-
cious metals, adsorbents, etc., and, more generally, certain recoverable pieces of equipment
or spare parts). Examples include platinum, iridium, palladium, cobalt, nickel, mercury, etc.,
and components of catalysts, electrodes, etc., which-allowing for various treatments and
certain losses+an be recovered and re-activated, once the system is shut down. This item,
which may account for a major portion of the working capital, represents a sum that can be
accurately defined on the basis of technical considerations. It thus differs from the current
allowance, whose amount is evaluated on the basis of an estimation of the risks incurred.

2.3.1.6 Grassroots Investments

The determination of this kind of investment is relatively uncommon, since it demands a good
prior knowledge of the chosen construction location, as well as all the accompanying param-
eters (geographic, geologic, climatic, financial, fiscal, social, etc.).
In fact, to determine the amount it is best to add the two following costs to the total invest-
ment for an industrialized site: land costs, and yard improvements.
 Next Page

184 2. The Elements of Economic Calculation

A. Land Costs
The cost of the land on which the production or processing facilities are to be constructed is
sometimes included among the project investments that a manufacturer has to make.
Although such outlays are far from negligible, they are generally rarely mentioned, espe-
cially when smaller installations are concerned. On the other hand, the matter is sometimes
addressed in the case of significant extensions, or the creation of entirely new complexes on
as yet un-serviced sites. It must be remembered that in most cases, for projects of moderate
size, the land purchase procedure is usually disconnected from the procedures affecting the
manufacturing units proper, since the site acquisition process is not conducted separately for
each unit, but is part of a much larger development program.
Nevertheless,when the question arises, the value of a site is mainly dependent on its particular
interest. Now, the parameters that can be applied to measure this are extremely numerous and var-
ied. Among the most important are not only the required surface area, which is easy to quantify,
but also several other factors that are more difficult to assess, because they are more subjective.
In particular, these include the geographic location (industrial area, open countryside, seashore),
the relief, the existence of port or river shipping facilities,railroad or highway systems,proximity
of urban centers, facilities for obtaining constructionmaterials, energy, drinking and other water,
ecologicalrestrictions,local conditions in terms of housing, socio-culturalenvironment,and edu-
cation, the availability,skills, and attitudes of local labor, etc. and generally speaking the existing
infrastructure-so many elements that may require a specific study of the problem, since in this
regard each project constitutes a special case. Outside factors must also be added to this list: they
can further complicate the situation and may prove decisive. These include local, regional, or
national incentive measures aimed at promoting job creation by means of tax exemptions, and
various development assistance measures, even offers of property. These measures may on the
other hand be deterrents, reflecting environmental,population balance, over-industrialization,or
other concerns: they can make any project that goes against them uneconomic.
Thus, unless we are able to estimate the respective weights of these various factors-
which amounts to knowing in detail all about the transactions in the area-it would seem
impossible to make even an approximate estimate of the land costs.

B. Yard Improvements
Besides the outlays for obtaining access to the physical property, projects on new sites require
specific preparation work that is likely to lead to additional investments, which, moreover,
precede any installation of equipment.
Depending on the characteristics of the site, these include in particular provisions for lev-
eling the surface, possible stabilization of the ground or protection against wind erosion,
access works, connections with highway, railroad, and waterway systems, etc., also for clean-
ing up, fire fighting, environmental protection, etc., and for supplying energy (oil and gas
pipelines, power lines, etc.) and water (water mains, connections to rivers, groundwater, puri-
fication or sea-water desalination stations, etc.).
In some cases, all or part of these infrastructures are missing. If they are required just for
the construction of the project concerned, they must be charged to it, which often leads to the
inclusion of major public works activities (port and highway facilities in particular).
Previous Page

2. The Elements of Economic Calculation 185

Here again, unless we are well informed about the geographic and structural contexts into
which a new manufacturing complex is to be inserted, it would seem difficult for an outside
observer to estimate the consequences in terms of investment costs. However, there is an
average empirical rule for the servicing of grassroots sites, which says that the amount of the
depreciable capital should be increased by 30%.

COMMENT
We must distinguish between the concept of construction on an as-yet unserviced
site, and that of the “first industrial plant”. The latter is a “prototype” unit, based on
new ideas, as yet untested over an extended period at full scale, and for which the
probability of incidents is higher than for proven technologies. To allow for such
possibilities, which can require improvements, hence additional investments, the
cost of the production or processing facilities proper is often increased by 20 to 30%,
depending on the degree of complexity.

2.3.2 ADJUSTMENTS TO INVESTMENT DATA

Available information about investments costs generally refers to very specific conditions.
Now, in practice, the questions that have to be answered usually refer to the following two
main situations:
0 Determination of the economic or profitability aspects of a facility that is technically

similar to the “reference” unit, but that operates under different conditions: production
capacity, date, site, possibly the source of raw materials, kind of catalysts, market factor,
yields, operating variables (temperature, pressure, etc.) and more broadly any parameter
likely to require a sensitivity study,
0 Comparison, on the same or “consistent” basis, of several processes or projects that lead

to the same products, but for which the available economic data, called “reference” data,
are mixed, and sometimes of differing credibility.
In such cases, one approach is to embark on complex procedures aimed at repeating all the
calculations for each modification, particularly those concerning the investment costs. A sec-
ond approach, more realistic because it is quicker although less accurate, emphasizes limiting
the number of parameters to be considered, and the use of simplified methods, usually empir-
ical ones. In this regard, three variables turn out to be essential in terms of finding the amount
of the battery limits investments:
0 The unit capacity

0 The date

0 The location.

2.3.2.1 Adjustment of Capacity

An overall, simple way of adjusting the available information on unit manufacturing capaci-
ties consists of applying the “scaling factor” or “six tenths factor” method. This procedure is
based on the observation that emerges from the statistical treatment and smoothing of histor-
186 2. The Elements of Economic Calculation

ical data, according to which the costs of two pieces of equipment-and more generally, of
two facilities-of the same kind but processing different volumes or quantities, are related by
an empirical expression having the following form:

1,and Zfl represent the erected costs of the equipment or units concerned, i.e., the battery
,
limits or comparable investments. Gland C’ are the processing or production capacities cor-
responding to these investments;f i s incorrectly called the scaling “factor”, even though it is
actually an exponent.
The use of such an expression was first advocated for calculating the costs of various kinds
of equipment by “similarity”. For example, Table 2.17 gives an idea of the values generally
recommended forf, according to the kind of equipment concerned: they are obviously only
rough guides, since it will always be possible to run into the exception that will invalidate the
rule. So, it is always worthwhile to confirm, before making any calculations, that the equip-
ment being considered has no special feature that would prevent any comparison with the
family to which, on the face of it, it is supposed to belong.

Table 2.1 7 Extrapolation factors for various items of equipment

Equipment Factor Equipment Factor

Pressure vessels: Crystallizers 0.65
columns and reactors 0.65-0.70 Basket centrifuges 1 .OO-1.25
tanks 0.60-0.65 Evaporators:
Exchangers: standard 0.55
standard 0.65 forced circulation 0.70
reboilers 0.50 Mills 0.65
Air coolers 0.80 Tanks:
Compressors 0.80-0.85 < 150 m3 0.30
Pumps: 150 m3 0.65
centrifigal 0.40-0.50 Pressure storage:
others 0.60-0.70 horizontal 0.65
Motors 0.90-1.00 spherical 0.70
Furnaces 0.80-0.85 Steam generation 0.80
Stirring reactors 0.45-0.50 Electrical generation 0.75
Steam ejectors 0.45-0.55 Cooling tower 0.60
Mixers: Refrigeration 0.70
propeller 0.50
turbine 0.30
Dryers 0.45-0.50
Filters 0.60-0.65
(Source: IFP)
 2. The Elements of Economic Calculation 187

The use of the equation for the units themselves was considered only later on. But here
again, before applying it, it is desirable first to confirm that there is no difference between the
design of the facility whose investment cost is required, and the reference facility; in other
words that the sequence of equipment is indeed the same in both cases. Figure 2.13 is a graph
on which the y-axis is the logarithm of the ratio of the investment costs, and the x-axis is the
logarithm of the ratio of the production capacities: it has a straight-line segment of slopef
with the following properties:
0 The value off is generally around 0.7. In theory, a production unit composed only of shells,
i.e., tanks, columns, exchangers, tubes, etc., or similar apparatus (Table 2.17), should be
capable of extrapolation at the power of 0.6667. This figure is explained by the fact that the
price of this kind of equipment is proportional to its weight, and that for the same thickness
the price increases with the total surface area (shell and heads), and decreases in inverse pro-
portion to the volume or the flow that this area defines. In other words:

surface area = k'-L~ = k'.L2/3

price = k
volume L3
wheref = 213 or 0.6667.
In practice, besides the static equipment mentioned previously, which can be extrapolated
with an exponent of 213, a facility includes other equipment such as pumps and drives, fur-
naces, etc., whose impact in investment terms is still modest, but whose extrapolation factors
are larger. All in all, the exponent resulting from weighting the exponents for the various
items of equipment concerned is found to be close to 0.7.
The factorf= 0.7 has a range of validity around a median value. Experience shows that,
roughly speaking, it can be applied to capacity ratios going from 1 6 to 5/1. On the other hand,
there is the problem of what values to choose when the ratio is above or below these limits.
Beyond the range where the factorf= 0.7 applies, discontinuities appear in the extrapolation
rules for certain pieces of the unit's equipment.
In fact, moving to a distinctly higher production capacity sometimes requires a doubling,
even a tripling of the number of pieces of apparatus, because their design makes it impossible
for them to function properly when their dimensions exceed a certain threshold. For example,
this is the case with solvent extractors, which consist of a shell with a central shaft for
mechanical stirring, equipped with rotating disks inside hollow circular baffles: they mix very
badly at the edges when the diameter exceeds 2.80 m. It is also the case for tubular catalytic
reactors, whose number of tubes per shell cannot exceed a certain limiting value, because of
congestion, poor thermal exchange, and homogeneous filling by the catalyst. Without wish-
ing to multiply the examples, it is important to be aware of this type of problem and to con-
sider the consequences: thus, in practice, when this situation arises, the choice is generally to
have two identical pieces of equipment rather than having one of maximal size and adding
another one to cover the required additional production. In fact, in the event that one of them
breaks down, we are sure to maintain a 50% level of plant production with the one that
remains. We also find that the weighted average of the extrapolation factors for this dupli-
cated equipment, and the factors for the apparatus that can be extrapolated to the power 0.7,
is a higher number, such as 0.8 to 0.9.
188 2. The Elements of Economic Calculation

/ -
0.8 0.9

//

/ -

1I5 1 5

Figure 2.1 3 Capacity extrapolation of battery limits investments.

(Source: IFP)

Below the range where the factor 0.7 applies, the battery limits investments are actually
found to be higher than those calculated according to the rule. To put it clearly, this means
that to obtain a credible result the value offmust be reduced, for example to 0.55-0.5. As will
be shown in Chapter 3, the explanation comes from the breaking down of the costs of facili-
ties into various categories, which, apart from the amount for the principal equipment itself,
include the amount for secondary equipment, the costs of erection and connection, indirect
construction costs, and contingencies. Now, it appears that when the production capacity is
significantly reduced, the relative weight of these additional categories in the unit limits
investments tends to grow in comparison to that of the apparatus, which forms the basis of
the calculation. By way of illustration, one can easily see that the use of a crane to position a
column on its base generates costs that are not proportional to the size of the shell thus
mounted, but are essentially a function of the period of utilization or rental of the crane, i.e.,
practically independent of the mass hoisted.
When all is said and done, it is just as if a certain portion of the curve in Fig. 2.13 were
likened to a straight-line segment of slopef= 0.7, the real value of the extrapolation factor at
each point of this diagram being represented by the slope of the tangent at that point.
 2. The Elements of Economic Calculation 189

Practical confirmation of the proposed use off values for calculating the battery limits
investments, by overall extrapolation based on the production capacity, is provided by
Table 2.18, which shows type examples from the refining, petrochemical, and inorganic
chemical sectors. These figures are based on mid-sized facilities currently operating in the
heavily industrialized regions (North America, Western Europe, and Japan). The lowest val-
ues correspond to extrapolations for low capacities, and conversely for the highest ones.

Table 2.1 8 Type examples of extrapolation exponents

Process Exponent
Refining industry
Atmospheric distillation 0.70-0.65
Vacuum distillation 0.70
Hydrodesulfurization
of naphtha 0.65-0.85
of residue 0.80
Catalytic reforming 0.70-0.85
Catalytic cracking 0.70
Steam cracking 0.75
Sulfur unit 0.60
Hydrogen unit 0.60-0.70
Petrochemical Industry
Steam cracking of naphtha 0.75-0.80
Acetic acid (carbonylation of methanol) 0.57-0.52
Methanol (steam reforming of natural gas) 0.75-0.93
Ethylene oxide (oxygen) 0.72-0.77
Phenol 0.58-0.66
Vinyl chloride (from dichlorethane) 0.60-0.70
Polyvinyl chloride (suspension) 0.66-0.68
Linear low-density polyethylene 0.70-0.75
Polypropylene (suspension) 0.66-0.74
Inorganic chemical industry
Nitric acid 0.77-0.85
Phosphoric acid 0.53-0.87
Sulfiuic Acid 0.75-0.80
Ammonia (steam reforming of natural gas) 0.76
Electrolysis (mercury amalgam) of sodium chloride 0.77-0.82

(Source: IFP)
190 2. The Nements of Economic Calculation

2.3.2.2 Investment Updating

A. The Problem

In different ways, this kind of information processing resembles both the deflation procedure
described in Chapter 1 (see Section 2.3.1. l), and the present value procedure discussed in the
present chapter (Section 2.1.2.3.A).
Although, it is incorrectly called “discounting”, investment updating does not consist of
calculating the present value of a sum that is usually paid later on, or sometimes earlier, but
rather of taking into account the variation over time of the prices of materials and the costs of
making or assembling equipment, i.e., both the productivity and the costs of labor (skills,
wage scales, benefits, etc.). This approach incorporates the effects both of currency depreci-
ation and of variations in the real value of the cost of raw materials and of the total wage bill.
Generally speaking, since in practice the observed overall drift is usually positive, the current
cost of a piece of apparatus or a facility will appear higher than the cost of the same system a
few years earlier. Thus, although there are similarities with the concept of present value, they
concern only that part of the idea that deals with historical data and the use of current money
calculations. However, such a definition seems not to apply to the methods referred to as price
revision, which often accompany estimates that the bidder guarantees for a limited period
only; after that the adjustment procedure is applied.
In fact, in this case it is not a matter of using existing data but of anticipating future infor-
mation. For all that, when the method is actually applied, the values to be used will be avail-
able and considered to be “objective”, i.e., observed or real, whereas estimating a discount
rate will always have a subjective aspect, related to the uncertainty of the future.
As for the deflation method, its special feature is that while it too concerns past informa-
tion, it also ignores currency depreciation proper. Its only connection with updating is that it
does consider the variation of the true value of the prices of “products” in the broad sense. In
this way, it expresses either source variations in the real prices of raw materials, which rise
and fall with the evolution of supply and demand, or technological improvements, which,
although their impact is more modest, generate a downward drift.
Thus, when compared with the related concepts discussed earlier, updating has two prac-
tical consequences:
Any informationabout an investment should be accompanied by the correspondingdate.
For a complete facility, this means stating the year and sometimes the quarter to which
the information applies (e.g., 4th quarter 1994 or mid-1990). For a piece of equipment
the month can be given (e.g., October 1985).
To determine the value of an investment at another date, or to make a uniform compar-
ison of two investments having different dates, we need a procedure for transposing the
available information from one year or one month to the other.

B. The Concept of an Updating or Cost Index

To respond to these two requirements, the most convenient solution is to use a coefficient,
incorrectly called a “discount index” or, better, a “cost index”.
 2. The Elements of Economic Calculation 191

By definition, an index is in fact a ratio, and therefore a dimensionless number, which mea-
sures the relative variation of a parameter between two different “states”, in time or space;
one of them may be arbitrarily taken as the reference (value 100) for the calculation. Eco-
nomic indices may concern flows (income, consumption, savings, sales, etc.), quantities
(products, consumables), activities (length of work, wage-earning employees, etc.), prices
(consumer, producer, wholesale, etc., deflated), or wages (hourly, monthly). Combining two
indices sometimes enables a third to be defined: for example, labor productivity from the pro-
duction and activity indices, or purchasing power from the income and price indices.
An index is said to be simple or elementary when it is based on the gross variation of a
single parameter, as compared to a given initial state. Thus, we speak of the elementary index
of the parameter Y, at time t, compared to the reference situation zero, given by:
V

It is obviously not necessary to refer to the zero state, and between two states t l and t2there
is a relation given by the number or coefficient:

It is also possible to summarize the variation of several parameters by calculating a syn-

thetic index, which is a weighted average of simple indices:

i=n
- arithmetic: 410= Ccxi4io
i=l

- geometric: 4/o= C(.tl/o)ui

i=l

where
n : number of parameters concerned
410 : elementary index of the parameter Yi at time t in comparison to reference state 0
ai : weight, expressed as a percentage of the parameter Yi.
The formulas most widely used for constructing price and quantity indices are those of
Laspeyres and de Paasche, which are the arithmetic means of simple indices, whose weight-
ing coefficients are proportional to the expenses or sales figures (quantity x price):
192 2. The Elements of Economic Calculation

0 Laspeyres index:

i=l i=l

0 De Paasche index:

i=l i=l

where
pP: price of product i at the reference date 0
pf : price of product i at date t
9: : quantity of product i at the reference date 0
9: : quantity of product i at date t.
Note that in the Laspeyres formula the reference structure comes from the base date (a con-
stant year), whereas in de Paasche’s the ratio for the sample concerned refers to the time when
it is made (the current year). This recalls the two approaches already discussed for deflation
and present value.
Instead of considering price variations, we can look at quantities and calculate volume
indices. When attempting to weight quantities by the price in the base year, we calculate a
“Laspeyres volume”:
n

When weighting by the prices for the current year, we calculate a Paasche volume:
n

i=l
 2. The Elements of Economic Calculation 193

COMMENTS
0 For convenience, Laspeyres-type indices are the ones most frequently used. In

fact, it is easier to refer to a past year than to examine the structure of a sample during
the current year. These can always be criticized for being based on a breakdown and
weighting from the past, which, over time, progressively lose their credibility. This
problem becomes crucial, for example, for a calculation of a retail price index: it does
not make sense to choose the same “shopping basket” in 2000 as in 1985.
0 To compensate for this drawback, one method consists of regularly changing the

standard state, so as to update the reference structure itself. This procedure produces a
succession of indices that cover short periods of time, so that, to get to the required
year, they all have to be multiplied together. This procedure, called the Laspeyres
chain-index, is used, for example, particularly in France, to calculate retail price indi-
ces. As mentioned in Chapter 1 it is also used to calculate the IPI (industrial production
index). It has the advantage of avoiding the progressive aging of the updating method.
0 Price revision formulas, which are found in many contracts, particularly in those

concerning the supply of raw materials or products, shed additional light on index
calculations and the uses to which they can be put. As a rough guide, taking natural
gas as an example, an expression such as the following one may be considered for
the periodic (monthly, quarterly, etc.) re-estimation of the sales price:

where
C
+0.2 -+0.1
CO
-+0.1 -
Go
GOO “1
NO

PandPo : price per m3 of gas, as delivered by the vendor company, on the cal-
culation t and reference 0 dates respectively (0 = date of contract
signature, for example)
FO and FOo : price per ton of fuel oil
CandCo : price per barrel of petroleum
NandNo : price per ton of naphtha,
each of the products considered as meeting precise supply specifications (quality,
delivery date, etc.).
The bracket does indeed represent a Laspeyres-type formula with weighting coeffi-
cients fixed at the outset. However, it is worth noting that these factors are generally
not obtained from calculations of prices and quantities for each of the component
products, but usually represent the result of a statistical study of historical data.
It may also be desirable to measure the variation of the price-quantity pair, and in
this way calculate a global value index (GVI) according to the following formula:
i=n

GVI,,, = i=n
,‘=I 100
194 2. The Elements of Economic Calculation

We thus find that: GVI = L(price) x P(quantity) = P(price) x L(quantity) 5 .

The GDP (gross domestic product) value index is calculated on this principle; it is
equal to the product of the GDP price index (Paasche price) and the GDP volume
index (Laspeyres volume).

C. Use of Updating Indices

Generally speaking, each company has its own updating procedure, whether it is the price
revision formula, announced to its customers and particularly popular with small- and
medium-sized companies and industries, or a true index, compiled and utilized in-house,
which usually means it is confidential. Official organizations have different motives, how-
ever, and tend to distribute the results of their work, if they can serve as reference bases for
the calculations made by companies in this regard.
However, it should be mentioned that certain journals regularly publish the variations of
updating indices for their respective activity sectors: in the principles of their design, these
resemble the indices constructed by companies for their own use.
Roughly speaking, these various indices have two essential features:
0 They carry a reference date, which generally corresponds to the month or year when they

were created, for which the arbitrary value of 100 applies.

0 They are “composites”, i.e., they are the result of a weighting among several component

indices, whose variation may or may not also be monitored by the authors of the total
index.
As for manufacturing facilities, the updating of battery limits investments is based on three
types of index, depending on the kind of components it contains:
Those whose constybents depend only o n variations in wage costs and the prices of
materials,
Those based on a weighting of component indices concerning equipment (pressure ves-
sels, exchangers, furnaces, compressors, etc.),
Finally, those that combine the two previous concepts and are therefore described as
“mixed”.

COMMENTS
Whatever type of index is concerned, two major features must be emphasized:
0 The kind of parameters used in the initial formulation, and their respective

weights, usually derive from statistical analysis and processing of available data.
Consequently, on the one hand there may be significant differences between the dis-
tribution represented by the index and that of the facility to which it is applied; so it
is of the highest importance to choose the updating method whose constitution
appears to be the most appropriate. On the other hand, this weighting may vary over
time, leading either to progressive obsolescence for this tool, or to periodic adjust-
ments, with changes in the reference date and development of different time series.

5. L = Laspeyres and P = Paasche

 2. The Elements of Economic Calculation 195

0 An index can always be reduced to two components, one concerning wages, the

other, materials. Obviously, this operation entails a loss of accuracy of several per-
centage points. On the other hand, for a country where information is hard to come
by, it offers the advantage of greatly simplifying the design of a formulation that is
better suited to local conditions. It is easy to achieve for indices of the first type,
because it only requires the grouping together of the parameters concerning wages
and the ones dealing with materials. It would seem to be trickier for indices of the
second type, since, before the grouping can be done, it is necessary to separate the
different equipment components into the part for boilermaking or shaping labor, and
the part for the basic manufacturing materials. For indices of the third type, the pro-
cedure to be applied derives from the two preceding ones.

D. Examples of Cost and Updating Indices

In most of the industrialized countries, periodicals compile and publish indices that are spe-
cific to the sector of activity that they serve. This kind of information has long been widely
available in the United States. The phenomenon has only recently reached other countries but
it continues to grow. Various examples can be presented to illustrate this trend.

a. France

Among the first indices to be published was R. Boulitrop’s (Techniques de I’lnginieur),

which suggested taking the arithmetic mean of:
The total weighted index S of wages in the engineering, electrical, and refractories indus-
tries,
The cost index for average sheet metal (open-hearth steel, 4 mm thickness).
The values of these two components are published monthly in the Bulletin Oficiel du Service
des Prix (Institut National de la Statistique et des Etudes Economiques) or in L ’Usinenouvelle.
At a more sophisticated level, M.N. Vandamme defined the following index for the chem-
ical industry:
chim = 0.615 S + 0.022 Cm + 0.08 Lmc + 0.015 Lma
IME wages cement rebars rolled-steel
Fe E 22 A 33
+ 0.011 Pta + 0.046Tma + 0.023 Ho + 0.026Tu
girders sheet metal cast iron steel pipe
E 24-1 A 33-2
+ 0.106 PsdB + 0.046 Cf + 0.032Af + 0.006Tc + 0.044Aidf
products, copper aluminum wire grain- cold-rolled
services, wire oriented Z6CNDT
misc. steel 17-12
Cat. B

The description of each component is summarized below it, in standardized terms. The
values are published regularly, notably in L ’Usine nouvelle.
196 2. The Elements of Economic Calculation

A comment must be made after considering these two examples, which belong to the first
type of index. It concerns the weight of the wage parameter, which exceeds 60%. Put another
way, too detailed a breakdown, depending on the kind of reference materials, may be
regarded as irrelevant, considering the precision with which the labor component is assessed.
Many other formulas could be given here, each representing a particular sector of activity
or even specific companies, if the related confidentiality problems could be ignored. When
localization is considered (see Section 2.3.2.3), there will be a more detailed analysis of one
of them, designed by AFITEP, a project management association.

b. USA
A lot of work has been done in this area, and several indices are publicly available, both in
mass-market publications and by subscription to consulting firms (for example SRI, the Stan-
ford Research Institute, which compiles an index called the PEP or Process Economics Pro-
gram index, for various countries). Among the better-known indices the following may be
mentioned

The ENR (Engineering News-Record) Construction Cost Index

This is one of the earliest, but it is still published weekly in the journal Engineering News-
Record, and twice a month in Chemical Engineering. It is a composite index of the first type,
with a reference date of 1913; its value represents the average of figures calculated for twenty
different U.S. cities. Its essential characteristic comes from its composition, which is based
on the costs for notional quantities of steel, structural timber, and cement sections, and two
hundred hours of labor.
Owing to its age, it now appears outdated, because it has not incorporated any technolog-
ical improvements since it was first set up: for example, when applied to the chemical indus-
try it indicates variations that are much larger than those actually observed. However, in
certain situations it may prove very useful, particularly when the construction of units for
industrial production involves the use of wood. The use of such a material may appear strange
or old-fashioned today, but to illustrate the continuing interest in this kind of raw material we
have only to cite Mobil’s construction, during the 1980s, of a facility for producing synthetic
gasoline from natural gas, via methanol. This was in New Zealand, where a coastal site and
the local availability of materials favored air-cooling systems based on a wooden rather than
a concrete structure.

The M&S (Marshall and Swift, Formerly Marshall and Stevens)

Equipment Cost Index
This is based on a periodic calculation of the detailed costs of equipment, plus correspond-
ing erection and connection expenses, performed by the engineering estimates department of
the Marshall & Stevens company.
Two indices are published:
The “all-industry’’ one is the arithmetic mean of forty-seven categories of industrial,
commercial, and construction equipment.
The other, specifically industrial one, called the “process industries and related indus-
tries” index is a weighted average of the unit cost indices for twelve particular industrial
 2. The Elements of Economic Calculation 197

fields. One of them concerns the petroleum sector, another is broadly chemical, others
more specifically cover the rubber, cement, glass, paint, paper, refractory, electrical
equipment, mining, crushing and grinding, refrigeration, and steam generationindustries.
These indices are published twice a month in the journal Chemical Engineering; their ref-
erence year is 1926.

The Nelson (now Nelson-Farrar) Refinery Construction Cost Indices

W.L. Nelson introduced various composite indices, which are weighted averages of the
costs of equipment or facilities, prices of materials, and labor expenses. In particular, there is:
An “inflation” index (the Nelson refinery inflation construction cost index), more usu-
ally and briefly called the “Nelson index”,
A “true cost” index (the Nelson refinery true construction cost index), which relates to
the “productivity” achieved in the construction of complete refining units (atmospheric
or vacuum distillation, catalytic reforming, catalytic cracking, etc., complete refinery),
0 And other “operating cost” indices, such as the Nelson refinery operating cost index, a

weighted average of the costs of fuel, labor, investments, and chemical products.
The most widely used is the inflation index. It belongs to the third, i.e., mixed, type, and
may be directly compared to various other total updating methods, compiled by various work-
ers, for adjusting battery limits investments by date. As its name indicates, it is intended for
the refining industry; it has three main components:
Various items of equipment (12%)
0 Materials (28%), including ferrous metals (20%) and non-metallic construction materi-

als (8%)
0 Labor (60%), including 39% for specialists and 21% for others.

Their variations in recent years, and the variation of the resulting composite inflation
index, are shown in Table 2.19. The item “equipment (various)” is itself the result of a
weighting of the costs of five particular types of apparatus: pumps and compressors, electrical
equipment (motors, transformers, welding machines, lighting, etc.), non-electric drives,
instruments, and heat exchangers.
The Nelson composite indices, especially the inflation index, are published in the first
issue of The Oil and Gas Journal for each month. A very detailed summary of the partial indi-
ces (Table 2.20) is given in the first issue for the months of January, April, July, and October:
the observed difference between the date of publication and the date the figures refer to is
about six months. These indices concern the costs of utilities, inorganic and organic products,
labor, and equipment and materials; each of them includes a reference, either a code number
from the U.S. Department of Labor’s Bureau of Statistics, or a description of origin such as
“manufacturer’s data” or “calculated”.
Table 2.20 shows examples of the variations of the cost indices for a certain number of
items of equipment that are widely used in the refining and chemical industries. The inflation
and true costs indices (the latter was dropped in 1976), as well as the labor, equipment, and
materials components, have 1946 as their reference date; the others have 1956 as their base
100 year.
198 2. The Elements of Economic Calculation

Table 2.1 9 Nelson Refinery (inflation) Construction Cost Index

Year Equipment (various) Materials Labor Nelson Index

-
1950 126.2 149.5 144.0 146.2
1951 145.0 164.0 152.5 157.2
1952 153.1 164.3 163.1 163.6
1953 158.8 172.4 174.2 173.5
1954 160.7 174.9 183.3 179.8
1955 161.5 176.1 189.6 184.2
1956 180.5 190.4 198.2 195.3
1957 192.1 201.9 208.6 205.9
1958 192.4 204.2 220.4 213.9
1959 196.1 207.8 231.6 222.1
1960 200.2 207.8 241.9 228.3
1961 199.5 207.7 249.4 232.7
1962 198.7 205.9 258.8 237.6
1963 201.4 206.3 268.4 243.6
1964 206.8 209.6 280.5 252.1
1965 211.6 212.0 294.4 261.4
1966 220.9 216.2 310.9 273.0
1967 226.1 2 19.7 33 1.3 286.7
1968 228.8 224.1 357.4 304.1
1969 239.3 234.9 391.8 329.0
1970 254.3 250.5 441.4 364.9
1971 268.7 265.2 499.9 406.0
1972 278.0 277.8 545.6 438.5
1973 291.4 292.3 585.2 468.0
1974 361.8 371.3 623.6 522.7
1975 415.9 421.0 678.5 575.5
1976 423.8 445.2 729.4 615.7
1977 438.3 471.3 774.1 653.0
1978 474.1 516.7 824.1 701.1
1979 514.4 573.1 879.0 756.6
1980 578.1 629.2 95 1.9 822.8
1981 647.9 693.2 1044.2 903.8
1982 662.8 707.6 1156.5 976.9
1983 656.8 712.4 1234.8 1025.8
1984 665.6 735.3 1 278.1 1061.0
1985 673.4 739.6 1297.6 1074.4
1986 684.4 730.0 1330.0 1089.9
1987 703.1 748.9 1370.0 1121.5
1988 732.5 802.8 1405.6 1164.5
1989 769.9 829.2 1440.4 1195.9
1990 797.5 832.8 1487.7 1225.7
1991 827.5 832.3 1533.3 1252.9
1992 837.6 824.6 1579.2 1277.3
1993 842.8 846.7 1620.2 1310.8
1994 851.1 877.2 1664.7 1349.7
1995 879.5 918.0 1708.1 1392.1
1996 903.5 917.1 1753.5 1418.9
1997 910.5 923.9 1799.5 1449.2
1998 933.2 917.5 1851.0 1477.6
999 (August] 913.4 882.4 1911.8 1500.1
iource: The G and Gas J )
 2. The Elements of Economic Calculation 199

Table 2.20 Partial Nelson Indices

Pressure Pumps and Heat Electric

Year vessels Furnaces Instruments
compressors exchangers motors
(code 1072) calculated (maker)
(code 1141) (maker) (code 1173)
1960 181.0 228.3 194.6 182.2 98.0 202.5
1961 177.3 228.8 188.1 172.8 95.0 207.5
1962 181.5 222.5 183.6 166.4 90.4 214.8
1963 183.1 224.3 189.1 165.1 91.1 224.5
1964 184.6 231.5 206.1 161.0 91.5 225.3
1965 186.0 245.6 218.4 159.8 90.2 22 1.5
1966 189.9 260.5 235.2 163.7 91.8 229.7
1967 192.7 272.2 237.9 171.0 93.0 232.3
1968 198.5 284.5 223.4 175.3 95.4 239.1
1969 207.8 298.4 235.8 184.7 100.8 252.8
1970 224.3 313.1 253.8 199.9 108.6 278.8
1971 237.0 330.6 268.0 206.4 118.0 306.5
1972 246.4 337.5 274.3 211.0 125.3 328.4
1973 262.5 346.9 313.7 219.8 130.0 338.0
1974 3 18.3 416.3 501.0 250.4 157.5 376.5
1975 357.9 509.8 553.1 302.8 179.5 420.5
1976 367.4 538.6 478.5 330.3 191.1 464.2
1977 397.2 575.0 454.0 347.0 206.9 484.1
1978 431.1 619.5 496.0 375.8 229.4 5 18.7
1979 471.5 673.9 557.4 405.2 250.7 541.9
1980 536.0 773.0 618.7 462.4 271.3 587.3
1981 589.3 880.6 727.3 513.6 300.9 623.4
1982 601.1 944.1 635.9 534.7 3 17.0 652.5
1983 591.0 944.5 541.7 563.5 324.0 675.2
1984 589.9 950.5 530.9 583.3 339.0 696.7
1985 587.8 969.0 520.0 601.8 340.5 713.7
1986 582.6 982.6 524.7 603.8 331.0 733.3
1987 58 1.2 995.0 575.2 613.0 339.0 744.2
1988 628.6 1027.7 664.1 645.8 361.2 748.8
1989 666.8 1086.3 737.8 685.8 371.6 770.3
1990 691.4 1125.6 755.7 707.4 369.6 81 1.4
1991 689.9 1177.8 772.6 717.5 366.2 844.7
1992 691.6 1216.4 746.6 726.9 361.5 865.5
1993 684.5 1254.6 704.1 736.9 367.8 879.3
1994 688.0 1278.2 690.7 746.7 383.7 887.6
1995 712.5 1316.7 758.6 764.6 402.7 904.4
1996 726.8 1354.5 793.3 776.5 395.7 932.3
1997 739.4 1383.9 773.6 768.9 400.1 956.9
1998 750.1 1406.7 841.1 772.5 397.6 981.3
999 (may) 763.1 1433.5 757.6 778.2 377.3 1002.8
;ource: The Oil and Gas J
200 2. The Elements of Economic Calculation

The CE (Chemical Engineering) Plant Cost Index

This is a more recent formulation, published in each issue of the journal Chemical Engi-
neering, i.e., twice a month. Its reference period is 1957-1959. The CE plant cost index is a
mixed index, representing the weighted average of four main components:
0 Equipment, rotating machinery, and structures (6 1%)

0 Erection (22%)

0 Buildings (7%)

0 Engineering (10%).

The first component is itself composed of the following elements: assembled equipment
(37%), pumps and compressors (7%), other machines (14%), piping, valves, and fittings
(20%), instrumentation and control (7%), electrical equipment (5%), structures, insulation,
and painting (10%).

c. Other Countries
It would seem both difficult and tedious to embark on a bibliographic study of the various
indices published worldwide, especially as, to do justice to their local features, it would be
necessary to revisit some of them. In the present case, the aim is not to add to the list of exam-
ples presented, but rather to analyze the principles of their design, and to show the develop-
ment of the ideas on which they are based.
However, by way of illustration, we should also mention some additional examples:
0 For the United Kingdom, the index published each month in the journal Process Engineer-

ing. The purpose of mentioning this publication is not so much to show the approach
employed, as to draw attention to its evolution over time. Thus, originally, the journal, which
until 1972 was called Chemical and Process Engineering, presented the “plant construction
cost index”, calculated by using the following expression:

0.075N. (Io)*
P = 0.46M + 0.09E + 0.18C+
V
in which
M , E, and C are the price indices for equipment and energy employed by the mechanical
engineering, electrical engineering, and construction industries
I0 is the average wage index for the construction industry
V represents the investment costs of the food processing, tobacco, refining, chemical,
paint, and ink industries
N is the number of employees in the construction industry.
This information is available from these official British organizations: the Department of
Trade and Industry, and the Department of Employment and Productivity. First established
with a base of 100 in 1965, it was modified and adjusted to a new base 100 in 1970. This same
publication also publishes a “productivity” index (Productivity index, manufacturing). In
January 1973, the journal’s change of name was accompanied by the appearance of a new
index: the PE (Process Engineering) index, based on the following expression:
 2. The Elements of Economic Calculation 20 1

1 4 . 3 7 ,1 -k 0.08 IE-k 0.lozc + 0.19 Zs + 0.26 Zo

in which
ZM represents the index for the mechanical engineering sector
ZE the index for electrical engineering
Zc for civil engineering
Is describes yard improvement engineering
lo refers to general services engineering.

For Germany, the monthly publication by Messrs. Von Kolbel and Schulze, in Chemische
Zndustrie, of a cost index specifically for chemical facilities. It is of the second type, and is
based on the following average distribution of investments for assembled units: static equip-
ment (26.7%), rotating machinery (13.3%), piping (16%), metal structures (5%), electrical
equipment (7%), instrumentation (7%), industrial buildings (12%), and indirect site costs
(13%). The values for these components can be found in documents published by the Statis-
chen Bundesamtes zur Verfugung, for a base of 100 in 1962.
0 For the Netherlands, the compilation by WEBCI (Werkgroep Begrotingsproblemen in der
Chemische Industrie) of a weighted cost index, as follows:
0 Reference year: 1966

0 Composition: principal equipment (33%), piping (8%), electrical installations (3%),

instrumentation (8%), civil engineering and metal structures (6%), erection (including
thermal insulation and painting), engineering, and overheads (42%)
0 Components: figures published by the Centraal Bureau Voor de Statistick.

F. Principal Features and Consistency of Updating lndices

The great variety of updating methods and their constituents tends to make potential users
doubtful, even skeptical, about the reliability of the results obtained from them. Nevertheless,
as long as the comparison is accompanied by minimal precautions, they do have a certain
number of shared characteristics.
0 The first, as shown by Fig. 2.14 for the Nelson-Farrar index, is that of the two main com-

ponents to which any updating procedure can be reduced, the one concerning labor tends to
rise more strongly than the one for materials. In the latter the prominence of plateaus and even
occasional drops should also be noted. However, it has to be remembered that the two ele-
ments do not have the same weight, since roughly speaking, and whatever the index con-
cerned, as long as it refers to industrial facilities the ratio of the labor to the materials item is
around 60140.
0 The second, shown in Figs. 2.15a, 2.15b, and Fig. 2.15c, is more difficult to bring out. It
assumes, in fact, that there has first been a certain normalization of values, in particular a
change in the reference year for the indices being compared. In addition, a comparison of this
nature has no significance unless it is restricted to updating formulas that relate to facilities
that are fairly similar in terms of their equipment.
202 2. The Elements of Economic Calculation

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000

Figure 2.1 4 Nelson Refinery (inflation) Construction Cost Index.

(Source: The Oil and Gas J )

Four indices were selected for this study, which, in view of the availability of information
and especially its variety, concerned the United States:
Chemical Engineering Plant Cost Index
PEP Cost Index from the Stanford Research Institute (SRI)
Marshall & Swft Index
Nelson Refinery Construction Cost Index.
If we compare the variation over time of these indices, reduced to the same base year 1958,
we see that they all rise after that time, but at different rates (Fig. 2.15a). The Nelson index
 2. The Elements of Economic Calculation 203

shows the largest gains (averaging 5.3% per year over the whole period), versus 4.9% for the
SRI’s PEP, 4.2% for the Marshall & Swift, and 3.7% for the Chemical Engineering index.
Figure 2.15b, which shows the annual percentage variations, confirms this visual impression
of differentiated rates, but on the other hand it brings out an overall similarity in the trends:
moderate variations until the early 1970%a good ten years of stronger growth, return to a
more stable situation at the end. Table 2.21 provides the figures for these three periods. Also,
on Fig. 2.15b, note the very strong changes that followed the petroleum crises of 1973 and
1980, as well as the almost negligible variation during the counter-crisis of 1986.
Although it is somewhat reassuring to find a basic consistency among the observed varia-
tions, nevertheless a little explanation is required in order to interpret, or even to justify, the
different amplitudes reflected by the curves.

1955 1960 1965 1970 1975 1980 B5 1990 1995 Do

+Chemical Engineering -D- PEP (SRI)

-e- Nelson -m- Marshall & Swift

Figure 2.15a Comparative variation of cost indices in the USA (base 100
in 1958). (Sources: The Oil and Gas 1,Chem. Eng., Slu)
204 2. The Elements of Economic Calculation

+Chemical Engineering + PEP (SAI)

+ Nelson + Marshall & Swift

Figure 2.1 5b Annual variations (%) in U.S. cost indices.

(Sources: The Oil and Gas .lChem.
, Eng., SRI,)

Table 2.21 Comparative variations of various North America cost indices

from 1958 to 2000 (average % per year)

1970-1982 1982-2000

CE 2.0 7.9 1.2

M&S 2.4 7.8 2.4
PEP 3.0 9.4 2.5
Nelson 4.6 8.6 2.8

To do this, it would appear necessary to take a closer look, particularly at the composition of
the synthetic indices concerned, and the actual definitionsof their components. For example, the
breakdown given by SFU for the PEP cost index is very similar to the one used by Chemical Engi-
 2. The Elements of Economic Calculation 205

neering. But although all the components of the first index include labor contributions at about
the same levels as the second one (the CE cost index), the accompanying description states that
the contributionsare adjusted “for constant productivity”. In other words, SRI compiles its index
as if there were no productivity gains over time, whereas Chemical Engineering corrects the
labor percentage by a “work improvement” factor, set at 2.5%per year up to 1982, and reduced
to 1.75%thereafter. In these circumstances, in order to check the consistency of the two indices
it seems appropriate to use this information to calculate a corrected CE cost index, as follows:
1958 :i=O
1959-1981 : 1 I i I 2 3
CE,,,,’ = 0.433 x CE, x (1.025)’+ 0.567 x CEi
1982-1994 : 24 I i I 3 6
CE,,,,= 0.436 x CE, x (1.025)23x (1.0175)’-23+ 0.564 x CEi
> 1995 : extension of the preceding expression.

The coefficients 0.433 and 0.436 represent the labor contributions used by the authors.
Figure 2.1% shows that when this correction is made, the new variation curves for the CE
and PEP cost indices are very similar. Although an even more detailed analysis would no
doubt be required before such an investigation could be considered rigorous, the result
obtained nevertheless tends to prove that the two approaches are consistent.
The problem is a little different with the Marshall & Swift index. In fact, regardless of the
productivity gain aspect, it must be noted that a smaller figure is used for the labor contribution.
This is explained by the fact that the M&S cost index is concerned with the installed costs of
equipment, not with complete production units. For this reason, it only considers the costs of
installation on an already improved site. Using the same approach as before, the variations of
the Marshall & Swift index can be brought into line with those of the corrected CE cost index,
but using weight factors of 0.15 and 0.85 respectively, since the labor share in the overall M&S
formulation is only 15%. In spite of a noticeable (but never greater than 10%)difference for the
period 1980- 1985,we once again observe a strong similarity between the representative curves.
Finally, the Nelson curve shows similar variations to those of the PEP, but with a stronger
average growth. This is mostly explained by the more rapid rise in construction labor costs in
the refining than in the chemical and, more generally, in other sectors of activity. For exam-
ple, from 1972 to 1992 they went up by a factor of roughly 3, whereas in the chemical sector
the ratio was closer to 2.5. This is no doubt due to the fact that, in comparison with other
industries, refining has two specific differences: it involves large unit capacities, and pro-
cesses products that are regarded as dangerous and that are often under pressure. Conse-
quently it was very quickly obliged to move to units that operate continuously, to adopt very
strict safety standards (minimum thicknesses for pressure vessels, a high level of automation,
centralization of information and control, special equipment, etc.), to employ more skilled
and thus better-paid personnel, to supervise operations more closely, etc. Looking at the size
of the “wage” component in the composition of the total index (>60%), it is easy to see why,
in this respect, refining has opened up a significant gap between itself and the other industrial
sectors, and maintained it over time. Nevertheless, if we wish to judge the consistency of the
Nelson index against the other indices, there is no reason why its figures should not be cor-
rected for productivity gains, which were not considered in its design.
206 2. The Elements of Economic Calculation

1955 1960 1965 1970 1975 1980 1985 1990 1995 2000

-c-
Q-
Chemical Engineering
Nelson
-+- Chemical Engineering
(corrected)
I
Figure 2.15~Comparative variations of the Chemical Engineering and
PEP indices (base 100 in 1958). (Source: Chem. Eng., SRI,)

2.3.2.3 The Location Effect

A. Background
It is often necessary to compare the costs of facilities constructed in different countries or on
different sites. The matter sometimes takes the form of a question, as follows: what is the cost
of an industrial unit under the local conditions prevailing in another geographical region,
given that we know all the costs of a “reference” unit that is identical in every way-design,
technology, and size-at a known location?
Experience shows, in fact, that just applying the currency exchange rate for a given date,
when two different countries are involved, leads to incorrect results: the difference from the
true figure may reach or even exceed 25%, depending on the period concerned. Although they
 2. The Elements of Economic Calculation 207

are smaller, disparities often also appear in comparisons within the same country, especially
when it is a large one or the development and operating conditions are very different. This
phenomenon has many explanations, not only technical but economic, social, political, and
financial: variations in the costs of labor, supervision, materials, etc., norms and standards
peculiar to each region (equipment characteristics, construction and safety regulations, etc.),
calculation methods that may differ from one engineering company to another, tax codes, reg-
ulatory restrictions, assistance, various incentives, etc., existing infrastructure, availability
and quality of labor, local customs, and so on.
But when put this way, the question implicitly assumes that the comparison concerns
extreme situations, in which all of the actions required to complete the project, from its design
phase up to the construction of the facilities, are entirely performed on one site or the other,
and under the corresponding conditions. Now, the localization effect can take much more
subtle and complicated forms, for example if we consider the successive locations in which
the preliminary studies, the production of the various pieces of primary and secondary equip-
ment, and the erection of the manufacturing unit proper take place. In heavily industrialized
countries the equipment is often manufactured in the same place where it is assembled, and
the same applies to its design and application; at the very least, even if this is not the case,
these activities occur in areas where industrial practices are much the same. Moreover, so as
to avoid possible disputes and especially to provide reference points, they refer to a type
installation that is readily identifiable.
In industrializing regions, it is necessary to distinguish between these different stages in
the construction of a manufacturing complex. In fact, equipment made of high alloy steel or
employing a special technology is usually imported; similarly, the specialized personnel
required to erect and start it up are brought in. In this case, it is necessary to allow for the addi-
tional costs involved in transporting the equipment and waiting for its delivery, as well as for
the use of skilled foreign labor. Conversely, local conditions may favor the use of worksite
personnel at very attractive rates of pay.

B. First A ffempfs
The development of a relatively systematic approach occurred only recently. Until then, most
authors were content simply to acknowledge the situation, and, at best, when they had enough
information, to make a limited statistical analysis. However, companies have proved to be
very helpful, since their work has allowed a progressive development of more rational meth-
ods of correlation, having a certain predictive character.
At first, information on local disparities was confidential, with each company using its
own figures, even its own approach, in this regard. Without trying to compile a complete his-
tory, some stages in making this kind of information available for general use nevertheless
deserve to be mentioned.
Thus, in 1963, W.L. Nelson connected the ratio of costs for facilities constructed overseas
and in the United States to the wages paid in the same countries. He showed that where there
was cheap local labor (South America, Southeast Asia, Middle-East, etc.) there were gener-
ally, at that time, high construction costs. Exceptions included Japan and Belgium, with free-
enterprise systems and highly skilled but poorly paid personnel, and Sweden and Spain, with
many state companies and wages that were high in one case, low in the other.
208 2. The Elements of Economic Calculation

In 1969, R.J. Johnson (using the Gulf Coast as a base) published comparative costs for
pieces of equipment and complete units in the principal industrialized countries (Tables 2.22
and 2.23).

Table 2.22 Relative costs of process equipment in 1969 (from R.J. Johnson)

United Germany
USA Vetherlands Belgium France Italy Japan
Kingdom (FRG)
Furnaces 1.oo 0.95 0.90 0.90 0.90 1.00 1.00 0.90
Tanks 1.oo 1.oo 1.10 1.oo 0.98 0.95 0.85 0.80
Exchangers 1.oo 1.10 1.oo 1.oo 0.90 1.10 0.85 1.15
MPS 1.oo 0.90 0.90 0.85 0.85 0.82 0.80 0.80
Compressors 1.oo 0.85 0.85 0.85 0.75 0.80 0.80 1.00
Piping 1.oo 0.91 1.oo 0.95 0.91 0.88 0.87 0.90
Structures 1.oo 0.90 0.95 0.95 0.90 0.90 0.85 0.85
Instruments 1.oo 1.05 1.05 1.05 1.10 1.05 1.00 1.00
Insulation 1.oo 0.95 1.oo 1.oo 0.90 0.90 0.90 0.95
Electricity 1.oo 0.95 0.95 0.93 0.93 0.90 0.88 0.85
Average 1.00 0.96 0.97 0.96 0.91 0.92 0.89 0.91
(Source: Chem. Eng.)

Table 2.23 Relative costs of process units in 1969 (from R.J. Johnson)

Equipment
Land
Engineering
USA

1.oo
1.OO
United
(Gulf Coast) Kingdom
1.oo 0.96
0.95
0.75
0.91
--l-rl
l 0.97
0.95
0.80
0.92
Wi-
Jetherlands Belgium

0.96
0.95
0.90
0.91
0.90
0.80
0.92
0.90
0.90
1 1 1
France Italy Japan

0.89
0.85
0.80
0.80

Average 1.00 0.94 0.88 0.91 0.86 0.83

The relative figures given for the erected units (battery limits investments) indicated lower
costs than those used in Nelson’s correlation. These differences arise as much from the lack
of precision in the methods applied and the statistical information available, as from changes
in the situations analyzed between 1963 and 1969.
In 1973, J. Cran presented the first principles of a method for comparing costs from var-
ious industrialized countries, taking the United Kingdom as reference. For a certain number
of refining and, especially, petrochemical facilities, with identical production capacities and
for which he had the same kinds of data, he used distribution curves to determine the repre-
sentative average costs, and then the corresponding localization coefficients. These resulted
 2. The Elements of Economic Calculation 209

in overall coefficients for each country. For example, for the United States he obtained a ratio
of 1.2 to 1.3 versus the United Kingdom, for April 1973. This analysis led to a dynamic
approach, whose main elements are presented in Section 2.3.2.3.C.
Also at this same time, P. Terris suggested order of magnitude scales for the breakdown
of erected costs (Table 2.24), which enabled investments made in France in 1973 to be com-
pared with those made in industrializing regions. For this purpose, he distinguished three cat-
egories of country, according to their level of development, particularly their ability to
undertake civil engineering and erection works.

Table 2.24 Relative levels of investment cost for petroleum and

petrochemical facilities in 1973 (from!I Tems)

I France
Principal equipment 38 42 40 42
Secondary equipment 19 21 20 21
Erection (including supervision) 26 31 44 54
Civil engineering 17 18 19 30
(including supervision)
Total 100 112 123 1
I

147
say, 110 to 115 say, 120 to 125 say, 140 to 150
(Source:AFTP)

His categories are:

0 Countries already having some industrial experience, especially in the petrochemical

area, where the foreign contribution consists of providing supervisory and training per-
sonnel (type l),
0 Countries where the local labor has good experience in civil engineering and public

works, and uses modern equipment for this purpose (type 2),
0 Finally, countries where the labor is unskilled, and where civil engineering and erection

equipment is non-existent (type 3).

In all these countries, the heavy-metal and mechanical industries, etc. are lacking. The
equipment itself is therefore mostly imported.
In a book published in the third quarter of 1974, K.M. Guthrie gave a list of location fac-
tors, which include the costs of equipment, labor, and indirect expenses. They were compiled
for mid-1970, and are shown in Table 2.25.
In 1973, and then in 1976, ten years after his first attempt, W.L. Nelson once again pub-
lished work concerning the effect of the location of refining facilities on their costs. These
studies successively addressed the impact of the access costs for materials, wage costs and
productivity, and finally a very detailed world survey, using the “Gulf Coast” American units
for the period 1971- 1972 as reference.
It would be time-consuming to present the full results of Nelson’s work here. As a rough
guide, Table 2.26 presents the recommended values for most of the countries considered.
210 2. The Elements of Economic Calculation

Table 2.25 Influence of location, mid-1970 (from K.M. Guthrie)

Country Coefficient Country Coeticient Country Coefficient

Canada 1.05 Brazil 1.10 Belgium 0.96
USA 1.oo Chile 1.08 Denmark 0.86
Mexico 1.08 Peru 0.86 Spain 0.76
Puerto Rico 1.04 Uruguay 1.06 France 0.98
Trinidad 1.08 Venezuela 0.92 N. Ireland 0.87
Argentina 1.06 Australia 1.07-1.1 1 Rep. of Ireland 0.79
Bolivia 0.98 Germany (FRG) 0.92 Italy 0.96
United Kingdom 0.89
(Source: Chem. Eng.)

Table 2.26 Effect of location on the cost of refining units during 1971-1972 (from W.L. Nelson)

Location Suggested value Location Suggested value

North America Middle East
USA: Iraq 1.020
Gulf Coast 1.OOO (reference) Iran 0.970
California 1.090 Kuwait( ) * 1.010
East Coast 1.250 Western Europe
Illinois 1.130 Germany 0.910
Louisiana 1.085 Belgium 0.890
Pennsylvania 1.ooo Denmark 0.970
Texas 1.ooo Spain 0.860
US average 1.066 Finland 0.970
Canada: France 0.880
Alaska 1.350 Greece 0.850
Canada (central) 1.025 Ireland 0.770
Canada (East) 1.060 Italy 0.820
Canada (West) 1.050 Norway 0.960
AK-Canada average 1.035 Netherlands 0.880
Latin America(') Portugal 0.850
Argentina 1.010 United Kingdom 0.890
Brazil 0.970 Sweden 0.970
Colombia 0.920 Turkey 0.860
El Salvador 0.910 Central Europe 0.890
Mexico 0.940 Asia-Pacific
Nicaragua 0.930 Australia 0.960
Puerto Rico 0.930 South Korea 0.840
Venezuela 1.030 India 0.950
Africa Indonesia 0.900
South Africa(2) 0.940 Japan 0.800
Gabon 1.020 Malaysia 0.840
Ghana 1.030 Pakistan 1.010
Libya 1.280 Singapore 0.830
Tanzania(2) 1.020 Taiwan 0.870
Zambia(2) 1.050 Thailand 0.930
(1) North American equipment. (2) European equipment.
(Source: The Oil and Gas 1)
 2. The Elements of Economic Calculation 21 1

Among recent publications providing specific accounts of location problems, the work of
C.A. Miller and A.V. Bridgewater must also be mentioned.
The first suggests an actual calculation procedure in the form of a matrix with four columns
for the most important parameters, permitting an assessment of the effect of the site (direct
and indirect labor costs, direct cost of materials, engineering and management expenses). The
rows give the “reference” (Gulf Coast) values generally accepted for these items, and then the
values of the assumed corrective coefficients, depending on the local conditions, productivity
and material costs, travel and subsistence expenses, accommodation, overtime, etc.
The second publication, from 1979, provides the relative costs for the USA and the United
Kingdom of chemical facilities with the same functions, making it clear that they are only
orders of magnitude, and that in the same country it is possible to find major fluctuations with
respect to these average values (Table 2.27).

Table 2.27 Location factors for similar chemical facilities (from A.V Bridgewater)

Values Values
Location United Location United
USA
Kingdom Kingdom
Australia 1.40 1.20 1.10
Austria 1.10 1.oo Netherlands 1.10 1.oo
Belgium 1.10 1.oo New Zealand 1.40 1.30
Canada 1.25 1.15 North Africa ( ) 1.20 1.10
China(‘) 1.20 1.10 (2) 0.80 0.75
(2) 0.60 0.55 Norway 1.20 1.10
Central America 1.10 1.oo Portugal 0.80 0.75
Denmark 1.10 1.oo South Africa 1.25 1.15
Finland 1.30 1.20 South America (N.) 1S O 1.35
France 1.05 0.95 ~ South America (S.) 2.50 2.25
Germany 1.10 1.oo Spain 0.80 0.75
Greece 1.oo 0.90 Sweden 1.20 1.10
India (l) 2.00 1.80 ~ Switzerland 1.20 1.10
(2) 0.70 0.65 Turkey 1.10 1.oo
Ireland 0.90 0.80 United Kingdom 1.oo 0.90
Italy 1.oo 0.90 1.10 1.oo
Japan 1.oo 0.90 1.oo 0.90
Malaysia 0.90 0.80
I) Imported equipment.
!) Locally-made equipment.
(Source: Process Economics International)

C. Development of Systematic Methods

Among the early attempts to put together a dynamic system for estimating the influence of
location, instead of just keeping track of it periodically, we must first mention the various
authors or publications who compiled composite indices that could be applied to different
212 2. The Elements of Economic Calculation

countries and bring out their specific features, especially in terms of wages and materials.
Whether they had many components or only two, these indices offer the great attraction of
taking into account two of the principal parameters that explain the observed disparities
between one geographic region and another, and especially of presenting formulations of a
universal nature, at least in most cases. On the other hand, they have the drawback of covering
only a limited number of countries, most of them having a level of industrialization or devel-
opment sufficient to allow access to the local values of the selected components. They have
also all adopted a common reference date, for which, whatever site is concerned, the index
arbitrarily takes the value 100. In this respect, the approach has not been carried quite far
enough in terms of location factor and currency exchange as to enable an easy conversion
from one country to another.
In this regard, J. Cran must first be mentioned: beginning in 1976 and throughout the life
of the journal Engineering and Process Economics, i.e., until 1991, he regularly published a
two-component index, the EPE index, which could be calculated for various industrial coun-
tries and thereby show their relative variations with respect to a given date, on which, what-
ever the site, it had the same value: 100 in 1970 (Fig. 2.16).
This approach can certainly be grouped with the publication, in 1979, of the PEI Plant Cost
Index, in the first issue of the journal Process Economics International, published by
D.H. Allen and A.V. Bridgewater. With a value of 100 for the first quarter of 1975, it has
the advantage of covering a greater number of countries than Cran’s index, with varying lev-
els of industrialization, situated not only in Western but also in Eastern Europe, North and
South America, Africa, the Middle East, the Pacific, and Southeast Asia.
In the same vein, the changes progressively made in the PE Index, published by the journal
Process Engineering, must also be noted (see Section 2.3.2.2.D.c). The reference year has
been changed on various occasions; in addition, this index was also calculated from 1983
onwards for other countries besides the United Kingdom, in particular Australia, Canada,
Japan, South Africa, and the USA. A final change took place in September 1991, with the
adoption of a base 100 for 1990, and an extension of the approach to the United Kingdom’s
principal trading partners: Australia, Canada, Denmark, France, Germany, Italy, Japan, the
Netherlands, New Zealand, Spain, Sweden, and the USA. In addition, the name was changed
to the “PREDICT” (Process Engineering Data and Information Cost/Time) Indices. The for-
mula remains the same for the United Kingdom, the components being drawn from the pub-
lications of H.M. Government’s Central Statistical Office. For the other countries, the indices
are reduced to two constituents, whose values are obtained from international publications,
or those of organizations such as the OECD; the calculation is effected monthly in the first
case, quarterly in the second.
It is always dangerous to make a claim to completeness when one is attempting to recon-
cile the chronology of the methods proposed and the successive advances to which they con-
tributed. Accordingly, we hope that the authors or publications that have not been mentioned
here will not be offended by the omission. Similarly, we trust that they will also not be upset
if the emphasis here is placed on two particular methods, one developed by N. Boyd, the other
by the IFP.
In both cases, the approach to location consists of first setting up a composite index, which
allows the relative variations for different countries to be tracked. Generally speaking, with
such methods the main problem lies in the choice of components, which have to be relatively
 2. The Elements of Economic Calculation 213

500

450

400

350

- 300

250

200

150

100
1970 1972 1974 1976 1978 1980 1982

Figure 2.1 6 Variation of the EPE indices (from J. Cran).

(Source: Engineering and Process Economics)

accessible for the largest possible number of different sites, while maintaining a certain level
of accuracy. This is why, when all is said and done, it is easiest to assign priority to indices
with two constituents: wages and materials.

D. N. Boyd’s Indices
The first merit of Boyd’s approach is that it provides a certain consistency with the observed
facts, beginning with the development of his updating formula, whose results were published
regularly in the journal Engineering Costs and Production Economics until a few years ago.
This periodical unfortunately ceased publication in 199 1.
214 2. The Elements of Economic Calculation

In practice, while there are many parameters capable of explaining the differences in costs
from one site to another, and thereby providing a location factor that can be applied for a
given date at the current exchange rate, some appear to be more relevant than others, and par-
ticularly these three:
0 Productivity, i.e., the number of hours required to complete the same work,

0 The costs of benefits, which vary from one country to another, producing wages and

hence hourly labor costs that are different,

0 The costs of obtaining materials, or in other words the access costs, depending on the

location of the installation of the processing facilities or the construction sites.

On the basis of this kind of working hypothesis, Boyd compiled the so-called “erected
costs of plants” indices (base 100 on 1/1/75) for 13 OECD countries (Australia, Belgium,
Canada, Denmark, France, Germany, Italy, Japan, Netherlands, Norway, Sweden, United
Kingdom, USA), whose component elements are based on only three differentiating criteria.
His first step was to conduct a statistical analysis of the available information, which showed
that on average the cost of a continuously operated chemical facility could be broken down
as follows:
0 7% of civil engineering and construction materials, with the following distribution:

- 3Y0cement
- 3% steel bars
- 1% steel structures,
63% of mechanical and electrical materials and equipment (including engineering and
purchasing), divided into:
- 1% steel structures
- 22% sheet steel
- 30% industrial labor,
0 30% of site assembly costs (construction labor).

Expressed differently, this distribution is equivalent to the following normalized break-

down:
0 10 tons of Portland cement

0 1 ton of steel bars (for reinforced concrete)

0 3 tons of steel (I-beam) structures

6 tons of sheet steel

0 300 hours of industrial labor (mechanical and electrical)

0 230 hours of construction (site) labor.

In a second step, Boyd displayed his originality by bringing together both the above break-
down and the three main explanatory measures from the location factor, in one and the same
formula, namely:
7 1
C = 10 4 + p2 + 3 p3 + 6 4 + 3 0 0 4 -
l+F
+ 230 L2 F
L.

where
P I: cost of a ton of Portland cement
P2:cost of a ton of steel rods
 2. The Elements of Economic Calculation 215

P3: cost of a ton of steel structure

P4: cost of a ton of steel sheet
L , : hourly cost of industrial labor
L2 : hourly cost of site labor
F : productivity factor for site labor relative to the United States.
The first four terms refer to the access costs for materials, which may thus have different
values, depending on the country; the last two allow for the incorporation of variations both
in benefits and in productivity. The use of this latter factor to correct the wage costs is based
on observations and statistical compilations made by the author and are his idea alone. How-
ever, it must be admitted that at the time when Boyd was studying the matter, he was not far
wrong; fortunately, things have changed since then. Thus, for France, it may at first seem sur-
prising that F should be 0.6 relative to the United States on January 1st 1975, which means
that on that date:
For industrial labor, where North Americans took 300 hours to manufacture a piece of
equipment, the French took 375 hours for the same work,
For site labor, where North Americans provided for 230 hours, the French needed to
allocate 380 hours for an identical task.
When the matter is examined a little more closely, the difference can be explained. Thus
in 1975, for example, in France forms were prepared by sawing and assembling planks along
each side of a framework of steel bars put together by hand onsite; at that time in the USA
they were already doing it, as is now also the custom in Europe, by using a crane and mechan-
ically positioning large metal sheets (several meters in size), with walkways on top, on each
side of complete framework units that are pre-assembled in the factory.
As for benefits, as a rough guide for comparing these same two countries on January 1st
1975, the following figures apply:
France:
L , = 1.77 S,, with S , = 11.59, or L1 = 20.51
L2 = 1.71 S2,with S2 = 10.08, or L2 = 17.24
USA:
L1 = 1.33 S,, with S1 = 5.49, or L , = 7.30
L2 = 1.28 S2,with S2 = 6.99, or L2 = 8.95
In a third step, Boyd next calculated the value of his composite index C for the 13 selected
OECD countries, at various dates (Table 2.28). He then turned them into “location factors”
relative to the United States, by arbitrarily assigning the value 1.O to that geographic region,
whatever the date concerned, and then for a given date dividing the figures for each country
by the one for the USA (Table 2.29). This type of calculation may be carried out monthly or
annually: one only has to obtain the values of the parameters from publications or specialized
organizations whose business it is to compile and maintain them. The author himself had
formed the habit of checking them on the first day of January each year.
As already mentioned above, this “location factor” acts as a corrective to the current mon-
etary parity. So if we refer to the published figures, it is the same as applying it, on January
1st of a given year, to the exchange rate for a particular country versus the USA on that date.
 Table 2.28 Indices for costs of erected facilities* on January lst, unless otherwise indicated (1975 = 1.00) (from N. Boyd)
-
1977 1978
-
1979 1980 1981 1982 1983
----
1984 1985 1986
1987 I 1987
onJulylst
1988
-
1988
on July 1st
Australia 1.33 1.46 1.57 1.71 1.93 2.15 2.58 2.77 2.91 3.15 3.36 3.45 3.55 3.60
Belgium 1.16 1.27 1.29 1.40 1.50 1.64 1.89 1.99 2.01 2.06 2.07 2.10 2.10 2.17
Canada 1.33 1.46 1.56 1.76 1.95 2.24 2.46 2.42 2.47 2.52 2.57 2.61 2.65 2.74
Denmark 1.06 1.16 1.27 1.38 1.54 1.64 1.99 2.14 2.19 2.32 2.41 2.59 2.63 2.75
France 1.25 1.34 1.46 1.59 1.82 2.05 2.50 2.73 2.84 2.97 3.10 3.17 3.21 3.26
Germany 1.07 1.11 1.14 1.21 1.29 1.38 1.53 1.52 1.56 1.60 1.60 1.61 1.62 1.66
Italy 1.43 1.67 1.92 2.27 2.67 3.15 3.89 4.38 4.71 5.12 5.3 1 5.53 5.60 5.76
Japan 1.26 1.39 1.so 1.58 1.70 1.79 1.84 1.84 1.84 1.89 1.87 1.86 1.90 1.92
Netherlands 1.14 1.22 1.18 1.24 1.31 1.42 1.54 1.54 1.57 1.59 1.63 1.61 1.65 1.65
Norway 1.22 1.35 1.45 1.53 1.69 1.79 2.07 2.18 2.28 2.40 2.73 2.99 3.16 3.23
Sweden 1.16 1.24 1.37 1.49 1.67 1.81 2.05 2.26 2.73 2.57 2.77 2.89 3 .OO 3.12
United Kingdom 1.45 1.61 1.76 2.04 2.35 2.56 2.78 2.92 3.10 3.34 3.56 3.67 3.75 3.84
USA 1.18 1.27 1.39 1.53 1.68 1.85 1.96 2.09 2.13 2.19 2.10 2.12 2.20 2.27
~ -
* for details of the method for calculating costs and indices, see Eng. Costs Prod. Econ., 6 (1982) 272, an 7 (1982
(Source :Engineering Costs and Production Economics)
 Table 2.29 Comparison of erected costs of facilities with US. costs (1.00) on January lst, unless otherwise indicated (from N. Boyd)

Australia
1976

0.89
1977

0.94
1978

0.85
1979

0.87
1980

0.84
1981

0.90
1982

0.89
7-
1983

0.86
1984

0.80
1985

0.76
1986 1987

0.67 0.68
1987
on July 1st
0.78
1988

0.77
1988
on July 1st
0.79
Belgium 1.oo 1.01 1.10 1.18 1.36 1.21 0.92 0.81 0.72 0.60 0.68 0.92 1.03 1.oo 1.05
Canada 0.95 1.05 0.98 0.87 0.90 0.90 0.92 0.92 0.86 0.81 0.76 0.80 0.85 0.84 0.89
Denmark 1.01 0.98 0.98 1.10 1.13 1.05 0.80 0.76 0.72 0.62 0.71 0.98 1.15 1.12 1.23
France 0.93 0.89 0.87 0.96 1.02 1.04 0.82 0.75 0.68 0.58 0.67 0.90 0.99 0.98 1.02
Germany 0.98 0.99 1.02 1.12 1.20 1.13 0.88 0.85 0.76 0.64 0.73 1.01 1.12 1.09 1.18
Italy 0.88 0.81 0.82 0.91 1.01 1.01 0.78 0.79 0.74 0.65 0.72 1.00 1.11 1.07 1.14
Japan 0.55 0.59 0.65 0.82 0.79 0.77 0.73 0.60 0.62 0.58 0.65 0.94 0.96 0.99 1.11
Netherlands 1.05 1.05 1.12 1.13 1.18 1.12 0.82 0.82 0.73 0.61 0.69 0.96 1.06 1.04 1.10
Norway 1.16 1.28 1.25 1.29 0.31 1.32 1.07 1.00 0.92 0.82 0.89 1.15 1.28 1.37 1.45
Sweden 1.18 1.24 1.13 1.19 1.26 1.29 0.99 0.90 0.76 0.70 0.79 1.03 1.14 1.16 1.26
United Kingdon 0.81 0.70 0.72 0.80 0.94 1.08 0.86 0.79 0.68 0.59 0.70 0.81R 0.90 0.92 1.03
USA 1.oo 1.00 1.00 I .oo 1.00 1.00 1.oo 1.00 1.00 1.oo 1.00 1.00 1.oo 1.oo 1.oo
-
(Source :Engineering Costs and Production Economics)
218 2. The Elements of Economic Calculation

In fact, and this is Boyd’s third original idea, the author suggested that instead of working
with such a monetary ratio, one should take an approximate 60/40 weighting of this ratio’s
values, three months and six months before the date concerned. Indeed, in order to deliver a
piece of equipment in operating condition on January 1st of any year, the required materials
have to be ordered, and their prices agreed to, on about the previous July 1st, and, considering
the time needed for their delivery, the manufacturing crews must be put to work on October
1st, which sets the manufacturing costs. Given, as we saw for example with the Nelson index,
that in a two-component system the weight of labor is of the order of 60%, while materials
are around 40%, we can better appreciate the value of Boyd’s practical rule. This also
explains why each location factor that is published on a given date is accompanied by a cal-
culated exchange rate that is different from the published one for the same date. Actually, the
weighting selected may vary over time; moreover, it is not exactly the one stated, which was
chosen in order to simplify the presentation.

E. The Concept of the Technical Dollar Exchange Rate (IFP)

This concept was originally based on the simple observation that applying the published
exchange rate between a given currency and the dollar at a given moment did not provide a
realistic evaluation of the investment costs for the same type of facility in the two countries
concerned. To express this fact, we need to introduce a theoretical exchange rate that, by def-
inition, is given by the following expression:
Investment in country X
Technical parity in the yearn -
-
in year n, in local currency
between and currency and the US$ Corresponding investment in the USA
in year n, in US$

This equation applies to two exactly similar units, but in one case the planning, the mate-
rials, the manufacturing and site labor, etc. are assumed to be done or produced only in coun-
try X , and in the other situation these same services are supposed to come entirely from North
American raw materials and personnel.
It can be expressed in another way by noting that, if we want to use the current exchange
rate to find the investment in year n in country X , when we know the investment for the cor-
responding facility in the USA, we must correct the rate by means of a coefficient which is
actually the location factor for the year concerned, i.e.:
Investment investment Monetary parity Location
incountryX -
-
intheUSA currency XIUS$ factor
in year n , in year n, in US$ in year n in year n
in local currency

In other words, if we combine the two expressions, we obtain the general rule:
Technical parity -
- Monetary parity Location factor
X
in year n in year n in year n
 2. The Elements of Economic Calculation 219

In fact, the first equation can be written in a more advanced form:

Value of the construction index
Investment in country X in year n in country X
X
in a reference year Value of the construction index
Technical parity -
- in the reference year in country X
in year n Value of the construction index
Investment in USA in year n in the USA
in same reference year Value of the construction index
in the reference year in the USA

which amounts to saying, when we generalize it:

Technical parity -
- Technical parity Ratios of values
X
in year n in a reference year of the construction index

To sum up, the procedure is therefore as follows:

0 Choice of a reference year

0 Creation of an index applicable to several countries

0 Determination of the technical parity between these countries in the reference year

0 Calculation of the technical parities for these countries at various dates

0 Identification of the corresponding monetary exchange rates

0 Obtaining the corresponding location factors.

As before, the problem comes down to first establishing a composite updating index. The
proposal that was made for France by IFP in 1984 is given in the following formulation:

I = 4.8- cM2 +2.4- LMc +8.7- 'a + 18.7-292.2 BTO 1

'Ma + 37,6-- Sime + 27.8 -
173.9 351.6 378.6 308.2 235.1
base: 100 in the year 1980.
The notations employed represent the following INSEE indices:
CM2: cement price index
LM, : Fe, E-40A steel rebars for concrete
PT, : E, 24.1(IPE) steel girders
TM, : hot rolled E, 24.2 sheet steel
Sime: labor cost for the Engineering and Electrical Industries
BTol: construction index.
In fact, the way in which this index, named "EVAL", is written incorporates, for another
reference date and in a different form, the original expression by N. Boyd; namely, for France
in 1975:
0 10 t of cement

0 1 t of steel bars for concrete

0 3 t of steel (I-beam) structures

6 t of sheet steel
0 375 hours of industrial labor (300/0.8)

383 hours of site labor (230/0.6).

220 2. The Elements of Economic Calculation

The difficulty that then arises is in applying this type of expression to other countries, par-
ticularly ones in geographic regions that are industrializing, where the required statistical
information is not always available. Bearing in mind Cran’s work, which tends to show that
a proper choice of the weighting between labor and basic materials can give a good idea of
the variation of a multi-component index, it appeared worthwhile to construct a two-compo-
nent index, called “BIEVAL”, which, for France, has the following form:

TM,
Sime + 0.33 -
I = 0.67 -
308.2 292.2
which thus includes 67% labor and 33% materials (steel).

COMMENT
0 By way of comparison, the index published by PEI (base 100: 1st quarter of 1980),
is given by the expression: I = 0.70s + 0.30M, where S = earnings in manufacturing,
and M = basic steel, obtained from the U ” s Monthly Statistical Bulletin.
0 Figure 2.17 displays the variation in monthly values of EVAL and BIEVAL, for

France since 1975. Generally speaking, we can see that over the whole period the
variations in the ratio between these two indices at the same moment very rarely
exceed 4% above and 2% below. The drift begins in 1983 and grows mainly until
1989; afterwards it stabilizes. To reduce the disparity and bring it down to less than
=k2.5%on average for the last fifteen years, while keeping the same components, the
weighting has to be changed, and brought back, in the case of BIEVAL, to a compo-
sition close to the following: 60% wages, 40% materials. Nevertheless, when all is
said and done this kind of analysis lends support to the idea that a two-component
index should be preferred only when getting hold of information sources proves to
be a problem. For all that, periodic readjustments of the original formulation are
found to be necessary if good consistency is to be maintained, in particular whenever
the period concerned exceeds ten years.

The second step is to determine the technical parity, on a particular date. If we take the case
of France vis-a-vis the United States as an example, the problem is the following, namely, to
determine the equivalent for France in July 1987 of US$1 000 in January 1984. With the tools
available at the current stage of the procedure, there are two possible routes for doing this
(Fig. 2.18):
The first consists of taking the exchange rate for January 1984,7.773 FF/US$, to go from
the USA to France, and then to apply the selected values of the updating index, namely 273
and 3 17, to obtain the equivalent of the original US$lOOO in France in July 1987, namely
9026 FF.
0 The second performs the updating right away, remaining in the USA from January 1984

until July 1987, using the corresponding values of the selected updating index, namely, in the
present case, 209 and 212, and then in July 1987 to go from the USA to France by applying
the average exchange rate for that date, namely 6.3 1 1 . The equivalent of the initial US$l 000
then comes to 6400 FF.
 .
2. The Elements of Economic Calculation 22 1

250 -

200.-

150.-

100.-

50

75
-
jan. jan.
n
jan.
79
jan.
81
jan.
83
jan.
85
jan.
a7
Ian.
89
jan.
91
jan.
93
jan.
95
jan.
97
jan.
99
jan.
01

Figure 2.1 7 Variation of the EVAL and BIEVAL cost indices (France,
base 100 on 1/1/1975). (Source: IFP)

Investments Investments
in the USA in France

x 7.773
January 1984 US$1000 b 7773 FF

x (212/209) x (317/273)

J. x 6.311
9026FF ?
July 1987 US$1014 b 6400FF

Figure 2.1 8 Influence of location:use of exchange rates.

If we accept the figures adopted for the indices as realistic, the disparity of the results
obtained can come only from the exchange rates that were applied-which means that we
have to change them, so that the two possible routes are made to converge on final values that
222 2. The Elements of Economic Calculation

are sufficiently close to one another (Fig. 2.19). The values thus obtained represent the
desired technical parities. In fact, the operation come down to defining a relationship between
the two technical parities x and y at two different dates, which, in the example concerned, is:

x -317
- 209
-- X-
y 273 212

Investments Investments
in the USA in France

x 5.28
January 1984 US$1000 b 5280FF

July 1987
x (212/209)

$.
US$1014
x 6.04
b
Ix (3171273)

6131 FF
6125FF
2

Figure 2.1 9 Influence of location: use of technical parities.

To solve this type of problem there are two possible alternatives:

To calculate the costs of the same facility in the USA in US$ and in France in FF, on a
given date, taken as a reference; the ratio will provide the base technical parity.
To proceed by approximation, working with a series of different dates and thus with cor-
responding index values. The results selected will be the ones that come closest to satis-
fying a pre-imposed correlation level (for example, a difference of less than 1%). In
practice, this approach was the one chosen.
The final step in the calculation is to determine the location factors. To do this, we have
only to divide the technical and monetary parities into each other for each date. The sole dif-
ficulty lies in choosing the values for the exchange rates. In this regard, the rule given by
Boyd has been retained: the calculated monetary parity for January 1st of any year is obtained
from a 60/40 weighting between the exchange rates on October 1st and July 1st of the pre-
ceding year.
 CHAPTER 3
The Determination of
Battery Limits Investments

As shown in Section 2.3 of the previous chapter, the battery limits investment is unquestion-
ably the one basic piece of information upon which all economic calculations are based. To
some extent, the same can be said of the complementary investments for general services and
storage facilities, the production of utilities being usually regarded as an independent matter.
The fundamental question that then arises is how to determine this cost.
Now, experience shows that in project development and management, economic and even
financial concerns have a tendency to show up at an increasingly early date, as a result of a nat-
ural and very creditable desire to avoid budgetary setbacks and slippages. It is therefore essential
to have a mechanism for finding this cost, as well as specialists capable of applying it correctly.

3.1 METHODS FOR APPLYING THE ESTIMATION PROCEDURE

3.1.1 BACKGROUND

The main problem, which is readily apparent, lies in the methods for applying this kind of
procedure, which can only function if the calculations have something to work on, i.e., data.
Obviously, depending on how advanced the project is, the information itself may be more or
less available and reliable. In these circumstances, the method chosen must first of all be flex-
ible, and designed to accommodate both the accessibility and accuracy levels of the reference
data, and the time allowed for a response. We must avoid compromising the result by reduc-
ing its accuracy still further, while at the same time not resorting to a system that demands
too much in terms of procedures, time, and personnel.
All this suggests that it may be difficult to create a method applicable to all cases: instead
we are likely to find a whole array of procedures. In addition, when it comes to deciding
which one should be utilized and how well it fits the problems to be faced, it would seem that
only a qualified person with some prior experience in the matter could provide a satisfactory
answer. This, then, is a genuine profession, and depending on the circumstances it may take
on various aspects, generally being referred to as that of estimator, sometimes evaluator, and
224 3. The Determination of Battery Limits investments

found both in engineering companies and in consulting firms, or the relevant departments of
the major manufacturers, R&D centers, etc.
In practice, the profession operates at the interface between the technical aspects of a
project and company management, between engineers and technical experts on the one hand,
and accountants and financiers on the other. Its role is to express in cost terms the operating
principles and the process data, such as-depending on the level of sophistication of these
elements-the operating conditions, the results, the material and thermal balances, the flow
diagram, the size of the apparatus, and so on. In this regard, it is important to emphasize the
essential feature of estimation, which is that it works in an a priori manner, i.e., with facilities
or construction projects that are notional, because they lie in the future. In the same way, two
main and quite different types of situation are often identified, on the basis of the kind of ini-
tial technical or even economic information they involve:
0 The first type relates to “research” projects: they rely on preliminary or confirmed exper-

imental results from laboratory or pilot and prototype development work, or from more frag-
mentary information coming from publications, patents, etc., that may furnish an idea for
research. To make use of these data it is necessary to resort to calculation methods that are
comparable to the sizing and costing procedures used by engineering companies, but which
differ from them both in the low level of reliability and detail of the available information,
and in the very short response times allowed. This situation usually corresponds to what are
referred to as preliminary or pre-project studies, and requires estimating methods that are rel-
atively rudimentary and which are grouped together under the heading of evaluation-but
this is far from being a general rule.
0 The second type concerns projects that can be described as “industrial”, based on operating

conditions of production or processing plants that are either similar to ones already in opera-
tion, or that arise from advanced studies in pilot or demonstrationplants, which express actual
behavior and provide a basis for the engineering studies that are essential for full-scale con-
struction. Such studies are applied both for the construction of a new factory and for modify-
ing an existing installation, and must thus meet specific technical and economic requirements
for the manufacture of products whose quality and price correspond to market demands; con-
sequently, their application demands very sophisticated sizing procedures and the use of ser-
vice providers as consultants for costing. The time needed to complete such work is obviously
much longer than that required for analyzing research projects. This approach is what is gen-
erally meant when the term estimation is used.
For all that, although these two situations may appear to be very different, they belong to
the same class. They are no more than the expression of two particular steps in the life of a
single project, when it is followed from its first beginnings to its concrete expression at the
industrial stage, as we shall see by examining the full sequence of events.

3.1.2 THE PRINCIPAL STAGES IN THE PROGRESS OF A PROjECT

Whatever its size, its degree of complexity, its location, or its purpose, the completion of a
project almost always involves utilizing the same approach. This means undertaking studies,
 3. The Determination of Battery Limits lnvestments 225
2. The Elements of Economic Calculation 165

purchases, and transport, then the actual construction and startup of the units. All these oper-
The investment cost for the facilities proper (unit investments), comprising:
ations are inter-related, and demand a proper scheduling over time, not just to enable each one
- The battery limits investments (on-site or on-plot investments)
of them to be performed in a technically and economically optimum manner, or to facilitate
- General services and storage (off sites or off plot),
the involvements of the various actors concerned, but also to give the decision-makers an
Fixed capital, comprising:
opportunity-at each critical step in the project-to halt it or to carry on, or to undertake cer-
- Investments related to the units
tain checks or make certain changes.
- Consulting and engineering fees
With this in mind, the timetable of activities that the client must initiate, manage, and mon-
- Supply of spare parts
itor in order to ensure the proper development of its project right up to the scheduled start of
- Contractors’ costs or licensor’s fees,
manufacture and marketing, must be of the kind shown in Fig. 3.1, which is drawn broadly
0 Depreciable capital, whose main components are:
enough to be applicable to most of the situations encountered. This diagram covers the exe-
- Fixed capital
cution of a project that extends over seven years: three of them are devoted to studies, the
- Initial loads
other four to the construction proper.
- Interest on construction loans
- Startup costs,

3.1.2.1 The Study or Pre-Study Stages

The first period actually includes two successive steps, during which the project is first iden-
tified, and then its feasibility is considered.
Figure 2.1 1 The various investment charges. (Source: ZFP)
226 3. The Determination of Battery Limits Investments

A. Identification
In principle, this consists of conducting an overall analysis of the techno-economic context
in which the project is placed, so as to fix its objectives, particularly in terms of production.
In practice, the approach is to assess the unit capacity or capacities of the production facilities,
in order to establish an order of magnitude for the amount of the investments required, and to
measure their profitability in several possible situations. More specifically, this kind of study
involves successively performing:
A market study, both for supply and demand, and in terms of quantities as well as price. This
work includes an examination of the current situation, of past data, and of a projection five to
ten years into the future under various scenarios: this must be done not only for the country
concerned, but also for neighboring states or even for the entire geographic region, which
means considering trade and consequently distribution systems, existing infrastructure, local
supplies, tax measures, etc. The procedures for doing this were discussed in Chapter 1. How-
ever, the use of them here certainly provides a better basis for evaluating their usefulness and
the way they are employed in the conduct, or, more broadly, in the management of projects.
An industrialplanningpre-study,to identify, first, the possible production targets in light
of the preceding observations. The next step involves evaluating any problems related to the
location of the facilities, calculating the total budget envelope required, considering the finan-
cial setup, defining preliminary operating costs and cost prices, undertaking general or spe-
cific profitability studies of certain steps, examining possible foreign exchange savings, and
so on. Here again, the methodology likely to be employed for this kind of analysis was dis-
cussed in detail in Chapter 2. In terms of investments, the corresponding provision is intended
to give the interested party what is usually called an “order of magnitude” budget.
The ranking of various possible alternatives, including the one of not undertaking the
project. In this regard, consideration of possible overcapacities, in anticipation of market
growth, or on the other hand the need to postpone the plant’s startup date for several years
while waiting for probable new market outlets, are just two examples of the potential situa-
tions that need to be envisaged. This kind of study is in principle based on sensitivity analysis:
usually it addresses only the economic components of the project concerned; its impact is
greatly improved when it is based directly on the most important technical parameters.

B. Feasibility Studies
These represent a major step, since the decision whether to invest in the project or not depends
on their conclusions. Their main objectives are:
To make it easier for the client to reach a decision, through a detailed examination of five
essential factors: the availability of raw materials and energy, the market outlets, the applica-
ble processes and technologies, the anticipated investment cost, and the personnel require-
ments. At this point, efforts must be aimed at identifying items that must be obtained from
third parties (supplies, foreign markets, access to technologies, outside contributions of cap-
ital, personnel training, etc.). As for the investments in particular, in this case the manufac-
turer will expect to receive what is usually called a “preliminary budget”.
 3. The Determination of Battery Limits Investments 227

To define the project’s technical and economicfeasibility. To this end, it is first of all nec-
essary to better define its technical aspects by developing manufacturing plans that fit the pro-
duction goals, while allowing a certain flexibility, particularly as regards raw materials; the
plans must also identify the essential processing stages (selection of the most suitable pro-
cesses or technologies) and constraints (accompanying services and treatments), and the
resources that the facilities must have in order to function. Next, it is desirable to consider the
infrastructure needed to begin the project, such as personnel requirements and their implica-
tions (accommodation, education, transport, etc.), both for the construction site and for oper-
ating the units, and also the arrangements for delivering equipment and materials via the
existing or planned highway, railroad, and port systems, etc. Finally a general development
timetable must be drawn up, and the project must be analyzed from an economic standpoint:
for this purpose more detailed estimates must be prepared, and annual budgets determined,
also operating costs, and probable financing arrangements and tax implications, so as to per-
form the profitability calculations. This work too may be accompanied by a sensitivity study.

C. The Transition Period

The pre-study phases also include a period of transition leading to the later stages: it begins
with the construction decision, and ends with the selection of a “contractor”, in fact an engi-
neering company, whose function is to carry out, or cause to be carried out, the actual imple-
mentation of the project, and to sign a contract with the client for that purpose. In this regard
there are two possibilities to consider: a flat fee, under which the contractor agrees to a turnkey
price and takes on most of the responsibilities, or a management or fee-for-service contract,
where the client remains the project manager but also becomes the main responsible party.
During this transition phase, a number of things have to be done; some are irreversible,
such as the choice of a site, the recommended processes, the type of contract, and definitely
the contractor; others can be revised, even though they may be necessary at the time, such as
approaches or activities of an administrative nature, the drawing up of supply or delivery con-
tracts, financial arrangements, etc. To the extent that the client is not capable of doing them
itself, these tasks are usually assigned to specialist firms, for example, consultants. This sort
of expertise is particularly employed to assist in the choice of the contractor, and in that regard
to help in drafting various bid documents and in analyzing the offers resulting from those
requests. There is first the pre-qualification document, which includes a certain amount of
information about the project, provides various instructions and restrictions, and also ques-
tionnaires concerning the qualifications of the engineering company involved, and details of
its operations and structure. The choice of a contractor is made by applying various criteria
to the replies provided: some apply to either kind of contract; other, additional ones are spe-
cific to the method selected (management or flat fee). The next step is the requestforpropos-
als, which governs the definitive choice of an engineering company and which generally has
three parts: the instructions to bidders (details concerning the preparation of the technical
proposal and the criteria for its evaluation), the draft contract and its appendices (services,
responsibilities, guarantees, special procedures, etc.), and the project description (technical
definition of the project and its setting, both during regular operations, and during its con-
struction). The conditions of the request for proposals may take account of comments or
restrictions made by the pre-qualified companies. The preparation of these documents takes
228 3. The Determination of Battery Limits Investments

about one month for management contracts, and three to five months for flat fee contracts;
their evaluation and the selection itself take about one month in the first case, and two or three
in the second.

3.1.2.2 The Implementation Phases

These include two main periods: the study and supply period, and the construction itself.

A. Studies and Supplies

These are undertaken by the headquarters staff of the selected engineering company. Three
steps must be distinguished at this stage of project development:
0 Preliminary studies (design stage), lasting six to nine months, and essentially covering the

following activities:
0 Drawing up timetables and plans, to define the scope of the project and establish the

design criteria,
0 Development of the project, the main goal being to approve and finalize the project

objectives (scope of documents, work plans, scheduling, budget revisions, etc.),

0 Drawing up a semi-detailed budget, also called the “target budget”, which on the client’s

side corresponds to the drafting of preliminary documents, in particular the signing of a

service contract.
0 Detailed studies, which can last from 12 to 15 months and whose objectives are to provide

the following services:

0 The approval for implementation, the final agreement to construct a project (orders for

equipment and materials, confirmation of construction plans, sub-contracts, manufactur-

ing schedules),
0 Drawing up of a detailed budget, also called the “initial budget”, for the service con-

tracts, or the “contract budget” for the turnkey ones, which will be used for cost control
and to define deadlines,
0 Production of documents (drafting and checking of plans and timetables; drawing up

schedules for shipping and testing, etc.; preparation of the operating manual, etc.),
0 Purchase of equipment and the signing of contracts for follow-up, checks, and the supply

of spare parts, etc.

0 The completion phase, during which the following operations are undertaken:
0 Monitoring of the final documents (plans, construction files, handover certificates, etc.),

0 Checking of costs and deadlines.

This latter period corresponds, in fact, to two notable changes in activity. First of all, it
marks the transition to the period when the budget envelope is roughly estimated, which
allows the project’s economic and financial status to be defined; it involves going from the
mere evaluation of investment costs to a more accurate estimation of them, so as to begin
making definite commitments. It then becomes the prelude to another form of business man-
agement, that of cost control, where it is no longer a matter of making forecasts, but of taking
steps to enforce them.
 3. The Determination of Battery Limits Investments 229

B. Construction Proper
This covers all the work performed on the site, and also comprises three distinctive phases:
0 Start of operations and site preparation, which begins when the preceding study and supply

period has not yet been completed, and mainly includes everyhng related to the constructionof
a working base and the development of the site, as well as foundation work, roads, and the civil
engineering work preliminary to assembly. This is also when most of the main apparatus is
delivered. At the end of this stage about 25 to 30% of the construction work has been completed.
0 The active constructionphase, which covers the whole set of activities related to the deliv-
ery and receiving of all the equipment and materials, the completion of civil engineering works,
the signing of contracts for mechanical erection (equipment and piping), electricity,instrumen-
tation, painting and thermal insulation (the final task), before transfer of ownership from the
engineering company to the client. At the end of this period the construction is 90% complete.
The completion and backj?tting,which include mechanical testing and acceptance, pre-
startup, startup, and performance testing of the facilities.

3.1.2.3 The Respective Contributions of Evaluation and Estimation

It always a tricky matter to attempt to add new shades of meaning to terms that are often
treated as related, even synonymous. Nevertheless, it is possible to view evaluation and esti-
mation as not covering exactly the same types of activity, and especially as not approaching
problems in the same way.
Generally speaking, the main differences are as follows:
0 Evaluation aims to judge the profitability of one or more projects, or various variants of

one of them. For this purpose, it has to have a three-point methodology, which is essentially
what Chapter 2 was about, but which also has the special quality of being based on a simpli-
fied approach using the amount of the battery limits investments, and which ultimately leads
to judging the comparative economic attractiveness of the alternatives considered, i.e., the
profitability of the facilities concerned. By contrast, estimation confines itself to the determi-
nation of the investment cost.
0 Estimation is an absolute matter, providing opinions on figures alone, usually obtained by

direct consultation with service providers, and required to be as exact as possible. In fact, for
example in the case of a turnkey contract, it must commit to all of the engineering work on
the basis of a single, un-revisable figure: in these circumstances, a calculating error can have
devastating effects, such as the loss of a business, the creation of deficits, compromising the
reputation of a brand name, etc. Evaluation, on the other hand, has an orienting or preparatory
role, offering a choice from several alternatives, in other words it “compares”. It thus works
in a relative manner, employing information of a “bibliographic” nature (databases); for this
reason, it is not necessary to know the investment costs with great accuracy, as long as this
uncertainty does not affect the ranking of the situations examined.
0 The accuracy of the result is meaningless if the starting information itself is not sufficiently

reliable. Estimation thus demands a priori a certain level of credibility, and consequentlya level
230 3. The Determination of Battery Limits Investments

of detail that is very markedly higher than that required for evaluation. Moreover, in the latter
case the methodology must be designed to require only a minimum of data, even to allow the use
of substitutesor approximations.Obviously,however, if the informationdoes exist the same pro-
cedure must be capable of incorporatingit, which will certainly improve the accuracy of the final
result. In other words, the techniques employed in evaluation must first of have great flexibility.
When all is said and done, the particular characteristics of evaluation and estimation are
such that in practice the pre-study period is based essentially on the concepts and procedures
of calculations arising from evaluation, especially for the definition of so-called “order of
magnitude” and “preliminary” budgets. In the construction phase, the methods used in esti-
mation become predominant, and progressively more detailed, complex, and accurate as the
date for the construction decision approaches, i.e., the time when the “target” and “contract”
budgets must be defined.

3.1.2.4 The Actors

The various phases in the development of a project demand many varied skills, which usually
require a client to interact with persons having very diverse occupations and concerns.
Figure 3.2 illustrates this, emphasizing the duration and the nature of the activities under-
taken by the principal actors or partners.
The client is obviously the key character, on stage during all the phases that lead up to
industrial activity, and continuing to play its role far beyond it, throughout the operating life
of the facilities.
The consultant usually plays the main supporting role, at least during the pre-marketing
phases, since in practice it acts as the client’s “gray eminence” during all this time. Indeed,
unless the manufacturer has departmentswithin its own company that are able to provide such
services, it generally has to call on the counsel and advice of a third party to protect itself.
Even if it does have in-house resources, it will be more convenient for its relations with its
other partners to proceed in this manner, particularly in conflict situations, when the consult-
ant can act as a referee or suggest alternative solutions.
Thefunding party or parties also have a major part to play, albeit a less prominent one,
particularly during the study or pre-study phases; on the other hand they are essential partners
when management or flat-fee contracts are signed, and in defining the timetables for carrying
out the correspondingbudgets. They are of course responsible for managing the pre-financing
of the construction proper.
Theparties who license the use ofprocesses or who own technologies, are mainly seen in
the second act, i.e., during the feasibility studies, when the consultant, in particular, must
recommend to the client a choice among the range of industrialized or developing techniques
that are likely to suit the project. These parties are involved in a much more specific and
detailed way when it comes to preparing documents for use by the engineering company.
The engineering company is in the spotlight as soon as it is hired. For practical purposes,
it becomes the lead actor, or at least the client’s favored representative, in everything that
affects the realization of the project. It acts as the client’s intermediary, substitute, or spokes-
person in dealings with third parties, and consequently leaves the client little to say. Its activ-
ities cease only at the end of the period of backfitting, startup, and testing of facilities.
 3. The Determination of Battery Limits Investments 23 1

Client
b
Consultant CaDital fundinq
b
Funding parties L Manaqement of pre-financina L
Process licensors Select deliver Pgcess Assist studies Assist end of
Engineering process. Data Book constr e
Suppliers
Companies
Reply to RFP
F
Do.
- Do.
Manufacture
Worksite
h
r

v b

Figure 3.2 Timetable of activities of main project actors. (Source: BF)

The suppliers of equipment and materials are solicited by the engineering company when
the tenders are called. They provide direct assistance during the supply phase. Their main
concerns are to meet both the specifications and deadlines required of them.
Finally, the service companies make their first contribution when they respond to the call
for tenders. Their participation becomes essential when work starts at the site, and is com-
pleted only when the factory is handed over. Their main constraints are similar to those of the
equipment manufacturers; however, it should be noted that these companies are exposed to
greater risks and proportionally more unforeseen events, particularly as regards weather,
stock shortages, strikes, etc.

3.1.3 CHOOSING AN ESTIMATION METHOD

As already mentioned, and as will be discussed in more detail later on, there are many meth-
ods of estimation and evaluation, whose very diversity proves that none of them is entirely
satisfactory. Nevertheless, some seem to be more effective than others, and in particular bet-
ter suited to the situation.
232 3. The Determination of Battery Limits Investments

3.1.3.1 Selection Criteria

Methods of estimation, in the broad sense of the term, have features of two main kinds:
0 The first concerns their use and includes the information required, the cost of the proce-

dure itself, etc.

0 The second relates to their operating principle and concerns archiving, i.e., the way in

which previous results are retained, and extrapolation,namely how the data are expressed
as costs, etc.
These properties are analyzed in detail, for each method, in Section 3.1.3.2.
In addition, it is tempting to judge the reliability of an estimation procedure only on the
“accuracy” it permits. Now, this characteristic,which should rather be called closeness to the
true value, to the extent that that can actually be known on the basis of initial calculations,
and which is discussed at greater length in Section 3.1.3.3, mainly depends on the quality of
the available information. In this case, it includes the reference informationjust as much as
information about the product being estimated. The level of uncertainty thus appears to be an
extremely insecure potential selection criterion: for the same method, it fluctuates, depending
on the situation to which it is applied.
In practice, it seems much more relevant to consider the following three characteristics:
0 Sensitivity, which determines to what extent a small change at the data level is expressed

in the result,
0 Traceability, which describes how easy it is to make the connection between the result

and the data,

0 Credibility, which takes into account the results previously given by the method, and

what the decision maker should know about the connection between data and results.
This criterion, which is a subjective one, is very important, since there is no point in per-
forming estimations if the person making the decisions has no confidence in the chosen
method.
In view of these considerations,how is the most appropriate estimation method to be defin-
itively selected?
By assuming that each method can be positioned within a segment that goes from the
detailed approach to the overall approach, the choice can be made in the following manner:
at the moment when a decision is made to undertake an estimation, we try to minimize the
cost associated with that operation. This consists of the total of the expenses resulting from
the forecast itself, and from the shortcomings of this procedure, costs shown on Fig. 3.3 by
the curves (C,) and (C2).
This then identifies the procedure to be followed, or rather the level of detail at which it
seems appropriate to work.
It is easy to see that curve (C,) does not depend on time, since the use of a given method in
different situations takes roughly the same effort, whatever the conditions. On the other hand,
curve (C2)varies with the progress of the project, which in practice means that there is not just
a single curve, but a whole system. In the initial period (C23, when the project is just getting
under way, it is relatively flat and the minimum point rn lies closer to the overall approach.
Later on, when the project is better defined, it curves more and more steeply (C2,,): employing
a detailed procedure allows us, as it were, to make better use of the available information,
 3. The Determination of Battery Limits Investments 233

cost

Detailed m, ma Overall
approach approach

Figure 3.3 Selection of an estimation method. (Source: IFP)

which is not usually the case with an overall approach, because by ignoring a portion of the
data it risks making a faulty prediction. The point m thus moves towards the detailed approach.
In other words, the choice of a method will therefore depend, in large part, on the phase in
which the project lies. When starting a study, the estimator must first settle the following
problem: given what is known about the project, and the assumptions that will have to be
made to arrive at a cost, which procedure is the best suited? Obviously, the answer will differ
from one project to another, but it will also change as the projects progress.
In conclusion, as concerns the quality of estimation methods, one is not better than another.
Each offers a certain number of features, and these determine which is the better tool at a
given moment.
However, many decision makers see little purpose in being able to apply an estimating pro-
cedure at the start of a project, though it seems self-evident that waiting until the last minute
to evaluate the cost of an operation is not a good way to operate, since in this case one might
end up evaluating a product whose cost price is greater than the average sales price the market
will accept. Even an inexact study would surely avoid such a situation, by drawing attention
at a much earlier date to problems, even to impossibilities, inherent in the project under con-
sideration.
In fact, it is best to make use of estimation as soon as possible, in particular before the
design phase, where 70% of the costs have already been settled. By way of example, although
the sums required for raw material purchases may not in fact be paid out until the moment
when the facilities actually begin working, the selection of this or that source of supply (and
consequently the commitment to possible additional costs resulting from the use of an expen-
sive starting product, as compared to another alternative), is decided well in advance.
234 3. The Determination of Battery Limits Investments

This interval between the decision and the actual correspondingexpense may be illustrated
by means of Fig. 3.4, which shows perfectly that the longer one waits, the less room there is
for maneuver.

cost

costs
-
70%
of expenses

design phase .

Figure 3.4 Settled costs versus committed costs. (Source:ZFP)

In addition, at the end of the project, not only does it become difficult, if not impossible,
to change the costs, but even when it can be done any last-minute adjustment carries heavy
economic penalties, as shown graphically in Fig. 3.5.

Time

Figure 3.5 Possibility of change, and associated cost. (Source: IFP)

 3. The Determination of Battery Limits Investments 235

3.1.3.2 Classification of Estimation Methods

There are many ways of classifying cost estimation procedures. The approach that a priori
seems most logical and most comprehensive consists of grouping them according to two
major principles: whether one looks at the activities necessary for the production of some-
thing, or, on the other hand, at the product itself. In the latter case, a hndamental distinction
must be made between the methods that perform a purely statistical processing of the
observed data (historical data), and those that employ a “mathematical” modeling of cost
behavior.

A. Activity or Analytical Approach

a. Principle

One way of looking at estimation is to view it from the following angle: “Doing something
is only a matter of repeating tasks, always the same ones, which differ only in their size”.
Based on this principle, it then becomes possible to undertake this kind of study if one has
both a description of the sequence of activities necessary for carrying out the action under
consideration, and a tool for quantifying each of the tasks in terms of cost.
In practice, this approach corresponds to the first way of beginning a cost estimate, i.e., by
asking the question: “What do you do to produce the product whose cost you want to deter-
mine?’.
In even more practical terms, it requires a two-stage process: first a list of the activities
must be drawn up; then a cost is assigned to each one of them, and the total gives the required
overall value. These costs are compiled in a database (catalog) and generally include a portion
for materials and another for labor, both of them derived from a size descriptor.
To completely describe one of the tasks concerning a product, it is necessary to know:
0 Its nature

0 Its unit of measurement

0 Its size, i.e., the quantity of the unit involved.

For example, to quantify a linear welding operation, the labor portion is essential, and
depends on the number of hours devoted to this work; the length of the weld often serves as
the descriptor, in view of its close connection with the working time required.

b. Application
The analytic approach is based on the following premise: the impact of the size of an activity
on the cost is linear. In other words, if the size of an activity is multiplied by a factor k, the
resulting cost is also multiplied by k.
This type of procedure can be applied either more or less precisely. Several things contrib-
ute to deciding the level of detail ultimately selected for the work. In the first place, the needs
of the company must be considered. Its strategy may in fact require the estimator to evaluate
an alternative solution that differs only in certain details. As a consequence, the level of gen-
eralization of the process is the determining factor: the more comprehensive the desired
approach, the more precise it must be. This has repercussions in terms of cost, since both the
236 3. The Determination of Battery Limits Investments

compilation and the maintenance of the catalog, and the estimation proper, take more and
more time. In practice, the lack of a certain number of data compels the estimators to accept
a less detailed approach than they might have wished.
The database itself constitutes a fundamental problem for the analytic approach. In the first
place, the amount of information to be compiled and stored is such that it leads to serious
management difficulties. Consequently, it is necessary to perfect a very precise nomencla-
ture. Moreover, it is essential to normalize the information contained, since it serves as refer-
ence for the estimations. To evaluate a new cost, two steps will be essential: to extract from
the database the data that correspond to the problem being studied, and to confirm that the
values selected are not contaminated by events of an exceptional nature. In the automobile
sector, for example, these “externalities” may be of various kinds, such as a strike at a pro-
duction center, an abnormally small production run, products made for unusual customers
with specific requirements, etc. This normalization leads immediately to a need’for an expert
capable of doing this kind of work. In practice, each specialist will establish a scale for his
own type of activity. For example, for welding an automobile chassis (for reinforcement), the
expert may fashion himself a rule such as: an operator makes 10 to 15 spot welds per minute;
the positioning of each of these with respect to the others largely determines the value to
employ (10 in a complex situation, 15 at most). Nevertheless, he always tends to think that
everything is a special case and he does not automatically try to establish general rules, even
if his approach seems to lead in that direction. This latter point is important, in fact it partly
explains why certain sectors, such as the automobile, have difficulty in getting away from
analytical methods. The second reason is linked to the number of items and activities
involved in a vehicle, and, consequently, the multitude of changes that are made during the
design phase. Whenever there is the slightest modification, the person in charge wants to
know what it costs: in this case an analytic approach seems desirable, if not necessary.
Generally speaking, applying such an approach takes time and therefore proves to be
costly; the complexity of the system prevents quick responses, and, in addition, traceability
is always poor: the very detailed breakdowns require a lot of work if the discrepancies
between forecasts and outcomes are to be explained.
Among the other characteristics of this kind of method the following favorable situations
for applying them must also be emphasized:
0 When the project is sufficiently advanced, their great sensitivity and the many points they
have in common with analytical accounting make them indispensable(possibility of allocating
charges by department,etc.); this feeling is strengthenedby the great credibility enjoyed by this
approach in such situations, since it fits in with what the decision maker expects, it has been
successfully tested over many years, and it leaves little room for the estimator’s subjectivity.
On the other hand, these tools may prove to be unsuited to certain situations. Firstly, at the
start of the project the data are generally insufficient for this approach to be used; later on,
analytical methods are poorly adapted to rapid technological development, which requires
frequent, expensive updating of the catalog and usually involves an incomplete definition of
the product, which is incompatible with this procedure.
Nevertheless, it must be noted that it is always possible to “imagine” a manufacturing
facility capable of making the product that one is trying to estimate, without ever intending
 3. The Determination of Battery Limits investments 237

to put it into practice: the work then consists of predicting the number and kinds of apparatus,
the persons required for the work (both for the operations themselves and for supervision,
maintenance, etc.), the nature and amounts of energy and raw materials, etc., so as to deter-
mine the factory’s operating cost, and thereby even the cost of the product. However, if one
considers the significant expenses involved in such a task, without having a fairly complete
description of all the activities, it is certainly preferable to adopt another approach. In fact,
the “accuracy” of these procedures is always very dependent on the setting in which they are
applied (factory, crew, etc.) and a change in assumptions automatically leads to a blurring of
the result.

9. Product Approach or Global Approach

The approach no longer consists of an analysis of the activities, but of a study of the product
itself. One attitude has been succeeded by another in the course of time, thanks to the devel-
opment of calculating aids and statistical tools: direct analogy and modeling.

a. Direct Analogy

Principle
This is a purely statistical method, applied to a homogeneous family of products, i.e., ones
that perform the same function and have been made by the same technology. By reduction of
past data, the estimator finds a link between the cost and certain physical variables, and then
utilizes the relation thus defined to assign values to the different sizes within the family.
In general, we may distinguish several kinds of method according to the statistical treat-
ment employed. The simplest is to use averages and thus to define “price lists”. Although this
way of working is useful in certain areas of activity (for example, the construction and public
works industry uses this kind of tool when it costs trenches in linear meters), as a general rule
it is best to be careful with it. In fact it is very restrictive, and is only useful in the case of a
linear relation between cost and size. Now, the cost per unit of size usually decreases when
the size increases: for this reason, in practice the average values are calculated for ranges of
size units. Consequently, it may seem convenient to apply such a tool to rapidly obtain an
“order of magnitude”: nevertheless, it is recommended that other, more “reliable” methods
be given priority for confirming the results.
In these circumstances, it is usual to apply Cost Estimation Formulas (CEFs), established
by linear regression (for example, using the least squares method: see Chapter 1) and taking
one of the following forms:
cost = ao. (descriptorl)al ... (descriptor,)”. (3.1)
cost = a0 + a1 . (descriptor,) + ... + a, . (descriptor,,) (3.2)
The first, exponential expression can be turned into a linear one by taking the logarithms
of the variables (cost and descriptor). In fact, (3.1) can also be written:
1, cost = 1, a. + a , . 1, (descriptor,) + ... + a,, . 1, (descriptor,)
238 3. The Determination of Battery Limits Investments

By way of example, the cost of a heat exchanger of a given type can be roughly calculated
by the following formula:
c=en. Sb
where
C : is the cost of the exchanger
S : the exchange area
a, b : are log-log regression coefficients.

COMMENT
In practice, the number of descriptors rarely exceeds five; for clarity of explanation,
only one descriptor is considered in what follows.

As already stated in Chapter 1, Section 1.2.3.1.c, the principle of the approach is to find
the best straight line fit for a number of known pairs on a cost-size graph, by minimizing the
sum of the square of the residuals, i.e., the sum of the squares of the distances, taken parallel
to they-axis, between each representative point and the straight line (Fig. 3.6). Then the con-
fidence limit (ymin; y,,) for a cost prediction is obtained by taking the value of y on the
regression line (D) for a given size x.
This confidence limit, which represents the area within which the cost has a probability p
of occurring (for example, p = 95%), is given by the two arms of a hyperbola. It is possible
to show that it is a function of three main factors:
0 The dispersion of data about the straight line, i.e., the spread of the residuals (which rep-

resents the information lost when the cloud of points is replaced by the CEF): the farther
the points are from the straight line, the greater the limit.
0 The spacing of the x-values of the reference sample about their average value (which is

given by the inverse of the variance of x in the confidence limit calculation): the bigger
it is, the closer the arms of the hyperbola are to each other.
0 The number of data points: as it increases, the limit tightens up.

It should be noted that these three factors are related: for example, increasing the number
of data points does not necessarily improve the regression. In fact, adding an aberrant point
also increases the dispersion. In practice, it is reasonable to consider that constructing a CEF
that is “credible” (having an acceptable confidence limit) requires about twenty points. In
view of the above considerations, it is evident that this is a minimum figure and that the het-
erogeneous character of the available information may demand a larger number.

Application
Given how quickly an estimation can be obtained from a CEF, it is often very desirable to
employ such tools, especially as they are generally very reliable if the family is properly
selected, which gives them a high credibility. On the other hand, it is essential to use them
with full understanding, particularly in areas of rapid technological development, where the
connection between current and earlier products is usually uncertain, in advance of the copi-
ous information required for the definition of a new “statistical model”. It should also be
noted that accuracy is not the best feature of this class of methods: in fact, unless there is a
very extensive database and the time allowed for a response is lengthy, it is best to limit the
number of parameters.
 3. The Determination of Battery Limits Investments 239

cost
I

I X Descriptor

Figure 3.6 Cost Estimation Formula (CEF). (Source: ZFP)

The so-called extrapolation factor method, discussed in Chapter 2, Section 2.3.2.1, which
is expressed in the following equation:

Cl/C2= (T,/T,)f (3.3)

where
Cj is the relative cost of a certain piece of equipment (or factory) of size 5,and
f is the extrapolation exponent, incorrectly called the extrapolation factor,
is only the most simplistic version of the analogical type of procedure. This expression (3.3)
actually amounts to a CEF of the type:
C = a . Tf
since, when applied to two items of respective costs C1 and C, and respective sizes T1 and
T2, it becomes:
on the one hand: C 1= a . T,f
on the other hand: C, = a . T2f
and then, by combining the two expressions, we obtain the original Eq. (3.3).
In practice, the observed values offare sometimes surprising. For example, if we consider
gyratory crushers, with the throughput as the size dimension, the extrapolation factor is close
to 1.2; this means that if the throughput is multiplied by 2, the cost is more than doubled. Even
though it may be necessary to perform a complete calculation, taking particular account of
the erection costs (which rise with the number of items of equipment), and producing an eco-
nomically optimum solution, nonetheless the initial approach considered in isolation tends to
favor the small apparatus. In other words, the effort that the maker must put in to produce the
equipment (the cost) rises more rapidly than that for the equipment itself (the crushing).
The situation seems obvious when one works with fairly simple items, which can be
described by a small number of parameters. However, this is no longer the case if we consider
a complete industrial complex (refinery, nuclear power station, etc.); in fact:
The amount of information is generally limited and thus does not allow the construction
of reliable CEFs,
240 3. The Determination of Battery Limits Investments

0 Its homogeneity appears highly doubtful, since in fact each construction is unique: for
example, depending on its structure, a refinery is capable of providing the various cuts
that it produces (liquefied petroleum gases, gasolines, diesel, etc.) in a range of given
proportions. Consequently, to simplify, the distribution of the demand for petroleum
products anticipated during the coming years largely controls the type and the relative
size of the facilities to be constructed.
There is, therefore, a double problem: lack of data, and inadequate homogeneity. To com-
pensate for these shortcomings, we “ ~ n l y ”need to descend to a more detailed level, specifi-
cally that of simple equipment. Their essential characteristic is that they are manufactured in
greater quantities, either because there are so many of them in the complex under consider-
ation, or because they are not specific to this complex, so that each item is itself a homoge-
neous family. In practice, we often calculate the investment cost required for the manufacture
of a product by using an analogous sub-unit (itself divisible into homogeneous components)
that is considered representative, and then using factors (which may be multipliers or not) to
arrive at the desired value (which in this case is the battery limits investment).

b. Modeling
History
Modeling applied to cost first appeared in the United States at the end of the 1960s, thanks
to the work of the American, Franck Freiman. Freiman was originally looking for an answer
for his superior, who wanted to have a tool that would allow him to quickly check the estima-
tions made by the relevant departments of his company, using traditional methods. The man-
ager, facing reports hundreds of pages in length, had no means at that time to confirm the
consistency of each study (was each one using the same assumptions?), their completeness
(was anything overlooked?), and so on. An overall approach was evidently needed.
Freiman engaged in a series of discussions with estimation experts from many different
sectors of activity. The observation that there were strong similarities between the approaches
gave him the idea that it might be possible to develop a model that would bring them all
together in a general theory of cost. In a certain way, then, it makes sense to speak of “mod-
eling how an expert thinks”.
A first prototype appeared in 1972 and, based on the good results obtained, it was marketed
in 1976 under the name: Price. The first generation of models dedicated to “equipment” was
born.
A number of drawbacks quickly appeared the need to know the mass, tricky estimation of
the complexity parameter owing to its eminently abstract nature, etc. In the face of these dif-
ficulties, Freiman took a fresh look at the original model, to eliminate its application prob-
lems. Although the core of the method remained unchanged, several basic improvements
were introduced and the Fast-E model (E for “equipment”), marketed in 1982, represented
the second generation, of which today’s Map-H is the direct descendant.
Since then, still following the same principle, estimating by using models has enlarged its
range of applications to include the development costs of software, construction and civil
engineering costs, etc., each one a variant on the basic theme of the “equipment” model.
 3. The Determination of Battery Limits Investments 24 1

Thus Map-H belongs to a family of models that continues to evolve, in France under the
initiative of Pierre Foussier (the 3F company). For example, the list includes Map-C for con-
struction and civil engineering costs, Map-S for software development, and Map-0 for car-
rying costs.
It should be noted that the Price family still exists, but that it has retained the original, less
advanced techniques: in particular, it does not incorporate the very important contributions
that emerged from the work of Freiman between 1976 and 1982.

Principle
Although they are often incorrectly grouped with direct analogy under the generic term
“parametric methods”, mathematical cost models are based on fundamentally different con-
cepts. In fact, like all scientific software, they aim to represent the behavior of a “system” in
mathematical terms. They are generally made up of several expressions that relate certain
variables, which are the ones being investigated, to certain others whose variations have
effects on the evolution of the first ones.
Developing a CEF is therefore a fundamentally different thing from developing a model.
In the first case, one reaches a conclusion by compressing data: the statistics are used a priori.
In contrast, in the second case, one tries to explain the observed behaviors of systems; conse-
quently, the construction proceeds in two distinct stages:
0 Research based on abstract concepts (Maxwell fields, Newtonian forces, etc.),

0 Trials of mathematical expressions until reality confirms them: the statistics are thus

applied a posteriori.
As in all models, there is a sharp distinction between the laws that govern the evolution of
a system, on the one hand, and the initial (or reference) conditions on the other. The two
aspects are complementary and essential.
Cost estimation software is also bound by this initialization and reference rule. Because of
its all-purpose nature it does not explicitly single out families of products, but places the item
to be estimated within a general system of curves. On the other hand, the specificity of the
company cannot be completely ignored, because although competition requires that the price
be around the “normal” value (a company that remained inefficient too long would disappear
from the market, another that offered lower prices would force its rivals to catch up), never-
theless there are differences in working methods. The initial conditions, called adjustment
variables, take into account the way in which the company designs its products.
The whole process is shown schematically in Fig. 3.7: it includes two distinct stages. The
first, called calibration, consists of acquiring the initial conditions. To do this, it is necessary
to have one or more reference points, each comprising a description of the product and the
accompanying cost. The model, operating in reverse, then deduces the adjustment variables.
Using the initial conditions and a description of the product to be estimated, it then becomes
possible to move on to the second phase of the operation: the cost estimation. In fact, it
involves identifying the “forces” that have acted on the reference system (changes in size, in
materials, etc.) and describing them “in numerical terms” to the model, which uses them to
deduce the cost of the final object. This comes down to extracting the common feature of the
pieces of equipment (the initial conditions), and then determining the differences between
them.
242 3. The Determination of Battery Limits Investments

Figure 3.7 Principle of mathematical cost modeling. (Source: ZFP)

The cost factors, which are used as much for the calibration as for the estimation itself, are
linked to the production context (quantity, degree of automation, economic conditions, etc.).
They are different from the description in that they do not directly refer to the product itself, but
they have a similar impact, since their variations also produce “forces” that act upon the cost.
The concept of calibration requires a special comment: it is central to the process of cost
estimation by mathematical modeling, so it is important not to make a mistake in understand-
ing concerning this step-a mistake that some persons, put off by the idea of having to calibrate
the procedure before each evaluation, commonly make. Now, it is not a matter of doing this on
the model proper, but on the processes and their environment. This approach is rendered essen-
tial by the following fundamental principle: a universal parametric method is not “absolute”:
by its nature, it is comparative. Estimations are made on the basis of previous experience, quan-
tified by various persons. The calibration thus improves the consistency of the description, but
not the accuracy of the model. It develops real measures of performance that serve as refer-
ences, so as to assure the link with the company’s capacities and to establish a credible base.
It also enables us to become familiar with the software’s parameters, and with the way in
which they can be used to describe the differences between the products, and between their
development possibilities. In this way, the estimator improves his ability to quantify the
assessments that are essential for comparing new products, processes, and environments with
existing ones.
 3. The Determination of Battery Limits Investments 243

Application
In a way, the principle consists of looking at the problem in a more abstract fashion than
when using CEFs. Formulas of the type:
cost = a. x(descriptorl)al ...(descriptor,,)’”
are still used, but the coefficients ( a j )are themselves estimated by relations that employ the
product’s characteristics, which have to be described to the model. Looked at in this way, the
universality of the procedure requires that the product description be simple, but above all not
specific to it alone. Accordingly, the parameters are the same, whatever the study undertaken,
even if they are understood in different ways. This feature is very important, since it allows
the application of the model to be modified to fit the progress of the project.
The estimation’s validity still depends on the values of the parameters entered by the user,
which the software simply translates into economic data. Next comes an essential training
stage, whose purpose is to find correspondences between the situations to be described and
the numbers derived from the parameters, and especially to determine their rates of change,
so as to be able to express the observed differences between the products correctly. This expe-
rience can only be acquired by using the model; as a result, it is quickly passed on. In fact, all
that is needed is to have an up-to-date computer file that covers the company’s history, and
includes each product for which an estimation was made, with a list of the values of its param-
eters, and especially the steps that led to their determination. In practice, passing the baton is
a little more complicated than that, since two different estimators may have slightly divergent
perceptions of the approach. It is necessary, therefore, to explain to one’s successor just how
one has employed the tool, which essentially means describing the conventions that were
established to ensure the greatest possible consistency between estimations.
Once the calibration has been done, making a prediction is quick and consequently cheap.
Moreover, as already noted, the model’s flexibility allows for multiple choices, in particular
the level of detail at which to work. Because of this, although the preferred place to apply
them is at the beginning of a project, these procedures can also be useful later on, and may
even allow activities to be evaluated, as long as they can be described in physical quantities.
Two additional properties of models should be noted. First, by their nature they establish
a very clear link between design and cost, and thus serve as a tool for dialog between the var-
ious actors, particularly the designers and the estimators. Next, since the estimators need to
consider only a fairly limited number of parameters, an increase in productivity is often seen.
However, it must be admitted that because the mathematical expressions cannot be examined,
the models have a “black box” character that sometimes leaves the estimator puzzled. In addi-
tion, one cannot become expert in applying this approach overnight: the training period,
which is fairly long, has to be considered as an investment that will bring future benefits, aris-
ing from the use of modeling. Finally, the overall approach generally does not allow expenses
to be distributed by department, but this is a very minor drawback at the outset of a project.
The possibility of using mathematical cost models to estimate the time it takes to perform
operations should also be mentioned. For example, following a principle very similar to the
one used previously, when the initial conditions have been defined, manufacturing times can
be evaluated on the basis of a certain number of parameters that take into account the com-
pany’s specificity in this area. This can be very valuable, since making a piece of equipment
available quickly can make a decisive difference.
244 3. The Determination of Battery Limits Investments

3.1.3.3 Selecting an Estimation Method

The selection of an estimation method can sometimes be difficult, since it depends on a large
number of factors:
0 The extent to which the project has been defined

The level of innovation

The time available
0 The actual cost of the estimation

0 The accuracy required

0 etc.

In practice, the unavailability of one procedure or another often simplifies the decision. If
not, the choice is governed primarily by how far the project has progressed, and secondly by
the product’s level of novelty.
At the start of a project, it is not well defined, which limits the level of detail attainable;
moreover several alternatives are generally competing. In fact, time and money are in short
supply, and a quick, cheap method is essential. This situation dooms the activity approach
because of its design principle. In this case, the technological stability of the field concerned
often influences the choice between the two options under the product approach. If develop-
ments seem to be gradual and not too rapid, direct analogy may appear to be the most suitable.
Indeed, nothing is simpler than a CEF, and there is enough time to accumulate the experience
and deduce robust rules from it. On the other hand, if the technology is constantly changing,
modeling is the only reliable solution. This is also the case when the number of known points
is too small to permit a linear regression to be performed.
When the project is well advanced, the activity approach is the.preferred one; in any event,
preparing for the construction phase requires detailed work and the objective is no longer to
make design decisions, most of them having already been made. Note that many companies
elect to develop their own methods, semi-detailed, for example, which attempt to reconcile
the advantages of the activity approach with those of direct analogy.
By way of illustration, Table 3.1 presents a summary of the various options, with their
respective principles, advantages, and shortcomings.
If we now draw a parallel between the various phases of project development, as shown in
Fig. 3.1, and the procedures to be applied to meet the various types of budget that a manufac-
turer must underwrite before making the decision to go ahead or not, and then to go on to the
construction stage, we are led to the choices recommended in Table 3.2.
The first two types of procedure belong to the product approach, the second explicitly
requiring CEFs; the fourth is purely analytic in nature and assumes prior consultation with
the equipment suppliers and service providers for all the apparatus and works that are likely
to be involved during the construction of the plant. As for the third, it is based on a mixture
of the two kinds of approach, consultation being involved in only a portion of the actions.
Modeling is not specifically mentioned in this presentation: it is mainly utilized in the prelim-
inary stages, when, as noted previously, reference to earlier constructions of the “same type”
is impossible, i.e., in situations whose elements make them virtually special cases, or when a
technological leap is achieved.
 3. The Determination of Battery Limits Investments 245

Table 3.1 Advantages and drawbacks of various estimation methods

Activity approach I Direct analogy Modeling

Principle Making an inventory of Place the product in a Determine the differences
all the tasks and putting homogeneous family between the product and
a value on each the reference and express
Cost = a . sizeb them for the model by
*Cost= Ci C ui N ti changing the parameters
Advantages Great sensitivity Quick, cheap Quick, cheap after
Company-specific Good traceability calibration
Accepted and credible Credible for a Flexible
homogeneous family Dialog tool
Drawbacks Detailed definition Number of data “Black box” aspect
Lengthy training
Requires stable Dubious homogeneity if Distributing charges by
technologies technology unstable department impossible
Poor traceability
(Source: IFP)

Table 3.2 Estimation methods for various types of budget

Periods 1 Pre-study Implementation

Project Identification Basic studies
Feasibility Development
phases Prefeasibility Design

Budget II 1
Order of magnitude
1
Preliminary
1
Target or initial
1
Contract
category
1 1 1
Estimation Modular or factor
Overall or similariq Semi-detailed Detailed
method method
1 1 1 1
Level of
+40/50% +25/30% k 15/20% k3/5%
precision
I
(Source: BF)

If we restrict ourselves to the so-called “evaluation” methods, in the sense defined earlier,
i.e., to the pre-study phases, the recommended procedures are likely to fall under direct analogy.
Nevertheless, a few words must be added concerning the degree of sophistication of these
approaches. Moreover, from this viewpoint the selection may be made partly on the basis of
chronological considerations, in that the accumulation of data over time, and the progressive
improvements in the methods of analyzing and interpreting information, have allowed them
to be perfected and given greater depth.
246 3. The Determination of Battery Limits Investments

To begin with, we may consider the distinction made earlier between overall extrapolation
methods, which are exponential in nature, and those based on a prior breakdown into ele-
ments that are calculated by analogy with cataloged data of the same kind, followed by a rein-
tegration of the whole using sets of coefficients. Looking at them in more detail, the first
category comprises:
0 The “extrapolation factor” procedure, an incorrect term although widely utilized by the
profession to express the determination of a new cost from another, known one, under iden-
tical conditions of date and site, by multiplying it by the ratio between two values of the same
size variable, raised to a certain power. It works just as well for apparatus (exchangers, col-
umns, furnaces, compressors, etc.) as for assemblies or complete units (in this case the size
parameter is usually the unit production or processing capacity).
0 The block, or notional operating unit methods. Their principle consists of a “Prtvert”

inventory, in which each of the constituents is assigned the same weight-r different
weights, depending on the designers-and which leads to a notional average piece of equip-
ment, whose cost is obtained by analogy (for example, R.D. Hill, F.C. Zevnik and
R.L. Buchanan, G.T. Wilson).
0 Flow methods, which attempt a weighting of the value of the size variable represented by

the unit flow through the various pieces of equipment in a plant (for example, E.A. Stallwor-
thy, D.H. Allen and R.C. Page).
0 Approaches based on identifying “significant” steps and calculating their respective costs
by direct analogy (for example, J.H. Taylor).
0 Procedures employing a prior breakdown of a plant into hnctional modules, i.e., ones that

have a physico-chemical reality, each one’s cost then being determined by analogy, on the
basis of one or preferably several significant size parameters (for example, J. Bergman and
A. Chauvel).
Next, as concerns the multiplying or the factorial methods, we may distinguish between:
0 The ones that use constant factors (for example, A.J. Lang, N.G. Bach, W.E. Hand), which

date from the decade following the second world war: they are based on a rather limited
amount of information concerning traditional materials, and fairly small unit manufacturing
capacities in comparison to the tonnages handled today.
0 Ones that recognize the variability of these factors and instead offer a range of possible
values (for example, C.H. Chilton, or M.S. Peters and K.D. Timmerhaus).
0 Finally, ones that endeavor to quantify these variations, taking the specific nature of each

project into account by means of a limited number of explanatory parameters; here we can
differentiate between:
0 Relatively comprehensive procedures (for example, J.H. Hirsch and E.M. Glazier)

0 Modular methods using type units (for example, C.A.Miller, or the AACE {American

Society of Cost Engineers} approach)

0 Methods using specific equipment modules (for example, K.M. Guthrie, A. Chauvel et

al., SRI).
 3. The Determination of Battery Limits investments 247

As shown in Table 3.3, these various possibilities reflect an almost chronological, twenty-
year evolution in the means available for collecting and processing information. Moreover,
the most recent developments in the field are a result of the enormous progress achieved in
computing in recent years; they mainly concern the marketing of estimation software, based
on reference to very large databases and the use of CEFs, and having nothing in common with
the cost models discussed above.

Table 3.3 Chronology of evaluation methods

I Exponential methods Factorial methods

1947 R. Williams 1947 H. J. Lang
1949 W.L. Wilson 1949 C.H. Chilton
1956 R.D. Hill 1958 N.G. Bach
1963 F.C. Zevnik and R.L. Buchanan 1958 M.S. Peters and K.D. Timmerhaus
1970 E.A. Stallworthy 1958 W.E. Hand
1971 G.T. Wilson 1960 J.E. Haselbarth and J.M. Berk
1975 D.H. Allen and R.C. Page 1960 J.H. Hirsch and E.M. Glazier
1975 J.H. Taylor 1965 C.A. Miller
1978 A.V. Bridater 1967 J.T. Callagher
1981 J.L. Viola 1969 K.M. Guthrie
1987 J. Bergman and A. Chauvel 1976 A. Chauvel et al.
1985 1.V Klumpar and S.T. Slavsky
1985 SRI
(Source: IFP)

3.2 METHODS FOR DETERMINING INVESTMENTS

In this section, we do not intend to make a detailed study of all the procedure that may or may
not have been published. The catalog would be far too tedious; and already some readers will
be complaining that it is long enough with just the ones selected so far. These two reasons
will suffice to explain why the analysis undertaken here is, on the contrary, far from being
exhaustive, and certainly partial as regards the authors who have not been mentioned. We
therefore trust that the persons who consider themselves to have been slighted will be kind
enough to believe that they have been omitted more out of ignorance of their contribution to
the development of these methods, than by any deliberate action.
In addition, so as to honor past descriptions of these procedures as far as possible, they are
mainly repeated in their initial form, in particular by retaining the original nomenclature and
notations. Sometimes this may lead to some confusion when going from one approach to
another, since the same acronyms may have completely different meanings; conversely, as
often happens, the same idea may be formulated in different ways.
248 3. The Determination of Battery Limits Investments

3.2.1 EXPONEN TIA1 METHODS

The main advantage of these approaches is that they generally enable a very quick evaluation
of investment cost, based on a minimum of information.

3.2.1.1 The “Extrapolation Factor” Procedure

A detailed description of this method was presented in Section 2.3.2.1 of Chapter 2. There is
therefore no point in repeating it here, except to recall that its application only requires us to
know the values of the same size variable for a reference case and for the one whose cost is
to be determined; we must also know the cost of the reference item. When it concerns an
industrial complex, this parameter is usually a unit processing or production capacity; it may
be a surface area, a weight, a power, a diameter, etc., when it concerns a specific piece of
equipment.

3.2.1.2 Notional Operating Unit Methods

To apply this type of approach, we must have access to a plant’s schematic diagram. All we
need to know is the sequence of unit operations (or blocks) that make up the production
concerned: it is not necessary to have detailed descriptions, particularly about sizes, or,
usually, the material balance. On the other hand, it is very important that all the different
stages of the technology employed, especially the ancillary operations, be listed and broken
down to provide a description that is as complete as possible, and that no “blocks” be over-
looked.
Various authors have attempted to progressively improve the level of accuracy of these
methods, while preserving their basic initial characteristic, i.e., that their application requires
very little information.

A. Method of R.D. Hill

This applies to facilities that operate at moderate pressure (<I00 psi, or 7 bar), in which the
only major products in circulation are fluids. Knowing the process diagram enables us to add
up the number of units or blocks as follows:
0 Carbon steel columns, reactors, evaporators, blowers, and cyclones count for one.

0 Stainless steel columns, furnaces, centrifuges, compressors, and refrigeration cycles

count for two.

0 One additional unit is added for each feed or discharge that requires intermediate stor-

age, two for each secondary solid or gaseous discharge.

The base cost of a unit is $30000 for a capacity of 10 million lb/year (4500 metric Uyr)
and a Marshall and Stevens updating index of 185 (1954).
The total cost of the erected equipment is obtained by multiplying this base cost by the
number of units thus determined. For different capacities, an overall extrapolation exponent
of 0.6 is utilized, and for higher pressures the following factor is applied to the equipment
 3. The Determination of Battery Limits Investments 249

concerned: @ressure in the vessel in psi)/lOO, or (corresponding pressure in bars)/7. We then

obtain the grassroots investment by employing various coefficients, such as those suggested
by C.H. Chilton for piping, instrumentation, construction, auxiliary services, engineering,
and contingencies (see Section 3.2.2.3).

B. Method of F.C. Zevnik and R.L. Buchanan

This approach, which is valid only for chemical plants that operate with fluids as their prin-
cipal products, takes into account the following characteristics:
The factory's production capacity,
The number of functional units or blocks: each functional unit is defined as the apparatus
required to perform a given operation, such as a distillation, a stripping, or a compres-
sion,
A complexity factor based on the process's range of operating conditions: temperature
(factor FT),pressure (factor Fp),and the quality of the materials employed (carbon steel,
specialty steels, alloys, etc., factor Fa).This complexity factor is expressed as follows:

where
FT= 1.765.104 T -0.053 for T > 300 K
FT = -2. T + 0.6 for T < 300 K
Fp = 3 . lo4 (loglo PIP,, - 1)
Fa = 0 for cast iron, carbon steel, and wood
0.1 for aluminum, copper, and AISI type 400 stainless steels
0.2 for type 300 stainless steels, monel metal, nickel, and Inconel
0.3 for Hastelloy
0.4 for precious metals.
The cost of the average notional functional unit for the diagram concerned (the unit cost)
is a function of the plant's production capacity and the complexity factor (Fig. 3.8). The bat-
tery limits investment is obtained by multiplying the number of unit blocks by the unit cost.
The total grassroots investment is deduced by applying a constant factor (33% extra). If nec-
essary, an update may be performed by using the ENR cost index.
The following expression is then obtained:

Zg = N . (CPFU) . 1.33 -
(E"
300
where
Zg is the total investment
N number of functional units
(CPFU) cost per functional unit
(ENR) value of the ENR index on the date of the evaluation.
250 3. The Determination of Battery Limits Investments

Capacity (106 ib/ysar)

Figure 3.8 Cost of a functional unit in $ lo6 (ENR Index 300).

Method of Zevnik and Buchanan.
(Source: Zevnik FC. and Buchanan R.L, Chem. Eng.)

The advantages of this method are to be found in the rapid evaluation that it enables, even
in the absence of accurate information. However, it provides acceptable results only for fairly
complex processes: the authors applied it successfully to the production of cyanhydric acid,
ammonia, acetylene, phthalic anhydride, etc. If the factory’s process diagram is simple, i.e.,
having only one stage of synthesis with practically total conversion, high selectivity, and easy
purification (for example, the hydrogenation of benzene to form cyclohexane),we obtain much
higher investments than those found in practice. The error may reach, or even exceed, 100%.
In fact, the value of the method depends on:
0 An accurate evaluation of the number of functional units, whose cost is directly propor-

tional to the total investment. This work is very sensitive, and demands a lot of concentration
and experience. Available process diagrams are often found to be incomplete as regards the
load preparation stages, the product finishing operations, the catalyst handling, the equipment
required for plant startup, and the regeneration or recovery of catalysts.
0 A proper determination of the complexity factor, in particular the coefficient Fa, which

depends on the kind of materials employed. This assumes that we either know at the outset
the quality of steel being used, or that we undertake a meticulous analysis of the corrosion
 3. The Determination of Battery Limits Investments 25 1

problems likely to be encountered in the process concerned. If this analysis cannot be per-
formed, for example by comparison with existing situations, the accuracy of the method will
be significantly affected, and its value greatly diminished.
In addition, a major drawback, which relates to the very principle of this way of evaluating
investments, consists of taking the same basic cost no matter what the function and relative
scale of the functional units, and looking only at the total production capacity. Because of
this, the approach does not allow their actual fluid circulations to be assigned separately to
each of the units that are counted: these circulations control their real sizes and consequently
their costs (recycling, by-products, reflux system, etc.).
To a lesser extent, the use of a single complexity factor, representing the whole range of
conditions, is also a built-in drawback. Thus, it is easy to understand why several authors have
subsequently attempted to correct these failings.

C. Method of G.T. Wilson

The procedure developed by G.T.Wilson is based both on the work of E.A. Stallworthy (see
Section 3.2.1.3), and on the adaptation of the so-called Lang's factor (see Section 3.2.2.1)
made by C.E. Miller, as part of his own method (see Section 3.2.2.4.F). This enables the
determination of battery limits investments for a given plant, expressed in 197 1 &, by using
the following equation:
Z=f. N . (AUC) . F M .Fp . FT
where
f is the investment factor. It is a function of the (AUC); its value, obtained from
Fig. 3.9, also depends on the kind of products circulated in the plant, according to
whether they are mostly fluids, mostly solids, or in between
N is the number of principal items of equipment, i.e., reactors, columns, exchangers,
apparatus for performing a unit operation, compressors, blowers, etc. Only the
pumps are omitted, in view of their small contribution to the total cost
(AUC) is the Average Unit Cost of the principal equipment. This cost, expressed in 1971
&, is a function of the average flow V, in t/yr, circulating through this equipment.
It is given by the following expression:

(AUC) = 21

FM is a factor specific to the type of material. Various values of FM are presented in

Table 3.4
Fp is a corrective factor for pressure. It is a function of the operating pressure,
expressed in psia, and may be obtained from Fig. 3.1 Oa
FT is a corrective factor for temperature. It is a function of the operating temperature
expressed in OC, and may be determined from Fig. 3.10b.
When the operating conditions vary significantly from one of the main pieces of equip-
ment to another, or employ different materials, weighted average values of F M ,Fp, and FT
must be used.
252 3. The Determinatioo of Battery Limits lnvesfments

Figure 3.9 Determinationof the investment factor.

Method of G.T. Wilson.
(Sources :British Chem. Eng. and Pmcess Tech.)

Table 3.4 Specific factors for types of material. Method of G.T. Wilson

Material FM Material FM
Carbon steel 1 .oo High-alloy steels 1S O
Bronze 1.05 Hastelloy C 1.54
Boiler steels 1.07 Monel metal 1.65
Aluminum 1.08 Nickel 1.71
Low-alloy steels 1.28 Titanium 2.00
Worthite (austenitic steel) 1.41
(Source:British Chern. Eng. and Process Tech.)

According to G.T. Wilson, the accuracy of the method is such that investments can be cal-
culated to *30%, as long as Vlies between l@ and lo6t/yr. In fact, it depends on how well
it is possible to know the plant diagram and the material balance, not just overall, but as far
as possible line by line, also the general operating characteristics of the principal equipment.
 3. The Determination of Battery Limits lnvesfmenfs 253

s Operating pressure (psia)

Figure 3.1 Oa Correction factor for pressure. Method of G.T. Wilson.

(Source: British Chem. Eng. and Process Tech.)

_r .
5 10-2 lo-’ 10‘ 102 103 104
e!
b
0
Operating temperature (“C)

Figure 3.1 Ob Correction factor for temperature. Method of G.T. Wilson.

(Source: British Chem. Eng. and Process Tech.)

3.2.1.3 Flow Methods

These take into account the influence on plant cost of the respective weights of discharges
other than the main product. In this way, it is possible to get closer to the true average size of
the equipment, and consequently to the amount of the investment. This is offset by the need
to have more detailed information, which takes something away from the initial purpose of
this kind of procedure: simplicity and ease of application.

A. Method of €.A. Stallworthy

The basic expression suggested by this author for calculating the battery limits investment,
expressed in 1967 5,is the following:

I = 0.00751 Az(
i= s

i=l
N . FM.Fp . FT .R)i
254 3. The Determination of Battery Limits Investments

in which we have
s the total number of primary and secondary products i
R the ratio of the output of product i concerned, to that of the principal product
N the number of steps (or component units) involving a product i
FM a factor specific to the type of material used in the equipment through which the dis-
charge i travels
Fp a factor related to the pressure in this same equipment
FT a factor related to the temperature, also for the same equipment.
The coefficient A, expressed in 1967&, is a function of the production capacity for the prin-
cipal product of the plant concerned. If V is this capacity, in tlyr, A has the form:
A = 0.62 . 10-5 (0-0.65
i may represent a secondary product as well as the principal product.
The values for the factors FM, Fp, and FT result from a weighting based on the operating
conditions of the main equipment involving the discharge i, and the type of material
employed.
Apart from the difficulties caused by having to describe the stages, i.e., the functional units,
and by determining their number, the drawbacks of this method are in particular related to:
0 The very definition of the coefficient A, which is calculated from the theoretical output

of the main product alone,

Obtaining the factor R for each product i, which assumes the availability of a sufficiently
detailed material balance for the plant.
However, when compared with Wilson’s proposal, it represents a significant advance. In
fact, an equivalent expression is recovered by assuming that all the factors that are a function
of i are constant, i.e.:
= k . ~0.6
to 0.7

a relation in which, moreover, k is assumed to be independent of V, and constant. Now, as

indicated in Section 3.2.2.4.F, Miller has shown that the coefficient k diminishes as the size
of the plant increases.

5. Method of D.H. AIIen and R.C. Page

This approach is even more detailed than Stallworthy’s procedure, but is based on some
parameters computed by G.T. Wilson. According to its authors, it enables calculation of
investments with an accuracy of between -20 and +25%. It applies only to plants that operate
with fluids.
The method’s principle consists of determining as accurately as possible the cost of the
principal equipment delivered to the site: the Delivered Equipment Cost (DEC). The battery
limits investment, or even the grassroots investment, is then deduced from it by applying
average statistical coefficients, such as those suggested by J.E. Haselbarth and J.M. Berk (see
Section 3.2.2.3), which vary with the kind of plant concerned. The DEC then represents
approximately 2 1% of the grassroots investment; this figure may vary between 15 and 30%
depending on the circumstances.
 3. The Determination of Battery Limits Investments 255

The cost of the principal equipment available onsite is obtained from the following rela-
tionship:
(DEC) = N . (SF) . (BIC)
where
N is the number of principal items of apparatus (including pumps), obtained by exam-
ining the complete plant diagram
(SF) is a complexity factor (the State Factor), reflecting the operating conditions and the
kind of materials employed. Thus, we have:
(SF) = FTmax ‘ FPmax ’ FMaver.

FT, Fp,and FMare obtained from Figs. 3.10a and 3. lob, and Table 3.4, compiled
by G.T. Wilson. According to Allen, it is not necessary to calculate the average
weighted temperature and pressure factors
(BIC) is the actual cost of a “standard” item of principal equipment (Basic Item Cost). It
is a function of the product’s average circulating flow rate through the unit, the
Throughput variable (TP), calculated from the following expression:

with
(BIC) in June 1972 $
(TP) expressed in lb mol/year
(BIC), $7 000 (June 1972)
(TP), 2.5 . lo6 lb mol/year
(EXP) extrapolation exponent for the plant concerned.
If an update is required, the authors recommend the use of the Marshall and Stevens index.
The average throughput (TP) is given by the following relationship:
(TP) = (CAP) . (FF) . (PF)
where
(CAP) is the total load in Ib mol/year
(FF) is the Flow Factor, defined as follows:

(FF) =
%‘ number of lines entering and leaving
each item of principal equipment
N

(FF) is never less than 2

(PF) is the Phase Factor, obtained as follows:
(VI)
(PF) = 0.0075 + -
N
where (VI) represents the proportion of the plant’s principal equipment whose size
depends essentially on the volume of gas or vapor, as compared to those calculated
256 3. The Determination of Battery Limits Investments

from the liquid fraction alone. The authors emphasize that if even 2% or more by
weight of a hydrocarbon is in the gaseous state in a given piece of equipment, then
its size parameters should be determined on the basis of the gaseous volumes pro-
cessed. The constant 0.0075 is the value of the ratio of the vapor and liquid densi-
ties of a type hydrocarbon product under average conditions. Expressing the
average circulating throughput in lb moyyear, i.e., in terms of volume, makes it
easier to keep track of the capacity effect.
The exponent (EXP), which applies to all the principal equipment, is calculated as the
weighted average value of the extrapolation exponents for each category of apparatus, i.e.:

(EXP)= ‘=’
5i=l
ci

where
ci is the base cost of the item of principal equipment i
ei is the extrapolation exponent for the item of principal equipment i.
Between i = 1 and i = N, there are a number of pieces of apparatus belonging to the same cat-
egory, and therefore having the same base costs and the same partial extrapolation exponents.
To calculate (EXP), the authors recommend using the values of ci and eisuggested by
K.M. Guthrie (see Section 3.2.2.4.F) and partly shown in Table 3.5.
Although the level of accuracy provided by Allen and Page’s method for quickly deter-
mining the cost of principal equipment is attractive, its impact is limited by having to apply
a constant factor to obtain the total investment (the erected cost) from it. In fact, it possesses
all the drawbacks of the methods that rely on multiple constant factors (see Section 3.2.2.2),
such as Lang’s method and those derived from it.

3.2.1.4 The Significant Steps Methods

This type of procedure requires only a simple block diagram, like that in Fig. 3.1 1, showing
the plant’s principal stages (reaction, filtration and washing, distillation, drying, etc.), as well
as the general operating conditions (input and output flows, temperature, pressure, residence
time, kind of steel utilized, etc.).

A. The Process Step Scoring Mefhod of J. H. Taylor

This is based on the principle that the cost of a production unit depends on a “costliness
index” (0, which expresses the complexity and the nature of the chemical conversion
involved in the process concerned, as well as its size, by means of an equation of the form:
investment cost = constant . costliness index . (production capacity)P
where p < 1.
 3. The Determination of Battery Limits Investments 257

Table 3.5 Extrapolation exponents and relative costs of various items of equipment, from K.M. Guthrie

Equipment I (;i:f 1
ExT:rt I
Base costs

I1
Equipment Exponent
(ei)
Base costs
( ~ 1968
i
$000)
Reactor furnace 0.85 135.0 11 Crusher:
Heating furnace 0.85 103.5 cone 0.85 12.0
Boiler: gyratory 1.20 3.0
15 psig 0.50 92.0 jaw 1.20 4.7
150 psig 0.50 101.2 pulverizer 0.35 23.4
300 psig 0.50 115.0 Mill:
600 psig 0.50 138.0 ball 0.65 4.4
“Packaged” boiler 0.70 60.0 rod 0.65 40.0
Tube-type exchanger 0.65 6.5 hammer 0.85 8.0
Reboiler 0.65 8.8 Evaporator:
U-tube heat exchanger 0.65 5.5 forced circulation 0.70 270.0
Cooler 0.66 6.8 vertical tube type 0.53 37.2
Cooling tower 0.60 9.9 horizontal tube type 0.53 30.4
Plate column 0.73 33.5 double shell 0.60 32.0
Packed column 0.65 35.2 Hopper:
Vertical drum 0.65 7.6 conical 0.68 0.1
Horizontal drum 0.60 5.0 silo 0.90 0.4
Storage vessel 0.30 6.0 Crystallizer:
Pressure storage: maturing 0.65 385.0
horizontal 0.65 4.8 forced circulation 0.55 276.5
spherical 0.70 8.0 batch 0.70 32.5
Centrifugal pump with: Filter:
motor 0.52 1.5 plate press 0.58 4.3
turbine 0.52 3.0 frame press:
Alternative pump: wet 0.58 5.3
with motor 0.70 6.0 dry 0.53 15.1
steam 0.70 1.1 rotating drum 0.63 17.5
Compressor: rotating disk 0.78 31.0
gas 1000 psig 0.82 85.0 Dryer
air 125 psig 0.28 36.5 drum 0.45 30.0
Blower and fan 0.68 9.5 tank 0.38 12.5
rotary vacuum 0.45 43.4
(Source: Chem. Eng.)

As for the costliness index, it is given by the following expression:

.,N
I (costliness index) = (1.3)Y
1
where
N is the number of significant steps in the process
y is the complexity score, determined for each step of the process, and taking various fac-
tors into consideration (throughputs, type of construction material, residence time,
operating conditions, number of flows, etc.).
258 3. The Determination of Battery Limits Investments

A 0.6 t . Reaction
B 0.6 t 1OO'C. 10 atm
S (recycling) 4 t
3.

Filtration Distillation
7
Wash water a-
ndzhing EO'C, ss
Filtrate 5 t
I I I

J.
Product C 1 t

Figure 3.1 1 Example of a block diagram. Method of significant steps.

(Source:Transactions of thejijh international Cost Engineering Congress)

The costliness indices generally lie between 10, for a plant designed to produce a relatively
simple chemical product (formol, for example), and 200, or even more, when it concerns a
more complex facility, for example one that produces a pesticide or a pharmaceutical constit-
uent.
The principle of the method is to make a direct correlation between the amount of the
investment for a production unit, and the chemical process employed, bypassing any consid-
eration of the engineering of the equipment involved. In this it differs fundamentally from the
other rapid cost calculation procedures. Another peculiarity is the fact that the complexity
score for each step is assessed separately for each of the determining factors (throughputs,
corrosion, operating conditions, etc.). Table 3.6 shows how such a calculation is performed.
The basic relation between the battery limits investment (C), expressed in k&,the costli-
ness index, and the size of the plant (S),expressed in kt/year and assumed to be built in the
United Kingdom, is therefore as follows:
c = 45 . I . 9 . 3 9
using an EPE (Engineering and Process Economics cost index) of 300 for that country. The
investment C includes the installation costs and engineering expenses.
The field of application claimed by the author covers a range of unit production capacities
going from 0.3 to 250 kt/yr, for traditional plants in continuous operation, processing both
organic and inorganic products. It can be extended to discontinuous operations with capaci-
ties up to 1 kt/yr, by using a modified equation similar to the previous one:
c = 45 . I . ~ 0 . 6
 Table 3.6 Determination of the complexity level and the costliness index. Method of J.H. Taylor

Highest temperature (“C) 500 1100 1700 2300

Lowest pressure (atm) 1 0.1 0.01
Highest pressure (atm) lo@ 50(a) 200 700 1500
Materials MS(b) SS@) ELMS(‘) Titanium Precious
Inconel Hastelloy metals
EbLMS(e) Nickel Tantalum
PVC Monel
PbLMS(g)
Number of flows 1 2 3 5 7 11
Special conditions:
- Explosion, dust, smell, toxicity: 1 if problem
- Reaction in fluidized beds: 1
- Distillation: 1 i f A < 5°C ; 2 ifA < 1°C
- Specifications (key product): 1 1 i f < 1Oppm
- Falling film evaporation: 1

Overall level (S) -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Costliness index(Z) 0.4 0.6 0.8 1 1.3 1.7 2.2 2.8 3.7 4.8 6.3 8.1 10.6 14 18 23 30 39 51 66
(a): Reaction in liquid phase only. 0 for all other levels. (b): MS = mild steel. (c): SS = stainless steel. (d): RLMS = Rubber Lined Mild Steel. (e): EbLMS = Ebonite
Lined Mild Steel. (f): ELMS = Enamel Lined Mild Steel. (g): PbLMS = Lead Lined Mild Steel.
(Source :Transactions of the F$h International Cost Engineering Congress)
260 3. The Determination of Battery Limits Investments

The procedure reaches its limits when it comes to conducting optimization studies, to eval-
uating the costs of modifying or extending existing plants, or dealing with operations that are
too specialized (electrolysis, plastic extrusion, spinning fibers, etc.).

B. Method of 1.1. Viola

Known as the “Rapid Estimating Technique”, this procedure, like the previous one, consists
of first determining a complexity factor K for the process whose cost is to be estimated. The
battery limits investment (in 1981 $) is then obtained by using Fig. 3.12, which provides a set
of curves corresponding to various production capacities. The value of the function K is cal-
culated from the following expressions:
0 For a facility processing both solids and fluids:
K = N . S . @ ( I I O . I / N ) .(1 - 0.6f,)
0 For a unit processing 100% fluids (i.e.,f, = 0):
K = N . S . @(IIO. IIN)
where
K is the complexity factor
N the number of significant steps in the process
S the correction factor for pressure and for the kind of material employed
I10 the input-output ratio (i.e., raw material-product)
@(I10 . 1IN) the correction factor for average throughput
f, the portion of significant steps operating with solid-fluid mixtures.

Production caDacitv (Mlb/an)

10

1
105 106 107
Battery limits investment ($)

Figure 3.1 2 Determination of the battery limits investment as a function

of the complexity factor and the unit capacity. Method of J.L. Viola.
(Source: Chem. Eng.)
 3. The Determination of Battery Limits Investments 26 1

Generally speaking, a significant step is one that enables the execution of a complete phys-
ico-chemical operation. It thus comprises not only the equipment essential for this operation,
but also all the peripheral apparatus required for it to function properly. In the absence of
exact information, N represents the total number of steps identified, each having a value of 1;
N is therefore usually a whole number. In some cases, knowledge of the plant diagram may
allow a judgment of the relative importance of the various treatments involved in this case,
we may assign a value of 1 to the more complex ones, and assess the others by comparison
(values < 1). As a result, the total may not be a whole number. This way of working offers
the advantage of getting closer to reality.
The value of the fimction may be obtained from Fig. 3.13, using the value of the ratio . The
ratio Z/O may be found directly by considering the input and output flows, and then simply
dividing by N.
S is obtained by using the set of curves in Fig. 3.14, which involve the pressure level,
expressed in psi, and the fraction Mfof equipment (in fact, the portion of significant steps)
that uses more sophisticated steels (actually, AISI type 304 stainless) than carbon steel. In
practice, S results from a weighting of the type:
N
S = x S i Ni I N
1

where Si is a function of M , including the fraction in carbon steel.

fi

COMMENT
The 100 psi curve refers to carbon steel; the ones for higher pressures apply to 304
stainless; this set of curves must be changed or adapted when the kinds of steel
employed are different.

0.1 1 10' 1
VON,inputloutput ratio by number of significant steps

Figure 3.1 3 Determination of the correction factor for average flow.

Method of J.L. Viola. (Source: Chem. Eng.)
 Next Page

262 3. The Determination of Battery Limits Investments

Figure 3.1 4 Determination of the correction factor for pressure and kind
of material employed. Method of J.L. Viola. (Source: Chem. Eng.)

f,is calculated from the number of significant steps in which solids appear, in comparison
to the total number of steps, under the assumption that they all have the same “weight”, or,
preferably, as a function of the total of “relative weights” assigned to them, compared to the
“total weight” of the whole, when such differences can be measured. It is limited to a maxi-
mum value of 0.45.
According to its author, the accuracy is f 15%. In practice, experience plays an important
role in assessing the kind, the number, and the respective weights of the various steps. N is in
fact the parameter that has the greatest influence on the final cost. In addition, as with
J.H. Taylor’s method, it is best to be cautious when the manufacturing diagrams include a lot
of specialized or modular processes involving electrolysis, membranes, crystallization,
adsorption, etc.
In 1984, T.J. Ward revisited the work of J.L. Viola, as well as that of other authors such as
F.C. Zevnik and R.L. Buchanan. In particular, he showed that the procedures suggested usu-
ally produced values that were too high, when compared to the reality. He therefore proposed
to change them, first of all by systematically calibrating them against known examples of
plants of similar design. He then recommended that, especially when applying J.L. Viola’s
method, all the ancillary operations should be left out of the equation: storage, packaging,
shipping, waste removal, solvent purification, heat recovery, etc. Moreover, he defined the
mathematical expressions for the curves in Figs. 3.12,3.13, and 3.14:
I =65 Co.6exp(Z/O. l/N)0.i067
Previous Page

3. The Determination of Battery Limits lnvestments 263

where
Z is the battery limits investment (expressed in 1981 $)
C the production capacity (lying between 2 and 50 million lb/year)
= 1,0167 (Z/O. l/N)0.1067
S = (0.8 + 0.00045 P + 2.5 lod P2)(A4+0.8) + 0.6
where P: pressure (expressed in psi).

3.2.1.5 Method of Functional Modules

(See Section 3.3.2)
In spite of the lack of precision inherent in the preceding methods, to which it is related,
this procedure, which was developed by the IFP during the years 1980-1985, nevertheless
deserves consideration because an evaluator is sometimes confronted by a field situation in
which he is completely cut off from his usual calculating tools.

3.2.2 FACTORIAL METHODS

3.2.2.1 The Structure of Battery Limits lnvestments

To gain a better understanding of the fundamental principles that underlie the factorial meth-
ods, it would seem appropriate to begin by re-examining the definition of battery limits
investments, in the broad sense of the term (manufacturing a product, whether chemical or
not, or a piece of equipment, or utilities) and, to a lesser extent, to re-examine general services
and storage. More specifically, it is not so much a matter of giving a different interpretation,
but rather of looking at their common structure.
Firstly, the argument for treating these two kinds of investment together, apart from the
fact that they essentially relate to plant costs, is that some authors make no distinction
between them as regards the manner of calculation. Thus, for example, H.J. Lang, H.C. Bau-
man, and others suggest a distribution of investments that assigns to the constituents of the
production units themselves a portion of the services that they directly use, another non-spe-
cific portion being assigned to the equipment as a whole. Such an approach is justified by the
fact that at the time these procedures were developed, the unit capacities employed were still
small and concerns about economic optimization and integration were not widely shared.
In addition, whether they concern production or processing plants proper, or the manufac-
ture of utilities or storage, the practical details of the calculations suggested for most factorial
methods are strictly similar. This is one reason why some persons prefer to speak of erected
costs, rather than describing the investment according to its destination or the function of the
equipment.
For these purposes, the amount of the investment can be broken down under various head-
ings, which are usually the following (Fig. 3.15):
0 The cost of the principal equipment, including particularly:

- Columns and drums

264 3. The Determination of Battery Limits Investments

- Tanks (for intermediate products) and storage

- Reactors and similar items (electric generators, cooling towers)
- Heat exchangers, re-boilers, condensers, evaporators (cooling cycles)

- Furnaces and boilers (steam production, dual cycles)

- Pumps and their dnves (turbines, motors, etc.)
- Compressors and their drives

- Specialized equipment: filters, centrifuges, steam ejectors, dryers, mills, crushers,

conveyor belts, etc.

- Instrumentation (optional, unless it concerns automatic control lines)
0 The cost of commodity material:
- Steel frameworks and structures
- Piping and valves

- Thermal insulation
- Instrumentation (if appropriate)

- Electrical system

- Buildings, including control room

- Painting, etc.
0 Cost of erection, connection, and civil engineering:
- Site preparation
- Foundations

- Laying concrete slabs

- Setting up equipment
- Hookups and connections

- Road work, etc.

0 Indirect construction costs and shipping costs:
- Rental and positioning of special hoisting equipment

- Temporary buildings

- Tax, insurance, miscellaneous worksite costs

- Transport of equipment from its delivery point to the site

0 Contingencies:
- Special costs arising for example from unforeseen delays due to strikes, bad weather,

etc.
- Last-minute technical changes

- Price changes, etc.

In some accounts, the engineering costs are sometimes included and set out in full for each
category of equipment, instead of being grouped together as shown in the calculation of
investment costs presented in Section 2.3.1 of Chapter 2.

3.2.2.2 Methods Employing Constant Multiplying Factors

Like all analogy methods, these approaches are based on the collection, examination, process-
ing, and statistical interpretation of information obtained from quotations submitted for actual
plants, or ones assumed to be real. The object of the operation is to deduce more or less empir-
ical correlations or rules, which can be used to make similar calculations in new case studies.
 3. The Determination of Battery Limits investments 265

Contingencies

Indirect construction expenses

hoisting equipment, temporary buildings, tax,
insurance, miscellaneousworksite costs,
transportingequipment, etc.

Erection
civil engineering, foundations, site preparation,
setting up equipment, connections, road work,
etc.

Secondary equipment
steel structures, structures, piping, valves, electricity,
insulation,buildings,control room, painting, regular
instrumentation,etc.

Principal equipment
columns, drums, tanks, storage, reactorsand similar items,
exchangers, evaporators, pumps, compressors, drives,
furnaces, specializedequipment (centrifuges, steam ejectors,
dryers, filters, mills, crushers, conveyor belts, etc.)

Figure 3.1 5 Typical structure of a plant’s erected costs. (Source: ZFP)

As far as the first attempts in the area of factorial procedures are concerned, the basic prin-
ciple obtained from the compilation and analysis of the data available at the time the work
was done consists of accepting that there is a constant relationship between the amount of the
principal equipment and its erected cost. In other words, if we refer to Fig. 3.15, and if we
accept that, given certain simplifying assumptions, the figures for the various headings are
valid for a large number of manufacturing units, we can see that as long as we know the cost
of the principal equipment, arbitrarily set to a normalized value of 100, then a multiplying
factor of 3.5 will give the erected value.

A. Factor Method of H.J. Lang

The first person to demonstrate the existence of such a relationship was H.J. Lang. However,
the constant factors determined in this way show a connection between the cost of the prin-
cipal equipment and the capital for the whole production complex, which in the terminology
266 3. The Determination of Battery Limits Investments

employed up to this point, is equivalent to the battery limits investment (II)and the general
services and storage (I2).Moreover, even at the time when he was doing the work, the author
evidently found that that this factor was not completely constant, since he was ultimately
forced to distinguish three major kinds of manufacturing units on the basis of their products,
as shown in Table 3.7.

Table 3.7 Lang factors

I Type of process 1 Langfactor

Processes for solid products 3.10
Processes for solids and fluids 3.63
Processes for fluids 4.74

Obviously, the results achieved are much more complex than might appear from the
summary figures in Table 3.7, and Lang evidently played a pioneer role in this area, since
this type of relationship is named after him. However, it must also be emphasized that at that
time, unlike today, the state of knowledge and of industrial development, especially con-
cerning unit production capacities, did not permit a differentiation of all the activities and
trades involved in constructing a commercial facility. For this reason, for processes that
involve fluids, the values suggested then seem too low when compared with the ones
employed today.

B. Factor Method of N.G. Bach (Modified rang Factor)

The analyses performed by Bach were similar to those of his predecessor, but he suggested a
different way of looking at the investments. For a particular application, he prefers to draw a
connection between the cost of only the principal equipment devoted to it, and the erected
cost of the corresponding facility. He thus separates the manufacturing unit proper from the
production of utilities and everything concerned with storage, by adopting specific erection
coefficients for each of these categories of facility. In fact, the author feels that these are inde-
pendent entities, whose cost calculation and management procedures are not connected. To
the extent that they are no longer being distinguished on the basis of the products being
processed, the result is not a single value for each, but a range of possibilities, as shown in
Table 3.8. These “constant” factors suggested by Bach do not include engineering costs.
In fact, the usual practice is to talk about one thing when meaning another, since, when the
Lang factor is involved, we actually imply the relationship between the battery limits
investment (I,)and the cost of the principal equipment, i.e., the Bach coefficient. For this
reason, when discussing the matter, it would be more correct to speak of the “modified Lang
factor”.
 3. The Determination of Battery Limits Investments 267

Table 3.8 Bach or modified Lang factors

I Type of unit I Range of Bach factor values I

Product manufacture 2.3 to 4.2
Utility production 1.7 to 2.6
Storage construction 2.8 to 4.8

C. Method of W.E. Hand

This procedure derives from the preceding one, and consists of generalizing the principle by
applying it to each category of equipment. In practice, this amounts to determining the equiv-
alent of an installed cost for each of the types of equipment concerned, by applying specific
Bach coefficients. The total of these various “erected costs” leads to the battery limits invest-
ment for the whole plant. In fact, Hand’s analysis confines itself to a fairly small number of
apparatuses, by concentrating on the most widely used ones (columns, drums, exchangers,
furnaces, pumps, compressors, instrumentation, and miscellaneous), which thus act as repre-
sentative elements, to which are then assigned the burden of secondary equipment, erection,
and indirect construction expenses. Nevertheless, it should not be inferred that this breakdown
is completely arbitrary, since it is based on a statistical compilation of existing data, and cor-
relates well with the facts. The overall results obtained by Hand are summarized in Table 3.9.

Table 3.9 Hand’s factors

I Type of equipment
I Hand’s facto
I

Distillation columns 4
Pressure drums 4
Exchangers 3.5
Furnaces 2
Pumps 4
Compressors 2.5
Instruments 4
Miscellaneous 2.5

As shown in Table 3.10, which contains some of the headings and sub-headings of
Fig. 3.15, these factors result from a much more sensitive and detailed analysis than might
appear from their overall values. Moreover, they allow us to anticipate certain characteristics,
which the methods that follow attempt to take account of by determining installation coeffi-
cients. Thus, we can see that the Hand factors for fixed equipment such as columns, tanks, and
 Table 3.1 0 Method of W.E.Hand

1. Principal equipment
2. Secondary equipment
Column
100
Drum
100
Exchanger
100
Furnace
100
Pump
100 1- Compressor Instrument

Piping 60 65 50 10 30
Electricity 5 5 3 5 73
Instrumentation(1) - - - - -
Sub-total 65 70 53 15 103
Total 1 + 2 165 170 153 115 203
3. Equipment and materials
Concrete 10 5 5 10 5
Structures 15 20 25 - -
Insulation 25 12 14 7 7
Painting 3 3 3 3 3
Buildings - - - - 10
Sub-total 53 40 47 20 25
4. Construction labor
Hoisting & installation of 10 10 3 - 10 10 10
principal equipment
Other construction 72 80 62 15 60 20 50
Sub-total 82 90 65 15 70 30 60
Total 3 + 4 135 130 112 35 95 60 110
5. Transport, insurance, tax (1) - - - - - - -
Total 1 to 5 300 300 265 150 300 190 300
6. Engineering and indirect costs 100 100 85 50 100
Total 1 to 6 400 400 350 200 400
Modular factor 4.0 4.0 3.5 2.0 4.0
(1) To be considered on a case-by-case bas] but must not be overlooked in the calculation.
(Source :Petroleum Refiner)
 3. The Determination of Battery Limits Investments 269

exchangers are high; for the others, especially those involving rotating items, such as com-
pressors and others (i.e., specialized equipment: rotary filters, centrifuges, etc.), these factors
are fairly low, around 2. Furnaces and pumps appear to be exceptions. In the case of furnaces,
it should be noted that, at the time when Hand defined his correlations, the scale of the facil-
ities allowed pre-assembled units to be brought directly to the site, which is no longer possible
today. As for pumps, it must be observed that, in spite of their modest cost when compared
with other equipment, they represent a compulsoryjunction point, generally located at ground
level, which requires the installation of a significant system of piping and valves.
The author also anticipated what other methods have since introduced, by noting that the
employment of specialty steels and the use of high pressures (P > 12 bar) represent two situ-
ations where the principal equipment accounts for a larger share of the total cost of the unit;
in these circumstances he recommends a reduction in the value of the factors usually applied.

3.2.2.3 Methods Employing Variable Multiplying Factors

The proceduresjust described are intended to remain simple and easy to use, and for this rea-
son they are based on average values. Their accuracy improves in proportion as the data to
which they are applied correspond more closely to those of a type facility. As soon as they
depart from this standard, they become less and less realistic. This is especially the case when,
for a particular unit, the cost of the principal equipment is abnormally high or low. Although
described differently, this situation, as W.E. Hand noted, is also encountered when a lot of
apparatus made of specialty steels is used to overcome corrosion problems, when operating
conditionsare severe (for example, high pressure), or when special technologiesare employed.,
Consequently, the need to refine and adapt the modified Lang coefficient arose very
quickly. In practice, the approach came about in two stages. First of all, it involved taking the
measure of the problem, by assessing the size of the variations and identifying the main
parameters responsible for these fluctuations. Attention then focused on the development of
methods capable of incorporating these phenomena and weighting them in a logical manner.
As concerns the first part of the exercise, various authors have found it necessary to break
down the battery limits investments, or more generally the erected costs, into their compo-
nents, following the example of Fig. 3.15, or of W.E. Hand’s approach in Table 3.10. The
original breakdown usually selected is not the same as the formulations eventually suggested
in either of these documents. The reason is simple: the categories listed, on which most of the
work was concentrated, are based on the various groups of trades involved in the construction
of an industrial plant. These particularly include the following:
0 Principal equipment, delivered onsite

0 Erection of principal equipment

0 Instrumentation

0 Erection of instrumentation

0 Underground piping

0 Above-ground piping

0 Structures

0 Buildings

0 Site preparation

0 Foundations
270 3. The Determination of Battery Limits lnvestrnents

Electrical systems
0 Thermal insulation
Access ways, fences, etc.
So as to have a uniform basis for discussion, most authors evaluate the respective weights
of the various items as a proportion of the principal equipment, whose value is arbitrarily set
to 100. In other words, the observed variations, which mostly arise from technological differ-
ences between the plants being studied, are expressed as percentages of the cost of the prin-
cipal equipment. For example, the expenses assigned to piping can range from 52% to 125%
of the cost of the principal equipment, depending on whether the technology is simple, with
few fluid transfer lines, or involves a process with a lot of large-diameter pipes, or a unit oper-
ating at high pressure or having many small-diameter lines of specialty steel.
By 1949 C.H. Chilton had already published ranges of variations, measured as percentages
of the erected cost of principal equipment, for a number of classifications, such as valves and
piping (10 to 40%), instrumentation (5 to 15%), buildings (0 to 8%), ancillary services (0 to
75%), engineering (30 to 40%), contingencies (10 to 40%), etc. He based his figures on a few
simple characteristics representing the degree of technological advancement or the level of
complexity of the plant concerned.
Subsequently, N.G. Bach and W.E. Hand, and later J.E. Haselbarth and J.E. Berk,
H.E. Bauman, and M.S. Peters and K.D. Timmerhaus,published fluctuation ranges for the coef-
ficients of various elements of the investment; depending on the particular case, and following
the original nomenclature used in the present work, this may be an erected cost, a battery limits
investment, fixed capital, etc. By way of illustration, Tables 3.1 1 and 3.12 present some of the
results obtained by M.S. Peters and K.D. Timmerhaus:the first concernsthe average distribution
of investments directly connected to the production unit and its associated facilities, according
to the kind of product processed (direct costs); the second concerns the variations and average
values for all the various cash flow items that make up the fixed capital (direct and indirect costs).

Table 3.1 1 Average structure of direct costs of units

Categories of direct costs I1 Percentage of the principal equipment cost,

Solids
by kind of material processed
1 Solids and fluids 1 Fluids
Principal equipment delivered onsite 100 100 100
Erection of principal equipment 45 39 47
Instrumentation and control (erected) 9 13 18
Piping (erected) 16 31 66
Electrical system (erected) 10 10 11
Buildings (including general services) 25 29 18
Site preparation 13 10 10
General services and provision of utilities 40 55 70
Cost of property (if required) 6 6 6
Total direct costs 264 293 346
(Source: McGmw-Hill)
 3. The Determination of Battery Limits Investments 27 1

Table 3.1 2 Fixed capital structure of a unit

Range of Average value (YO)

Categories of fixed capital variation compared to
(”/) fixed capital principal equipmen1
Direct costs
Principal equipment 20.0-40.0 22.8 100
Erection of principal equipment 7.3-26.0 8.7 38
Instrumentation and control (erected) 2.5- 7.0 3.0 13
Piping (erected) 3.5- 15.0 6.6 29
Electrical system (erected) 2.5- 9.0 4.1 18
Buildings (including general services) 6.0- 20.0 8.0 35
Site preparation 1.5- 5.0 2.3 10
General services and provision of utilities 8.1 -35.0 12.7 56
Cost of property 1.0- 2.0 1.1 5
~

Indirect costs
Engineering and supervision 4.0- 2 1.O 9.1 40
Construction costs 4.8-22.0 10.2 45
Contractor costs 1.5- 5.0 2.1 9
Contingencies 6.0- 18.0 9.3 41
Fixed capital 100.0 439
~~ ~

iource: McGraw-Hill)

From one author to another, there are obvious differences both in the title and in the con-
tent of the categories, and also in the respective weights assigned to them. However, once the
procedures have been edited and normalized, it is possible to see a certain convergence in the
results, and a good overlap in the ranges of variation suggested. This is particularly apparent
when a rearrangement is made on the basis of a distribution such as the one presented in
Fig. 3.15.

3.2.2.4 Methods that Consider the Particular Nature of Each Project

Having clearly demonstrated that multiplying factors, which had originally been taken to be
constant, showed major fluctuations resulting from variations-also considerable-in their
components, various authors turned their attention to developing logical procedures for han-
dling these situations.

A. Identifying the Most Relevant Criteria of Variation

In principle, a given situation can always be described by means of a finite number of param-
eters, but this number is often very large, which makes it practically impossible to represent
the phenomenon in its full reality, using every one of the criteria. However, some of them are
272 3. The Determination of Battery Limits Investments

more significant than others, that is to say they have a greater influence on the final result. If
we accept a certain loss of precision, for example, to come no closer to the truth than some
predetermined level of correlation (>90%), we can considerably reduce the number of vari-
ables to be investigated. The problem then comes down to finding a way to identify the most
relevant criteria. The tools of modern mathematics and computing, such as principal compo-
nent analysis and factor analysis using a distance table, generally allow us to find answers to
such questions. Obviously, during the preceding decades, when the methods for calculating
production battery limits investments and erected costs were slowly being put together, these
tools were both difficult of access and largely unknown. The identification of the most sig-
nificant parameters thus resulted mainly from a heuristic approach, based on experience.
Among the most representative of these explanatory variables, we may mention the fol-
lowing:
The kind of material employed. The presence of corrosive products requires the use of
sophisticated materials or specialty steels, either for all or part of the equipment itself, or as
internal cladding (reactors, exchangers, etc.). Knowing, for example, that the cost for a unit
weight of AISI type 304 stainless steel is 2.5 to 3 times higher than for a carbon (285 C) steel,
then all other things being equal, we may measure the potential effect on the erected cost of
an apparatus of using the same value for the modified Lang coefficient in both situations. In
fact, by comparison with a standard state where the factor is directly applicable, the more the
kind of material used improves, the smaller the modified Lang factor should be. We can see,
in fact, that for hoisting a carbon steel reactor or a specialty steel equivalent of the same
weight, the hoisting expenses (rental and utilization) are exactly the same. The same is true
for erection: the costs of welding or brazing are similar, in that the expense depends essen-
tially on the time taken by the specialist worker who carries out the operation (working time
and wages practically the same in both cases). Turning this concept into a general rule, most
authors have concluded that the coefficient that relates the value of the principal equipment
to its erected cost should be applied only to the base price of the equipment in carbon steel,
or steel assumed to be so, as long as the equivalent sum for this material is recalculated (for
certain pieces of apparatus, centrifuges for example, the reference may be already be a high-
alloy steel).
The unitproduction capacity. As this increases, experience shows that the cost of second-
ary material, erection expenses, and indirect construction costs do not rise in proportion to the
cost of the principal equipment. In fact, certain items are changed little if at all (instrumenta-
tion, buildings, etc.), others vary somewhat (civil engineering, electrical systems, etc.), and
others do grow, but not proportionately (structures, piping, connections, etc.). By way of
example, if the size of a column is doubled, its price is multiplied by approximately 1.5: the
expense of renting a crane for positioning the two kinds of equipment is the same; handling
costs may be higher for the heavier apparatus, but are still negligible in comparison to the
rental cost. A similar reasoning applies to pipe welding: the larger diameters require more
welding and take longer, but in the long run what counts is the number of jobs that a welder
can complete in a day, and this is effectively the same in both cases. Altogether, when the unit
capacity grows, the modified Lang coefficient should diminish. This is another critical mea-
sure that the authors very quickly adopted in developing their methods.
 3. The Determination of Battery Limits Investments 273

The severity of operating conditions. The most critical operating parameter is pressure. In
fact, all other things being equal, when it rises from 50 to 200 bars, the thickness of the vessel
concerned must be multiplied by approximately 4: the same applies to weight, and thus to
price. Keeping the same value for the modified Lang factor in both situations leads to an
erected cost four times as large, which is not in accord with reality. In other words, some
items do not vary in proportion to the amount of the principal equipment: this is particularly
true of hoisting charges or connecting expenses, for exactly the same reasons as those men-
tioned above for the other criteria. Consequently, an increase in the severity of operating con-
ditions should be accompanied by a relative reduction in the modified Lang factor. Only the
latest procedures incorporate this concept.
The cost structure ofprincipal equipment. As shown by W.E. Hand, the factors for calcu-
lating the erected costs of different categories of equipment are a function of their nature: gen-
erally speaking, high values for coefficients should be applied to static equipment, but low
ones to rotating machinery. In other words, first of all the data and the calculation itself should
be detailed enough to enable a specific average factor to be applied to each major family of
equipment; then, by comparison with a reference situation, where all the classes of equipment
have the same weight in terms of erected costs, a correction factor can be applied that takes
into account their respective actual contributions to the production battery limits investment.
Certain authors have developed methods that to some extent incorporate these two features.
The oldest approaches make do with only two or three different types of equipment, for exam-
ple by using a series of curves; the latest ones utilize appropriate coefficients and make sys-
tematic “relative size” corrections with respect to a standard situation covering all equipment.
The type of equipment. The influence of this criterion on variations in the modified Lang
factor is certainly much greater than that due to either of the two previous characteristics, on
the respective weights of the various categories of equipment. It arises from considerations
that are similar to those chosen to reflect the kind of material used. In fact, the same family
of equipment, for example, exchangers, may be composed of elements of different types
(condensers, reboilers, cross-flow exchangers with or without phase change, etc.). The mod-
ified Lang coefficient applicable to this group (which is actually a Hand coefficient), is cal-
culated in reference to a standard equipment item (for example, TEMA exchanger type AES).
In practice, the cost of secondary material, erection expenses, and indirect construction costs
are more or less independent of the type of apparatus concerned. In this situation, we must
calculate the equivalent amount for equipment in the same category, not just in the reference
steel (carbon steel), but also in the basic type, and apply the factor for obtaining the erected
cost only to the standardized equipment. The difference between the actual figure and the
total used for calculation is added at the end. If we apply the same procedure when consider-
ing the severity of use of the equipment, we can see that this difference represents a sort of
“complexity increment”. This approach is employed in the majority of current methods, par-
ticularly the ones used in commercial software.
The list of criteria to be considered could certainly be extended; nevertheless, a compro-
mise has to be found between improvement of the level of accuracy on one hand, and on the
other, the accessibility of information, which rapidly becomes pointless and devoid of inter-
est. In fact, the main qualities expected of a procedure are that it should be both easy to apply
274 3. The Determination of Battery Limits Investments

and flexible. This latter characteristic is particularly popular with engineering companies,
who use analogy methods as a preliminary to consultations, which are often very time-con-
suming, and which also employ a very high level of detail. In line with the French saying: “he
who can do most, may do least”, they prefer to utilize the means at their disposal, rather than
waste them; in this case, when the corresponding data are not available, the procedure must
nevertheless be capable of application, using, by default, the standard values.

B. Methods that Consider the Effects of Materials

Apart from W.E. Hand, who had already raised the question, among those who paid particular
attention to this matter we must first of all mention J. Clerk. In the method he recommends,
he suggests that for each major category of apparatus comprising the principal equipment
(columns, exchangers, pumps, compressors, furnaces, instruments), a graph be constructed
with, on the y-axis, a factor for converting it to the corresponding erected cost (including indi-
rect costs, but excluding engineering), and, on the x-axis, the kind of material utilized. The
value of the independent variable is determined by comparing the price of a piece of equip-
ment constructed in a given stainless steel or some other specialty material, to the price of the
same apparatus if made from carbon steel. Accompanying tables provide examples of the rel-
ative costs of vessels, exchangers, and centrifugal pumps; in some cases, corrections are pre-
sented to take into account the effect of thickness.
J.T. Callagher should also be mentioned: he returned to this approach to enable a rapid
evaluation of the impact of using corrosive reagents on the economics of a system.

C. Methods that Consider the Effect of Capacity

The New York section of the American Association of Cost Engineers developed this proce-
dure; it was named by its authors the “Module Estimating Technique”, and published in the
association’s bulletin. It consists of using graphs to assign to each element of the principal
equipment the expenses required for its erection and its connection to the other apparatus in
the unit.
For the various items (foundations, structures, piping, electrical system, thermal insula-
tion, erection and connection labor), the curves show the value of these expenses in terms of
the cost of the piece of principal equipment concerned. Given that this parameter was itself
determined by the size of that equipment, we are thus able to reflect the plant’s production or
processing capacity.
All that is then required is to add up the expenses for each activity, for all the different
pieces of equipment, and we obtain a breakdown of the expenses for the whole unit, by cate-
gory and sub-category. Such methods enable a rapid analysis of the cost-capacity relation-
ship. However, its utilization is limited:
0 To the types of equipment given by the authors: columns, tanks, exchangers, and pumps.

Furnaces and compressors among other items are not discussed. It is true that quoted
prices (or published information) for such equipment very often concerns erected costs,
To equipment made from carbon steel.
 3. The Determination of Battery Limits Investments 275

D. Method of J. H. Hirsch and E.M. Glazier

This is based on a breakdown of the modified Lang factor (“indirect costs” not included, i.e.,
excluding contractor costs, engineering costs, and contingencies) into three separate items:
Thefieldfactor, FL, which reflects site expenses, excluding supervision
The piping factor, Fp, which reflects the expenses for piping, valves, supports, etc.
0 The miscellaneousfactor, F M ,which covers expenses for foundations, steel structures,

buildings, electrical connections, instrumentation, thermal insulation, site supervision,

etc.
The total production battery limits investment I is therefore defined by the following equa-
tion:
I = E [A(1 f FL + Fp f FM) + B + c]
This approach allows us to make a distinction between the equipment whose erection costs
are not included in the price for onsite delivery, the equipment for which it is in fact included,
and the additional expense occasioned by the use of special materials for certain apparatus
instead of carbon steel.
In this case:
0 A represents the cost of the equipment, assumed to be entirely of carbon steel, whose erec-

tion and connection expenses are not included in the delivered price, and which are, conse-
quently, not explicitly known. This equipment includes items that, because of their relatively
small size, arrive onsite pre-assembled, but still need additional adjustments, as well as ones
that, because of their large size, require onsite assembly.
0 B concerns the equipment whose erection, positioning, and connection costs are included
in the onsite delivery price. At the time when the authors were developing their procedure,
these were mainly furnaces, sometimes compressors, and some intermediate storage; today,
when the unit capacity of plants has very greatly increased, this possibility has almost disap-
peared, which usually comes down to using B = 0.
0 C is the additional price that arises from the use of specialty steels in the manufacture of

certain equipment. It represents the difference between the actual cost of the equipment to be
assembled, and the cost, A, of this same equipment if it were made from carbon steel.
0 E is a coefficient, near 1.4, which reflects the “indirect costs”. J.P. O’Donnel suggested a

more accurate way to determine this factor. In the nomenclature used hitherto, the ratio I/E
corresponds to the definition given earlier for the production battery limits investment I1,
which explains the modifier “total” added to the description of I in the previous equation.
FL, Fp, and FM are obtained from empirical equations that contain, on the one hand, the
value of A itself, and on the other, the ratios to the cost A of the cost of exchangers in carbon
steel (e),the cost of vessels assembled onsite cf)l, the cost of pumps and motors (p), and the
cost of columns, or more precisely the tank shells alone (t).

1. Columns and reactors with diameters greater than 4 m, excepting very high pressure, multiple
bed reactors.
276 3. The Determination of Battery Limits Investments

When the value of the equipment is expressed in thousands of dollars, we obtain the fol-
lowing equations:
logloFL= 0.635 - 0.154 logloA - 0.992 (e/A)+ 0.506 ( f / A )
logloFp= 4 . 2 6 6 - 0.014 logloA - 0.156 (e/A)+ 0.556 @ / A )
FM = 0.344 + 0.033 logloA + 1.194 ( t / A )
FL, F p , and FM may also be obtained directly by using the graphs in Figs. 3.16a, 3.16b,
and 3.16c, expressed in thousands of dollars for the year 1957.
The method of Hirsch and Glazier allows corrections to be made to the modified Lang fac-
tor for the effects of the kind of material used, by means of C, and of the unit manufacturing
capacity and the relative weight of certain categories of equipment, by means of FL, F p , and
FM, which depend on the value of A, and thus on the size of the plant.

E. Method of C.A. Miller

This procedure consists of determining the average cost of a notional piece of equipment that
is representative of the principal equipment concerned. To do this, the total sum for the prin-
cipal equipment is divided by the number of its components, i.e., by the sum of the columns,
tanks, exchangers, pumps, compressors, furnaces, etc.; the instrumentation is not included in
the equipment counted for this calculation.
The average price thus obtained, which characterizes the plant considered, its complexity,
and its size, is then compared to the figures in a real price list, which the author’s experience
has allowed him to compile, and which contains several price brackets (seven in all). Each of
these carries ranges of percentages that apply to a certain number of items, other than the prin-
cipal equipment, that make up the manufacturing battery limits investment: these ranges are
obtained from statistical analyses and thus give lower and upper limiting values.
Within the ranges thus defined, we can select multiplying factors which, when applied to
the average price of the type component of the principal equipment for the corresponding
bracket, enable us to calculate the following items:
0 Erection of the principal equipment

0 Foundations and structures

0 Piping

0 Thermal insulation for the principal equipment

0 Electrical systems

0 Instrumentation

Miscellaneous: site preparation, painting, etc.

0 Buildings related to the principal equipment

0 Other buildings.

When the average cost of the representative component rises, i.e., when we go from the
lowest price brackets to the highest ones, the value of the percentages and consequently the
multiplying factors diminishes: this approach allows us to see the effect of capacity. In fact,
since the intervals that could be assigned to these factors were relatively wide, Miller endeav-
ored to improve the accuracy of his method by defining various possible cases for each item,
which allowed him to make a more detailed breakdown.
 3. The Determination of Battery Limits Investments 277

exchangers
_t_ I vessels assembled onsite
basic principal equipment A , basic principal equipment

Figure 3.1 6a Determination

10 102 103

A: basic principal equipment $000

4 104 0.3 1
FL 410 of the factor FL. Method of
J.H. Hirsch and E.M. Glazier.
(Source: Chem. Eng. Pnog.)

I P pumps and motors

e
-
A
exchangers
pnncipal equipment
basic principal
0
I A basic pnnctpal
I
principal equipment

,- 0.2
- -- -- P
L
0.4 I
0.6

0
AL
0.2

0.4

0.6

Figure 3.16b Determination

of the factor F p Method of

FM 1.2

0.8

0.4

Figure 3.1 6c Determination

0 of the factor FM. Method of
10 102 103 104 105 J.H. Hirsch and E.M. Glazier.
A basic principal equipment $000 (Source: Chem. Eng. Progr)
278 3. The Determination of Battery Limits Investments

Table 3.13 illustrates this method for the calculation of the expenses attributable to foun-
dations and structures, an item offering four possibilities. The same approach is used for the
other items; certain slight differences are sometimes introduced, particularly for building
costs. In this way, by adding the results obtained for the various items, we can determine the
manufacturing battery limits investment.
The method also allows us to obtain the total expenses required for general services, and
the provision of utilities and storage. Here the basis for calculation is no longer the principal
equipment, but the production battery limits investment. A distinction must be made between
plants set up on pre-existing industrial sites and those constructed on new sites (“grassroots”
units), for which the “offsite” portion is necessarily higher.
It must be recalled that in practice Miller’s method presents only ranges of percentage, and
that, where it is not possible to make a choice, the calculation actually provides extreme val-
ues of the partial or total multiplying factors obtained by addition (the modified Lang factor).
The author therefore recommends immediately adding 10% to the lowest figure, and deduct-
ing 10% from the highest one. The Lang factor is thus situated within a smaller range, whose
most probable intermediate value is not-unless by a lucky chance-the arithmetic mean of
the upper and lower limits.
One of the main attractions of this procedure is that it immediately provides results
together with their confidence limits. But it is very demanding of the user’s industrial expe-
rience, which makes it harder to apply, and thereby reduces its impact.

Table 3.1 3 Multiplying factors (in %) for obtaining production battery limits investments.
Example: the item “foundations and structures”. Method of C.A. Miller.

3 000 5 000 7 000 10 000 13 000

4verage price of the representative
-=3000 to to to to to >17000
component expressed in 1958 $
5 000 7 000 10000 13 000 17000
High values:
Units including compressors or
carbon steel equipment requiring
substantial foundations. - - 17-12 15-10 14-9 12-8 10.5-6
Moderately high values:
Plants containing a majority
of carbon steel equipment. - - 12.5-7 11-6 9.5-5 8-4 7-3
Moderately low values:
Units mainly of stainless steel
or including costly equipment. 7-3 8-3 8.5-3 7.5-3 5.5-2.5 5.5-2 4.5-1.5
Low values:
Plants composed of light
equipment not requiring 5-0 4-0 3-0 2.5-0 2-0 1.5-0 1-0
substantial foundations.
- -
Comment: an increase of 25 to 100% should be made when there are special problems during site preparation: for
example, pilings or excavations in rocky ground. Only a detailed study will provide an appropriate value.
(Source: Chem. Eng.)
 3. The Determination of Battery Limits Investments 279

F. Method of K.M. Guthrie

This procedure, like the previous ones, is based on statistical data derived from the analysis
of investments obtained from a number of manufacturing projects.
It is relatively complex, and considers the effects of the following parameters:
The structure of the principal equipment: its components are classified by major categories
of apparatus. Each of these is assigned a multiplying factor that enables calculation of the
erected cost, or if necessary, the various items that make up this cost. Using the fraction of
the principal equipment expense represented by a particular category, we thus obtain, by
applying these coefficients, the amount of the fractions corresponding to the other items, and
then by addition we obtain the investment for the unit itself.
0 Capacity: its effect is considered at two levels. Firstly the author recommends using a set
of curves or equations that provide a “base price” for each piece of equipment proper, as a
function of its dimensions. Secondly, going back to one of the principles of Miller’s method,
discussed in the preceding section, there is a breakdown of the costs of the various categories
of equipment into price brackets, and multiplying factors that vary from one bracket to
another.
0 Operating conditions: the “base prices” obtained for the various component items of equip-

ment can be corrected by means of adjustment coefficients, depending on the conditionsunder

which they f i c t i o n (pressure, temperature, etc.), their size characteristics, their type, etc.
0 The kind of material utilized: correction factors also allow changes in the “base prices”,

which are generally established under the assumption that carbon steel is used.
The method described by Guthrie can equally well be applied to calculating manufacturing
battery limits investments, or those for general services and storage, or even to determining
the various supplementary expenses required for grassroots plants.
For this purpose, it comprises six “modules” within the total investment: five are related
to “direct costs”, the last to “indirect costs”. They are:
0 For “direct costs”:
0 Standard chemical equipment, i.e., for the principal equipment: furnaces, exchangers,

vertical vessels, pumps and their drives, compressors, built-in storage; for the secondary
material: piping, foundations, steel structures, instrumentation, electrical systems, ther-
mal insulation, and painting
0 Solids handling, i.e., for the principal equipment: mills, mixers, centrifuges, conveyors,

crushers, dryers, evaporators, filters, hydraulic presses, sieves, hoppers, equipment for
hoisting and transport; for the secondary material: piping, foundations, steel structures,
instrumentation, and electrical systems
Site development: the property purchase including the costs of topographic surveying
and fees, dewatering and drainage, clearing, excavations, earthmoving, sewers, fencing
and pilings, roadways, footbridges, parking lots, landscaping, walls, firefighting equip-
ment
0 Industrial buildings: offices, laboratories, community health services, shops, ware-

houses, garages, cafeterias, and various mechanical constructions (sheds, etc.)

280 3. The Determination of Battery Limits Investments

0 Ancillary services: steam and electricity generation and their distribution systems, cool-
ing towers and the circulation system for this water, distribution systems for liquid and
gaseous fuels, blow-downs and blow-offs, pollution control facilities, fire hydrants,
purification and settling tanks, lighting and telecommunication systems, storage for
receiving and shipping (trucks, railroad, boats: platforms and docks).

COMMENT
In contrast to the approach followed in the methods considered hitherto, where the
principal equipment constituted a non-dissociable whole for reference purposes,
Guthrie distinguishes two major types of equipment, which he treats separately.

The five modules of “direct costs” include not only the expenses for equipment (principal
and secondary), but also the erection and connection costs. They consequently give us the
erected costs.
0 For “indirect costs”:
0 Freight, tax, and insurance

0 Indirect construction costs (profit margins, welfare costs, site supervision, temporary

facilities, construction materials, work tools, etc.)

Engineering.
The expenses connected to this module are a direct function of those of the other modules.
They are totally determined when the various erected costs are added up. Thus, in practice,
the expenses related to the “direct cost” modules are calculated independently; from their
total we deduce the “indirect costs”; by adding the contingencies and the contractor costs we
arrive at the total investment.
It is not possible to present Guthrie’s method here in its entirety. However, in order to gain
a better understanding of its application, we can show the working of a few modules in spe-
cific cases, for example standard chemical equipment, solids handling, and indirect costs.

a. “Direct Costs” Module: Standard Chemical Equipment

The approach proceeds in two stages.

Calculating the Cost of the Component Equipment

The calculation is based on the determination of the “base price” of each of the component
pieces of equipment that make up the principal equipment. To do this, the author presents a
number of graphs that give a “so-called base price” for each type of standard equipment,
assumed to be made from carbon steel, as a function of the processing capacity, i.e., of the
value of one (or more) representative and characteristic size dimensions (power, exchange
surface, diameter and height, etc.).
Correction factors are then applied to obtain the “real price” of this equipment: in particu-
lar they reflect:
The actual type of device to be utilized, as opposed to the one adopted as a reference to
establish the “base price”,
The operating conditions (pressure, temperature, etc.),
 3. The Determination of Battery Limits Investments 28 1

0 The kind of materials employed,

0 A possible updating, since the “base prices” are obtained from graphs expressed in mid-
1968 $.
As an example, we can consider the case of tube-type heat exchangers. A curve on log-log
coordinates portrays the following linear function:
loglo [base price, tax free, ex-factory (FOB) in $0001
= 0.65 loglo [exchange surface in sq ft] - 1

This allows us to determine, for a known exchange surface, the “price” of an exchanger:
0 With floating head

0 In carbon steel

0 Operating at a maximum pressure of 150 psi,

which serves as a “base price”.

The actual cost of the equipment concerned is then given by the following expression:
exchanger cost in $ = base price in $ . (Fd + Fp). F,. (update)
in which Fd, Fp, and F, are correction factors obtained from Tables 3.14a, 3.14b, and
3.14.c

Table 3.1 4a Method of K.M. Guthrie. Exchangers.

Correction factors for type

Reboiler
Floating head
U-tube 0.85
Fixed tube sheet 0.80
(Sources: Crajsman, Chem. Eng.)

Table 3.14b Method of K.M. Guthrie. Exchangers.

Correction factors for pressure

Calculation pressure (psi) Fd

<I50 0.00
150-300 0.10
300- 400 0.25
400- 800 0.52
800- 1000 0.55
282 3. The Determination of Battery Limits lnvestments

Table 3.14~Method of K.M. Guthrie. Exchangers.

Correction factors for material

Exchange
surface CSI CSI stainless1 CSI MoneY Ti/
CSITi
(sq ft) Mo stainless stainless Monel Monel Ti
400 1.60 1.54 2.50 2.00 3.20 4.10 10.28
100-500 1.75 1.78 3.10 2.30 3.50 5.20 10.60
500- 1000 1.82 2.25 3.26 2.50 3.65 6.15 10.75
1000-5 000 1.30 2.15 2.81 3.75 3.10 4.25 8.95 13.05
5 000- 10000 2.50 3.52 450.00 3.75 4.95 11.10 16.60

(sources: Craffsman, Chem. Eng.)

Determination of the So-called Gross Module Costs

The component equipment is grouped by major categories: reaction furnaces, heating fur-
naces, tube-type exchangers, air coolers, vertical pressure vessels, horizontal vessels, recip-
rocating and centrifugal pumps and their dnves, compressors and drives. For each of these,
we determine an actual cost and a standard base cost by adding the corresponding “prices”
obtained for each piece of equipment.
The amounts of the various items making up the secondary material, and the costs of erec-
tion, setting up, and also the “indirect costs”, although these usually come from a module that
is treated separately, are obtained by using factors expressed in percentages of the “standard
base price” of each category of equipment. In addition, the coefficients vary as a function of
the size, i.e., the absolute value, of this reference, for which the author found it necessary to
present a breakdown by price brackets. In practice, these divisions are presented in the form
of tables that accompany each of the curves for determining the “base prices” of the compo-
nent equipment.
For tube-type heat exchangers, already mentioned above, the results are shown in
Table 3.15.
Using documents of this kind, the user is able to calculate the following items for every
category of standard chemical equipment, on the basis of ordinary steel construction and
using reference systems:
0 Direct costs of equipment and labor, including:

- Direct equipment costs:

Principal equipment
Secondary material
- Direct site costs:

Erection
Set-up.
0 Indirect costs (freight, insurance, and tax; indirect construction costs; engineering).
 Table 3.1 5 Method of K.M. Guthrie. Factors for determining the gross module cost
for tube-type heat exchangers, in % of principal equipment cost

Base cost brackets $ lo5 <2 2to4 I 4to6 I 6to8 I 8to10

Cost of principal equipment 100.0 100.0 100.0 100.0 100.0
Cost of secondary material 71.4 70.5 69.9 69.5 69.3
Piping 45.6 45.1 44.7 44.4 44.3
foundations 5.1 5.0 5.0 5.0 5.0
steel structures 3.1 3.0 3.0 3.0 3.0
instrumentation 10.2 10.1 10.0 9.9 9.8
electrical systems 2.0 2.0 2.0 2.0 2.0
thermal insulation 4.9 4.8 4.7 4.7 4.7
painting 0.5 0.5 .0.5 0.5 0.5
Direct equipment costs I 171.4 170.5 169.9 169.5 169.3

I
Direct construction costs 61.2 60.1 59.4 59.0
erection 55.4 54.7 54.2 53.9 53.8
set-up 63.0 7.6 6.5 5.9 5.5 5.2
Direct costs of equipment and labor I 234.4 231.7 230.0 228.9 228.3

I
Indirect costs(') 86.8 83.9 83.5 81.0
freight, insurance, tax 8.0 8.0 8.0 8.0 8.0
other 94'7 86.7 78.8 75.9 75.5 73.0
Gross module cost I 329.1 318.5 I 313.9 I I 312.4 I 1 309.3 I
(1) Unlike other factors, which are the result of direct statistical compilation, the indirect cost coefficients are given by a particular formula that is shown under the cor-
responding module.
(Sources: Craftsman, Chem. Eng.)
284 3. The Determination of Battery Limits Investments

The sum of the direct and indirect costs constitutes the “gross module cost”. If necessary,
by adding contingencies and contractor costs, we obtain the “total module cost”. The result
obtained from these tables assumes that only reference equipment is utilized. To allow for the
complexity of the equipment actually employed, and the kind of materials really used in their
manufacture, and also for updating, the author defined a “final module cost”, calculated as
follows:
final module cost = updated gross module cost, for carbon steel
+ standard base cost (corrected for type, operating conditions,
and materials) and updated
- uncorrected base cost, updated.

b. “Direct Cost” Module: Solids Handling or,

More Generally, Special Equipment
The approach adopted for treating this module is much simpler than that for standard chemi-
cal equipment, but it is also less accurate and less general.
The main items of equipment for handling solids are presented in a list that also shows, for
the same piece of equipment, several possible variants, depending on type, dimensions, appli-
cation, quality of materials employed, etc. For each alternative, the author first identifies the
kind of representative parameter, which may be a power, a throughput (weight or volume), a
dimension (length, height, etc.), a surface area, or a distinctive feature. Next the price per unit
of the measurement parameter being counted is given in mid-1968 dollars, with the respective
“extrapolation factor”$ In this way, the cost of the equipment concerned can be calculated,
using the following expression:
cost of equipment in mid-1968 $ =
cost of the counting unit . (actual value of the representative parameterjf
Finally the overall multiplying factor is shown, which gives the direct cost of equipment
and labor, as shown in Table 3.16 for standard chemical equipment. The table also shows the
ratio between equipment and labor at this stage, from which the indirect costs can be calcu-
lated, following the general procedure for this particular module.
As an example, in the case of crushers and crystallizers the presentation has the format
shown in Table 3.16.
From the coefficients provided we obtain the actual gross module cost, once the updating
is done. There is no need here to consider the complexity of the equipment concerned, partic-
ularly the kind of materials employed: everything is included in the unit value of the repre-
sentative parameter, and the distinction is made when selecting the kind of equipment to be
used.

c. “Indirect Cost” Module

As already noted, the expenses considered at this stage are directly connected to the invest-
ments for the “direct cost” modules. This approach involves multiplying factors (or rather,
percentages) applied to the price of the principal equipment. An example of this was given in
the discussion of standard chemical equipment.
 3. The Determination of Baffery Limits lnvestments 285

Table 3.1 6 Method of K.M. Guthrie. Determination of the gross module cost for special equipment

Factor for
Kind of cost Labor/
Extrapolation direct cost
Equipment unit of unit equipment
factor of equipment
Barameter 3arameter ($) ratio
and labor
Crushers:
cone ton/hr 750 0.85 1.57 0.23
gyratory tonnu 55 1.20 1.57 0.23
jaw ton/hr 85 1.20 1.57 0.25
pulverizers Ibh 520 0.35 1.59 0.25
Crystallizers:
maturing tonlday 5 500 0.65 1.75 0.38
forced circulation todday 7 900 0.55 1.75 0.38
batch gall 170 0.70 1.60 0.26
(Sources: Craffsman, Chem. Eng.)

In fact, the coefficients come either from direct consideration of the installation require-
ments for the unit concerned, or from formulas that involve the direct costs of the equipment
and the labor.
Thus, for the three main items that make up the module, the author suggests the following
approach:
The transport, tax, and insurance costs are strongly dependent on the place where the units
are installed; in particular, they are a function of the proximity of the factories manufacturing
the equipment, of existing tax provisions and incentives, and the risks incurred. It is conse-
quently impossible to give a precise rule for determining the expenses for this item. For facil-
ities in the United States, or more generally in industrialized countries, Guthrie reckons them
at 8% of the cost of principal equipment.
0 Indirect site costs are calculated from the following expression:
Indirect construction costs in mid-1968 $ = direct costs of equipment and labor
in mid-1968 $ x 0.178 x F,, x F,, x (update)
F,, is a correction factor based on the labodequipment ratio:

Laborlequipment ratio 0.1 0.3 0.5 0.7 1.0

FCO 0.3 1.0 1.3 1.6 2.0

F,, reflects the size of the direct cost of equipment and labor:
Direct cost of equipment and labor 1 3 5 7 10
(mid-1968 $ million)
Fmo 1.07 1.02 1.0 0.97 0.95
286 3. The Determination of Battery Limits Investments

0 Engineering costs are determined by using the following equation:

Engineering costs in $ = direct costs of equipment and labor
in mid-1968 $ x 0.10 x F,, x F, x FPtx(update)
F,, and F,, have definitions similar to those of F,, and F,, i.e.:

Labodequipment ratio 0.1 0.3 0.5 0.7 1.0

Fce 1.6 1.1 0.8 0.7 0.6
Direct cost of equipment and labor 1 3 5 7 10
(mid-1968 $ million)
Fme 1.15 1.05 1.0 0.95 0.90

Fptis a function of the type of unit concerned: 1.4 for chemical complexes; 1.O for com-
mon chemical facilities; 0.8 for mits processing solids and liquids; 0.6 for those processing
solids, and 0.4 for buildings.
The indirect site and engineering costs are then expressed in percentages of the cost of the
principal equipment.
The above analysis of the three modules of Guthrie’s method focuses on the increased
accuracy that it can offer. When it is fully applied, it more closely resembles estimation tech-
niques than those used in preliminary evaluation; in such a context it no doubt requires some
simplification, and this is what the approach developed by the IFP is intended to do, using
factorial procedures (see Section 3.3).
However, we should point out that the validity of the method essentially depends on the
reliability of the curves used to determine the base prices of equipment (CEFs). Now, it is
impossible to summarize on a single graph, using a few correction factors, all the costs that
are actually applied to a whole category of equipment. For this reason, it seems best to apply
this method either directly to prices obtained from discussions with manufacturers, or by
using a more detailed set of curves.

C.The IFP “Prd-Estime” Method

(See Section 3.3.3 of the present chapter.)

H. Method of I. V. Klumpar and S. T. Slavsky

This procedure, called the factor method, assumes that the installed cost (i) of a piece of
equipment includes four elements:
i = E + M + L, + L,
where
E refers to the purchased equipment, i.e., the cost of the principal equipment
M corresponds to the acquisition of supplies for the insertion into the plant, i.e., the cost
of the commodity material
L, covers the labor cost for equipment setting
L, finally, represents the installation-labor cost for commodity material.
 3. The Determination of Battery Limits Investments 287

Roughly speaking, L, and L, cover both the erection and connection costs, as well as the
site costs.
The total costs (i)give the equipment battery limits investment for a complete installed
unit, which is equivalent to I1under the breakdown and nomenclature used earlier.
0 The purchase price E for the principal equipment is obtained from graphs, compiled for

each category of equipment and containing one or more curves, based on a representative size
parameter. Among the equipment included we find traditional items such as columns, tanks,
compressors, pumps, exchangers, etc., as well as components intended mainly for handling
solids, such as filters, cyclones, centrifuges, conveyors, dryers, etc. The independent vari-
ables, depending on the particular case, are power, surface area, throughput, sizes, volume,
etc. The mathematical expressions describing these curves are of the form:

log E = A + a log S + a ' (log q2

where, when a' = 0:
E = Eo (S/So)'

The price M of the commodity material is deducted from the cost of the principal equip-
ment E, either by applying an overall coefficient C, or by adding up the results obtained in
the same way by applying the partial factors ci,in particular for foundations and civil engi-
neering, piping, steel structures, instrumentation, thermal insulation, electrical systems, and
painting, i.e.:
M = E C C=~E . C

As an illustration of this, Table 3.17 gives the values suggested by the authors for some items
of equipment.

Table 3.1 7 Method of 1.V Klumpar and S.T. Slavsky. Cost factors for secondary material
--
1 I
Zentrifuge Conveyoi :yclone hyer Wtei Cxchanger Column Zompressot
--
Foundations c1 9 12 2 9 13 9.0 11 6
Civil
engineering
Piping c2 25 - 9 10 59 49.0 55 20
Steel c3 10 25 22 10 15 7.0 9 7
structures
Instruments c4 10 7 4 6 5 6
Thermal c5 4 - 10 - - 3
insulation
Electricity c6 27 20 47 20 10 26
Painting c7 1 1 1 1 5 0.5 1
--
Total C 86 I 65 95 56
--
107 79.0 103 69
(Source: Chem. Eng.)
288 3. The Determination of Battery Limits Investments

The cost L, of setting up the principal equipment is given by the following equation:
L , = E . H e . (pg/p). we
where L, and E are defined as before, and the other parameters are respectively:
He: the setting-up factor for the equipment concerned, expressed in hours of work per $100
of principal equipment
Pg:the relative productivity factor for the geographic area considered
p: the productivity factor for the reference area (Southern California, mid- 1984), arbi-
trarily set to 1
we:the hourly cost of setting up the principal equipment under the reference conditions,
expressed in dollars per hour worked.
0 The cost L, of installing the commodity material is obtained from the following expres-
sion:
L, = Me (pg/p) . C (h,j . w,j)
in which
L,, M, andp are defined as before
hMjrepresentsthe number of hours worked per $100 of commodity material, as required
by the involvement of a given trade i in installing the ancillary equipment necessary
for the operation of a component of the principal equipment
wCiis the average wage cost in the reference situation, expressed in dollars per hour
worked, for the trade i.
This expression can be simplified by employing the overall average values for all the dif-
ferent trades: it then takes the following form:
L , = M . H M . ( p g / p ) .W
where H M = ChMjand W = the average hourly cost of all the trades involved. If we now return
to the principal equipment, the preceding equations become:

and:
since on the one hand: M=E.C
and on the other we take: hMj= hEjlC
or: HM=HEIC

In this case, hEi represents the number of hours worked per $100 of principal equipment,
as required by the involvement of a given trade i in installing the ancillary equipment neces-
sary for the operation of a component of the principal equipment, and HE= ChEj. Table 3.18
gives an outline of the values of these factors for various pieces of equipment. Among the
problems that may arise in applying Klumpar and Slavsky's procedure is that of obtaining the
relative productivity values, and the hourly costs of services. To address this, the authors sug-
gest that the calculation may be simplified by using the following expressions:
 3. The Determination of Battery Limits Investments 289

L e = E . D e and L c = E . D ~
in which De and DEare factors, expressed directly in dollars per $100 of principal equipment,
relative on the one hand to the setting up of an apparatus, and on the other hand to the execu-
tion of ancillary civil engineering, erection, and connection work. Obviously, this simplifica-
tion takes a toll on the accuracy of the results, even though the coefficients in $/$ are not
affected by the effects of currency inflation, in contrast to those expressed in hours worked/$,
since in this case inflation does not act equally on both elements of the ratio. Table 3.18 also
provides examples of the values of these factors.

Table 3.1 8 Method of 1.V Klumpar and S.T. Slavsky. Cost factors for installing principal equipment
and secondary material. Southern California, mid- 1984

Hours workedB100 of principal equipment $/$loo of

principal
Secondary material equipment
-
n
W a
e
.I
U '5
c)

:
h v)

.
8e
I
W
.
a
M
e
I ie 2
E!
.n
$
8 .
I
h c)
k a
n
.I
> H
I e
.

u
I
8
4
- -
Factor hEl hE2 hE3 hE4 hE7 He DE De
- -
Centrifuge 0.4 0.9 0.4 0.4 0.2 1.0 0.0 3.3 0.5 50.5 10.1
Column 0.5 2.5 0.4 0.5 0.3 0.2 0.1 4.5 0.5 70.3 8.7
Compressor 0.2 0.8 0.3 0.4 0.1 1.0 0.0 2.8 0.2 44.1 3.6
Conveyor 0.9 0.0 1.5 0.3 0.0 1.3 0.1 4.1 0.8 61.0 15.1
Crystallizer 0.4 1.4 0.4 0.6 0.2 0.3 0.0 3.3 0.4 51.5 7.6
Cyclone 0.2 1.9 0.4 0.2 0.1 0.1 0.0 2.9 0.4 46.0 8.2
Exchanger(') 0.3 1.7 0.1 0.4 0.2 0.1 0.0 2.8 0.1 43.9 0.9
Evaporator(2) 0.8 7.8 0.1 1.2 0.2 0.0 0.0 10.1 0.5 160.0 9.0
Filted3) 0.5 2.9 0.4 0.4 0.0 0.3 0.1 4.6 0.4 73.0 7.4
Pump(4) 0.2 1.6 0.0 0.1 0.0 0.9 0.0 2.8 0.6 45.0 9.9
Dryer 1.1 0.9 0.9 0.5 0.0 1.7 0.1 5.2 0.7 79.6 13.0
Blower 0.4 0.0 0.7 0.2 0.1 0.7 0.0 2.1 0.2 31.6 3.0
Fan 0.4 0.0 0.5 0.2 0.0 0.9 0.0 2.0 0.5 29.6 9.6
(I): with tubular shell and bundle. (2): with long vertical shell and tubes. (3): rotary, with pump, larger than 10 m:
(4): centrifugal,rotary, or geared.
(Source: Chem. Eng.)
290 3. The Determination of Battery Limits Investments

COMMENTS
0To remain consistent with the observations in Section 2.3.2.3 of Chapter 2, if the
location effect is the subject of an overall adjustment, there is no need to correct for
productivity when looking at the setting up of equipment, or at civil engineering,
erection, or connection.
0 However, the suggested method of calculation does allow for a better distinction

to be made, if intermediate situations should require it, between local labor and out-
side personnel.

I. Method of SRI (Stanford Research Institute)

The Stanford Research Institute has developed a program called PEPCOST (Process Eco-
nomic Program) for rapid computer calculation of the cost of a plant. Without relying on the
definition and costing of certain basic elements, particularly for processes under development
(outlines, balances, sizing of equipment, etc.), it determines the investment by employing a
series of coefficients similar to those suggested by J.H. Hirsch, and E.M. Glazier, and
C.A. Miller and K.M. Guthrie, for their methods.
One modification consists of allowing for the effect of the materials used when calculating
the portion to be assigned to the piping item. Its cost is generally a fixed percentage of that of
the principal equipment in carbon steel (methods of Hirsch and Glazier, also more specifi-
cally by Guthrie). Taking this approach ignores the fact that the greater the amount of equip-
ment made from specialty steel, the greater is the length of piping that is also made from
sophisticated materials. Accordingly, it is desirable either to increase the increment reflecting
the kind of steel utilized (see parameter C in the method of Hirsch and Glazier), or to alter the
coefficient applicable to the corresponding item.
SRI therefore suggests that a modified piping factor be calculated as follows:
flp =f [ 1 + ( X - A)/A] =f (XlA)

where
fp is the initial piping coefficient
f IP the modified piping coefficient
A the cost of the equipment in carbon steel
X the actual cost of the equipment.

J. The Principal Commercial Evaluation Software

In recent years, every engineering company has developed computer databases, correlations,
and investment calculation programs for estimation purposes. On the commercial front, many
organizations provide their customers either with the results of applying such software, or the
software itself. In the latter case, these are essentially tools for evaluation proper, because
they are directly inspired by the multiplying methods described above, and not built up by
consideration of the successive activities of various trades. SRI is an example of this. In
addition, the American Association of Cost Engineers, mentioned previously, has itself sug-
gested a similar procedure, but there are also many firms that use a similar calculation method
 3. The Determination of Battery Limits Investments 29 1

internally, to provide answers to very specific requests for the comparison of processes, or of
variants of the same technology.
Without claiming to be exhaustive, we can identify two kinds of commercial tools in the
area of evaluation:
0 Chemical engineering software. These programs have the common characteristic of pro-

viding a compact means for sizing the principal equipment. Some of them also enable the
costing of equipment, even a complete economic calculation. Examples of this are “Aspen
Plus” from Aspen Technology, and “Process” from Simulation Science Inc. (SIMSCI).
According to available information the new version of “Process”, called PROAI (with PROAI
Provision) only provides sizing for some equipment. Prosim SA’s “Prosim” goes as far as
providing quotes for principal equipment; Belsim SA, with its “Simu 11” and “Vali 11” stops
at sizing.
Cost calculation software. Some of these are specific to types of equipment, especially in
the United States. Among the ones that can handle the full range of principal equipment we
may single out PDQS and Icarus:
0 PDQS (Price and Delivery Quoting Service), marketed by PDQS Inc., is a relatively

low-cost software program that enables the cost of equipment to be evaluated, as long as
there is a rough and ready preliminary sizing, or a way to approximate the result of the
estimation calculation by specifying the program’s default parameters. In addition, in
this case, the source of the information is given.
0 Icarus (Investigating Costs and Reliability in Utility Systems), from Icarus Corporation,

is a relatively expensive software program that offers a complete range of options, from
sizing of equipment right up to the final economic calculation. However, the user is
expected to provide certain basic data.

3.3 METHODS DEVELOPED BY THE IFP

3.3.1 BACKGROUND

In the course of their regular activities, evaluators will in future have to face two very differ-
ent kinds of situation:
Either they will work at their offices, and in this case will have access to all the computing
assistance they may need. The advances made in this area in recent years enable them to
employ very elaborate models and sophisticated methods, especially ones based on solid
databases and very detailed correlations (CEFs).
Or they will be located far from their usual working base, with no proper computing facil-
ities at their disposal, and nevertheless be obliged to provide an outline of a budget envelope.
In this case, a pocket calculator alone, a few graphs, or just their experience must suffice to
produce an answer.
292 3. The Determination of Battery Limits Investments

In the first case, like most organizations currently engaged in evaluation, IFP usually
employs existing commercial software. It is necessary to observe, however, that these instru-
ments do not always give satisfaction, especially when it is a matter of considering equipment
that has only recently been industrialized, or that has been significantly improved in the last
few years in terms of productivity, size, design, etc. In addition, it must be admitted that the
early tools tend to be preferred, because they are easier to learn about and have a track record,
over the more popular, widely used pieces of equipment. Now, we are currently seeing a dou-
ble evolution, which first of all concerns mass-market products, whose standards are chang-
ing, with the aim of bringing new kinds of equipment to market (plate exchangers, turbines,
falling film evaporators, etc.); on the other hand there is a general tendency in the equipment
field that is directed more towards a greater degree of specialization and increased sophisti-
cation (multiphase reactors, multiphase pumps, simulated counter-current adsorption, etc.).
To address these two concerns, it is necessary, in the first place, to keep up the traditional
evaluating system (based on the use of CEFs), and to update it, if need be, as new generations
of standardized equipment appear. In addition, we must acquire the means to determine the
investments generated by the use of specialized equipment, in other words to employ mathe-
matical models of costs. While taking action on the first point, it may be possible to draw up
a list of the advances in the field, given that it is an instrument developed in-house, but it is
much more tricky to provide a detailed description of the procedures applied in the second
case, since that involves using commercial software.
In a field situation, without any support, only exponential methods can provide any relief.
In these circumstances we must endeavor to preserve the elements that give them their main
attraction and their special nature, i.e., their simplicity of application and especially their min-
imal requirements concerning data, namely:
A general diagram of the plant
0 An overall material balance

An idea of the average values for temperature and pressure

An outline of the kinds of material employed.
Nevertheless, we must try to make up for their intrinsic shortcoming to the extent possible,
by making the most of previous attempts. J.H. Taylor and J.L. Viola have already shown that
significant improvements may be obtained by focusing attention on “significant” groups of
equipment, rather than on some average notional facility. Following this approach, we may
break the plant diagram down into modules, each of them having a functional reality, i.e., exe-
cuting a particular physico-chemical unit operation. In this way of working, a “significant”
stage may include several functional modules of different kinds, or even of the same kind.
Their costs are calculated separately, and the total investment is obtained by addition.

3.3.2 “FUNCTIONAL MODULES“ METHOD

In this section, we will simply describe the principal operating rules on which this method is
based. All the information for actually applying it, in its current state of development, is laid
out in full detail in Appendix A1 .
 3. The Determination of Battery Limits Investments 293

The term module covers not just the basic equipment essential for the physico-chemical
operation concerned, but also all the peripheral equipment.Thus a distillationmodule includes:
the column itself, the reflw tank, the condenser, the re-boiler, the reflw and feed pumps, and,
if necessary, the exchangers for preheating the feed and the complementary cooling of the dis-
tillate and draw-off. A plant diagram may comprise several distillation modules in series. The
same principal applies to all the other parts of the unit.
The method developed by the IFP more specifically addresses the study of the following
modules: reaction, compression, distillation, crystallization, and evaporation, as well as ancil-
lary operations designed to make certain calculations easier (such as the conversion of a
weight material balance into its molar equivalent).
The approach used for finding the cost of the various kinds of modules is based, like most
other methods, on the collection and mathematical and statistical processing of information
of a historical nature. After confirmation against known examples, the correlations obtained
are applied to new cases. What is really different about the method used here has to do with
the vast possibilities now offered by computerization, which makes it easy to perform princi-
pal component analysis, factor analysis using a distance table, and multiple linear regression.
Under these circumstances it becomes possible in principle to consider all the variables that
may be involved in a cost determination, without even having to be concerned about their
interactions. The object of the exercise is then to rank them, and select only the most relevant
ones, i.e., the ones that have a decisive effect on the cost. In practice, this amounts to keeping
only a minimum number, as long as they allow an estimate of the true value that satisfies a
pre-selected correlation level, for example above 90%. Moreover, so that the method shall be
as simple to apply as possible, these parameters must be easily available or, if they are not,
they must be made easy to estimate by some calculating or graphic device.
To illustrate this approach, it is best to look at its application to a particular module: the
relatively complex distillation module seems the obvious choice. The first step consists of
assembling the largest possible sampling of unprocessed data, i.e., information in whatever
form it is available, and covering all the physical parameters involved in a “distillation”,
namely 25 variables including the cost. To ensure that the method is universally applicable,
a representative number of samples must be collected (more than 900 in this case), and they
must cover the various fields (straight-chain alkanes from C1 to (220, branched alkanes, ole-
fins from C2 to C,, diolefins, acetylene compounds, cycloparaffins and cyclo-olefins, alco-
hols, chloroparaffins, aromatics, carbon monoxide and dioxide, hydrogen, water, nitrogen,
acetic acid and acetates, hydrogen sulfide, acetone, ammonia).
Processing this raw information by principal component analysis, i.e., by orthogonal projec-
tion on a plane, did not at first bring out a principal axis, in other words a particular axis along
which the projected images of the points representing the samples in a 25-dimensional space
have a tendency to cluster. To achieve this, the data had to be modified. Experiments having
shown that the kind of material employed was the variable with the strongest influence on cost,
it was decided to homogenize the data set being processed by recalculating the ones correspond-
ing to alloy steel into their equivalent in carbon steel. This time, the analysis of this mixed sam-
ple group, composed of items that are either natural or artificial, led to the identificationof a first
principal axis, then a second one at right angles, in other words a principal plane defined by
these two axes. During this operation, it proved necessary to drop a certain number of samples
for lack of information: the total for the remaining group was reduced to about 600 samples.
294 3. The Determination of Batrery Limits Investments

The resulting factor analysis using a distance table enabled the most relevant ones to be
selected: the ends of the 25 unit vectors were projected along the line defined by the origin of
the 25-dimensional space and the origin of the two principal axes, into a circle of radius 1
(Fig. 3.17). In this case, the selection involved keeping only the ones whose unit vector pro-
jections fell closest to the first principal axis and, whether on the positive or negative side,
closest to the circumference of the circle. The pre-selected correlation level was achieved
with only four variables (pressure, content of lights in the feed, output, and relative volatility).

f Unitvectors

Figure 3.1 7 Principle of factor analysis using a distance table.

(Source: IFP)

The last stage consisted of defining a linear regression, which had the following from:
In (cost) = 4.3588 + 0.5848 x In (DEDIST)
+ 0.7540 x In [In (VORM)] - 0.3477 x In (XLEC) - 0.1138 x In (PTE)
where
(DEDIST) is the molar output of distillate
(VORM) is the average relative volatility
(XLEC) is the proportion of light constituents in the feed
(PTE) is the pressure.
Such equations can also be expressed in graphic form (Fig. 3.18).
 D
f- 50000.0
50000.0:
100.0- - -
-- -
- 20000.0- - 20000.0
-
50.0- C Separation
10000.0-
- -
- - 10000.0
20.0- very difficult 5000.01
-
- - 5000.0
10.0--
-
2000.0- IA-X-B 1
-
-
difficult
/C-Y-D I
-
5.0-
1ow.o=
-
-
-
- 2000.0 lx-2-Y I
m.0-
- 1.75 -
- Material factors
2.0-
-- 1000.0
- -
200.0 -
A - SA 203
1.o 1.0- 100.0 1 -500.0 304 stainless -1.57
--- - 316 stainless -1.99
Pressure
0.75
50.0 7 -
0.5 -
20.0-
-
-200.0
. 150.0
! Erection factor
ISBL -3.39
20.0

50.0 10.0 1
-
-
0.25 Relative
volatility 5.0 -
-
Distillate
' Basecost
(FF 000: Feb. '85)
0.1

Percentage
of lights
0.05 in the feed
(molar X)

Figure 3.1 8 Graphical presentation of the linear regression. Example: distillation module.
Method of functional modules. (Source: ZFP)
296 3. The Determination of Battery Limits hvestments

COMMENTS
In view of the results, one may doubt whether the values of the parameters can be
easily obtained. In practice some changes need to be made, if we are to answer the
initial desire for a simple tool that is easy to apply:
0 The conversion of an overall weight material balance, covering a plant, into more
detailed balances, confined to each module, can be performed in an approximate way
by first considering the assumed yield of a chemical conversion or an equivalent sep-
aration; we then only have to identify the light ends, and make the arbitrary assump-
tion that it is roughly equivalent to a separation into a binary mixture, with a 100%
yield, ignoring impurities. In this approach, an azeotrope is treated like a pure con-
stituent.
We can go from a weight material balance to its molar equivalent by using an aux-
iliary graphic, based on the molecular weights of the constituents of a stream, and its
approximate assumed composition.
0 The content (molar percentage) of light constituents in the feed can be obtained

approximately by considering the distillate/feed ratio.

0 The relative volatility is a variable that is generally hard to obtain. The suggested

graphic allows the actual value to be used when it is known, or to make an estimation
by indicating whether the distillation is more or less difficult to perform. To assist
the user in doing this, an auxiliary table provides some significant examples
(Table 3.19).

Table 3.1 9 Examples of average relative volatilities

Difficulty of the separation Average value Example

Very difficult 1.167 p-x ylenelo-xylene
Difficult 1.338 hexanelbenzene
Average 2.046 benzeneltoluene
Easy 4.602 butanehexane
Very easy 12.327 butaneltoluene
(Source:IFP)

The graphic gives an un-erected cost for the module, assumed to be in carbon steel. To
obtain the final cost, it is necessary to apply a multiplying factor, which in the case of distil-
lation is 3.39, to which is added an increment that depends on the kind of steel actually used
(for example, 1.57 for 304 stainless).
There is a similar approach for each type of module, and an equivalent regression, so that
the production battery limits investment may be found by adding up the results thereby
obtained.
 3. The Determination of Battery Limits Investments 297

The accuracy of the method, in comparison with more sophisticatedprocedures such as the
ones based on the systematic application of multiplying factors, is f 15%. Its interest to some
extent lies in providing a compromise between the extrapolation procedures, and those based
on defining cost estimation formulas (CEFs).
Its major drawback is that, in spite of everything, it demands a certain amount of practice
and knowledge in the field of industrial engineering.

3.3.3 THE PRE-ESTIME METHOD

3.3.3.1 Current State of Development

This procedure, which was developed in-house in the early 1970s, was described at length in
the 1975 edition of the Manuel d 'dvaluationdconomique desprocddds, published by Technip.
At the time, it was based on the rapid sizing of principal equipment with the help of a series
of short-cut methods and graphics grouped in various appendices for each of the major equip-
ment categories. Since then, progress in computing and the marketing of s o h a r e that is both
flexible and user-friendly has rendered the use of such tools obsolete for making simplified
determinations of the dimensional characteristics of principal equipment, to be used for sub-
sequent quotations. In other words, the graphical resolution of that generation of calculations
is no longer of interest, in view of the performance and accessibility of the methods available
now. This evolution will explain why the appendices of the new version of the Prd-Estime
method no longer have a sizing section, and refer users to the usual models for doing this.
However, it must be remembered that this situation does not necessarily eliminate the need
to have a minimal description of the facilities before undertaking this approach. It is therefore
essential to draw up as complete a process diagram as possible, even if it is limited to the basic
physico-chemical operations, and to define material and thermal balances using the available
data and the nominal production capacity selected. At this stage of the design, which is dis-
tantly related-because it is necessarily broader and more approximate-to what the process
specialists have to produce, it involves taking advantage of both the information coming from
R&D or from publications on the subject, and of the basic culture of process engineering. The
main concern at this point is to follow the process that leads from laboratory experiments to
the pilot plant, but then not to forget to make a logical, progressive transition into industrial
equipment. However, it is also desirable to avoid excessive detail. A thorough analysis of the
fundamental mechanisms of the procedure to be industrialized is therefore recommended, so
as to properly identify potential problems and suggest solutions to overcome them, or to adapt
them to the latest commercial production practices. Nevertheless, it is best to concentrate only
on the major points and the elements that, a priori, are the most important in economic terms,
or that could constitute insurmountable obstacles.
The suggested approach then consists of assuming that the results of the design and sizing
stages for the principal equipment are available. From that point on, the Prd-Estime method
comes down to obtaining quotations for this equipment, using on the one hand price curves
for so-called standard equipment, usually based on a single decisive dimensional variable,
and on the other, correction factors that reflect the actual operating conditions and the intrin-
sic, true characteristics of the devices utilized.
298 3. The Determination of Battery Limits Investments

3.3.3.2 Development and Application of the Method

The Pri-Estime method belongs in the category of factorial or multiplying procedures; in its
principle it resembles the method developed by K.M. Guthrie, although it differs from it on
a good number of points. It is meant to be flexible, and capable of being applied equally as
well by engineering companies or consulting firms possessing very detailed data, as by per-
sons or departments responsible for making preliminary analyses and thus not having access
to sufficiently sophisticated sources of information.
To obtain a better idea of its operating principle, it would seem desirable to demonstrate it
on a concrete example: that of tube-type heat exchangers.
In this particular case, where the characteristic dimensional variable is the surface area of
heat transfer, the procedure assumes for the purpose of application that:
0 All the exchangers in the plant have been previously identified and listed

E101, El023 El039 .*.,E201, E202, ...

0 And that for each of them the exchange area, expressed in m2, is known:

Using a graph showing the cost of the equipment as a function of the value of its dimensional
parameter ((Fig. 3.19) for tube-type heat exchangers), and under very specific conditions:
0 European Union mid-2000 €, tax-free, ex-factory

0 Standard material: in the example selected, we specifically assume that the shell and the

tubes are of carbon steel, the apparatus is of TEMA type AES, the operating pressure is
no greater than 10 bar, etc.,
it is then possible to assign a so-called base price to each exchange surface:

and the corresponding real prices (PR)are obtained from the base prices (PB)by using a set
of correction factors cf;), such that:
PReal =PBase *fi‘k!... ’ f p
’ ’

In this case, the expression becomes:

PRE=PBE.fd * f q.A .fnp *fp o f t .fm
where
f d is a factor characteristic of the type of exchanger
f q is a correction factor for diameter and pitch
A is a correction factor for the tube length
fnp is a factor reflecting the number of tube passes
f p is a correction factor for the pressure in the shell and tubes
f t is a correction factor for temperature
f , is a factor reflecting the kind of materials employed.
 3. The Determination of Battery Limits Investments 299

1 000000

100000

i
L

w
10 000
;
I

E
I
c
0
I

h
1 000
1 10 100 lo00
Exchange surface (m2)

Figure 3.19 Determination of the base price of tube-type exchangers.

Prk-Estime method. (Source: IFP)

The values to be assigned to these factors are obtained from Tables 3.20a through 3.20g.

COMMENTS
Not all of the correction factors have the same weight: some have a much stronger
influence on the final result than others. In the case of tube-type heat exchangers, we
can see that three of them make a decisive contribution: the kind of material utilized,
and the type and pressure of operation.
This approach already allows us to judge one of the goals of the suggested method,
which is to be flexible and to be capable of dealing with the greatest possible variety
of situations. In the case of tube-type heat exchangers, when all the equipment is
specified and their characteristics are known (as with an engineering company), the
procedure may be applied in full, by giving each correction factor its corresponding
value. When a preliminary (evaluation) study is concerned, where it would be an illu-
sion to expect every detail to be available, it is sufficient to try to roughly estimate
only two or three of the coefficients, and to adopt a value of 1 for the others.

After making this preliminary calculation, we now have the total actual and base prices of
the exchangers. Obtaining the erected cost for the “exchanger” category is done by using a
reference assembly factor& exchangers, which can if necessary be corrected, using the proce-
dures to be described in Section 3.3.3.3. This applies to the total base prices. To this we add
the difference between the totals of the real and base prices, in other words a complexity
increment that is not affected by the erection.
300 3. The Determination of Battery Limits Investments

Table 3.20 Correction factors

fa Distribution Shell Tubes

~

AES 1.oo A 1.oo E 1.oo L 1.01

AEM 0.90 B 0.95 F 1.03 M 0.90
AEU 0.90 C 1.04 G 1.03 N 0.97
AKT 1.30 J 1.01 S 1.oo
BES 0.95 K 1.25 T 1.05
BEM 0.86 U 0.90
BEU 0.86 (Source: IFP)
BKT 1.25
Thermosiphon 1.20
(Source: IFP)

b. Effect of diameter and pitch c. Effect of tube length

Diameter
Pitch (in) fP
I Length
(in) (mm)
314 19.2 Square: 1 1.oo 8 2.4 1.35
314 19.2 Triangular: 15116 0.95 12 3.7 1.13
314 19.2 Triangular: 1 0.97 16 4.9 1.oo
1 25.4 Square: 114 1.07 20 6.1 0.92
1 25.4 Triangular: 1 1/4 0.97 24 7.3 0.90
(Source: IFP) (Source: IFP)

d. Effect of number of tube passes

Number of passes fnP
2 1.oo
4 1.02
6 1.04
8 1.06
12 1.08
(Source: IFP)
 3. The Determination of Battery Limits Investments 301

Pressure
(bar) Tubes Tubes and shell
1.oo 1.oo
10 1.07 1.12
20 1.17 1.29
30 1.27 1.48
40 1.45 1.81
65 1.63 2.13
85 2.03 2.85
130 3.25 3.25
- ~~

(Source: IFP)

f. Effect of temperature

I Temperature ("C) I ft I
t < 350 1.oo
350 < t < 550 1.08

g. Effect of type of material

f,: material of shelYtubes
Surface area
CSI MoneY
CS1CS CS1Cu CSMo CS1304 3041304 CS1316 Monel Monel
csmi Ti/Ti

1.00 1.50 2.00 2.10 2.80 2.40 3.45 4.80 5.80 13.00
lOO<S< 1000 1.00 1.65 2.45 2.25 3.06 2.40 3.55 5.30 5.80 13.00
Cu:brass. Mo: molybdenum steel. Ti: titanium
(Source: IFP)

If:
-
C A exchangers - PRexchangers
-
'
PBexchangers

We obtain:

erected cost of the exchanger category = fg exchangers corrected + c A exchangers

C PBexchangers
302 3. The Determination of Battery Limits Investments

Since this calculation can be extended to every category of principal equipment, and as
long as we have the corresponding erection factors, we thus obtain:
corrected + C A
pBcol"ms . fgcolums

C P'compressors . fgcornpncssors corrected + C A compressors

C PBexchangers . fgexchangers corrected + C A exchangers

' 'Bpumps . fg pumps

corrected + C A pumps

The total then provides the production battery limits investment:

corrected + ...

+ Z ( X A columns + Z A compressors + ...)

3.3.3.3 Description of the Method

Given a sufficiently detailed process diagram, and knowing the values of the characteristic
dimensional variables for all the principal equipment, the appendices can provide the follow-
ing information for major categories of equipment:
0 A reference price, called the base price, obtained from a specific graph,

0 A real price, calculated from the base price by using correction factors obtained from

tables accompanying the above graphic.

Tables 3.21 and 3.22 then provide the erection factorsfg, called reference factors, for families
of equipment. They are applied to the total of the base prices for each family. The second table,
which more specifically concerns what is called special equipment (centrifuges, dryers, mills,
crushers, rimmed conveyors, etc.), gives these erection coefficients directly, but the first table,
which is much more detailed and deals with mass-produced equipment (columns, tanks,reac-
tors, compressors, tube-type exchangers, furnaces, etc.), mentions the coefficients only on its
bottom line, and shows that they are the simplifiedresult of an intermediateprocedure. This con-
trasting treatment will be explained later, following the complete description of the method.
Thef, factors obtained in this way actually represent a manufacturing plant in which every
category of equipment has the same weight in terms of investment cost, i.e., in real prices. In
practice, when compared with a situation of this type adopted as a reference for calculation pur-
poses, the distribution is usually completely different: the various families of equipment do not
have the same weights, nor are they the same as the standard, except by chance. To reflect this
reality, fg can be modified by means of an adjustment or correction coefficient f,. This is
obtained from the family of curves in Fig. 3.20, which give the values for four major categories
of equipment, as a function of the total realprices of the equipment falling under each category.
 Table 3.21 Coefficients v,)
for converting principal equipment to
production battery limits investments, by type of equipment

Centrifugal
Pumps and
Pressure vessels Heat exchangers compressors(*) Furnaces
drives
Type of equipment

Principal equipment (PB)

Reactor
columns(')
100
ITanks

100
Tube-type

100
Air-cooler

100 100
and drives

100
reaction

100
heating

100
Secondary material vms) 103 63 78 50 56 48 43 42
piping and valves 60 40 45 14 23 17 19 17
civil engineering 10 6 5 2 3 10 11 11
steel structures 8 - 10 20 - - 7 7
instrumentation 11 6 10 4 2 7 4 5
electrical equipment 5 5 2 9 25 11 2 2
thermal insulation 8 5 5 - 2 2 - -

painting 1 1 1 1 1 1 - -
Erection 97 58 60 29 64 54 36 35
Indirect site costs v;) 105 77 84 62 77 70 62 62
Total 405 298 322 24 1 297 272 24 1 239
Multiplying factor v,) 4.05 !.98 3.22 2.41 2.97 2.72 2.41 2.39
(1) Plate assembly included.
(2) For reciprocating compressors the factor& is 2.95. The items to be increased are piping (reduction of vibration and hammering), civil engineering (more massive
foundations), and erection.
(Source: IFP)

w
s
304 3. The Determination of Battery Limits Investments

Table 3.22 Coefficients u,) for converting principal equipment to production

battery limits investments, for various types of special equipment

Equipment Coefficient&
Agitators 1.85
Mills 2.00
Centrifuges 1.80
Conveyors 1.90
Crystallizers 2.00
Steam ejectors 1.20
Evaporators 2.25
Filters 2.00
Dryers 2.00
Vibrating screens 1.50

1.2

1.1

z
c
8
4.0

.2

0.9
0 1 2 3 4 5 6 7 8

Real price (fR)

in mid-2000 000 for each category of principal equipment

Figure 3.20 Correction factor for relative weight of various categories of

principal equipment. Prk-Estime method. (Source: IFP)

COMMENTS
0 The family of curves is defined for a particular date, so that the real reference

prices-for which& takes the value 1 for every category of equipment, when they
have the same weight-vary over time as a function of the depreciation of the cur-
rency. This graph must therefore be updated periodically.
 3. The Determination of Battery Limits Investments 305

The fact that f, is calculated from real prices and not from base prices must be
underlined.
0 The variation of the correction factor is represented by:

- a single curve for every family of equipment, whenf, is less than 1,

- but it takes four different curves whenf, is greater than 1.
In particular, curves 3 and 4 leave room for some interpretation,by allowing the eval-
uator to make possible comparisons to columns, reactors, or more generally to verti-
cal vessels with internal structures, or, failing that, by using the miscellaneous
equipment option. However, this term should preferably be applied to specialized
equipment, often devoted to the treatment of solids (centrifuges, filters, crushers,
etc.).

The procedure described here applies not only to the determination of battery limits invest-
ments for manufacturing or processing, but also to determining the costs of producing utilities
and constructing storage. The principle is the same: it is based on knowing the base and real
prices of the principal equipment for these two types of plant, which can be obtained from
curves and adjustment coefficients presented in an appendix. Then, we have only to apply the
erection factors& from Table 3.23 to the base prices, and if necessary correct them by usingf,,
taken from Fig. 3.19 as a function of the real prices of the categories of equipment concerned.

Table 3.23 Coefficients (f,) for converting the cost of complete equipment
to battery limits investments, for utility and storage production

Equipment Coefficient&
~~

Steam generation:
package boiler 1.50
erected boiler 1.60
Electricity generation 1.40
Cooling towers 1.50
Refrigeration 1.40
Storage tanks:
<I50 m3 1.65
>I50 m3 2.00
Pressurized storage:
horizontal 1.40
spherical 1.85
Source: IFP)

3.3.3.4 Validity and Application limits

Generally speaking, the erection factors& are smaller for miscellaneous equipment than for
commonly used items, because of the greater effect of scale. Apart from this first explanation,
we can also point to the following main reasons:
306 3. The Determination of Battery Limits Investments

0 Most specialized equipment has moving parts: they tend to belong more to the rotating
system class than the static equipment one. Now, as W.E. Hand showed, the modified
Lang factor is smaller in this case.
0 The maximum unit processing capacities are relatively limited for methods involving
diffusion, hydrodynamics, etc. In other words, this kind of equipment can generally be
pre-erected in the workshops of the company that designed it. When it reaches the con-
struction site, it needs only to be positioned and connected.
0 In matters of wear, maintenance, reliability, etc., the standard of reference for special-
ized equipment represents to a more elaborate product than is the case for static devices.
The kind of material employed is often superior (stainless steel); the electrical systems
for operation, measurement, and control are more sophisticated.
Assigning a higher erection coefficient to static equipment only means that it has a greater
weight in the evaluation of site costs, and consequently, that more attention should be paid to
determining its contribution. This is the main reason why a much more detailed breakdown
of its constituent headings was given in Table 3.21, when describing this process. The sec-
ondary material, erection, and indirect construction costs are shown separately, and can be
adjusted by changing their respective sizes according to the conditions encountered: for
example, the principal equipment and the secondary material may in some cases be imported,
while the site labor is local. Going even further, the secondary material heading is broken up
into piping, civil engineering, steel structures, instrumentation,electrical equipment, thermal
insulation, and painting. Doing this makes it possible to work on each item by changing, if
need be, the value of the coefficient that was initially indicated, when it is known that it cor-
responds to a plant that was set up in typical average conditions. To illustrate this possibility,
two examples can be cited
The case of a plant constructed on marshy ground: instead of pouring a concrete slab at
ground level, which is the hypothesis selected for the suggested reference figures, it will
be necessary to place it on shell or ram piles. This requires a correction, which can be
performed by multiplying the standard value by the ratio of the cost of one m2 of slab
seated on such pilings to the cost of one m2 of regular slab.
0 The case of a food-processingplant: all the equipment is at least stainless steel, including

the piping and valves. The suggested reference assumes that, for economic reasons, the
use of superior materials is limited to only that equipment, or those portions of the equip-
ment, which actually require it. In other words, traps or similar devices are provided to
prevent corrosion products from spreading. In these circumstances, most of the piping is
of carbon steel. In the food business, the regulations do not permit such an approach. To
allow for this, the weight of the piping and valve item must be increased by multiplying
the corresponding reference value by the ratio of the cost of a kilogram of stainless steel
(304, for example) to that of a kilogram of carbon steel (235 C, for example).
Here too, the procedure described answers one of the initial concerns, which was to allow
for a certain flexibility, and to make it easy to extend it into situations where more precise
information is available, particularly about plant construction. In the opposite case, as stated
earlier, one only needs to go straight to the all-inclusive value for the factorfg.
 Abbreviations and Acronyms

AACE American Association of Cost Engineers

ABS Acrylonitrile-Butadiene-Styrene
AES Straight tubes, internal clamp-ring floating head cover (TEMA standard)
AFITEP Association francaise du management de projet (French project management
association)
AIE Agence Internationale de 1 'Energie (see IEA)
AISI American Iron and Steel Institute
A M Amsterdam, Rotterdam, Antwerp
ASTM American Society for Testing and Materials
AUC Average Unit Cost
BF Beicip-Franlab
BIC Basic Item Cost
BIEVAL Two-component IFP index for updating investment evaluations
BIP Bulletin de 1'industriepitrolidre
BP British Petroleum
BTS Low sulfur content
Btu British Thermal Unit
BWG Birmingham Wire Gauge
CA Depreciable capital
C&F Cost and Freight
CAP Capacity
CE Chemical Engineering
CEF Cost Estimation Formula
CF Cash Flow
CF Complexity Factor, or Fixed Capital
CHA Heat exchanged
CIF Cost, Insurance, Freight
COMP Compression factor
CPFU Cost Per Functional Unit
cs Carbon Steel
cv Power
DCF Discounted Cash Flow
DDT Dichlorodiphenyl Trichloroethane
2 Abbreviations and Acronyms

DEB Flow
DEC Delivered Equipment Cost
DEDIST Distillate or Molar distillate
DT Temperature differential
EAPF Plate joint exchanger
EbLMS Ebonite Lined Mild Steel
ECMRA European Chemical Marketing Research Association
(now European Chemical Marketing and Strategy Association)
ECN European Chemical News
ELMS Enamel Lined Mild Steel
EMIP Equivalent Maximum Investment Period
ENR Engineering News-Record
EO Ethylene Oxide
EPDM Ethylene-Propylene-Diene-Monomers
EPE Engineering and Process Economics
EVAL n (> 2)-component IFP index for updating investment evaluations
EXP Plant Exponent
FAS Free Alongside Ship
FAST-E Freiman Analysis Systems Technique-Equipment
FCC Fluid Catalytic Cracking
FF Flow Factor
FF French Francs
FOB Free On Board
FPA (shipping contract terms) Free of Particular Average
GDP Gross Domestic Product
GNP Gross National Product
GVI Global Value Index
HC Hydrocarbons
HD High density
HDS Hydrodesulfuration
HDT Hydrotreating
HDPE High Density Polyethylene
hr Hour
HTS High sulfur content
ICARUS Investigating Cost and Reliability in Utility Systems
IEA International Energy Agency
IFP Institut FranGais du Pitrole
IME Institute for Market Economics
INSEE Institut national de la statistique et des itudes bconomiques
I/O Input/Output ratio
IPA Isopropyl Alcohol
IPE International Product Exchange
IPI Industrial Production Index
IRP Interest Recovery Period
IRR Internal Rate of Return
 Abbreviations and Acronyms 3

ISBL In Site Battery Limits

LASL Los Alamos Scientific Laboratory
lb Pound
LD Low Density
LDPE Low Density Polyethylene
LHSV (volume) Liquid Hourly Space Velocity (volume)
LHSV (weight) Liquid Hourly Space Velocity (weight)
LLD Linear Low Density
LLDPE Linear Low Density Polyethylene
LMTD Logarithmic Mean Temperature Difference
MAP-H Mathematical Analysis of Projects-Hardware
(C, E, S) (Construction, Excavation, Software)
M&S Marshall and Stevens, subsequently Marshall and Swift)
MEK Methyl Ethyl Ketone
MF Million Francs
MMF Mkthode des Modules Fonctionnels
MS Mild Steel
MTBE Methyl Tertio Butyl Ether
MTOE Million Tons Oil Equivalent
NRMC Net Raw Material Cost
NYMEX New York Mercantile Exchange
OECD Organization for Economic Cooperation and Development
OLADE Organizacidn Latinoamericana de Energia
PbLMS Lead Lined Mild Steel
PCI Low heat value
PCS High heat value
PDQS Price and Delivery Quoting Service
PE Process Engineering
PEI Process Engineering Index
PEP Process Economics Program
PF Phase Factor
POT Payout Time
PP Polypropylene
PREDICT Process Engineering Data and Information Cost/Time
PRES Pressure
PRICE Parametric Review of Information for Costing and Evaluation
psi (a, g) Pounds Per Square Inch (Absolute, Gauge)
PTE Pressure Top (column)
PVC Polyvinyl Chloride
R&D Research and Development
RCE Relative Capital Enrichment
RFP Request For Proposals
RLMS Rubber Lined Mild Steel
ROI Return On Investment
SBR Styrene-Butadiene Rubber
4 Abbreviations and Acronyms

sc Steam Cracking
SCN S y s t h e des comptes normalisb (Standardized accounting system)
SEC Syst2me europten des comptes kconomiques normalists (Standardized European
economic accounting system)
SECN SystGme klargi de comptabiliti nationale (Expanded national accounting system)
SF State Factor
SIMSCI Simulation Science, Inc.
SMI Small- and Medium-sized industries
sq. ft. Square feet
SRI Stanford Research Institute
ss Stainless Steel
TEFC Total Enclosed Fan Cooled
TEMA Standards of Tubular Exchanger Manufacturers Association
TI Tightness Index
TML Logarithmic Mean Temperature
TP Throughput
TPX Commercial acronym: Polymethyl-4 pentene- 1
TS Residence time
tlyr tons per year
VAT Value Added Tax
VI Volume Items
VORM Mean relative volatility
WEBCI Werkgroep Begrotingsproblem in der Chemische Industrie
XLEC Amount of light constituents in the feedstock
 Bibliography

To help the reader find information on a given subject, the bib-

liography has been grouped by topic. Under each heading, the
references are listed in chronological order.

1. General interest publications

Chilton CH (1960) Cost Engineering in the Process Industries. McGraw-Hill, New York.
Page JS (1963) Estimator’s Manual of Equipment and Installation Costs. Gulf Publishing Co,
Houston.
Bauman HC (1963) Fundamentals of Cost Engineering in the Chemical Industry. Reinhold
Publishing Corp., New York.
Grant EL, Ireson WG (1964) Principles of Engineering Economy. The Ronald Press Co,
New York.
Aries RS, Newton RD (1965) Chemical Engineering Cost Estimation. McGraw-Hill, New
York.
Rase HF, Barrow MH (1966) Project Engineering of Process Plants. John Wiley and Sons,
New York.
Jelen FC (1970) Cost and Optimization Engineering. McGraw-Hill, New York.
Popper H (1970) Modern Cost Engineering Techniques. McGraw-Hill, New York.
Stallworthy EA (1973) The Control of Investment in New Manufacturing Facilities. Gower
Press Ltd., Epping, England.
Park W R (1973) Cost Engineering Analysis. John Wiley and Sons, New York.
Wright MG (1973) Discounted Cash Flow. McGraw-Hill, New York.
Guthrie KM (1974) Process Plant Estimating Evaluation and Control. Craftsman Book Co of
America, Solana Beach, Cal.
Ludwig EE (1974) Applied Project Engineering and Management. Gulf Publishing Co,
Houston.
Bussey LE (1975) The Economic Analysis of Industrial Projects. Prentice-Hall Inc., Engle-
wood Cliffs, N.J.
Happel J, Jordan DG (1975) Chemical Process Economics. Marcel Dekker Inc., New York.
436 Bibliography

Wilson RMS (1975) Cost Control Handbook. Gower Press Ltd., Farnborough, England.
Baasel WD (1976) Preliminary Chemical Engineering Plant Design. American Elsevier Pub-
lishing Co, New York.
Nelson WL (1976) Guide to Refinery Operating Costs. The Petroleum Publishing Co, Tulsa,
Okla.
Guthrie KM (1977) Managing Capital Expenditures for Construction Projects. Craftsman
Book Co of America, Solana Beach, Cal.
Clark FD, Lorenzoni AB (1978) Applied Cost Engineering. Marcel Dekker Inc., New York.
Wilson RMS (1978) Cost Control Handbook. Gower Press Ltd., London.
Jelen FC (1979) Project and Cost Engineers' Handbook. American Association of Cost Engi-
neers, AACE, Morgantown, W. Va.
Kharbanda OP (1979) Process Plant and Equipment Cost Estimation. Craftsman Book Co,
Solana Beach, Cal.
Chemical -Engineering Magazine (1979) Modern Cost Engineering: Methods and Data.
McGraw-Hill, New York.
Peters MS, Timmerhaus KD (1980) Plant Design and Economics for Chemical Engineers.
McGraw-Hill, New York.
Humphreys KK, Kate11 S (198 1) Basic Cost Engineering. Marcel Dekker Inc., New York.
Stewart RD (1982) Cost Estimating. John Wiley and Sons, New York.
American Association of Cost Engineers, AACE (1984) Project and Cost Engineers' Hand-
book. Marcel Dekker Inc., New York.
Humphreys KK (1984) Project and Cost Engineers' Handbook. Marcel Dekker Inc.,
New York.
Page JS (1984) Conceptual Cost Estimating Manual. Gulf Publishing Co, Houston.
Ulrich GD (1984) Economic Analysis. A Guide to Chemical Engineering Process Design and
Economics. John Wiley and Sons, New York.
Gutman N (1985) How to Keep Product Costs in Line. Marcel Dekker Inc., New York.
Westney RE (1985) Managing the Engineering and Construction of Small Projects. Marcel
Dekker Inc., New York.
Axel1 0, Robertson JM (1986) Economic Evaluation in the Chemical Process Industries.
John Wiley and Sons, New York.
Kerridge AE ( 1986) Evaluate Project Cost Factors. Engineering Construction Project Man-
agement. Gulf Publishing Co, pp 217-238, Houston.
Humphreys KK, Wellman P (1987) Basic Cost Engineering (revised and expanded). Marcel
Dekker Inc., New York.
Spear M (1991) Defining home office engineering costs. Process Engng.
Part 1 (Apr.) p 17.
Part 2 (May) p 19.
Part 3 (June) p 2 1.
Part 4 (July) p 17.
Part 5 (Oct.) p 16.
 Bibliography 437

Spear M (199 1, 1992) Estimates of estimating. Process Engng.

Part 1 (Nov. 1991) Terminology, p 18.
Part 2 (Dec. 1991) Statements of error, pp 18-19.
Part 3 (Mar. 1992) Available data, p 18.
Part 4 (Apr. 1992) Correlating the data, p 18.
Part 5 (June 1992) A general correlation, p 18.
Part 6 (July 1992) The cost of cost estimates, p 20.
Part 7 (Sept. 1992) Performance and statistics, p 20.
Part 8 (Oct. 1992) Montecarlo or bust, p 20.
Part 9 (Nov. 1992) The risk analysis, p 20.
Puyhardy-Moreau L, Barrio1 JP, Blonde1 C, Bourdichon P, Estoup J, Garnaud R,
Mouchel JC, Rodi C (1995) Estimation des coQtsd’un projet industriel. AFITEP, AFNOR,
Paris.

2. Market data

Methodology
Duchesnes D (1964) Le r6le de l’ttude de march6 dans l’entreprise chimique. Annales de la
Socittt belge de l’industrie du pttrole 14, p 27.
Twaddle WW, Malloy JB (1966) Evaluating and sizing new chemical plants in a dynamic
economy. American Association of Cost Engineers, AACE, 10th National meeting,
Philadelphie, Penn., June 20-22.
Comgan TE, Dean MJ (1967) Determining optimum plant size. Chem. Engng. 74, 17, p 152.
Amstutz AD (1967) Computer Simulation of Competitive Market Response. The MIT Press,
Cambridge, Mass.
Bright JR (1968) Technological Forecasting for Industry and Government. Prentice Hall,
Englewood Cliffs, N.J.
Box GEP, Jenkins GM (1 970) Time Series Analysis. Forecasting and Control, Holden-Day,
San Francisco.
Infor. Chimie (1970) Les matrices Input-Output europtennes. 80, janv., pp 3 1-38.
Saglin I (1970) How to Price New Products. Modern Cost Engineering Techniques. McGraw-
Hill, New York.
Stobaugh RB (1970) Chemical Marketing Research. Modem Cost Engineering Techniques.
McGraw-Hill, New York.
Jolson MA, Rossow GL (197 1) The Delphi process in marketing decision making. Journal of
Marketing Research VII, Nov., pp 443-448.
Azoulay P, Ponthus P (1974) Les moddes de dtcision dans l’entreprise. Presses Universi-
taires de France, pp 318-335, Paris.
Malloy JB (1974) Projecting chemical product prices. Chem. Engng. Progr. 70,9, pp 77-83.
Selden RA (1978) How forecasters forecast, Chemtech., Apr., pp 204-208.
43 8 Bibliography

De la Mare RF (1978) Long term forecasting. A novel time-trend analysis technique. Engi-
neering and Process Economics 3,2, pp 129-140.
Guyot F (1979) Elkments de macro-kconomie. Editions Technip, Paris.
Stevenson D (1984) Analysing marketing information: a macro view of consumption. Pro-
ceedings of the International Conference of the ECMRA, Amsterdam, 15-17 Oct., pp 189-
215.
Dayan A, Blon J, Cadix A, de Maricourt R, Michon C, Ollivier A (1985) Marketing. Presses
Universitaires de France, Paris.
Chauvel A (1987) Apport de l’evaluation a l’orientation des projets de recherche. Rev. Inst.
FranG. du Pktrole 42, 1, pp 121-134.
Schmeder G (1988) Prkvision technologique: rktrospective critique. Problkmes
Economiques, 2103,14 dkc., pp 2-8.
Sedriks W, Carmichael LA (1993) Predict petrochemicals’ price behavior simply. Chem.
Engng Progr. 89, 10, pp 61-67.

Quantities and prices

Consultants:
Chemical Market Associates Inc. (CMAI), Houston.
Chem Systems Inc., Process Evaluation Research Planning, New York.
Dewitt & Co Inc., Houston.
Frost & Sullivan, Mountain View, Cal.
Parpinelli Tecnon, Milan: World Petrochemical Industry, World Petrochemical Feeds-
tocks.
Poten & Partners Inc., New York.
Purvin & Gertz Inc., Dallas.
Stanford Research Institute, Menlo Park, Cal.: Chemical Economics Handbook, Process
Economics Program, Specialty Chemicals.
SFA Pacific Inc., Mountain View, Cal.
The Freedonia Group Inc., Cleveland, Ohio.
Prices, capacities, production, consumption:
Bulletin de l’industrie pktrolikre, Bulletin des Communautks europkennes, Bulletin men-
sue1 des prix des produits de base Gentve, Cedigaz News Report, Chemical Market Repor-
ter, Chemical Week, Chimie actualitks, Conjoncture - bulletin kconomique mensuel de la
banque Paribas, CPDP - bulletin analytique pktrolier global (Paris), Economic and Energy
Indicators (Washington DC), Enerpresse, European Chemical News, Hydrocarbon Pro-
cessing, ICIS-LOR Reports, Informations chimie, Japan Chemical Week, Kunststoffe,
The Oil and Gas Journal, The Petroleum Economist, Platt’s Oilgram Price Report, Process
Engineering, Revue de Rexecode, Salomon Smith Barney Oil Service Global, L’Usine
nouvelle.
Statistics:
AIE et OCDE - statistiques annuelles de l’knergie, Annales d’kconomie et de statistiques
(INSEE, Paris), Bulletin mensuel de statistiques des Nations-Unies, Bulletin mensuel de
 Bibliography 439

statistique - indices de prix de vente a la production, INSEE - bulletin mensuel de statis-

tiques, Japanese Oil Statistics Today, Statistiques automobiles mondiales (comitt des
constructeurs fianCais d’automobiles), Statistiques en bref tnergie et industrie (Eurostat).
Chambre de Commerce Internationale (1990) Incoterms. ICC Publishing S.A., Paris.
Chambre de Commerce Internationale (1995) Comparaison internationale des prix de l’ilec-
tricitt. Enerpresse, document 6376,28 juillet, pp 1-4.
Chambre de Commerce Internationale (1995) Les tarifs publics en Europe: electriciti, gaz,
eau. Enerpresse, document 6391,22 aoiit, pp 1-4.

3. Indices

Engineering News Record Index

Engineering News Record (1949) 143,9, p 398.

Nelson Cost Index

Nelson WL
(1949) A refinery cost index. The Oil and Gas J. 48,32, p 91.
(1956) How the Nelson index is computed. The Oil and Gas J. 54,74, p 110.
(1957) Construction costs. The Oil and Gas J. 55,32, p 101.
(1962) How to construct cost indexes from components and labor productivity. The Oil and
Gas J. 60,46, pp 153-154.
(1963) New cost indexes for this journal feature. The Oil and Gas J. 61, 14, pp 119-120.
(1965) Tabulated values of Nelson refinery construction cost index. The Oil and Gas J. 63,
14, p 185.
(1965) Where to find yearly indexes? The Oil and Gas J. 63,27, p 117.
(1966) True cost indexes of refinery construction. The Oil and Gas J. 64,2, p 76.
(1967) How Nelson index is computed. The Oil and Gas J. 65, 19, p 97.
(1969) Comparing construction costs abroad. The Oil and Gas J. 67,50, p 76.
(1970) Here are “true” and other refinery cost indexes. The Oil and Gas J. 68,46, p 165.
(1971) Labor cost remain key factor in determining inflation index. The Oil and Gas J. 69,5,
pp 74-75.
(1971) Explaining the Nelson indexes. The Oil and Gas J. 69,34, p 78.
(1972) How to use the Nelson productivities. The Oil and Gas J. 70,38, p 84.
(1973) Refinery construction costs continue upward. The Oil and Gas J. 71,5, p 114.
(1973) Wage ratios and productivity. The Oil and Gas J. 71, 3 1,p 134.
(1973) Effect of location on refinery costs. The Oil and Gas J. 7 1,34, pp 54-56.
(1974) What are the Nelson refinery indexes. The Oil and Gas J. 72,44, pp 105-106.
(1976) Refinery costs compared throughout the world. The Oil and Gas J. 74, 17, pp 102-
103.
440 Bibliography

Marshall and Swift or Marshall and Stevens Cost Index

Stevens RW (1947) Equipment cost indexes for process industries. Chem. Engng. 54, 11,
pp 124-126.
Chem. Engng. (1962) Cost recover from downward dip. 69, 5, pp 125-126.

Chemical Engineering Cost Index

Chem. Engng. (1963) CE’s plant cost index shows the least rate of growth. 70,4, pp 143-152.
Norden RE3 (1969) CE cost indexes: a sharp rice since 1956. Chem. Engng. 76, 10, pp 134-
138.
Ricci LJ (1975) CE cost indexes accelerate 10-year climb. Chem. Engng. 82,9, pp 117-118.
Caldwell DW, Ortego JH (1975) A method for forecasting the CE plant cost index. Chem.
Engng. 82, 14, pp 83-85.

Process Engineering (and Predict) Cost Index

Process Engng. (1972) Productivity and construction costs. Jan., p 13.
Process Engng. (1973) Cost indices. Jan., p 18.
Process Engng. (1973) Plant construction cost indexes. How to compare with the U.S.? Mar.,
pp 108-109.
Bond A (1979) Plant cost indices change with the times. Process Engng., June, pp 10-11.
Bridgwater AV (1988) PEI plant cost indices. Process Economics International 7, 3 and 4,
pp 97-99.
Spear M (199 1) Forecasting plant cost indices. Predict indices. Process Engng., Sept., pp 18-
19.

Process Economics International Cost Index

Allen DH, Bridgwater AV (1979) The PEI plant cost indexes. Process Economics Interna-
tional l , l , pp 4-5.
Brigdwater AV (1979) International construction cost location factors. Chem. Engng. 86,24,
pp 119-121.
Bridgwater AV (1980) Location factors for capital cost estimating. Process Economics Inter-
national 1,3, pp 3-4.
Gerrard AM (1984) PEI two-component indices: a comparison with other established indices
for West Germany. Process Economics International 4,4, p 4.
Allen DH, Bridgwater AV (1985) PEI plant cost indices: a ten year summary. Process Eco-
nomics International 5, 3 and 4, pp 4-5.

Engineering and Process Economics Cost Index

Cran J (1973) Location index compares costs of building process plants overseas. Process
Engng., Apr., pp 109-111.
 Bibliography 44 1

Cran J (1976) Cost indices. Engng. and Process Economics 1, 1, pp 13-23.

Cran J (1976) EPE plant cost indices international. Engng. and Process Economics 1, 4,
pp 321-323.
Cran J (1976) EPE plant cost indices. Engng. and Process Economics 4, 1, pp 3-4.

Boyd Cost Index

Boyd W)
(1981) Cost and price indices. Engng. Costs and Production Economics 5,4, pp 245-254.
(1982) Cost and price indices. Engng. Costs and Production Economics 7, 1, pp 13-18.
(1982) Cost and price indices. Engng. Costs and Production Economics 7,2, pp 97-105.
(1983) Cost and price indices. Engng. Costs and Production Economics 7,3, pp 175-186.
(1984) Cost and price indices. Engng. Costs and Production Economics 7,4, pp 249-255.
(1984) Cost and price indices. Engng. Costs and Production Economics 8, 1 and 2, pp 3-14.
(1984) Cost and price indices. Engng. Costs and Production Economics 8,3, pp 137-144.
(1985) Cost and price indices. Engng. Costs and Production Economics 8,4, pp 237-248.
(1985) Cost and price indices. Engng. Costs and Production Economics 9,4, pp 329-333.
(1986) Cost and price indices. Engng. Costs and Production Economics 10, 1, pp 1- 12.
(1986) Cost and price indices. Engng. Costs and Production Economics 10,3, pp 189-201.
(1986) Cost and price indices. Engng. Costs and Production Economics 10,4, pp 267-270.
(1987) Cost and price indices. Engng. Costs and Production Economics 11, 1, pp 1-12.
(1987) Cost and price indices. Engng. Costs and Production Economics 11,3, pp 125-131.
(1988) Cost and price indices. Engng. Costs and Production Economics 13,2, pp 77-88.
(1988) Cost and price indices. Engng. Costs and Production Economics 13,3, pp 157-168.
(1988) Cost and price indices. Engng. Costs and Production Economics 14,4, pp 247-258.
(1989) Cost and price indices. Engng. Costs and Production Economics 16,4, pp 233-244.

French cost indices and technical parity

Boursin JL (1979) Les indices de prix. Presses Universitaires de France. Collection: Que sais-
je?, Paris.
Pigeyre A (1985) ParitC financiere et parite technique. La Cible 4, 3, pp 17-18.
Benchecroun N (1985) RCflexions sur la parite technique franc-dollar. La Cible 4, 3, pp 18-
21.
Benchecroun N (1987) Indice d’tvolution des coiits des installations industrielles chimiques
et petrochimiques en France. Rev. Inst. Franq. du PCtrole 42, 3, pp 401-407.
ThiCblemont R, Gerdeaux D, Benchecroun N, Gousty Y (1987) Evolution des coiits de
construction dans l’industrie chimique: Claboration d’un indice global rkduit. Rev. Inst.
Franq. du PCtrole 42, 5, pp 617-625.
Benchecroun N (1988) ParitC technique: mythe et rCalitC. La Cible 7,4, pp 16-17.
Montastruc P (1989) Estimation intemationale d’un projet: facteur de localisation. GCnie
Industriel 179, sept.-oct., pp 23-27.
442 Bibliography

Cost indices in other countries

Heller KB (1974) A cost index for plant construction in the Netherlands (A Webci study).
Third International Cost Engineering Symposium, Ref. FS., Oct. 6-9, London.
Schulze J (1974) Development of a cost index for chemical plants in Western Germany. Third
International Cost Engineering Symposium, Ref. F.6., Oct. 6-9, London.
Gallo R, Gasparri M (1977) An italian approach to the cost indices of industrial construction.
Engng. and Process Economics 2,4, pp 28 1-294.
Gerrard AM (1985) Regional variations of plant cost indices in Canada. Process Economics
International 5 , 3 and 4, p 6.

Localization (miscellaneous)
Yen-chen Yen (1972) Estimating plant costs in the developing countries. Chem. Engng. 79,
15, pp 89-92.
Bauman HC (1962) Costs for chemical process plants abroad. Industrial & Engineering
Chemistry 54,9, pp 40-43.
Johnson RJ (1969) Costs of oversea plants. Chem. Engng. 76,5, p 146.
Callagher JT (1969) Efficient estimating of world wide plant costs. Chem. Engng. 76, 12,
p 196.
Alonso JRF (1971) Estimating the cost of gas cleaning plants. Chem. Engng. 78,28, pp 86-
96.
Tems Ph (1973) R61e des pays producteurs de petrole dans l’industrie pktrochimique. Revue
de 1’AssociationFranCaise des Techniciens du PCtrole, nov.-dCc., pp 186-189.
Catry JP (1975) La localisation industrielle (confkrence presentee au stminaire: Evaluation
des avants-projets, au Centre de perfectionnement des industries chimiques de Nancy,
30/05/75).
Miller CA (1979) Converting construction costs from one country to another. Chem. Engng.
86, 14, pp 89-93.
Massa RV (1983) International composite cost location factors. Cost Engng. 25,5, pp 15-21.
Wagialla Kh4 (1984) The use of location factors in estimating the total fixed investment of
chemical plants. Process Economics International 5, 1, pp 37-41.

Inflation
Hirschmann WB (1971) Has the cost of building new refineries really gone up? Chem.
Engng. Progr. 67,8, pp 39-47.
Burford CL, Liles DH, Dryden RD (1977) Effects of inflation on capital investments.
Chemtech, Feb., pp 129-134.
Jones BW (1982) Inflation in Engineering Economic Analysis. John Wiley and Sons,
New York.
 Bibliography 443

4. Methods for calculating investments

General
Zwingelstein G (1984) Panorama des mkthodes d’identification de processus. Le Nouvel
Automatisme, Oct., pp 65-71.
Foussier P (1988) Pourquoi? ou les besoins en matiere d’estimation. Les methodes d’estima-
tion de coQt.Conference AFITEP, Paris, pp 5- 16.
Foussier P (1988) Comment? Les methodes d’estimation de coQt.Conference AFITEP, Paris,
pp 17-40.
Kharbanda OP, Stallworthy EA (1988) Capital cost estimating for the process industries. But-
terworths, London.
Perrin JL, Chauvel A (1994) Projets de R&D: methodes de pre-estimation des cotits.
lre partie, La Cible 53, pp 18-23. 2e partie, La Cible 54, pp 9-13.

Accuracy of estimating methods

Nichols WT (1951) Capital cost estimating. Ind. Engng. Chem. 43, 10, pp 2295-2298.
Hackney J W (1960) Estimating methods for process industry capital costs. Chem. Engng. 67,
5, pp 113-131 et 67,7, pp 119-134.
Stockton DJ, Middle JE (1982) An approach to improving cost estimating. International Jour-
nal of Production Research 20,6, pp 74 1-751.

Extrapolating investments
Williams R, Jr. (1947) Six tenths factor aids in approximating costs. Chem. Engng. 54, 12,
pp 124-125.
Nelson WL (1 949) Overall plant costs cracking. The Oil and Gas J. 48,29, p 9 1.
Nelson WL (1964) How to scale plant costs to other sizes. The Oil and Gas J. 62, 39, pp 84-
85.
Woodier AB, Voolcock J W (1965) The ABC of the 0.6 scale up factor. European Chem.
News, large plant supplement, Sept. 10, pp 7-9.
Chase JD (1970) Plant cost vs capacity, new way to use exponents. Chem. Engng. 77, 7,
pp 113-118.
Guthrie KM (1970) Capital and operating costs for 54 chemical processes. Chem. Engng. 77,
10, pp 140-156.
Earl PH (1977) Specifying escalation indexes to aid in process plant cost forecasting. Engng.
and Process Economics 2,2, pp 119-126.

Exponential methods for calculating investments

Hill RD (1956) What petrochemical plants cost. Petroleum Refiner 35, 8, pp 106-110.
Zevnik FC, Buchanan RL (1 963) Generalized correlation of process investment. Chem.
Engng. 59,2, pp 70-77.
444 Bibliography

Stallworthy EA (1970) The viewpoint of a large chemical manufacturing company. The

Chemical Engineer, June, pp 182-189.
Wilson GT (1971) Capital investment for chemical plant. British Chem. Engng. and Process.
Tech. 16, 10, pp 931-934.
Carmichael A, Sood BM, Muller RG (1972) Capital cost estimates for process screening.
164th ACS Meeting, Aug. 28-Sep. 2, Washington.
Bridgwater AV (1974) Rapid cost estimation in the chemical process industries. Third Inter-
national Cost Engineering Symposium, Oct. 6-9, Ref. A.6, London.
Allen DH, Page RC (1975) Revised technique for predesign cost estimating. Chem. Engng.
82,3, pp 142-150.
Taylor JH (1978) The process step scoring method for making quick capital estimates: expe-
rience in use. Transactions of the Fifth International Cost Engineering Congress, Oct. 30-
Nov. 1, Utrecht (Netherlands), pp 10-14.
Viola JL (1981) Estimate capital costs via a new shortcut method. Chem. Engng. 88,7, pp 80-
86.
Chauvel A, Benchecroun N, Raimbault C, Bergman J (1987) lhaluation rapide du coQt
d’investissement: methode des modules fonctionnels. Rev. Inst. FranG. du Petrole 42, 4,
pp 465-479.

Factorial methods for determining investments

Lang HJ (1947) Cost relationships in preliminary cost estimation. Chem. Engng. 54, 10,
pp 117-121.
Lang HJ (1948) Simplified approach to preliminary estimates.Chem. Engng. 55,6, pp 112-113.
Chilton CH (1949) Cost data correlated. Chem. Engng. 56,6, pp 97-106.
O’Donnell JP (1953) New correlation of engineering and other indirect project costs. Chem.
Engng. 60, 1, pp 188-190.
Bach NG (1958) More accurate plant cost estimates. Chem. Engng. 65, 19, pp 155-159.
Hand WE (1958) From flowsheet to cost estimate. Petroleum Refiner 37, 9, pp 33 1-334.
Hackney JW (1960) Capital cost estimate for process industries. Chem. Engng. 67,5, pp 113-
130.
Haselbarth JE, Berk JM (1960) Chemical plant breakdown. Chem. Engng. 67, 10, pp 158-
160.
Hirsh JH, Glazier EM (1960) Estimating plant investment costs. Chem. Engng. Progr. 56, 12,
pp 37-43.
Walas SM (1961) Plant investment costs by the factor method. Chem. Engng. Progr. 57, 6,
pp 68-69.
Clerk J (1963) Multiplying factors give installed costs of process equipment. Chem. Engng.
Progr. 70,4, pp 182-184.
Hellenach LJ, Johnson EJ, Carnegie DR, Dodge WJ, Bych W, Jr., O’Connell EF (1963) Mod-
ule estimating technique as an aid in developing capital cost. American Association of
Cost Engineers, AACE Bulletin 6,4, pp 182-184.
 Bibliography 445

Miller CA (1965) New cost factors give quick accurate estimates. Chem. Engng. 72, 19,
pp 226-236.
Guthrie KM (1969) “Rapid calc.” chart. Chem. Engng. 76, 1, pp 138-144.
Montfoort AG, Meijer FA (1983) Improved Lang factor appoach to capital cost estimating.
Process Economics International 4, 1, pp 20-2 1.
Ward TJ (1984) Predesign estimating of plant capital costs. Chem. Engng. 91, 19, pp 121-
124.
Blecker HG, Smithson D (1985) Detailed cost estimating during the process engineering
phase of a project. Process Economics International 5,2, pp 17-28.
Klumpar IV, Slavsky ST (1985) Updated cost factors:
Part 1. Process equipment. Chem. Engng. 92, 15, pp 73-75.
Part 2. Commodity Materials. Chem. Engng. 92, 17, p 76-77.
Part 3. Installation Labor. Chem. Engng. 92, 19, pp 85-87.
Klumpar IV ( 1986) Updated capital cost estimation factors. Process Economics International
6 , 3 and 4, pp 40-53.
Klumpar IV, Brown RF, Fromme JW (1988) Rapid capital estimation based on process mod-
ules. Process Economics International 7, 1, pp 5- 10.
Meisl CJ (1988) Techniques for cost estimating in early program phases. Engng. Costs and
Production Economics 14,2, pp 95-106.

Mathematical models and so-called parametric methods

Cochran EB (1976) Using regression techniques in cost analysis. Parts 1 and 2. International
Journal of Production Research 14,4, pp 465-487 and pp 488-5 11.
Freiman FR (1976) Price: a cost-predicting model. Design to Cost Conference Proceedings,
San Francisco, pp 593-602.
Freiman FR (1982) An explanation and discussion of the Fast methodology. 4th ISPA
Conference Proceedings, Virginia Beach, pp 5 1-53.
Freiman FR (1983) The fast cost estimating models. American Association of Cost Engi-
neers, AACE Transactions, Philadelphia, pp G.5.1-G.5.13.
Mauro CV (1984) Overview of the Fast methodology. 6th ISPA Conference Proceedings,
San Francisco, pp 548-575.
Freiman FR (1984) New capabilities in Fast methodology. 6th ISPA Conference Proceed-
ings, San Francisco, pp 576-589.
Freiman FR (1985) New capabilities in the Fast methodology. 7th ISPA Conference Proceed-
ings, Orlando, pp 3 10-334.
Gasperow LA, Hackney JW, Hudson KK (1987) Parametric cost estimating: a guide. Journal
of Parametrics 7,2, pp 67-87.
Foussier P (1988) Illustration de la mkthode parametrique: le modele Fast-E. Les methodes
d’estimation de coiit. Conference AFITEP, Paris, pp 47-92.
Beltramo h4N (1988) Beyond parametrics: the role of subjectivity in cost models. Engng.
Costs and Production Economics 14,2, pp 131-136.
446 Bibliography

Park RE, Korda PB (1989) Making estimates credible or why should I believe what you are
telling me? Journal of Parametrics, 9,2, pp 9-37.
Gottardi G, Coppola L (1990) Elaboration des modeles parametriques d’estimation de coQt.
Actes du 1le congres AFITEP, Paris, pp F.3.1-F.3.5.
XPAR (1992) ExposCs de la journCe d’etude ESCP-XPAR. Le Journal des Paramttriciens,
Janv., pp 1-16.
Melin JB, Jr. (1994) Parametric estimation. Cost Engng., Janv., pp 19-24.
Fournier G (1997) Application des modbles mathtmatiques de coQt a la determination des
investissements dans l’industrie pttrolikre. These, UniversitB de Bourgogne, Dijon - Insti-
tut franCais du petrole, Rueil-Malmaison (soutenue le 20 mai 1997).

Modeling
Enyedy G, Jr. (1979) PDQ$ (Price and Delivery Quoting Service for CPI Equipment). A
computer-based cost estimation service. Transactions of American Association of Cost
Engineers, AACE, Gates Mills, Ohio, July, pp 139-143.
Koehret B, Joulia X, Le Lann MV (1985) Developpement des programmes de simulation des
prockdts chimiques. Le simulateur PROSIM. L’actualitC chimique, juin-juillet, pp 63-70.
Lui A, Branch J. (1985) Costing and economic evaluation using ASPEN PLUS. Process Eco-
nomics International 6, 1, pp 30-37.

5. Project profitability
(1 960) Petrokhem. Engineers, Aug., C8-C44:
Happel J - Economic evaluation.
Newton RD - Selection of plant capacity.
Martin JC - How to measure project profitability.
Clark RW - Consider these factors affecting plant costs.
Schraishuhn EA, Kellett J - How to obtain speed and accuracy in preliminary appraisals.
Happel J, Kapfer WH - Evaluate plant design with this simple index.
Sniffen TJ - The right cost can reduce pumping costs.
Hackney J W (1961) How to appraise capital investments. Chem. Engng. 67, 11, pp 145-164.
McLean JG (1963) Techniques for the appraisal of new investments, 6e Congres mondial du
petrole, Francfort, section VIII, juin, pp 47-63.
Early WE (1964) How process companies evaluate capital investments. Chem. Engng. 7 1,6,
pp 111-116.
Weinberger AJ (1964) Improving R&D’s batting average; How to estimate required invest-
ment; Calculating manufacturing costs; Estimating sales and markets; Methods for esti-
mating profitability; Post audits and qualitative factors. Chem. Engng. 71, 9, pp 123-168.
Urban WJ, Holland FA (1966) How to determine optimum plant size. Chem. Engng. 73, 6,
pp 103-108.
Sinclair CG (1966) Measurement of investment worth. Chem. Process Engng. 47, pp 137-
139.
 Bibliography 447

Walton PR (1966) Cost estimating at the research level. American Association of Cost Engi-
neers, AACE, 10th National Meeting, June 20,21 and 22, Philadelphie.
Twaddle WW, Malloy JB (1966) Evaluating and sizing new chemical plants. Chem. Engng.,
62,7, pp 90-96.
Nitchie EB (1967) Accounting data and methods help control costs and evaluate profits.
Chem. Engng. 74, 1, pp 87-92.
Herron DP (1967) Comparing investment evaluation methods. Chem. Engng. 74,2, pp 125-
132.
Allen DH (1967) Two new tools for project evaluation. Chem. Engng. 74, 15, pp 75-78.
Allen DH (1969) Economic evaluation and taking decision. British Chem. Engng. 14, 6,
pp 790-793.
Chilton TH (1969) Investment return via the engineer’s method. Chem. Engng. Progr. 65, 7,
pp 29-34.
Depallens G (1970) Gestion financihre de l’entreprise. Sirey, Paris.
Malloy JB (1969) Instant economic evaluation. Chem. Engng. Progr. 65, 11, pp 47-54.
Mapstone GE (1971) The present value of plant allowances. Chem. and process Engng.,
April, pp 66-69.
Leibson I, Trischman CA, Jr. (1 97 1- 1972):
Part 1 (1971) Spotlight on operating cost. Chem. Engng. 78, 12, pp 69-74.
Part 2 (1971) How to cut operating costs: evaluate your feedstocks. Chem. Engng. 78, 13,
pp 92-95.
Part 3 (1971) How to get approval of capital projects. Chem. Engng. 78, 13, pp 95-102.
Part 4 (1971) Avoiding pitfalls in developing a major capital project. Chem. Engng. 78,
18, pp 103-110.
Part 5 (1971) A realistic project development case study. Chem. Engng. 78,20, pp 86-92.
Part 6 (1971) Case study shows project development in action. Chem. Engng. 78, 22,
pp 85-92.
Part 7 (1971) Final stage in the project. Chem. Engng. 78,25, pp 78-85.
Part 8 (1971) When and how to apply discounted cash flow and present worth. Chem.
Engng. 78,28, p 97-105.
Part 9 (1972) Decision trees: a rapid evaluation of investment risk. Chem. Engng. 79, 2,
pp 99- 105.
Part 10 (1972) Should you make or buy your major raw material? Chem. Engng. 79,4,
p 76-83.
Part 11 (1972) Zeroing in on “make or buy” decisions. Chem. Engng. 79,6, pp 113-118.
Part 12 (1972) How to profit from product improvement and development. Chem. Engng.
79, 8, pp 103-112.
Ohsol EO (1972) Commercial evaluation of new projects. American Chemical Society, 164th
meeting, Aug.-Sept., New York.
Allen DH (1972) A guide to the economic evaluation of projects. Department of Chemical
Engineering, University of Nottingham, The Institution of Chemical Engineers, London.
448 Bibliography

Holland FA, Watson FA, Wilkinson JK (1972- 1974):

Part 1 (1973) Engineering economics for chemical engineers. Chem. Engng. 80, 15,
pp 103-107.
Part 2 (1973) Capital costs and depreciation. Chem. Engng. 80, 17, pp 118-12 1.
Part 3 (1973) Profitability of invested capital. Chem. Engng. 80, 19, pp 139-144.
Part 4 (1973) Time value of money. Chem. Engng. 80,21, pp 123-126.
Part 5 (1973) Methods of estimating project profitability. Chem. Engng. 80,22, pp 80-86.
Part 6 (1973) Sensitivity analysis of project profitabilities. Chem. Engng. 80,25, pp 115-
119.
Part 7 (1973) Time, capital and interest affect choice of project. Chem. Engng. 80, 27,
pp 83-89.
Part 8 (1973) Statistical techniques improve decision making. Chem. Engng. 80, 29,
pp 61-66.
Part 9 (1974) Probability techniques for estimates of profitability. Chem. Engng. 8 1, 1,
pp 105-110.
Part 10 (1974) Estimating profitability when uncertainties exist. Chem. Engng. 81, 3,
pp 73-79.
Part 11 (1974) Numerical measure of risk. Chem. Engng. 81,5, pp 119-125.
Part 12 (1974) How to estimate capital costs. Chem. Engng. 81, 7, pp 71-76.
Part 13 (1974) Manufacturing costs and how to estimate them. Chem. Engng. 8 1, 8, pp 9 1-
96.
Part 14 (1974) How to budget and control manufacturing costs. Chem. Engng. 81, 10,
pp 105-110.
Part 15 (1974) How to allocate overhead cost and appraise inventory. Chem. Engng. 8 1,
12, pp 83-87.
Part 16 (1974) Principles of accounting. Chem. Engng. 8 1, 14, pp 93-98.
Part 17 (1974) How to evaluate working capital for a company. Chem. Engng. 81, 16,
pp 101-106.
Part 18 (1974) Financing assets by equity and debt. Chem. Engng. 8 1, 18, pp 62-66.
Part 19 (1974) How to assess your company’s progress. Chem. Engng. 8 1, 19, pp 119-124.
Part 20 (1974) Inflation and its impact on costs and prices. Chem. Engng. 81,23, pp 107-
112.
Jelen FC, Cole MS (1974) Methods for economic analysis. Hydr. Process.
Part 1,53,7, pp 133-139.
Part 2,53,9, pp 227-233.
Part 3,53, 10, pp 161-163.
Bussey LE (1978) Basic concepts. The economic analysis of industrial projects. Prentice-
Hall, London, pp 1-185.
Babusiaux D (1990) DCcision d’investissement et calcul Cconomique dans l’entreprise. Edi-
tions Technip, Paris.
Bellus S (1990) La compCtitivitC par la maitrise des cotits. AFNOR, Paris.
This second edition is based on the experience acquired by the French Petroleum
Institute’s Evaluation Department since its founding in 1964. It was written by this
Department’s engineers:

Alain CHAUVEL
Engineer, ENSIC, Nancy
Doctor of Engineering, Paris
Professor at the IFP School
Deputy Director, Strategy & Corporate Planning Division, IFP
Senior Director, Scientific Division, IFP

Gilles FOURNIER
Engineer, ENSIGC, Toulouse
Doctor of Economics, Dijon
Engineer, Strategy & Corporate Planning Division, IFP

Claude RAIMBAULT
Degree in Science, DES in Physics
Engineer, IFP School
Senior Engineer, Strategy & Corporate Planning Division, IFP

In association with Alain PIGEYRE

Engineer, IFP School
International Petroleum Consultant, BEICIP-Franlab
Head of the Evaluation and Project Inspection Department
 APPENDICES

Functional Module Method A1

Prb-Estime Method A2
Evaluation of Pressure Vessels 1

EvaI uation of Reactors A2

2

EvaI uat ion of Heat Exchangers A2

3

Evaluation of Pumps and Compressors A2

4

Evaluation of Furnaces A2
5

Evaluation of Steam Ejectors A2

6

Evaluation of Special Equipment A2

7

Evaluation of Units for Providing Utilities A2

8

Evaluation of Storage A2
9
 APPENDIX1
Functional Modules
Method (FMM)

Although the FMM resembles other modular methods, it differs from them in that its princi-
ple is not to look at each item of equipment in a plant, but rather at the groups of equipment
required to perform a particular physico-chemical operation (distillation, reaction, etc.).
However, it has proved possible to establish a direct relationship between the prices of such
groupings, called “modules”, and the basic physico-chemical parameters that control their
functioning.
The choice of the most important technical parameters (the key variables) for each of the
functional modules emerges from a prior statistical analysis. In practice, the base price of a
module is obtained from regressions or graphics, by assigning each representative parameter
its actual or approximate value. The erected cost (or the battery limits cost) is obtained by
using correction factors that are characteristic for each module, and which themselves come
from a statistical correlation.
The overall battery limits investment is obtained by adding up the erected costs of the var-
ious modules identified on the plant’s flowsheet l .

Al.l DATA ANALYSIS

The statistical analysis performed on the selected samples had three objectives:
0 Confirmation of the consistency and representativeness of each module’s database,

0 Selection of the most significant technical parameters in each case, i.e., the key vari-

ables,
0 Definition of regression models.

The collection of information involved more than nine hundred different sets of equip-
ment. After being accepted into the sampling, each one was recalculated to identify possible

1. The flowsheet may include several successive modules of the same kind. This is particularly
true for reaction, when there are reactors in series, or for distillation when this operation includes a
series of columns, each with a specific function.
 3 10 Appendix 1. Functional Modules Method (FMM)

Interval Frequency Percentage

Values 5 10 15 20 25 30 Ink Cum. Int. Cum.
._
__._ __ _
___ . _____..
1.2 xxxxx xxxxx xxxxx x 21 21 28.0 28.0
2.0 xxxxx xxxxx xxxxx xxxxx xxx 26 49 37.3 65.3
3.5 4 53 5.3 70.7
5.0 xx 7 60 9.3 60.0
10.0 xx 7 67 9.3 89.3
20.0 4 71 5.3 94.7
25.0 4 75 5.3 100.0
...._
45.0 xxx 8 8 10.7 10.7
50.0 4 12 5.3 16.0
75.0 xxxxx xxxx 14 26 18.7 34.7
100.0 xxxxx xxxxx 15 41 20.0 54.7
125.0 xxx 8 49 10.7 65.3
A1 150.0
180.0
xxxxx xxxxx
xxxxx X
Temperature (“C) 15
11
64
75
20.0
14.7
85.3
100.0
_..__ _ _ ____ _ _._ _ . - _ _
60.0 xxxxx X 11 11 14.7 14.7
90.0 xxx 8 19 10.7 25.3
120.0 xxxxx xxxxx x 16 35 21.3 46.7
240.0 xxxxx xxxxx xxxxx 20 55 26.7 73.3
1000.0 xxxxx xxxxx x 16 71 21.3 94.7
2000.0 Distillate flow (kmohr) 2 73 2.7 97.3
2100.0 2 75 2.7 100.0

0.8 xxxxx 20 20 26.7 26.7

1.5 xxx 8 28 10.7 37.3
2.0 xxxxx XXX 13 41 17.3 54.7
5.0 XXXXX xxxxx xxxxx x 21 62 28.0 82.7
7.5 xx 7 69 9.3 92.0
12.0 Reflux rate 3 72 4.0 96.0
14.0 3 75 4.0 100.0

1.16 4 4 5.3 5.3

1.45 XUX 9 13 12.0 17.3
2.01 xxxxx xxxxx XXXxx xxx 23 36 30.7 48.0
3.40 xxxxx xxxxx xxxxx 20 56 26.7 74.7
5.75 xxx 8 64 10.7 85.3
11.58 xx Relative volatility 7 71 9.3 94.7
17.00 4 75 5.3 100.0
___
- _ _ _ _ _--_- _.__ ___ _____
20 xxxxx X 11 11 14.7 14.7
30 xxxxx XXXX 14 25 18.7 33.3
50 xxxxx xxxxx xxxxx xxx 23 48 30.7 64.0
75 xxxxx xxxxx 15 63 20.0 84.0
100 xxxx 9 72 12.0 96.0
130 Number of plates 2 74 2.7 98.7
140 1 75 1.3 100.0
_____ _---- -____ ---__ .__-_
750 xx 7 7 9.3 9.3
1 500 xxxxx xxxxx xxxxx x 21 28 28.0 37.3
3 000 xxxxx xxxxx xxxxx 20 48 26.7 64.0
5000 XXMX X 11 59 14.7 78.7
7 750 XXXXX 10 69 13.3 92.0
10 000 Total cost ISBL (FF 000 - 02/85) 4 73 5.3 97.3
12 250 2 75 2.7 100.0
_ .. - - - _ _ __
_ _ __ _ _ __ ___.__
~

10 15 20 25 30

Figure A1 .I Distillation module. Histogram of principal variables in the

selected sample. (Source: IFP)
 Appendix 1. Functional Modules Method (FMM) 311

disparities that were too large and/or impossible to explain in terms of the model, which was
being developed in parallel. At the end of this primary selection process, just over six hundred
sets were retained.
In addition, the choice involved equipment for treating a wide range of chemical products.
For this purpose the range selected was designed to be wide enough for the method to have
the broadest possible coverage. The following items were included
n-paraffins, from c1to C20
Branched paraffins
Olefins, from C2 to C9
Diolefins
Aromatics
Cycloparaffins and cyclo-olefins A1
Alcohols
Acetylene and butyne
Chloroparaffins and aromatics
Carbon dioxide and monoxide
Hydrogen, nitrogen, water, and ammonia
Hydrogen sulfide and acetone.
Similarly, in compiling the sample, a special effort was made to obtain a representative
grouping, providing the best possible coverage of the various industrial operating conditions
normally encountered in operating this equipment. Thus in Fig. A1 .1 for example, we see that
although the range for the number of plates in a distillation column covers values as extreme
as 12 and 138, in practice most of the numbers fall between 20 and 40. This same procedure
was also followed for the other variables.
The data analysis techniques that had already been applied in studying the consistency of
the sampling proved to be particularly valuable in selecting the key variables, especially in
the case of unit distillation and reaction operations, given their complexity and the large num-
ber of interdependent parameters. The construction of predictive models, using the appropri-
ate regression techniques, was also made easier in both cases by the prior application of these
techniques. For “reaction”, for example, it was quickly shown that a single model could not,
by itself, reflect the diversity of the situations encountered.

Al.2 DESCRIPTION OF THE METHOD

The models that have been developed, which can be applied by using regressions or nomo-
grams that enable determination of a total erected price, take the following headings into
account:
The cost of principal equipment (columns, drums, tanks, heat exchangers, evaporators,
ovens, boilers, compressorsand drives, control instruments and various accessories,etc.),
The cost of secondary material (valves, piping, thermal insulation, steel structures, instru-
mentation and electrical systems, buildings including the control room, painting, etc.),
Erection and civil engineering expenses (site preparation, foundations, setting up equip-
ment, hookups and connections, road construction, etc.),
 312 Appendix 1. Functional Modules Method (FMM)

0 Indirect construction and transport expenses (rental and positioning of special hoisting
equipment, temporary site buildings, tax, insurance, and miscellaneous site costs, etc.),
0 Contingencies.
The cost of the principal equipment is determined at the level of each module. It represents
what we will call here the base price. The application of assembly factors for each type of
module provides for inclusion of the other components of the erected cost, as mentioned
above.
To extend the method to materials that are different from the (carbon steel) ones taken as
references for the models, we employ another series of multiplying factors. In the same way,
certain operating parameters, whose effect is not dominant but which may have a significant
weight, may also be considered.
A1 The utilization of the method does not necessarily require that the exact structure of the
modules concerned be known in advance. However, when this information is available,
Table Al. 1 presents a partial list of the equipment likely to be used in each one.

Table A l . l The functional modules and their operations

Functiona1 Partial list of equipment involved:

Module
1 module D operation equipment
Distillation D Standard, fractionation, stripping.
Shell, bottoms, skirt, reflux drum, plates and/or packing,
accessories, heat exchangers and/or fiunaces.
Compression D Centrifugal, reciprocating, and rotary.
Motor, reduction gear, base, intermediate exchangers, fans,
ancillary equipment.
Reaction D Tube-type, multitubular, fixed bed, fluidized bed, furnace, 3
absorption, Grignard.
Shell, bottoms, skirt, plates and/or packing, heat exchangers and/or
furnaces and/or coils and/or jacket, motor, reduction gear, external
pump system, turbine or propeller agitator, compression gland,
mechanical packing, and accessories.
Crystallization D Forced (external), forced (internal), cold cycle soaking (batch).
Crystallizer and accessories such as motor, reduction gear, vacuum
equipment, baffle, stirrer, external exchangers, coils, jacket,
scraping devices.
Evaporation D Forced circulation, long vertical shell and tubes, thin film, standard
(shell and tubes), falling film and jacket (horizontal or vertical).
Evaporator and accessories such as baffle, motor, reduction gear,
external drum, shaft with moving scrapers.
(Source: IFP)
 Appendix 1. Functional Modules Method (FMM) 313

The five modules examined in this way cover, by themselves, almost all of the equipment
involved in current petrochemical and organic chemical processes. These are the ones that are
generally used in the evaluation of a plant for which we have only a block diagram, a material
balance for the main production lines, and approximate knowledge of the operating condi-
tions (pressure and temperature levels).
Nevertheless, we sometimes have to allow for additional equipment such as tanks, drums,
furnaces, or exchangers that cannot be included in any of the modules considered. In this case
the method provides for the use of auxiliary modules. They are listed in Table A1.2.

Table A l . 2 Auxiliaj modules

A1
Module Auxiliary
no. module
A. 1 Flash/separating/refluxdrums
A.2 Intermediate storage drums
A.3 Blenders
A.4 Heating furnaces
AS Heat exchangers
(Source: IFP)

In addition, it is recommended that the total investment be increased by 5 to 10 percent of

the base cost, which represents the portion of the investment covering pumps.

Al.3 APPLICATION OF THE FUNCTIONAL MODULE M E T H O D

In order to make the method easier to use, we have emphasized the presentation of the equa-
tions themselves rather than their expression in graphical form. In fact, the accuracy of the
calculations is improved by doing this, in that it is freed from the constraints imposed by
drafting and printing the diagrams. Moreover, currently available personal computing
devices allow the equations to be programmed into a pocket calculator without difficulty.
With this in mind, Table A1.3 shows the regressions that enable determination of the base
prices of the modules considered here. Accompanying each one are the factors that are
required for obtaining the corresponding actual prices, depending on the kind of material uti-
lized, in particular, and in some cases factors for other parameters (pressure, type, etc.), and
also for “erection”. In addition, the recommended limits of variation are shown for all the key
variables involved. Thus, it is strongly recommended not to go above or below the maximum
and minimum limits specified, even when applying the equation gives a final total within the
range of variation of the base price parameter. A more realistic solution would then be to
adopt the minimum price as the base price in the low case, or in the high case, to consider
doubling the offending module.
 Table A l . 3 Functional modules method. Determination of base and real prices of modules

Module Range of Correction factors*

- Regression Key variables
No.
-
variation Materials Other I Erection
1 %till ation In (PB) = 4.3588 PB=base price (lo3 FF), early 1985 100.0-50000.0 Carbon 0.00
steel
M.5848 In (DEDIST) DEDIST = distillate output 5.0-5000.0 SA203 0.22
w o w
4 7 5 4 0 In (In (VORM)) VORM = average relative volatility 50.0- 1.1 304stainless 1.57
-0.3477 In (XLEC) XLEC = mole % light ends in the 5-100 316stainless 1.99
feed
4 . 1 138 In (PTE) PTE = column top pressure (bar) 1.O-100.0
-
2 ,ompression In (PB)= 3.6950 Pg=baseprice(IO3FF), early 1985 300.0-10000.0 Carbon 0.00 Pressure Centrif. 2.91
steel
+0.6246 In (DEB) DEB = flow ( k m o k ) 20.0-10000.0 304 stainless 0.28 <75 bar 0.00 Recip. 2.95
+0.6246 In (In (COMP)) COMP = compression factor 1.15-5.0 3 16 stainless 0.49 75-100 0.09 Other 2.71
100-150 0.13
150-200 0.21
200-300 0.33
3OMOO 0.50
10&500 0.60
5OMOO 0.80
-
3 baction In (PB) = 2.9386 PB= base price (lo3 FF), early 1985 20.0-15000.0 Carbon 0.00 External pumping
fixed bed homo- steel
5eneous reactors M.2908 In (DEB) DEB = flow (kmovhr) 1.&20 000.0 SA203 0.30 - with withou
Nith external +0.2426 In (TS) TS = residence time (s) 1.0-15 000.0 304stainless 1.80 shell 4.20 3.75
lumping) +0.2789 In (PFES) PRES =pressure (bar) 1 .0-250.0 316 stainless 2.30 quench 4.80 4.00
-
A. 1 h m s Same Same Same Same Det./Sep. 2.72
and - Inter. Storage 2.25
A.2
-
 Table Al.3 (cont.) Functional modules method. Determination of base and real prices of modules
~~~~

Module Range of Correction factors*

Regression Key variables
Kind variation Materials Other I Erection
Reaction (fluidi- In (PB) = 2.5776 PB = base price (lo’ FF), early 1985 25.0-20000.0 Carbon O.O(
zed bed reactors) steel
+0.2839 In (DEB) DEB = flow ( k m o l h ) 1.O-20000.0 SA203 0.3(
+0.2369 In (TS) TS = residence time (s) 10.0-50000.0 304 stainless 1.8(
+O. 1606 In (CV) CV = fluidizer power (kW) 1.0-100.0 316 stainless 2.3(
N . 1 4 4 1 In(PRES) PRES = pressure (bar) 1.0-250.0
Reaction In (PB) = -727.230 PB =base price (10’ FF), early 1985 50.0-4000.0 Carbon O.O(
(multitubular steel
reactors) +5.6956 (PRES) PRES = pressure (bar) 0.0-250.0 SA203 0.3(
+184.290 In (DEB) DEB = flow (kmovhr) 50.0-5 000.0 304stainless 1.8C
+0.23191 (TS) (11) TS = residence time (s) 100.0-7200.0 316 stainless 2.3C
+8.5134 (CHA) (12) CHA = heat exchanged 0.0- 100.0
(kcaVg mol)
I , and I, = bivalent integer variables 0 or 1
Reaction In (PB) = 2.990 PB = base price (10’ FF), early 1985 35.0-30000.0 Carbon -0.58
(Grignard type steel
reactors) +0.3684 In (DEB) DEB = flow ( k m o l h ) 1.O-5 000.0 Enameled - 0.2 1
steel
+O. 1442 In (TS) TS = residence time (s) 1.O-20000.0 304 stainless 0.00
+0.0178 In (PRES) PRES = pressure (bar) 0.0-150.0
~

Blenders Same Same Same Same - 2.02

Reaction In(PB) =5.1610 PB =base price (10’ FF), early 1985 i90.0-4500.0 Carbon 0.00 Pressure Pre-heating 2.1 1
(reaction furnaces steel <30bar 0.00
+OM12 In (Q) Q = heating power (Mkcalihr) 4.0-40.0 Cr/Mo 0.50 30-70 0.10 Pyrolysis 2.1;
steel
Stainless 0.75 70-100 0.15 Other 2.14
steel 100-150 0.25
150-200 0.45
200-250 0.60
Heating furnaces Same Same Same Same Same 1.31
 Table Al.3 (concl.) Functional modules method. Determination of base and real prices of modules
Module Range of Correction factors*
Regression Key variables
Kind variation Materials Other Erection
Zrystallization In (PB) = 7.0047 Pg=baseprice(103 FF), early 1985 250.0-10000.0 Carbon 0.00 2.12
steel
+0.5993 In (CAP) CAP = processing capacity (t/hr) 0.1-40.0 Rubber 0.30 -
coated
304 stainless 1.25
Evaporation In (PB) = 10.8470 P, =base price (lo3 FF), early 1985 100.0-50000.0 Carbon -0.40 2.25
steel
+0.5336 In (Q) Q = quantity of heat exchanged 0.1- 10.0 Cast iron - 0.25
(MkcaVhr) and Cu
-0.5336 In (T) T = maximum temperature of the 100.0-1.0 304 stainless 0.15
fluid ("C) -
-0.5336 In (DT) DT = temperature differential, if 100.0- 1.o 316 stainless 0.25
useful (evaporator to shell and
tubes)
Ni 1.15
Hastelloy 2.15
I

Jeat exchangers In (PB) = 10.3530 PB=baseprice (lo3FF), early 1985 10.0-2 000.0 Shell/ Pressure 2.84
tubes
+0.5880 In (Q) Q = quantity of heat exchanged 0.1-10.0 cs/cs 0.00 <lobar 0.00
(MkcaVhr)
-0.7375 In (U) U = overall transfer coefficient 1000.0-100.0 cs/cu 0.30 10-20 0.08
(kcalhm*."C)
-0.5365 In (TML) TML = log mean temperature ("C) 100.0-1.0 CS/304 1.60 20-30 0.20
304/304 2.40 3WO 0.35

CS/316 2.00 4065 0.75

CS/Monel 4.35 65-85 0.90
CSKi 9.40 85-130 1.45
130-180 2.15
* The correction factors are additive: only the total is used for multiplying the base price to obtain the real price of the module.
(Source: IFP)
 Appendix 1. Functional Modules Method (FMM) 317

Another general observation concerns the problem posed by a missing piece of informa-
tion. For example, in the case of distillation, while it is often easy to find the distillate output,
the concentration of the key light component in the feed, and the column’s top pressure, it is
not so simple to find the relative volatility of the principal constituents, which moreover is a
particularly sensitive variable. To deal with problems of this nature, a dual parameter
approach can be employed: the first stage consists of using the actual values if they are avail-
able; the second is aimed at defining a reference, in practice a qualitative scale which repre-
sents the degree of difficulty in performing the operation.
Thus, in the specific case of the distillation module, a statistical analysis has produced the
results presented in Fig. Al.2, which is expressed as an auxiliary calibration by means of five
pointers that identify the relative difficulty, measured qualitatively, of separating the feed’s
constituents. A1

Interval Frequency Percentage

5 10 15 20 25 30 35 40 Int. Cum. Int. Cum.
.- - .. - .... ..- .. - ..- . .- ... - ...- ..- .. .- ..
V. DIFFICULT AAAAA AA 7 7 9.3 9.3
DIFFICULT BBBBB B 6 13 8.0 17.3
AVERAGE CCCCC CCCCC CCCCC CCCCC CCCCC CCCCC CCCCC CCCC 39 52 52.0 69.3
EASY DDDDD DDDDD DDDDD D 16 68 21.3 90.7
V.EASY EEEEE EE 7 75 9.3 100.0
- ..- . .- ... ..- - . .- ..- ..- - . .- ... - ...- ..- ..
5 10 15 20 25 30 35 40

Notes:
a) Each symbol represents an observation.
b) Symbol Number Mean Std. Deviation Corresp. Value
V. DIFFICULT A 7 -1 367 0.178 1.167
DIFFICULT B 6 -1.235 0.209 1.338
AVERAGE C 39 4.335 0.163 2.046
EASY D 16 0.423 0.192 4.602
V.EASY E 7 0.921 0.100 12.327

Figure Al.2 Histogram for relative volatility. (Source: ZFP)

COMMENTS
a) Each symbol represents an observation

Std. Corresponding
Separation Symbol No. Mean
deviation volatility
-
Very difficult A 7 -1.867 0.178 1.167
Difficult B 6 - 1.235 0.209 1.338
Average C 3.9 -0.335 0.163 2.046
Easy D 16 0.423 0.192 4.602
Very easy E 7 0.92 1 0.100 12.327
(Source: IFP)
 318 Appendix 1. Functional Modules Method (FMM)

For modules where the calculation can be based either on the residence time or on the
quantities of heat to be exchanged (heat of reaction), it was decided to leave the users to select
the values most widely used in their fields of activity, either directly, or by using a scale sim-
ilar to the one employed to allow for relative volatilities in the case of distillation.
However, in the case of the reaction module (multitubular reactors), a dual parameter
method was suggested, using two exclusive variables I I and Z, that are bivalent integers (0 or
l), so that the calculation can be made using either the residence time or the quantity of heat
exchanged. In practice, when these two parameters are both available, the user should per-
form the calculation with each of them and select the result that gives the most unfavorable
result as regards base price. In physical terms, this approach amounts to favoring the variable
that happens to be decisive, depending on whether reaction kinetics or heat transfer is con-
A1 cerned.
In conclusion, two comments should be added. First of all, given that the throughputs are
always expressed in kilomoles in the regression, and that in addition the unit commonly avail-
able is the ton or the kilogram, it may be useful to program the pocket calculator with the
expression for relating them via molecular mass.
Throughput (kmolh) = throughput (kghr) / molar mass
(1.0to 10000) (100.0 to 100000) (300.0 to 1.0)
Finally, it would seem desirable to emphasize that, when the functional modules are being
defined by means of a plant flowsheet, their principal equipment consists of the items consid-
ered to be essential for the achievement of the intended operation. Exchangers, for example,
will only be considered to be auxiliary modules as long as they do not directly belong to the
whole module for distillation, reaction, or other unit operation. To obtain a better understand-
ing of this matter, Table A 1.1 should be consulted.
 APPENDIX 2
Pr6-Estime Method

As already mentioned in Chapter 3 (Section 3.3.3.1), this method, which was first described
in the 1975 edition of the Manuel d '&valuationdconomique desproc&dds,published by Tech-
nip, was based on the prior application of rapid sizing procedures for principal equipment,
using graphics grouped in various appendices for each of the major equipment categories. A2
Since then, progress in computing and the marketing of user-friendly software have greatly 1
reduced the attraction of this type of tool. As a result, with certain exceptions, the appendices for
the new presentation of the Prd-Estime method no longer include the rapid, pre-sizing approach
for equipment, especially as concerns the graphics: they assume that the principal dimensional
characteristics (diameter, height, area, power, etc.) are known, or determined elsewhere.

A2.1 EVALUATION OF PRESSURE VESSELS

By pressure vessel we mean any system including a shell with bottoms, enclosingcertain contents.
These include:
0 Various kinds of columns: with trays, with packing

0 Drums

0 Related reactors (tubular reactors or ones having a heating jacket and stirred beds are not

included in this equipment category).

The columns are designed to perform these operations:
0 Distillation

0 Absorption-stripping

0 Solvent extraction.

A2.1.1 THICKNESS CALCULATION

A2.1.1.1 Basic Formula

PR (A2.1.1)
eb =
at - 0.6P
 320 Appendix 2. Pr4-Estime Method

where
f?b is the base case thickness of the vessel (mm)
P is the relative operating pressure (bar)
R the radius of the vessel (mm)
t the maximum allowable stress (bar)
a the welding coefficient.

A2.1.1.2 Values of Constants

A. Welding Coefficient
a = 1.OO if there is a complete X-ray
a = 0.85 if there are spot X-rays
a = 0.70 if there are no X-rays.
A2 For a first approximation, take a = 1.OO.
1
B. Maximum Allowable Stress
Use the values in Tables A2.1.1a and A2.1.1b. They depend on:
0 The kind of material

0 The operating temperature.

Table A2.1 .la Value of the maximum allowable stress (bar) as a function
of utilization temperature, for commonly used steels

ASTM designation 399 427 454 482

----
Carbon steels
SA 201 B 910 759 608 457
SA212 B 1037 844 650 457
SA 285 A 721 633 545 457
SA 285 B 879 784 675 566 457
SA 285 C 967 93 1 847 717 587 457
Low alloy steels
SA203 A and B 974 801 629 457
SA 302 B 1046 1343 1181 93 1
SA 357 92 1 900 872 808
SA 387 C 1055 1055 1012 92 1
--
(Source: Perry b Chemical Engineers 'Handbook)
 Appendix 2. Pr6-Estime Method 32 1

Table A2.1.1 b (concl.) Value of the maximum allowable stress (bar) as a function
of utilization temperature, for commonly used steels

ASTM-AISI
designation
1- r r r 149 204 260 316 343 371 399 427 454 482 510

ij
--
Stainless
alloy steels
SA 240-304 1318 1195 1125 1086 1044 1040 1033 1023 1005 984 942
SA 304 L 1230 1195 1125 1055 879 844 808 773
SA 310 S 1318 1318 1301 1279 1188 1167 1142 1104 1047 970 879
SA 316 1318 1318 1258 1230 1199 1195 1188 1178 1160 1125 1062
SA 316 L 1230 1230 1111 1037 1047 946 93 1 914 893 861
SA 317 1318 1318 1258 1230 1209 1199 1195 1188 1178 1160 1125 1062 984
SA 321 1318 1318 1195 1111 1069 1044 1040 1033 1016 1005 991 974 949
SA 347 1318 1318 1195 1111 1069 1044 1040 1033 1016 1005 991 974 949
SA410 1142 1097 1062 1026 995 963 942 92 1 896 851 773 619 450
(Source: Perry k Chemical Engineers 'Handbook) A2
1
A2.1.1.3 Application
Determine the thickness of the vessel using Eq. (A2.1.1). Add an extra thickness se for cor-
rosion, usually 3 m.The total thickness then becomes:
e = eb + se
The value of e must not be less than:
0 6 mm, including corrosion addition, for drums

0 8 mm, including corrosion addition, for columns and related reactors.

A2.1.2 DETERMINATION OF THE PRICES OF COLUMNS, DRUMS,

AND RELATED REACTORS

A2.1.2.1 Principle
The method consists of determining the prices, as a function of the weight and the material
used, for:
0 Externals:

- Shell and bottoms, for columns, drums, and reactors

- Skirt, especially for columns (sometimes for reactors)
- Accessories, for columns, drums, and reactors; in this case the overall price is
obtained directly from the total weight of the other externals.
0 Intemals:
- trays
- packing
depending on the requirements of the columns and reactors.
 322 Appendix 2. Pr6-Estime Method

Depending on the thickness of the shell and bottoms, economics may dictate different
materials:
0 Homogeneous steel (carbon or stainless): thickness less than 8 mm

0 Carbon steel with cladding: thickness between 8 and 20 mm

0 Coatings or linings on carbon steel: thickness greater than 20 mm.

In all three cases, the calculation is based on carbon steel; if necessary, correction factors
are applied for stainless steels. The cladding or coating is treated as a supplement. Generally
speaking, the skirt is of carbon steel, the accessories and trays of homogeneous steel.

A2.1.2.2 Weight Determination

The required data are:
0 The diameter D' of the shell (m)

0 The length or height H of the shell (m)

0 The thickness e of the shell and bottoms (mm).

A2
1
A. Weight of the Shell
We use the following expression:
shell weight (kg) = 24.7 . D'. H . e (A2.1.2)

B. Weight of the Skirt

Skirts are required for distillation columns, and sometimes for reactors. Their characteristics
are as follows:
0 Diameter (m): take the diameter of the shell

Height (m): take:

- 3 m when there is no bottoms pump
- 5 m when there is one
Thickness (mm): this depends on the diameter and height of the shell.
For a first approximation, we may take:
0 8 mm for columns with 10 trays or equivalent heights

0 10 mm for columns with 20 trays or equivalent heights

0 12 mm for columns with 35 trays or more, or equivalent heights.

The expression (A2.1.2) then enables calculation of the weight of the skirt.

C. Weight of the Bottoms

There are two of these bottoms: they include the top and the bottom, or the two ends of the
pressure vessels.
Assuming that they are elliptical, with a diameter equal to that of the shell, their total
weight (i.e., for both of them), for a unit thickness (1 mm), is given by curve 2 in Fig. A2.1.1.
 Appendix 2. Pr6-Estime Method 323

100 1000

g
- 10 100
m
f
a,
m
5
rn 1

A2
1
Figure A2.1.1 Determination of surface area and weight for
two bottoms. (Source: IFP)

Accordingly, the total weight is as follows:

weight of bottom (kg) = unit weight (kg) . thickness (e) (mm)

A2.1.2.3 Determination of Prices of Externals

A. Price of Shell and Bottoms

a. Homogeneous Material

This applies when:

0 The vessel is entirely of carbon steel

Or when the thickness is less than 8 mm.

Determining the Base Price

This price, in €kg, is obtained from the curve in Fig. A2.1.2 as a function of the diameter
(mm) of the shell, for a thickness of 8 mm, and for carbon steel.

Correction factors
Effect of thickness
The correction factorf, is obtained from Fig. A2.1.3.
 324 Appendix 2. Pr6-Estime Method

5.5

0
8
cy
.-6 3.5
E

f 3.0
.-u
t
2.5
A2 0 1 000 2 000 3 000 4 000 5 000 6000 7 000
Diameter (mm)
1
Figure A2.1.2 Determination of base price for shell and bottoms.
(Source: IFP)

0.8 0.85 0.9 0.95 1 1.05 1.1 1.15 1.2

Correction factor (re)

Figure A2.1.3 Determination of the thickness correction factorf,.

(Source: IFP)
 Appendix 2. Prb-Estirne Method 325

Effect of material
Table A2.1.2 gives the correction factorf
, for the various materials.

Table A2.1.2 Correction factors for boilermaking materials

Code designation
fm
ASTM AISI Other
SA 285 C - - 1.o
SA 203 A and D - - 1.3
SA 357 - - 1.8
SA 240 304 - 2.3
SA 240 304 L - 2.3
SA 240 310 S - 3.8
SA 240 316 - 3.0
SA 240 316 L - 3.2 A2
SA 240 316 (Ti) - 3.1
SA 240 32 1 - 2.6 1
SA 240 347 - 2.9
SA 240 410 - 2.4
- - Uranus 50 3.6
- - Uranus B6 4.8
- - Monel400 6.5
- - Inconel 600 10.0
- - Inconel 625 16.0
- - Hastelloy G 12.5
(Source: IFP)

Totalprice
corrected price of shell and bottoms = (€/kg). (weight of shell + weight of bottoms. ff).f,.f,
where
ff = 2.8 for a shell diameter less than 1 m
2.5 for a shell diameter between 1 and 1.5 m
2.0 for a shell diameter between 1.5 and 2 m
1.5 for a shell diameter greater than 2 m.

b. Carbon Steel with Cladding

This applies when the thickness of the vessel is between 8 and 20 mm, i.e., when the exterior
is of carbon steel for a thickness compatible with the shell diameter and the relative pressure;
an interior stainless steel cladding provides protection against corrosion. The price then
includes:
0 The price of the carbon steel portion

The cost of the cladding.

 326 Appendix 2. PrSstime Method

Determining the base price of the cladding

This price is obtained from the following expression:
base price of cladding (mid-2000 €) = (3.14 . D' H + &,fioms)
* . 1 180
where
D' is the shell diameter (m)
H is the shell length (m)
Sbofioms is the surface area of the bottoms from Fig. A2.1.1, curve (1).

This expression is valid under the following conditions:

0 Thickness of the surface to be cladded: 8 mm

0 Percentage of cladding: 10%

0 Cladding material: 304 stainless.

Correction factors
A2 Effect of thickness
1 The corresponding correction factorfk is obtained from Fig. A2.1.4.

1.9

1.8

-g 1.7

8 1.6

d
-=
0
1.5

g 1.4
s
0 1.3

1.2

1.1

1
8 10 12 14 16 18 20 22
Thickness of surface for cladding (mm)

Figure A2.1.4 Thickness correction factor for claddingsfe.

(Source: IFP)

Effect of material utilized and percentage of thickness cladded

These correction factorsf & andfc are given in Table A2.1.3.
 Appendix 2. Pr6-Estime Method 321

Table A2.1.3 Cladding correction factors

'Qpe of stainless steel

10% 20%
304 1.oo 1.oo 1.15
304 L 1.05
310 S 1.25 1.00 1.25
316 1.20 1.oo 1.15
347 1.05 1.oo 1.15
410 0.95 1.oo 1.25
Monel 2.40 1.oo 1.25
Inconel 2.70 1.oo 1.25
Nickel 2.70 1.oo 1.10
Titanium 1.40 1.05
Cu-Ni alloy 2.70 1.oo 1.05
A2
(Source: IFP)
1

Price of cladding
.f En .fc
corrected price of cladding = base price *ye

c. Carbon Steel with lining

The calculation is exactly the same as for cladded steel. The coating is applied to carbon steel
with thicknesses greater than 20 mm. The price thus includes:
0 The price of the carbon steel portion

0 The cost of the coating.

Base price of the lining

base price (mid-2000 €) = (3.14 . D' H + Sbottom) . 300
This expression is valid under the following conditions:
0 Thickness of the lining: 3 mm

0 Lining material: 304 stainless.

Correction factors

Effect of lining thickness

The corresponding correction factory: is obtained from Fig. A2.1.5.

Effect of material utilized

Table A2.1.4 gives the correction factorf & for the various materials.
 328 Appendix 2. Pr6-Estirne Method

A2 1 2 3 4 5 6 7
Thickness of lining (mm)
1
Figure A2.1.5 Thickness correction factor for liningsf;.
(Source: IFP)

Table A2.1.4 Correction factors for lining material

w p e of stainless steel
304 1.oo
304 L 1.10
310 S 1.35
316 1.25
347 1.10
410 0.95
Monel 2.50
Inconel 2.80
Natural rubber 0.30
Synthetic rubber 0.40
(Source: IFP)

d. Special Cases (mid-2000 Um2)

Aluminizing 250
Bricks 750
Gunite (cement spraying) 630
 Appendix 2. Pr6-Estime Method 329

Price of lining
corrected lining price = base price -fl -yA

B. Price of Skirt
Generally speaking, the skirt is:
0 Of homogeneous steel

0 Of carbon steel.

Consequently:
corrected skirt price = skirt price . ( a g ) .f,
where
€/kg is the price given by the curve in Fig. A2.1.2 for a thickness of 8 mm
f, is the correction factor for thickness, given by the curves in Fig. A2.1.3 for thick-
nesses greater than 8 mm.
A2
C. Price of Accessories 1
Determination of Base Price
This price (in mid-2000 €) is obtained directly from the curves in Fig. A2.1.6, which are
based on carbon steel:
0 For columns and reactors, as a function of the total weight of externals: shells, bottoms,

and skirt
0 For tanks, as a function of the total weight of the externals: shell and bottoms.

Figure A2.1.6 Determination of prices of accessories for reactors,

columns, and drums. (Source: IFP)
 330 Appendix 2. Pr@-€sfirneMethod

Correction Factor
Correction factors are applied according to the machining qualities of the various steels
(Table A2.1.5).

Table A2.1.5 Material correction factors for machining

Code designation
ASTM AISI fam
designation designation
SA 285 C 1.o
SA 203 A et D 1.2
SA 357 1.6
SA 240 304 2.5
SA 240 310 S 3.5
A2 SA 240 316 3.0
1 SA 240 32 1 2.5
SA 240 347 3.0
Monel 7.0
Inconel 8.5

Price of accessories
corrected price = base price *fa

A2.1.2.4 Determination of the Price of the lnternals

A. Price of Trays
The main kinds of tray are:
Bubble cap trays
0 Valve trays, where the valve can be tared as required, depending on the steam output

0 Sieve trays, with an overflow, used to give low load losses, but unsuitable for corrosive

or dirty products; there are versions without an overflow, the corrugated iron type for
example
Jet trays.
Determining the base price
The price of trays, established for valve trays, is a function of:
Diameter (m)
Material
Tray thickness (mm)
Number of fluxes.
 Appendix 2. Pr4-Estime Method 33 1

It is obtained from Fig. A2.1.7, on the basis of the following:

0 Material and thickness:

- Carbon steel: thickness 3.5 mm

- 410 stainless: thickness 2 mm
- 304 stainless: thickness 2 mm
- Monel: thickness 2 mm
0 Number of cross flows: 1.

A2
1

0.1 1 10
Diameter (m)

Figure A2.1.7 Prices of valve trays. (Source: IFP)

Correction factors

Effect oftray type41 (Table A2.1.6a)

Table A2.1.6a Relative prices of trays, by type

Bubble cap
Valve
Sieve
(Source: IFP)
 332 Appendix 2. Pr6-Estirne Method

Effect of tray thicknessf, (Table A2.1.6b)

Table A2.1.6b Tray prices: thickness correction factor

Thickness (mm) f,
2.0 1.oo
3.5 1.25
6.0 1.60
12.0 2.50

Effect of number of crossflows fpa

To obtain the price of a double cross flow tray from the price of a simple cross flow one,
A2 multiply by 1.12.
1 To obtain the price of a quadruple cross flow tray from a simple cross flow one, multiply
by 1.35.

Effect of type of material fa,

For materials that differ from the base case, for a rough calculation take the coefficients
used for accessories, for equal thicknesses and the same number of cross flows.
Note, however, that the price of material proper and the price of labor work against each
other: the correction factor is weaker for the small and large diameters, and is stronger for
medium diameters.

Effect of number of identical traysfpn

Depending on the number of identical trays, reductions should be made as shown in
Table A2.1.6~.

Table A2.1.6~ Tray price variation, by number of trays

Number of trays fpn

0-15 1.oo
16-50 0.95
51-75 0.92
>76 0.90
(Source: IFP)

Price of trays
corrected price = base price .fpl .f,'fpa * f a m ' f p n
 Appendix 2. Pr&Estime Method 333

B. Price of Packings

Table A2.1.7 gives an approximate idea of the prices for a number of packings, assumed to
be sold in lots of 20 m3.
0 Reduce the amount by 10% for quantities greater than 50 m3.

0 Reduce the amount by 15% for quantities greater than 100 m3.

Table A2.1.7 Prices of packings (mid-2000 a m 3 )

Dimensions: (in] 1 1112 2

m
) -1 25 35 50
Raschig rings
ceramic 725 665 665
carbon steel 1795 1360 1105 A2
304 stainless 4 140 3 220 3 000 1
Pall rings
carbon steel 1735 1105 1010
304 stainless 3 255 2 545 2 095
3 16 stainless - - 2 380
monel - - 4090
polypropylene 740 545 440
Intalox saddles
ceramic 880 725 695
polypropylene 835 545 545
de Berl saddles
ceramic 850 760 695
Packing cups
carbon steel 870 675 610
304 stainless 3 620 2 825 2 590
(Source: IFP)

A2.1.2.5 Determination of Final Price

Add the prices:

Of the externals
And the internals.
Multiply the total by:
1.15 for columns and reactors
0 1.10 for drums,
 334 Appendix 2. Pr6Estirne Method

to take into account annealing, X-ray, packaging if required, possible extras, etc. Then:
total price (mid-2000 €) = (price of shell and bottoms + price of cladding or lining
+ price of skirt + price of accessories + price of trays or packings) . 1.10 to 1.15
COMMENT
When the diameter of the shell is large (greater than 5 to 6 m), assembly is done
onsite; the cost of erecting the shell itself must be added to the setting-up cost nor-
mally employed. As a rough guide, the price of the equipment alone should be dou-
bled.

A2
1
 Appendix 2. Pr6-Estime Method 335

A2.2 EVALUATION OF REACTORS

In most cases, reactors are comparable to columns or even to simple drums. The determina-
tion of the corresponding prices is similar to the calculation procedure used for pressure ves-
sels, i.e.:
Calculation of the thickness of the shell and the bottoms, as a function of diameter and
pressure,
0 Determination of the components of the total price:

- Price of the shell, the bottoms, and the skirt, if required, from the weight and kind of
material utilized,
- Price of accessories: for a rough estimate, use a comparable column,
- Price of internals: trays, coils, baffles, etc. (see columns, exchangers, etc.).

For agitated reactors with heating jackets, of Grignard type, and for simple mettler-mixers,
use the graph in Fig. A2.2.1, which gives the total cost, including stirrer and motor.

1 o0oooo I I I I I I I I I I I I I I I I I

I I I I I I I I I I I I I I I I I

A2
2

6
2 100000
a
W
0
0

-z
-E
C

."""" . 10
0.1 1
Capacity (m3)

Figure A2.2.1 Determination ofthe prices of agitated reactors with

heating jacket (mechanical seal, tamped, anti-explosion motor,
reducing gear, agitator, variable speed, etc. included). (Source: ZFP)

For the other types of agitated reactors, the price of the shell with bottoms, accessories, and
internals is calculated separately, as discussed above, and the cost of the agitator is given by
Fig. A2.2.2, for 304 stainless steel and an average viscosity on the order of 1000 CPO.
Depending on the material employed, use the correction factors from Table A2.2.1, and
the following expression: corrected price = base price .f,.
 336 Appendix 2. Pr6-Estime Method

I I I I I I I I I I I I I I l l l /
1 10 100
Power (HP)

Figure A2.2.2 Determination of agitator prices. (Source: ZFP)

A2
2 Table A2.2.1 Material correction factors for agitated reactors

Material
304 stainless
3 16 stainless
Rubber coated 1.25
Monel 1.30
(Source: IFP)

As a rough guide, the power ranges for the motors utilized with dispersers and reactors
with heating jackets are presented in Table A2.2.2.

Table A2.2.2 Motor powers for agitated reactors

Capacity (m3) Power (HP)

0.150 4-8
0.300 10-15
1.o 3- 10
4.0 5-15
6.0 7-20
10.0 10-30
 Appendix 2. Pr6-Estirne Method 337

A2.3 EVALUATION OF HEAT EXCHANGERS

There are three main categories of heat exchanger:

0 Those with tube bundles

0 Air-coolers

0 Plate exchangers.

A2.3.1 TUBE-TYPE HEAT EXCHANGERS

A2.3.1.1 Additional Data

The simulation software employed in chemical engineering generally provides the product
S . U, where S is the surface area in m2 and U the exchange coefficient ( k c a l h . m2 * "C).
Table A2.3.1 and Fig. A2.3.1 provide the data needed to calculate U.
The consumption of utilities is provided by the software; however, an appropriate pressure
level must be selected.
For this purpose, Table A2.3.2 shows the variation of the latent heat of vaporization of
water.

A2
Heavy hydrocarbon - residue 3
(without phase change)

Medium hydrocarbon - heavy distillate

(without phase change)

Light hydrocarbon - C3C4gasoline

(without phase change)

Light hydrocarbon
(with phase change)

Cooling water

Condensing or vaporizing steam

0 100 200 300 400 5M) 600 700 loo0

U (kcaVhr/m2/"C)

Figure A2.3.1 Determination of the overall gross heat transfer

coefficient. (Source: IFP)
 338 Appendix 2. Pr4-Estirne Method

Table A2.3.1 Overall gross transfer coefficient (kcal/h.m2."C)

Fluid
~~
U
'hbe side Shell side
1. Cooling exchanger
Butadiene Steam 60
Olefinic C4 Propylene (vaporization) 65-90
Ethylene vapor Condensate and steam 450-600
Ethylene vapor Cold water 250-400
Liquid ethylene Ethylene vapor 50-100
Propane vapor Liquid propane 30-75
Oleiinic light hydrocarbons, Steam 50- 100
CO, CO2. H2
Light and chlorinated Steam 60- 150
hydrocarbons
Ethanolamine Steam 75-125
Solvent Propylene (vaporization) 150-200
Solvent Solvent 170-200
Solvent Cold water 170-350
Oil Oil 300-400
Condensate Propylene 300-600
A2 Calcium chloride 25% Chlorinated C, 200-300
3 Steam Air (mixture) 50-100
Steam Styrene and tars 250-300
Cold water Freon 12 500-600
Cold water Recycle oil 200-350
Water Treated water (35-45°C) 500-600
Water Treated water (100-35OC) 800-1 100
Water Light chlorinated C2 hydrocarbons 30-50
Water Heavy chlorinated C2 220- 150
hydrocarbons
Water Perchloroethylene 270-170
Water Air-C12 40-90
Water HCl 35-75
Water Air and water vapor 100-170
Water Adsorption oil 400-560
 Appendix 2. Pr6-Estime Method 339

Table A2.3.1 Overall gross transfer coefficient (kcalhm*."C)

Fluid
U
'hbe side Shell side
2. Condenser Propylene (cooling) 330-390
Butadiene Propylene 290-330
Olefinic C4 Propylene 300-450
Ethylene Propylene 250-300
Olefinic light hydrocarbons Propylene 300-550
HCl Propylene 75-125
Light ends and chloroethanes Water 450-600
Olefinic chlorinated hydrocarbons Water 100-150
Chlorinated hydrocarbons Water 125-75
Solvent and non-condensables Propylene vapor 650-750
Water Propylene 300-500
Water Steam 600-1,100
Water Steam 100-150
Treated water Steam 350-550
Oil Air-Cl, 40-75
(partial condensation)
Cold water Propylene (cooling and
condensation)(')
A2
Water Light hydrocarbons 200-450 3
(cooling and condensation)(')
Water Ammonia 700-800
Air-water vapor Freon 50-250
3. Reboiler
Olefinic C4 Steam 450-550
Chlorinated hydrocarbons Steam 170-120
Olefinic chlorinated hydrocarb ns Steam 500-700
Dichloroethane Steam 350-450
Heavy solvent Steam 350-550
Mono- and diethanolamine Steam 750-1,000
Organic acids-water Steam 300-500
Amine-water Steam 600-700
Steam Naphtha 75-100
Propylene Ethane-ethylene 600-700
Propylene-butadiene Propylene-butadiene 75-90
(1) Depending on the temperature level: 550-750 (-10 to -5OC); 125-250 (0 to 7°C).
(Source: Perry k Chemical Engineers 'Handbook,IF8 etc.)
 340 Appendix 2. Pr.6-Estime Method

Table A2.3.2 Latent heat of vaporization of water (kcalkg)

Pressure Temper- Temper- Latent Pressure Temper- Latent

bar abs.) ture ("C: ItUre ("c heat bar abs.' iture ("C heat
0.1 45.58 574.75 4.4 146.27 503.48 8.7 172.95 484.15
0.2 59.76 564.84 4.5 147.09 502.89 8.8 173.43 483.80
0.3 68.74 558.53 4.6 147.90 502.31 8.9 173.91 483.45
0.4 75.47 553.81 4.7 148.69 50 1.73 9.0 174.38 483.1 1
0.5 80.90 549.99 4.8 149.47 501.17
0.6 85.48 546.75 4.9 150.24 500.62 9.1 174.85 482.77
0.7 89.47 543.94 5.0 150.99 500.07 9.2 175.3 1 482.43
0.8 93.00 541.44 9.3 175.77 482.09
0.9 96.19 539.20 5.1 151.73 499.54 9.4 176.23 48 1.76
1.o 99.09 537.15 5.2 152.47 499.01 9.5 176.68 48 1.43
5.3 153.19 498.49 9.6 177.13 481.10
1.1 101.76 535.26 5.4 153.90 497.98 9.7 177.57 480.78
1.2 104.24 533.50 5.5 154.59 497.47 9.8 178.01 480.45
1.3 106.55 53 1.86 5.6 155.28 497.07 9.9 178.45 480.14
1.4 108.72 530.33 5.7 155.96 496.48 10.0 178.89 479.82
1.5 110.76 528.47 5.8 156.63 496.00
1.6 112.70 527.49 5.9 157.29 495.52 10.25 179.96 479.03
1.7 114.54 526.18 6.0 157.94 495.05 10.50 181.01 478.27
1.8 116.29 524.84 10.75 182.04 477.51
1.9 117.97 523.74 6.1 158.59 494.58 11.oo 183.05 476.77
2.0 119.57 522.60 6.2 159.22 494.12
6.3 159.85 493.67 11.25 184.05 476.04
A2 2.1 121.11 521.50 6.4 160.47 493.22 11.50 185.03 475.32
3 2.2 122.59 520.46 6.5 161.08 492.78 11.75 185.99 474.62
2.3 124.02 5 19.43 6.6 161.68 492.34 12.00 186.99 473.92
2.4 125.40 5 18.44 6.7 162.28 491.91
2.5 126.73 517.49 6.8 162.87 491.48 12.25 187.87 473.24
2.6 128.02 516.57 6.9 163.45 49 1.06 12.50 188.78 472.57
2.7 129.26 515.68 7.0 164.03 490.64 12.75 189.69 471.90
2.8 130.48 514.81 13.00 190.57 47 1.25
2.9 131.65 513.97 7.1 164.60 490.22
3.0 132.80 513.15 7.2 165.16 489.82 13.25 191.45 470.61
7.3 165.72 489.41 13.50 192.3 1 469.97
3.1 133.91 512.35 7.4 166.27 489.01 13.75 193.16 469.34
3.2 135.00 511.57 7.5 166.82 488.62 14.00 194.00 468.73
3.3 136.06 510.81 7.6 167.36 488.22
3.4 137.09 5 10.07 7.7 167.89 487.83 14.25 194.83 468.12
3.5 138.10 509.35 7.8 168.42 487.45 14.50 195.64 467.52
3.6 139.09 508.67 7.9 168.94 487.07 14.75 196.45 466.92
3.7 140.05 507.95 8.0 169.46 486.69 15.00 197.24 466.34
3.8 141.00 507.27
3.9 141.92 506.61 8.1 169.97 486.32 16.00 200.32 464.07
4.0 142.82 505.96 8.2 170.48 485.95 17.00 203.20 461.83
8.3 170.98 485.58 18.00 206.07 459.81
4.1 143.71 505.32 8.4 171.48 485.22 19.00 208.75 457.82
4.2 144.58 504.70 8.5 171.98 484.86
4.3 145.43 504.08 8.6 172.47 484.50 20.00 211.34 455.89
(Source: Perry S Chemical Engineers 'Handbook)
 Appendix 2. Pr6-Estime Method 341

A2.3.1.2 Selection of Tube-Type Exchangers

A. Determination of Type
Table A2.3.3 provides a rough guide.

Table A2.3.3 Selection of tube-type exchangers

I Type I Characteristics Applications Limitations

Fixed tube sheet The two tube sheets are Condensers, coolers, Maximum temperature
part of the shell. reboilers. difference around
100°C due to
expansion.
~~~

Floating head The cap is removable Temperature difference Risk of leaking from
and reveals the ends of greater than 100OC. joints. Corrosion in the
the tubes: May be used with floating parts.
- the head and the tube dirty fluids.
pull through sheet are bolted
together;
- the tube sheet is
split ring floating; the tube
bundle may be
removed or not. A2
U-tubes One single tube sheet Big temperature hTk of failure at b e n d s
3
with U-tube bundle. difference. due to erosion at high
Removable bundle. Considerable speeds. Particle-free
expansion. Clean fluids.
fluids.
Kettle Kettle. Removable Reboiling or Bulk. High price.
bundle of U-tube or evaporation in the
floating head type. shell. High rate of
vaporization: 60% of
weight.
Thermosiphon Immersion reboiler with Reboiling in the tubes: Low vaporization rate:
natural circulation(') vertical. 35 to 40% by weight.
Reboiling in the shell:
horizontal.
Double pipe Concentric tubes Small exchanger Similar performance
surface. High provided by fin tubes.
pressure. Lower bulk and price.
I I
(1) Forced circulation reboiler: add the pump.
(Source: IFP)
 342 Appendix 2. Pr6-Estirne Method

The most common types are the following:

(1) Double pipe exchangers, which have a maximum exchanger surface of 10 m2.
(2) Exchangers with tube bundles, whose exchanger surface is at least 3 to 5 m2:
- Condensers: single pass, floating head, removable tube bundle,
- Xooling or cross-flow exchangers: single pass, U-tube exchangers; single pass,
removable bundle; single pass, fixed tube sheet,
- Peboilers: thermosiphon, immersion forced circulation, kettle.

B. TEMA Standard Designation

Use Fig. A2.3.2.

The designation employs three characteristic letters:
0 For type of stationary head (1 st column)

0 For type of shell (2nd column)

0 For the rear head and tube sheet (3rd column).

Examples:
0 Condensers: AES, BES, etc.

0 Cross-flow/cooling exchangers: BEU, BES, etc.

0 Reboilers: AKU, etc.

The thermosiphons are sometimes incorrectly designated by the acronym TS.

A2
3
A2.3.1.3 Prices of Tube-Type Exchangers

A. Exchangers with Tube Bundles

a. Principle

Basic expression:
corrected price = base price .fd .fi .Ap.fp - f ; .f
,
where
fd is a factor characteristic of the type of exchanger
fi is the correction factor for tube length
fnp is a factor for the number of tube passes
fp is the correction factor for the pressure in the shell and tubes
f; is the temperature correction factor
f, is a factor characteristic of the kinds of material used.

Base Price Conditions (Fig. A2.3.3)

Exchanger type: single pass, floating head, removable bundle.
 Appendix 2. Pr6-Estime Method 343

-
Front end stationary
head types
Shell types I Rear end head types

Fixed tube sheet

A (like type A)

One pass shell

Channel with
removable cover
Fixed tube sheet
(like type B)

Two pass shell

B with longitudinal baffle
Fixed tube sheet
(like type C)
Bonnet
(integral cover)

Split flow
Ip Outside packed
floating head

________
-ft-:-:
......... A2
u. ....,...
...,:~:,~.
.......
.
.I

3
Floating head
C with backing device

Double split flow

Pull through
floating head
Channel integral
with tube sheet
and removable cover

ID" Divided flow

U-tube bundle

D ::.
Ill-
Special high-pressure Externally sealed
- closure "Kettle type" reboiler I floating head tube sheet

Figure A2.3.2 Conventional exchanger nomenclature.

(Source: Perry k Chemical Engineers 'Handbook)
 344 Appendix 2. Pr6-Estime Method

Bundle characteristics:
- tube diameter: 3/4";BWG: 14
- square pitch
- length: 16 8,or = 4.9 m.

Selected pressure: 110 bar

Selected temperature: 135OOC
Material: carbon steel (CS) for the shell and tubes.
Erection of the exchanger itself is included in the price.

1 oooo00

6
H
c

100 000

loo00

A2
3 .-8
p'
1 000

Figure A2.3.3 Price of tube-type exchangers. (Source: IFP)

Data Required for Determining the Price of an Exchanger

To determine the base price, we need only know the exchanger surface.
To obtain the correction factors, we need to know, in order of importance:
Thetype
0 The kind of material

0 The maximum pressure

the other coefficients have a lesser effect; these are: the tube diameter and pitch, the length of
the bundle, the number of tube passes, and the utilization temperature.
 Appendix 2. Pr6-Estime Method 345

b. Values of the Correction Factors

These are presented in Tables A2.3.4a through A2.3.4g.
The coeficients may also be calculatedas a hnction ofeach constituentelement ofthe type.

Table A2.3.4 Base price correction factors for tube-type exchangers

5Pe fd
AES 1.oo
AEM 0.90
AEU 0.90
AKT 1.30
BES 0.95
BEM 0.86
BEU 0.86
BKT 1.25
Thermosiphon 1.20

b. Effect of type (detail values)(')

Distribution Shell Bundle A2
A 1.00 E 1.00 L 1.01 3
B 0.95 F 1.03 M 0.90
C 1.04 G 1.03 N 0.97
J 1.01 s 1.00
K 1.25 T 1.05
U 0.90

c. Effect of length of bundle

Length
A
(ft) (m)
8 2.4 1.35
12 3.7 1.13
16 4.9 1.oo
20 6.1 0.92
24 7.3 0.90
 346 Appendix 2. Pr6-Estirne Method

Number of passes fn,

2 1.00
4 1.02
6 1.04
8 1.06
12 1.08

+
e. Effect of maximum pressure
I I

Pressure (bar) Tubes and shells

<10 1.oo 1.oo
10 to 20 1.07 1.12
20 to 30 1.17 1.29
30 to 40 1.27 1.48
40 to 65 1.45 1.81
65 to 85 1.63 2.13
85 to 130 2.03 2.85
130 to 180 3.25 3.25
A2 (Source: IFP)
3

Temperature ("C) ft
t < 350 1.oo
350 < t < 550 1.08
(Source: IFP)

Material of shellkubes fm
CSICS 1.o
CSl304 L stainless 1.9
CSl3 16 stainless 2.2
CSICU 1.3
csmo 2.0
CSIMonel 4.0
CSlTi 10.0
 Appendix 2. Pr6-Estirne Method 347

COMMENT
When the parameters that define the tube bundle (diameter, pitch, length, number of
passes) are unknown, it is best to adopt standard dimensions, i.e., to set the corre-
sponding correction factors equal to 1.OO.

c. Price Determination

Use the graph in Fig. A2.3.3 to obtain the base price, and apply correction factors from
Tables A2.3.4a through A2.3.4g. Add 5% to cover possible extras.

B. Double Pipe Exchangers

a. Basic Expression

This takes the same form as that utilized for exchangers with tube bundles. There is only one
type of exchanger: double pipe, sofd = 1.OO.Similarly, numerous other correction factors can
be taken as 1.OO;only the effects of pressure and the kind of materials employed are consid-
ered.
The base price conditions are also identical: only the type of reference changes (double
pipe exchanger).

b. Values of the Correction Factors (Table A2.3.5).

A2
3
Table A2.3.5 Base price correction factors for double pipe exchangers

Pressure (bar) fP
4 1.oo
4 to 6 1.10
6 to 7 1.25
(Source: IFP)

b. Effect of kind of material

I Material of shelYtubes I fm I
CSICS 1.o
CS1304 L stainless 1.9
CS13 16 stainless 2.2
(Source: IFP)
 348 Appendix 2. PrcCEstirne Method

c. Price Determination

Use the graph in Fig. A2.3.4 to obtain the base price. Add 5% to cover extras.

10000

E
B
c

E
d
?!
*;

m
e
s
8
8
zE
.-
-8
t
1 000
0.1 1 10
Exchanger surface (m2)

A2 Figure A2.3.4 Price ofdouble pipe exchangers with fins.

(Source: IFP)
3

C. Special Case
It may be useful to know the price of exchanger tubes, when they are used separately in a shell
of particular size, or as a coil. Figure A2.3.5 gives the prices of these tubes for carbon steel.
As a first approximation, we may use the correction factors for double pipe exchangers,
according to the kind of material. The cost of welded and extruded tubes is strongly affected
by supply and demand. In 1974, it reached a particularly high level, as also shown by the
Nelson index for exchangers. Current prices have fallen back strongly: Fig. A2.3.5 therefore
presents average values. On this subject, the Nelson index for exchangers does not seem to
give as accurate a picture as other kinds of index, such as those from the LASL (Los Alamos
Scientific Laboratory).
Figure A2.3.5 was based on lots equivalent to a length of 20000 m, and carbon steel tubes.
Correction factors must be applied when conditions are different.
For unwelded tubes, we may use the following factors, depending on the kind of material
employed:
1.2 for low alloy steel (ASTM A 199)
2.5 for 304 stainless
4.0 for 3 16 stainless.
 Appendix 2. Pr6-Estime Method 349

10

0
P
c)

(D
c
K
0

i$
1
4.0

6
e l
W

QY
rr
E
.-E
.-0
h
0.1
0.4 1 2
Internal diameter (inches)

Figure A2.3.5 Prices of exchanger tubes. (Source: IFP)

A2.3.2 AIR-COOLERS A2
3
Some chemical engineering software does not give methods for sizing air-coolers: to provide
an answer in case one is needed, we present here a rapid procedure for obtaining the correct
sizing.
Air-coolers are used for the cooling and condensation of certain fluids, which can also be
done with tube-type heat exchange equipment. We therefore have a choice between these two
kinds of apparatus. Our preference for one or the other will primarily depend on the availabil-
ity of utilities, and secondly on economic considerations. In evaluation, where it is difficult
to optimize equipment or even individual units, only the first point is generally considered,
unless additional information is provided.
The usual kind of air-cooler has stacked rows of fin tubes. A fan provides forced air circu-
lation around the tubes. The sizing and costing of this equipment may be performed in greater
or lesser detail.

A2.3.2.1 Simplified Method: Principle

The method consists of determining the equipment’s characteristics, in particular the fan’s
power requirement and the cost, using coefficients (K and C ) that are a function of:
0 Q, the quantity of heat to eliminate

0 T I and T2, the inlet and outlet temperatures of the fluid to be cooled
 350 Appendix 2. Pre-Estime Method

ta, the inlet temperature of the air

0 r, the overall resistance due to the heat-transfer and to fouling.

A2.3.2.2 Preliminary Calculation

In this method, the sizing of an air-cooler and the determination of its cost assume that the
following parameters are known or can be calculated:
0 The thermal ratio:

0 The reduced duty:

an expression in which Q (kcalh) is a function of the hourly throughput of the fluid to be

cooled from T I to T2, and the change in enthalpy between Tl and T2.
0 The overall resistance: r = internal film coefficient rl + fouling coefficient rd. Values for

r, rl, and rd for various fluids are presented in Table A2.3.6.

A2 Table A2.3.6 Air-coolers: value of r (hr.m2."C/kcal) for various fluids

3
Film Fouling r
coefficient ri coefficient rd (hr-m2.0C/kcal)
Cooling:
water 0.000 16 0.0002 0.00036
aqueous solutions 0.0002 0.0004 0.0006
GPL 0.0004 0.0002 0.0006
light hydrocarbons 0.0005 0.0003 0.0008
medium hydrocarbons:
average viscosity 1 cPo 0.0010 0.0004 0.0014
average viscosity 5 cPo 0.0032 0.0006 0.0038
average viscosity 10 cPo 0.0050 0.0008 0.0058
Condensation:
steam 0.0002 0.0001 0.0003
ammonia 0.00028 0.0002 0.00048
GPL 0.0006 0.0002 0.0008
light hydrocarbons 0.0008 0.0003 0.001 1
light naphtha 0.0010 0.0004 0.0014
heavy naphtha 0.0014 0.0004 0.0018
gasoline 0.0008 0.0002 0.0010
diesel 0.0014 0.0004 0.00 18
(Source: IFP)
 Appendix 2. Pr6-Estime Method 35 1

0 The overall heat-transfer coefficient U ( k c a l h . m2 . "C) for the air-cooler, where:

1
-
U
= ri + rd + Y, + ra
an expression in which
r, is the transfer resistance of the metal, usually equal to 0.00015 (hr . m2 . OCkcal)
ra is the external film coefficient, i.e., the coefficient for air. Table A2.3.7 gives its val-
ues as a function of the number of rows of tubes, which itself depends on the ratio:

T1 -ta
U
A very quick calculation by approximation thus becomes necessary.
We utilize the corrected Logarithmic Mean Temperature Difference (LMTD). To obtain it
from Fig. A2.3.6 we must not only know TI, T2, and ta, but also th, the outlet temperature of
the hot air; th is a function of the power required by the fan and the face velocity of the air Vf
(given in Table A2.3.7); it can therefore be calculated only after having obtained a figure for
this power. The same applies to the correction factor, which, as in the case of a tube-type
exchanger, is determined by knowing:

A2
and: 3

Table A2.3.7 Air-coolers: external film coefficient and superficial air velocity

External film Superficial air

T1 - * a Number
velocity Vf
U of rows, N

(Ti-ta)lU<0.13 3 0.00096 3.20

0.13<(Tl-t&lU<0.17 4 0.00 102 3.02
0.17<(T1 -ta)/U<0.22 5 0.00107 2.87
0.22 < (Ti- t&/U < 0.28 6 0.001 12 2.74
0.28 < (Ti- ta)/U < 0.36 7 0.001 18 2.58
0.36 <(Ti - t&/U< 0.46 8 0.00121 2.48
0.46 < (Ti- t.J/U < 0.58 9 0.00125 2.38
0.58 < (Ti - ta)/U < 0.73 10 0.00 128 2.26
0.73 < (Ti - t&/U 11 0.00132 2.16
(Source: IFP)
 352 Appendix 2. Pr6-Estime Method

100
90
80
70
60
50

40

30

20

15

A2
3

At, or A t 2
At, or A t p

Figure A2.3.6 Determination of the LMTD. (Source: ZFP)

For this, we use the curves in Fig. A2.3.7,which are valid for one or more tube passes.
Since the flows have perpendicular instead of parallel directions, the correction factors for
air-coolers are different from those applicable to tube-type exchangers.
 Appendix 2. Pr&Estime Method 353

8 0.9
‘ii
e
-
C
4-

e
8
0.8

0.7
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R factor

A2
8
0.9 3
H
r

.-s
c)

e
b
0.8

0.7
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
R factor

Figure A2.3.7 LMTD correction factor for air-coolers.(Source: IFP)

 354 Appendix 2. Pr&Estime Method

A2.3.2.3 Sizing Air-Coolers

A. Calculating the Power Requirement

Knowing the values ofR and r, we first determine the coefficient K from Fig. A2.3.8.

A2
3

Figure A2.3.8 Air-coolers: determination of K as a function of r and R.

(Source: IFP)

The power requirement PCV is then calculated from the following expression:
Pcv = K . S

B. Determining the Width of the Air-Cooler

For the bundles commonly employed, whose length is 10 m, the width (1) is obtained approx-
imately as a function of Pcvby using the following equation:
I = 0.1 12 Pcv
where 1 is expressed in meters.

1. There are various methods for air-cooler calculations. Since the common sizing software does
not necessarily include one, it seems desirable to make an exception and suggest such a procedure here.
 Appendix 2. Pr&Estime Method 355

C. Determining N, the Number of Rows

N is obtained directly from Table A2.3.7, knowing (TI - ta)/U.

D. Calculating the Hot Air Outlet Temperature (th)

This temperature is obtained as a function of the power requirement and the face velocity of
the air, using the following equation:

t - Q
- 1061 Vf .Cv + ta
where
Q is in k c a l h
Vf is i n d s
P,, is in CV
th and ta are in "C.

E. Calculating the Exchange Area for Bare Tubes

This exchange area A is given by the following expression:

A= Q A2
U .corrected LMTD 3
where
A is in m2
Q is in k c a l h
U is in kcalhr. m2. "C
LMTD isin"C.

A2.3.2.4 Determining the Price of Air-Coolers

A. Setting the Base Price

The base price is determined for the following conditions:

0 Tubes: length 10 m, outside diameter 1 in (25.4 mm), carbon steel material, of 12 BWG

wall thickness, aluminum fins of type "G" (grooved tubes with embedded fins), six rows
Operating pressure: < 10 bar
Bolted return-type end cover, in carbon steel
Manually adjustable fan blade angle
Driver motor of type TEFC (Total Enclosed Fan Cooled) with V-belts.
We first determine the cost coefficient C by using Fig. A2.3.9, knowing R and r.
 356 Appendix 2. Pr6-Estime Method

The tax-free, ex factory base price in mid-2000 f is given by the following expression:
base price (mid-2000 f ) = 0.35 . C . S

100000

loo00

lo00
0.0001 0.001 0.01

A2 r factor

3
Figure A2.3.9 Air-coolers: determination of Cas a function of Y and R.
(Source: IFP)

B. Correction Factors
Depending on the technical characteristics of the air-cooler concerned, we multiply the base
price by one or more of the correction factors shown in Tables A2.3.8a through A2.3.8e. We
then have:
air cooler price = base price .S, .fp .A . f N . f ,
where
f , is the correction factor for tube thickness
fp is the pressure correction factor
fi is the correction factor for tube length
f N is the correction factor for number of rows
f , is the correction factor for the kind of material employed.
 Appendix 2. Pr6-fstime Method 357

Table A2.3.8 Price correction factors for air-coolers

a. Effect of tube thickness

Tube thickness (mm)

2.11
1.65 0.8
1.24 0.7
(Source: IFP)

b. Effect of pressure
Pressure (bar) fP
4 0 1 .oo
10 to 20 1.03
20 to 30 1.06
30 to 50 1.10
50 to 75 1.13
75 to 100 1.15
100 to 150 1.20

A2
3
Tube length (m) A
12 0.90
10 1.oo
8 1.05
6 1.12
5 1.15

Number of rows fN
3 1.25
4 1.15
5 1.05
6 1 .oo
8 0.90
10 0.85
 358

'hbe material
(aluminum fins) fm

Carbon steel 1.oo

3 S aluminum 1.20
Aluminum brass 1.15
304 stainless 1.60
321 stainless 1.65
3 16 stainless 1.75
Monel 2.00

A2.3.2.5 Validity of the Method

The method does not apply to vacuum exchangers. Its accuracy depends essentially upon the
accuracy of the determination of r (or U).

A2 A2.3.3 PLA TE EXCHANGERS

3
A2.3.3.1 General

Plate exchangers were designed for the purpose of improving heat management. Their overall
transfer coefficients are generally much higher than those obtained for tube-type exchangers.
They have several other advantages, notably:
0 Their mass and footprint are smaller than those for traditional exchangers: for this reason

they are very useful for offshore platforms, for example,

0 Fouling is one tenth of that for tube-type exchangers: the flow is highly turbulent, the

surfaces are better machined, and dead spots are rare,

0 Their economics are superior to those oftraditional systems; however, these are still irre-

placeable in some situations.

They can be divided into the following families:
0 Plate and joint exchangers (PJE), the commonest kind, whose principle consists of

a stack of plates in a frame, sealed by intervening joints that are secured by tie beams;
they offer great flexibility of operation, since the number of plates can easily be
changed,
0 Part-welded or chambered exchangers, a modification of the first category: every other

joint is replaced by a weld; this type of equipment is very useful when one of the two
fluids is likely to cause joint deterioration,
 Appendix 2. Prk-Estime Method 359

0 Welded exchangers, including:

- Fully welded plate exchangers, which are PJEs in which every joint is replaced by a
weld; they no longer offer any operational flexibility, but are indispensable for chem-
ically active fluids,
- Lamellar or flat plate exchangers, which are like traditional exchangers with flattened
tubes,
- Spiral exchangers, consisting of two strips of sheet metal wound together and held
“parallel”,
- The Packinox exchanger, whose principal innovation is the shaping of its plates by
explosion; the stack of plates, welded together, forms a bundle which is placed inside
a shell, which allows high pressures (up to 200 bar) to be applied; it should be noted
that this technology is suited only for plates of considerable size, with large exchange
surfaces (at least 500 m2),
0 Exchangers with crimped plates, consisting of flat metal sheets with fins placed between
them,
0 Exchangers with brazed plates, contiguous stacks of embossed (wavy) and flat sheet
metal.
The categories most often encountered in the petroleum industry are:
0 PJE type (in various refining and petrochemical units)

0 Part-welded exchangers (idem)

0 Fully-welded exchangers (idem)

0 Packinox exchangers (reforming, HDS, HDT).

This discussion will be confined to these four families.

A2
3
A2.3.3.2 Sizing Plate Exchangers

The defining size is the exchanger surface and it is calculated in a similar fashion to the size
of tube-type exchangers, i.e., by using the following formula:

where
Sexch is the exchanger surface (m2)
Q is the quantity of heat exchanged (kcal/hr)
U is the overall heat-transfer coefficient (kcal/hr. m2. “C)
LMTD is the Logarithmic Mean Temperature Difference (“C).
The difference as compared with traditional exchangers lies in the value of the coefficient
U.As a rough guide, Table A2.3.9 provides an outline of the ranges given by manufacturers
(Baniquand in particular).
 360 Appendix 2. Pr6-Estime Method

Fluids utilized U (kcaVhr.m2.0C)

Watedwater 4 000 to 6 000
Viscous aqueous solution (50 cPo)/water 1000 to 1500
Mineral oil/water 500 to 1000
Mineral oil/mineral oil 250 to 500
Organic products (0.1 kcal/hr.m2."C, (10 cPo)/water) 1 800 to 2 500
Organic products (10 cPo)/water 800 to 1200

A2.3.3.3 Determining the Price of Plate Exchangers

Basic expression
corrected price = base price .fp (A2.3.1)
where4 is the pressure correction factor.
The base price and the pressure correction factor depend on the material utilized (stainless
steel or titanium) and are obtained from Figs. A2.3.10 and A2.3.11 respectively.
For exchangers with big exchanger surfaces (Packinox), Figs. A2.3.12 and A2.3.13 apply.

A2
3 1 000 000

100 000

loo00

1 000
10 100 1wO
Surface area (m2)

Figure A2.3.10 Determination of plate exchanger prices.

(Source: IFP)
 Appendix 2. Pr&Estime Method 36 1

1.8

1.7

1.6
L

s
m
1.5
-
0
c.

1.4
8
E
3 1.3
a
g!
n
1.2

1.1

1
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Pressure (bar)

Figure A2.3.11 Pressure correction factors for plate exchangers.

(Source: IFP)

A2
1ooooo00 3

I Oooo00

loo 000
100 1000 10000
Surface area (m2)

Figure A2.3.12 Determination ofprices for large area (Packinox)

plate exchangers. (Source: Packinox)
 362 Appendix 2. fr6-Estime Method

1.3

1.25

b
‘i, 1.2
d
E

e 1.15
e
E
n
l.l

1.05

I
10 20 30 40 50 60 70 80 90 100
Pressure (bar)

Figure A2.3.13 Pressure correction factors for Packinox exchangers.

(Source: Packinox)

A2
3
Base price conditions
This is a tax-free, ex factory price, expressed in mid-2000 €, for a pressure of 10 bar and
nitrile rubber joints.
There are now hardly any exchangers with joints in a material other than nitrile rubber or
EPDM (the use of this latter material has little effect on the price: 5 to 10%).
Although a fully welded exchanger is generally more attractive from a technical view-
point, we sometimes find P E s with Viton joints, which can withstand temperatures up to
170°C. In this case, a good approximation of the price may be obtained by adding together
the corrected price and the value of the base price in stainless steel for the calculated
exchanger surface, i.e., by multiplying the base price by 1 + f p instead offp.

A2.3.3.4 Comment

All these graphs have been compiled by using the MAP-H modeling software, without which
the approach would not have been possible. In fact, only a dozen points were available and
the use of linear regressions would not have provided answers.
In terms of the figures, thanks to the “universal” equations and the functional descriptions
that it employs, the model was able first of all to determine the form of the various curves,
and then to position them appropriately by using the available points.
 Appendix 2. Pr6-Estime Method 363

A2.4 EVALUATION OF PUMPS A N D COMPRESSORS

A2.4.1 PUMP PRICES

The price includes the electric motor driver (or turbine driver, if the appropriate correction
factor is applied).

A2.4.1.1 Centrifugal Pumps

Basic Equation
This takes the form:
corrected price = base price -fd.f,.f, .ft .fp
where
fd is a factor characteristic of the type of pump
f, is a factor characteristic of the type of driver
f, is a factor characteristic of the kind of material employed
ft is the temperature correction factor
fp is the correction factor for suction pressure.

Base Price Conditions

Type horizontal “process” pump with vertical seal
Material cast steel casing, internals cast iron
Operating temperature 0-150°C
Maximum outflow 650 m3/hr at 2900 rpm
A2
350 m 3 / b at 1450 rpm 4
Maximum head 230 m at 2900 rpm
Suction pressure <40 bar.

Price Determination
Use the graphs in Figs. A2.4.la and A2.4.lb for process pumps, supplied with driver. If
necessary, apply the corrections from Tables A2.4.la through A2.4.le. For vertical centrifu-
gal pumps with high outflows (axial flow and mixed types), use Fig. A2.4.2, which gives their
prices, complete with driver.
 3 64 Appendix 2. Pr6-Estime Method

Table A2.4.1 Base price correction factors for centrifugal pumps

a. Effect of type of pump

Type fa
“Process”: horizontal with vertical seal 1.0
“Process”: with double lining 1.5
Vertical “in line”: (bar) &<20 1.5
2O<AP<50 0.9
0.7
Chemical 0.5 to 0.7
(Source: IFP)

b. Effect of type of driver

Electric motor
Turbine: power rating between 8 and 10 kW
Turbine: power rating between 11 and 30 kW
Turbine: power rating between 3 1 and 100 kW 1.50
Turbine: power rating greater than 100 kW 1.20

m e fm
Cast steel 1.oo
Bronze 1.15
Cast steel. 3 16 stainless 1.10
A2 3 16 stainless 1.20
4 Monel 1.45
3 16 stainless 1.20
Monel 1.45

Temperature (“C) ft
450 1.oo
150-250 1.15
>250 1.30

Suction pressure (bar) fP

<20 1.oo
20 to 40 1.05
 Appendix 2. Pr6-Estirne Method 365

100000

f
zm

lo00 1

1 10 100 loo0
Outflow (m3hr)

Figure A2.4.1 a Price of centrifugal pumps (AP=0 to 25 bar)

(electric motor driver included). (Source: ZFP)

1 oo0000

A2
4

1OOo00
100
Outflow ( m 3 h r )

Figure A2.4.1 b Price of centrifugal pumps (AP= 50 to 150 bar)

(electric motor driver included). (Source: IFP)
 366 Appendix 2. Pr6-Estime Method

1 o00000

6
ti
3
X

5f 100000

6
e
w

1000
100 1 000 loo00
Outflow (m3hr)

Figure A2.4.2 Price of high-outflow vertical centrifugal pumps

(driver included). (Source: IFP)

A2.4.1.2 ReciprocatingPumps
A2 Use the graph in Fig. A2.4.3, which gives the cost of reciprocating pumps for two pressure
4 domains, as a function of the hydraulic power (the product of outflow and manometric
height). It applies to duplex (2-cylinder) cast iron pumps, with manual regulation of outflow
for suction pressures below 20 bar. Various corrections must be applied, depending on the
following characteristics:
0 Number of cylinders utilized,

0 How the piston is controlled,

0 The lubrication system,

The material employed,

0 The discharge pressure.

Only the correction factors for the type of material utilized and the effect of discharge pres-
sure (Tables A2.4.2a and A2.4.2b) can easily be defined.
Figure A2.4.4 gives the electrical power requirements of centrifugal and reciprocating
pumps, based on the required hydraulic power.
 Appendix 2. Pr6-Estime Method 367

1o
oom

1oOooO

1om

1 10 100 1MX) 1om

Outflow' A f (m3/hr*bar)

Figure A2.4.3 Price of reciprocating pumps (driver included).

(Source: IFP)

Table A2.4.2 Base price correction factors for reciprocating pumps

a. Effect of kind of material utilized A2

4
MateriaIs I f*
Cast iron 1.oo
Vulcanite covered cast iron 1.15
Cast steel 1.10
304 stainless 1.30
3 16 stainless 1.50
(Source: IFP)

b. Effect of suction pressure

Suction pressure (bar) fP
<20 1.o
20 to 40 1.3
40 to 70 1.5
(Source: IFP)
 368 Appendix 2. Pr6-Estime Method

loo00

I I 1 1 1 1 1 1 I I 1 l l l l I I I I 1 1 1 1 1 I 1 I I 1 1 1 1
1
I I I

lo00
/I

t
I I I I I I I I I I I I 1 1 1 1 1 1 I I I 1 T l I I I I I l l l l

I I I I I IIII I I 1 1 1 1 1 1 1 I 1 I111111/ I I I I1111

100
F
p' 10

0.1
10 100 lo00 loo00
Outflow (m3hr) * AP(bar)

Figure A2.4.4 Power of centrifugal and reciprocating pumps.

(Source: IFP)

A2.4.1.3 Rotary and Metering Pumps

A2 Figure A2.4.5provides, as a function of output:

The price of rotary lobe pumps, entirely in 3 16 stainless steel, with driver,
4
The price of screw-type rotary pumps, in cast iron with stainless steel rotor, with driver,
The price of metering pumps with plunger piston and diaphragm, whose essential parts
(body, piston, valve housing and valves) are in 3 16 stainless steel, with crank-case and
base in aluminum alloy. The explosion-proof electric motor driver is included in the
price given by Fig. A2.4.5,along with the price of the head, which can be adapted to the
required output.

A2.4.2 COMPRESSOR PRICES

A2.4.2.1 Types of Compressors

The main types of compressors are as follows:
Centrifugal compressors:
0 With horizontal seal

With vertical seal

 Appendix 2. Pr6-Mime Method 369

100OOO

P
B
e

-8
t
lo00
1 10 100 1OOO loo00
Outflow (m3hr)

Figure A2.4.5 Prices of reciprocating metering pumps of diaphragm

and piston type, and rotary lobe- and screw-type pumps (driver included).
(Source: IFP)

Axial compressors: used with gas turbines for high outflows:

Potary compressors
0 Screw type: helico-compressor A2
Vanetype, 4
Reciprocating piston compressors: single acting or double acting, lubricated or not, hori-
zontal or vertical,
Diaphragm compressors: used for moving small amounts of gas (10 to 50 m3/hr) at very
high pressures (100 to 1000 bar),
0 Superchargersor blowers:
Rotary compressors:
- Lobe type: AP,,: 0.8 bar, outflow,,: 25 000 m 3 h
- Sliding vane type: discharge pressure <8 to 9 bar, outflow,,: 5 000 m3/hr
- Liquid piston type: AP,,: 5 bar at low outflow, outflow,,: 10000 m3/hr for a AP
of 0.5 to 1 bar
Centrifugal compressors: turbo-blowers,
Fans: used for very low discharge pressures, on the order of 30 g/cm2 actual:
Centrifugal, with curved or straight blades
Axial flow, disk or propeller type
 370 Appendix 2. Pr@-EsfimeMethod

A2.4.2.2 Determining Compressor Prices

Figure A2.4.6 gives the base prices of reciprocating and centrifugal compressors for power
ratings from 100 to 6000 kW. Beyond this figure, it is preferable to double up the equipment,
each item being sized to provide at least 60% of the capacity theoretically required. Some
high powered compressors, like those used for steam cracking units, come in only one size
and are made to order: their real cost is considerably less than the figure given by exponen-
tially extrapolating the data in Fig. A2.4.6.

A2
4
Figure A2.4.6 Compressorprices (electric motor driver included).
(Source: IFP)

Figure A2.4.7 gives the cost of lower powered compressors (particularly reciprocating
compressors for air and hydrogen).
The cost of turbo-blowers, superchargers, and fans may be evaluated using the data in
Fig. 142.4.8,for the range of conditions defined above.
The data in these figures cover the cost of equipment that is unassembled, but that includes
reducing gear, base, ancillary equipment, and electric motor driver. For turbine driven compres-
sors we apply the correction coefficienth, defined in Table A2.4.3a; the effects of maximum
pressure and of materials are corrected by the coefficients& max and f, from Tables A2.4.3b
and A2.4.3~.
(Note that the manufacturers may include the costs of erection of some of the equipment
in the ex factory cost.)
 Appendix 2. PrkEstime Method 371

1 o00000

100ooO

loo00
10 100 lo00
Power (kw)

Figure A2.4.7 Prices for air and hydrogen compressors.

(Source: IFP)

1mm -
A2
4
1OOo00

loo00

1OOO
lo00 loo00 100OOO
Intake volume (m3/hr)

Figure A2.4.8 Prices of fans and superchargers(driver included).

(Source: IFP)
 3 72 Appendix 2. Pr6-Estime Method

Table A2.4.3 Base price correction factors for compressors

Power (kw) Correction coefficientft

<I00 1s o
100 to 1 000 1.15
>1 000 1.07
(Source: IFP)

b. Effect of maximum discharge pressure

fp max
Maximum pressure (bar)
Centrifugal compressor Reciprocating compressor
<75 1.oo 1.00
75- 100 1.08 1.05
100-150 1.16 1.10
150-200 1.25 1.18
200-300 1.38 1.30
300-400 1.52 1.43
400- 500 1.68 1.57
500-600 1.86 1.72
(Source: IFP)

c. Effect of kind of material employed

A2
4 Material
fm
Compressors Blowers
Cast steel 1.o 1.o
304 stainless (moving parts) 1.4 2.0
3 16 stainless (moving parts) 1.7 2.5
(Source: IFP)
 Appendix 2. Pr&Estime Method 373

A2.5 EVALUATION OF FURNACES

A2.5.1 GENERAL OBSERVATIONS ON FURNACES

We may distinguish:
Heating furnaces
0 Reaction furnaces (see Reactors, Appendix A2.2).

The first kind are utilized:

0 For heating a fluid without phase change. This may be the provision of heat to a fluid

agent that then transmits its heat to a product that cannot be directly heated, the heating
of feeds in certain plants, or the intermediate heating of certain outflows.
For heating a fluid with partial vaporization. These are, for example, distillation fur-
naces.
In the second type, not only are the feeds brought to the desired temperature, they are also
maintained for a certain time at that level so that the desired reaction may take place (furnaces
for steam cracking, thermal reforming, etc.).
Sizing a furnace is a very complicated matter, best left to specialists. However, the heating
power to be provided is a characteristic dimension that can easily be calculated, and by itself
is sufficient to determine the price.
We note however that a distinction is often made between furnaces that are delivered pre-
assembled, generally for plants of smaller capacities, and ones that have to assembled in
place, and which are of larger scale.

A2.5.2 FURNACE PRICES

A2.5.2.1 Determining the Base Price A2

5
The curves in Fig. A2.5.1 provide an approximate value for the prices of un-assembled heat-
ing and reaction furnaces, as a function of the power required.
They were determined on the following basis:
0 Material employed for the tubes: carbon steel

Operating pressure: < 30 bar

The most common uses (distillation and pre-heating of feeds).
 Appendix 2. Pr6-Estirne Method

0.1 1 10 100
Heating power (106 kcallhr)

Figure A2.5.1 Furnace prices. (Source: IFP)

A2.5.2.2 Correction Factors (Table A2.5.1)

Correction factors must be applied for the other cases, according to the following equation:
corrected price = base price . (1 +fd +fm +fp)
where
A2 fd is the correction factor for type
fm is the correction factor for material
5 f p is the pressure correction factor.

Table A2.5.1 Correction factors for base prices of furnaces

Type fd

Heating furnaces Cylindrical furnaces 0.00

Furnaces for heating the heat transfer fluid 0.15

Pre-heating furnaces 0.0

Reaction furnaces pyrolysis furnaces 0.1
Other types of furnaces >0.3
 Next Page

Appendix 2. Pr6-Estime Method 375

Material fm
Carbon steel 0.00
Chrome-molybdenumsteel 0.50
18-8 stainless steel 0.75
(Source: IFP)

c. Effect of pressure
Pressure (bar) fP
<30 0.00
30-70 0.10
70- 100 0.15
100-150 0.25
150-200 0.45
200-250 0.60

A2.5.2.3 Determining the Energy Cost

The quantity of fuel required is calculated by adopting a caloric yield of 80%. The heating
value of the fuel utilized then enables us to find the quantity to be employed. For example:
0 For heating fuel oil burned in a refinery, the net heating value is roughly 10000 kcal/kg,

0 For gaseous fuels, we have the following values (Table A2.5.2):

Table A2.5.2 Heating values of various fuels

A2
Fuel

Synthesis gas (coal or gasoline)

I Low heating value
(th/m3)
4.5
5

Butane 30.0
Propane 23.0
Natural gas:
Lacq 9.6
Hassi R’Mel 10.6
Groningue 8.4
Previous Page

Appendix 2. Pr6-Estime Method 3 77

A2.6 EVALUATION OF STEAM EJECTORS

Steam ejectors, used for vacuum operations, do not in themselves require a very accurate
evaluation, since their price is usually low, even negligible, in comparison with the other
components of a unit’s principal equipment. But it is desirable to determine the consumptions
of utilities, particularly steam, that their use entails. The performance of ejectors, i.e., the vac-
uum that they produce, depends on the number of stages (Table A2.6.1).

Table A2.6.1 Ejector performance

Number of stages 1 2 3 4 5
Absolute suction pressure (mmHg) 276 10-100 1-25 0.15-4 0.05-0.20

In industry, one or two stages are generally used.

When the mixture to be extracted contains condensable components, we utilize:
0 A pre-condenser To reduce the load on the ejector and the consumption of

0
Intermediate condensers
} steam, and if necessary to recover these components.
A final condenser before venting to the atmosphere, if necessary.
The cooling fluid is cold water or brine.

A2.6.1 EJECT0R CA1CULATIONS

To the extent that the most commonly used software programs do not spell out the procedures
for obtaining the characteristic dimensions of ejectors, it seems desirable to provide some
guidance on methods for supplying possible deficiencies or answering specific needs.
A2
A2.6.1.1 Determining the Maximum Suction Flow 6
The calculation has the following steps:
0 Selection of the absolute suction pressure in millimeters of mercury

0 Determination of the total volume of equipment to be exhausted

0 Determination of the maximum suction flow, using Fig. A2.6.1.

The values, provided by the Heat Exchange Institute, are for perfectly sealed commercial
systems. In practice, the value obtained from Fig. A2.6.1 is doubled.
 378 Appendix 2. Pr6-EstirneMethod

Figure A2.6.1 Ejectors: determination of suction flow.

(Source: IFP)

A2.6.1.2 Determining the 2OoC Equivalent Dry Air Flow

A. Calculating the Content of Condensable Components (if necessary)

Determine:
The partial pressure of condensable components at the operating temperature:

Pv (for a mixture: Pv= C Pvi)

A2
6 The partial pressure of noncondensable components (air):

P,, =n - P, where n = selected vacuum pressure

The number of kmol of noncondensables:

flow of feed from Fig. A2.6.1, doubled

Nnc =
molecular weight

For air, the molecular weight is 29. For a mixture, calculate the average molecular weight,
or consider each constituent separately and add them up.
 Appendix 2. Pr4-Estirne Method 379

0 The number of kmol of condensable components:

0 The flow of condensable components:

N, * molecular weight = D, (kg/hr)

For a mixture, take the average molecular weight or consider each constituent separately

and add them up: N P

V . and DVi= Nvi. molecular weight of component i; conse-
.I
v i = N nc
Pl,
.
quently: 0, = Dvi

B. Determining the 20°C Equivalent Dry Air Flow

Use Fig. A2.6.2, which enables the correction coefficients to be determined according to:
0 The temperature at which the operation takes place:

The molecular weight of the components extracted:fM.

Temperature(“C)
200 300 400
1.8

1.6

1.4
c
$ 1.2
E
8 1
c

0.8
IE
5
W
0.6

0.4

0.2

0
0 20 40 60 80 100 120 140 180 180
Molecular weight

Figure A2.6.2 Ejectors: determination of dry air flow at 20°C.

(Source: IFP)
 380 Appendix 2. Pr6Estime Method

In an actual calculation we consider five cases:

(1) Air only: for temperatures above 20°C divide the amount of air to be exhausted by the
temperature coefficient for the operating temperature.
(2) Water vapor only: divide the amount of vapor to be exhausted by the temperature coef-
ficient for the water vapor and by 0.80, which is molecular weight coefficient for water
(A4= 18).
(3) Air-water vapor mixtures: divide the amount of air to be exhausted by the temperature
coefficient of the air. Divide the amount of steam in the load by the temperature coef-
ficient of the water vapor and by 0.80. Add the results together.
(4) Mixtures of gases without water vapor: determine the average molecular weight of the
gases. Divide the amount of gas to be exhausted by the temperature coefficient of the
air (or the gases) and by the molecular weight coefficient.
(5) Mixtures of gases with water vapor: calculate the amount of air equivalent to the gases
alone (case 4). Calculate the amount of air equivalent to the water vapor (case 2). Add
the results together.

C. Determining the Consumption of Utilities

The utilities consumed are:

0 Without intermediate condensers: water vapor only

0 With intermediate condensers: water vapor and cooling water (or brine).

COMMENT
Only the intermediate condensers that are integral parts of the ejector system are con-
sidered. The pre-condensers and the final condenser are calculated separately.

a. Without Intermediate Condensers

A2 Use the water vapor consumptions given in Table A2.6.2 as a function of the suction pressure.
Interpolate if necessary.
6
When the water vapor is used at a different pressure from that selected for the reference
case (7 bar), apply the correction factor given by Fig. A2.6.3.

COMMENT
In Table A2.6.2 we consider only two stages. In manufacturing, this number is rarely
exceeded.
 Appendix 2. Pr6-Estime Method 381

Table A2.6.2 Steam requirement (kgkg of air) for ejectors without intermediate condenser

Suction pressure kg of steam at 7 bar per kg

Number of stages
(mmW of equivalent air at 2OoC
1 100 6.90
125 4.90
150 3.70
175 3.00
200 2.60
225 2.30
250 2.05
2 12.5 34.50
25.0 16.50
50.0 9.30
75.0 6.75
100.0 5.30
125.0 4.60

1.3

1.2

1.1

0.9

0.8 A:
6
0.7

0.6
4 6 8 10 12 14 16 18 20
Steam pressure (bar)

Figure A2.6.3 Ejectors: correction factor for pressure of steam utilized.

(Source: IFP)
 382 Appendix 2. Pr6-Estime Method

b. With IntermediateCondensers
Determine the percentage of noncondensables in the load (wt. %).
Use the utility consumptionsgiven in Table A2.6.3 (water vapor at 7 bar and cooling water
at 20°C) for the various suction pressures and equally various noncondensable contents.
If necessary, interpolate for the various suction pressures or noncondensable contents that
are different from those shown in Table A2.6.3.

Table A2.6.3 Requirements of steam (kg/kg of air) and cooling water (m3ikg of air)
for ejectors with intermediate condensers

Suction kg of 7 bar steam m3 of cooling water

Noncondensables
pressure per kg of equivalent at 2OoCper kg
(wt.%)
(mmHg) dry air at 2OoC of equivalent dry air at 2OoC
12.5 10.00 0.200
25.0 6.30 0.125
100 50.0 4.45 0.075
75.0 3.50 0.057
100.0 3.00 0.047
12.5 8.70 0.172
25.0 5.45 0.122
75 50.0 3.80 0.080
75.0 3.00 0.065
100.0 2.55 0.057
12.5 7.05 0.155
25.0 4.45 0.112
50 50.0 3.05 0.085
75.0 2.40 0.070
100.0 2.05 0.062
12.5 5.20 0.165
25.0 3.20 0.110
25 50.0 2.10 0.077
A2 75.0 1.65 0.067
6 100.0 1.45 0.062
(Source: IFP)

If necessary, correct the steam consumption by the pressure factor from Fig. A2.6.3.

COMMENT
There are two kinds of intermediate condenser, mixture or surface type. The first are
less and less frequently used; the second are traditional tube-type exchangers.
 Appendix 2. Pr6-Estime Method 383

A2.6.2 EJECTOR PRICES

Use Figs. A2.6.4a through A2.6.4c, according to whether the ejector has:
0 One stage only

0 Two stages without intermediate condenser

0 Two stages with intermediate mixture condenser.

To determine the price, the following data are required:

0 Suction pressure (mmHg)

0 Total feed to be exhausted ( k g h ) or the doubled maximum flow calculated previously.

The base conditions for which the price curves were compiled are as follows:
0 Exhaust pipe in stainless steel, ejector body and diffuser in cast iron
0 Reference year: mid-2000.

COMMENT
To determine the price of a two-stage ejector with intermediate surface condenser,
simply take the cost of an ejector without intermediatecondenser and add the amount
for the required tube-type exchanger.

5M)

450

400

350

: 300
5 250
u 200
150

100
A2
6
50

I I I I I I
0.5 1 1.5 2 2.5 3 3.5 4 4.5
Mid-2000 price in 103 Euro. tax-free. ex factory

Figure A2.6.4a Price of one-stage ejectors. (Source: IFP)

 3 84 Appendix 2. Pr6-Estirne Method

3oc

250

- 200

55 150

P 100

50

0
1 2 3 4 5 6 7 0 9 10
Mld-2000 price in 103 Euro, tax-free, ex factory

Figure A2.6.4b Price of two-stageejectors without

intermediate condenser. (Source: IFP)

350

300

250

A2
6 100

50

0
5 10 15 20 25
Mid-2000 price in I@ Euro, tax-free, ex factory

Figure A2.6.4~ Price of two-stage ejectors with intermediate

mixture condenser. (Source: IFP)
 Appendix 2. Pr6-Estirne Method 385

A2.7 EVALUATION OF SPECIAL EQUIPMENT

Generally speaking, it is very difficult to present simple, practical rules for sizing special
equipment, even for quick sizing, in view of the range of apparatus on the market, and the
multiplicity of uses for which they are designed. For this reason, the current software does
not include any general options that are altogether satisfactory.
This is also why the generally available price curves refer to a parameter that is character-
istic of the type of equipment employed, but difficult to determine by means of simplified cal-
culations.
In estimation, consulting a manufacturer will easily solve the problem; in evaluation, we
must proceed in an empirical fashion, by similarity, given that even the basic information that
the vendors of special equipment require for defining their sizing and their prices is usually
unknown.
Accordingly, this appendix presents various pieces of information obtained from a variety
of sources, as well as dimensional items that may help in gaining access to the most recently
available correlation parameters.
The following special equipment is discussed:
A2.7.1 Dryers A2.7.6 Crushers and grinders
A2.7.2 Crystallizers A2.7.7 Cyclones
A2.7.3 Evaporators A2.7.8 Conveyors
A2.7.4 Filters A2.7.9 Vibrating screens
A2.7.5 Centrifuges A2.7.10 Instrumentation

A2.7.1 EVALUATING DRYERS

A2.7.1.1 Dryer Selection

The principal methods of drying and their applications are summarized in Table A2.7.1. The
selection, sizing, and price of dryers depend essentially on the characteristics of the product
to be treated. Only a specialist can provide satisfactory answers; for a proper study the man-
ufacturers should therefore be consulted.

A2.7.1.2 Dryer Sizing

A2
Nevertheless, for a rough guide, it is sometimes possible, if need be, to make some quick cal- 7
culations by referring to other appendices (absorption, exchangers, refrigeration, etc.). Here
we will limit the discussion to the kinds of drying shown in Table A2.7.1.
 Table A2.7.1 Dryer applications

Kinds of
drying
Applications 1 Liquid
gases I Suspensions
Continuous
6lms
Adsorption, absorption
Cooling, compression, coalescence
Flash drying X X x x x x x x
spray drying x x x x x x X -

Convection or Fluid bed dryer X X X x x x x

direct heat TunneVconveyor dryer - - x x x x x x
Tray dryer - X X X X
Continuous sheeting dryer - - - - -

Cylinder dryer - - - - -

Drum dryer X X X - -
Steam tube rotary dryer - X X x x x x
Conduction or
Screw conveyor dryer X X x x x x x x
indirect heat - -
Vibrating tray dryer - x x x x
Vacuum rotary dryer - X X x x x x
Cabinet (vacuum) (batch) - X x x x x x x
X X: commonlyused
X : limiteduse
- : rarelyused
(Source: IFP)
 Appendix 2. Pr6-Estime Method 387

A. Dryers Using Alumina or Molecular Sieves (Adsorption)

The sizing of the vessels is done in the same way as for reactors. The quantity of the adsorbent
determines the volume of the shell and hence its dimensions.
Since this kind of drying proceeds in a cyclical fashion, we may need to provide two, if not
three identical columns to allow for the regeneration phase as well as the essential ancillary
equipment (furnace, exchanger, pumps, etc.).
The initial load of sieves or alumina depends on the adsorbent power and the length of the
cycle. In fact, the duration of the cycle is determined by an economic optimization, involving
the utilization time of the auxiliary equipment (particularly the furnace).
As for the adsorbent power, it is much higher during the first few cycles than during the
final ones. The overall lifetime, measured by the number of cycles, depends on the water con-
tent of the product being dried.
Roughly speaking, we may take the following figures:
0 Length of cycle: 24 to 48 hours

0 Lifetime:

- 150 cycles for alumina

- 400 cycles for molecular sieves
0 Average adsorbent power over the lifetime:

- 10 kg of water for 100 kg of alumina

- 20 kg of water for 100 kg of sieve.

B. Surface Exchange Drying (Convection, Conduction)

Considering how many ways of drying there are, it is impossible not only to give a detailed
discussion of how the sizing is to be done in each case, but even to define the information that
the manufacturers need for such calculations.
Obviously, the physico-chemical characteristics of the product to be dried are important
parameters; but since dryers have standardized dimensions, we have to consider the volume
of the feed to be treated, and the duration of the operation. Roughly speaking, this depends
on the exchanger surface available (direct or indirect heating), the average transfer coeffi-
cient, the temperature difference between the product to be dried and the heat transfer fluid
circulating in the vessel (gas, steam, fumes, etc.), and the quantity of water to be removed
(evaporation capacity).
Accordingly, we return to the general expression for heat transfers:
Q = U . S . At (A2.7.1) A2
where
7
Q is in kcal/hr
U is in kcal/hrm2."C
s is in m2
At is in OC.
For a solution of salts, U lies between 100 and 350 kcal/hr.m2."C (generally 200). For
sludges, powders, and granules, U is on the order of 25 to 200 kcal/hr.m2."C.
 388 Appendix 2. Pr6-Estime Method

Since it would be difficult to assemble the characteristics and general performances for
every type of dryer, the information given in Tables A2.7.2a through A2.7.2f concerns only
the equipment most commonly used in the chemistry of the major organic intermediates, cat-
alysts, and sometimes inorganic products.

a. Spray Dryers

The heat requirement of spray dryers is proportional to their maximum thermal efficiency,
i.e., to the ratio:

where
T, is the inlet temperature of the drying gas
Ts is the outlet temperature of the drying gas.
The consumption of electricity is a function of the flow of drymg gas; it also depends on
the temperature.
By way of example, Table A2.7.2a shows the energy requirements for a sludge containing
40% water, a gas inlet temperature of 4OO0C, and an outlet temperature of 120OC.

Table A2.7.2a Characteristics of spray dryers under particular conditions

1 Evaporating capacity (kg/hr) I 750 I 1000

Heat duty ( 1O6 kcal/hr)
Electricity requirement (kW)
(Source: IFP)

The inlet gas temperature generally lies between 150 and 350"C, the outlet temperature
between 80 and 90°C.

A2 b. Flash Dryers
7 For a sludge containing 40% water, a gas inlet temperature of 400°C and an outlet tempera-
ture of 1OO"C, Table A2.7.2b shows the energy requirement for flash dryers. These figures
are only indicative, and depend on the characteristics of the product being dried.
 Appendix 2. Pr4-Estime Method 389

Table A2.7.2b Characteristics of flash dryers

Evaporating capacity (kg/hr) 750 1000 1500 2500 3000

Heat duty (1O6 kcal/hr) 0.80 1.10 1.70 2.80 3.30
Electricity requirement (kw) 45 60 80 130 140

c. Vacuum Rotary Dryers

Table A2.7.2~gives the main characteristicsof vacuum rotary dryers, which operate in batch
mode. We distinguish:
0 Traditional dryers, which include a horizontal shell with a central agitator shaft equipped

with paddles, and a jacket for circulation of the heat transfer fluid (hot water, steam,
etc.).
Bicone dryers whose double shell pivots around a horizontal axis (perpendicularto the
cones’ axis of symmetry).

Table A2.7.2~ Characteristics of vacuum rotary dryers

Diameter (m) 0.5 0.7 1 1.3 1.5 1.5

Length (m) 1.2 2 3 6.5 8.5 10
Usable capacity (m3) 0.1 0.3 1.5 3.5 6.5 8.0
Heating surface (m2) 2 4 16.5 30 50 60
Motor power (CV) 1 1.5 5 10 20 25

I Bicone rotary dryers I

Diameter (m) 1 1.3 1.5 1.8 2.1 2.4
Usable capacity (m3)(’) 0.25 0.60 0.95 1.50 2.50 3.80
Heating surface (m2) 2.8 4.8 6.7 9.5 13.1 17.9 A2
Motor power (CV) 2.2 3 4 5.5 11 15
7
(1) About 60% of total capacity.
(Source: IFP)

Apply Eq. A2.7.1 to obtain the heat transfer fluid requirement (low pressure steam: 6 bar
in the jacket) from the quantity of heat required (see Appendix A2.3).
 390 Appendix 2. Pr6-Estirne Method

d. Rotary Drum Dryers

These are actually rotating furnaces in which hot gas circulates: the gas may be air when the
required temperatures are not too high, or it may come from the combustion of fuel oil in a
burner when more extreme operating conditions are desired. In this case, the inner walls of
the dryer are lined with a refractory material.
The principal characteristics of these two kinds of rotary drum dryer are presented in
Table A2.7.2d.

Table A2.7.2d Characteristics of rotary drum dryers

I Hot air heating

Diameter (m) 1 1.2 1 -7.4 1.5
~~~
1.8
Length (m) 7.5 7.5 9 10
Lateral surface area (m2) 28 33 42 57
Driver motor power (CV) 3 5 7.5 10
Air fan power (CV) 5 7.5 7.5 10 30
(Source: IFP)

Heating by fuel oil combustion in a burner

Diameter (m) 2.45 I 2.75 I 3.05
Length (m)
Lateral surface area
(m2)
Motor power (CV) 20
Diameter (m)
Length (m)
Lateral surface area
50
530
1 I
75
790
3.35
1101 1251 75
1160 1310 870
I
1 3.70
1001 125 140
1160 1450 1630
(m2)
Motor power (CV) 60 100 125 125 125 150 200 200
7 (Source: IFP)

To calculate the heat requirement, we use Eq. A2.7.1.

 Appendix 2. Pr6-Estime Method 391

e. Steam-Tube Rotary Dryers

This type of dryer consists of a horizontal shell rotating about its central axis, with internal
walls covered with concentric rows of tubes for circulating steam or hot water.
The main characteristics are presented in Table A2.7.2e.

Table A2.7.2e Characteristics of steam tube rotary dryers

Diameter (rn) 1 1.4

-
Length (m) 9
Exchange surface area 95.5 207 260 313 424 530 635

Free surface area (m2) 0.6 1.3 2.4 2.4 2.4 4.0 4.0 4.0
Motor power (CV) 5 7.5 15 20 20 20
-
(Source: IFP)

To calculate the heat requirement we use Eq. A2.7.1; the overall transfer coefficient in
steam tubes ranges from 25 to 75 kcal/hr.m2."C (m2 of tube).

f. Multipurpose Dryers (Cabinet)

These are standard drying cabinets with electrical pulsed air heating and a maximum temper-
ature of 450°C: their principal characteristics are shown in Table A2.7.2f.

Table A2.7.2f Characteristics of multipurpose dryers

Capacity (m3) 1 1.5 4 7

Tray surface area (m2) 1 1.5 2.7 4
Maximum energy requirement (kW)(l) 18 18 48 60
(1) Heating and agitation.
(Source: IFP)

A2
7
A2.7.1.3 Dryer Prices

A. Base Prices
Figures A2.7. la and A2.7.1 b give an order of magnitude for the base prices of various types
of convection and conduction dryers, whose characteristicswere presented in Tables A2.7.2a
through A2.7.2f.
 392 Appendix 2. Pr6-Estirne Method

1 10 100
Surface area (m*)

Figure A2.7.1 a Dryer prices. (Source: IFP)

ISteam or water tube I

A2
7
-8
h

10 100 loo0 loo00

Surface area (mZ)

Figure A2.7.1 b Dryer prices. (Source: IFP)

 Appendix 2. Pi&-Estime Method 393

Corrections must be made according to the kind of material employed or the operating
conditions.

B. Price Corrections

a. Spray and Flash Dryers

The base prices are for 304 stainless steel. To obtain the prices for other materials we may,
as a rough guide, take the figures in Table A2.1.5.

b. Vacuum Rotary Dryers

The base prices for vacuum rotary dryers are for 304 stainless steel. To obtain the cost of this
equipment in carbon steel or enameled steel, apply the coefficients in Table A2.7.3a.

Table A2.7.3a Effect of kind of material on the cost of rotary vacuum dryers

Material 'Qpe of dryer Usable capacity (m3) fm

Carbon steel Standard All 0.65

Bicone All 0.70
0.25 1.25
0.60 1.13
Enameled steel Bicone 0.95 1.05
1.50 1.oo
2.50 0.95
3.80 0.90
(Source: IFP)

c. Rotary Drum Dryers (Table A2.7.3b)

The following equation may be used:
corrected price = base price . (1 +fg +f,) A2
where 7
fg is a factor reflecting the fluid utilized
f, is a correction factor for the material.

d. Steam-Tube Rotary Dryers

The price curve is based on carbon steel. To obtain the cost of this equipment in 304 stainless
steel, use a correction coeficient,f,, of 1.75.
 394 Appendix 2. Pr6-Estime Method

Table A2.7.3b Price correction factors Cf, andf,) for rotary drum dryers

I Heat transfer fluid I fg I Material I fm

Hot air 0.00 Carbon steel 0.00
Flue gas (direct contact) 0.12 304 stainless plating - 20% 0.25
Flue gas (indirect contact) 0.35 3 16 stainless plating - 20% 0.50

e. Multipurpose Dryers (Table A2.7.3~)

The necessary corrections are found in the following equation:

corrected price = base price -4.f,
where
4
f,
is the pressure correction factor
is a correction factor for the material.

Table A2.7.3~ Correction factors Cf, andf,) for multipurpose dryers

Pressure fP Material fm

Atmospheric pressure 1 .o Carbon steel 1.o

Vacuum 2.0 304 stainless steel 1.4
(Source: IFP)

A2.7.2 EVALUATING CRYSTALLIZERS

A2.7.2.1 Crystallization Efficiency

A2 The theoretical crystallization efficiency is given by the equation:
7
loowo-S(Ho-E)
C=R
100- S ( R - 1)

where
C is the weight of crystallized product obtained (kg of hydrated or anhydrous crystals)
R is ratio of the molecular weight of the hydrated product to the molecular weight of the
anhydrous product
 Appendix 2. PrcSEstirne Method 395

S is the weight solubility (parts of anhydrous product per 100 parts of solvent) at the
final drying temperature
W, is the weight of anhydrous product in the initial solution (kg)
Ho is the weight of solvent in the initial solution (kg)
E is the quantity of solvent evaporated during the drying (kg).
When E is unknown it can be calculated from the expression:

(W, + H o ) . C . At .[lo0 - S . ( R - l)] + qc *R.(lOOWo- S . Ho)

E=
L, . [lo0 - S (R - l)] + qc .R . S

where
C is the specific heat of the initial solution (kcal/kg*"C)
At is the temperature differential between entry and exit ("C)
qc is the crystals' latent heat of crystallization per unit of weight (kcalkg)
L, is the solution's latent heat of evaporation (kcaVkg).
When the required production is known, these equations enable calculation of the feed to
the crystallizer.

A2.7.2.2 Crystallizer Selection

Crystallization may be obtained by:
0 Cooling without evaporation (or limited evaporation):

- By ambient air: simple vessel, with a hot feed

- By an intermediate fluid: batch agitated with coils, jacketed vessels in continuous
operation with scrapers if required, etc.
0 Evaporation without cooling: evaporators

0 Evaporation and cooling: crystallizers.

The main types of industrial crystallizers are the following:

0 Crystallizers with external forced circulation:
- Operating either by evaporation or by cooling
- With or without baMes
- With cooling agents applied in a tube-type exchanger or by direct contact
0 Soakers (Oslo type):

- Operating by evaporation or cooling

- With cooling agents applied in a tube-type exchanger or by direct contact A2
0 Surface exchangers with scraper
7
0 Vessels with or without cooling, agitated or not, for batch operations.
 396 Appendix 2. Pr4-Estime Method

A2.7.2.3 Crystallizer Prices

The choice, sizing, and price of a crystallizer essentially depend on the characteristics of the
solution to be treated and the product sought, particularly the size of the solid particles. Con-
sequently, one must consult a manufacturer to obtain satisfactory answers. However, various
publications provide price curves, which can be used as rough guides.
Thus, Fig. A2.7.2 gives the cost of crystallizers-xternal and internal forced circulation,
seeded, and batch vacuum types-as a function of the processing capacity in m3 or tkr.These
base prices include the cost of accessories and the equipment required to produce the vacuum;
they are based on the use of carbon steel. The curve for seeded crystallizers corresponds to
90% recovery of product larger than 16 mesh; the internal forced circulation crystallizer
curve corresponds to 90%recovery of product larger than 20 mesh.

E
s
0
e

Figure A2.7.2 Crystallizerprices. (Source: IFP)

A2
7
Corrections have to be made according to the kind of material employed or the perfor-
mance required (Table A2.7.4a and A2.7.4b):
corrected price = base price .film
.fm
 Appendix 2. Pr6-Estime Method 397

Dimensions fiim
90% + 16 mesh 1.oo
90% + 20 mesh 0.90
90% + 30 mesh 0.75
(Source: IFP)

b. Effect of kind of material on crystallizer price

I m e I Material I fm

Forced circulation crystallizers

Carbon steel
Stainless steel (304) 1 1.o
2.5
Carbon steel 1.o
Batch vacuum crystallizers Rubber coated 1.3
Stainless steel (304)
(Source: IFP)

A2.7.3 EVA1UATING EVAPORATORS

A2.7.3.1 Evaporator Selection

The principal kinds of evaporators and their applications are as follows:

Climbing-film evaporators: concentration of clear solutions, with few solids, well below
saturation level,
0 Falling-film evaporators, same applications as above, also concentration of viscous solu-

tions that are sensitive to heat (food processing industry),

0 Shell-and-tube evaporators. This type of equipment includes the following variants:

- Standard vertical and horizontal, short-tube evaporators A2

- Vertical long-tube evaporators
- Forced-circulationevaporators. 7
Traditional shell-and-tube evaporators allow concentration of solutions that are nearly sat-
urated, with high contents of solids, but not viscous or toxic. Some of them are equipped with
agitators to improve fluid circulation.
Forced circulation, by means of a large-flow pump, makes it possible to evaporate solu-
tions that produce scaling, or that are viscous, foamy, contain solids, or that may precipitate
them.
 398 Appendix 2. Pr6-Estime Method

0 Jacketed evaporators: treatment of high viscosity or heat-sensitive materials. They consist

of a jacketed cylindrical horizontal or vertical shell with a rotating shaft equipped with mov-
able scrapers.
0 Thin-film evaporators: these are characterized by extremely short treatment time and very

small production. They consist of a vertical tube whose lower portion is fitted with a jacket,
with a high-speed rotor inside.

A2.7.3.2 Sizing Evaporators

A. Principle
As a rough guide, the exchanger surface can be taken as the principal parameter for determin-
ing a price. In fact, numerous other factors are involved in defining the essential characteris-
tics of an evaporator (dimensions, agitator power, etc.).
The basic expression that is employed has already been utilized for heat exchanger calcu-
lations (Appendix A2.3), namely:

where
S is the surface area (m2)
Q is the total quantity of heat transferred (kcal/hr)
U is the heat transfer coefficient (kcal/hr*m2."C)
At is the average temperature difference ("C).

B, Determining the Overall Heat Transfer Coefficient

Here too, it is hard to obtain sufficiently accurate figures for U, the heat transfer coefficient
of the various kinds of evaporators encountered. Consequently, it is necessary to consult the
manufacturers to get a satisfactory answer on this topic. However, various authors have sug-
gested values that are functions of the temperature and viscosity, or the flow rate, of the prod-
uct being processed. Thus, between 30 and 120°C, the following equations may be used
0 For agitated, shell-type evaporators:
A2 u = 2 1 t-355
7 0 For forced-circulationevaporators:
U = 16 t + 515 for a flow rate of 1.5 m/s
U=21 t+865 foraflowrateof 2 m/s
U = 2 1 t + 1030 foraflowrateof 3 m/s
0 For falling-film evaporators:
U=33t+215
 Appendix 2. Pr6-Estime Method 399

where
t is the temperature of the fluid ("C)
U is in kcal/hrm2."C.
In addition, Table A2.7.5 gives the heat transfer coefficients for thin-film evaporators,
according to their intended application.

Table A2.7.5 Heat transfer coefficients (U) for thin-film evaporators

U
Applications
(kcaYhr.m2-oC)
Steam heating:
Concentration of aqueous solutions 1 100-1 600
Dehydration of organic solutions 700-950
Distillation of organic liquids 700- 1 000
Stripping of organic liquids 400-600
Operation as a reboiler 850- 1 000
Recovery of solvents 750-800
Heating by heat transfer fluid:
Distillation of liquids with high boiling points 350-500
Stripping of organic liquids 200-350
Operation as a reboiler 500-550
(Source: IFP)

C. Technical Characteristics of Various Types of Evaporator

a. Jacketed Evaporators with Movable Scrapers (Table A2.7.6a)

Table A2.7.6a Technical characteristics of jacketed evaporators with scrapers

Horizontal evaporators Vertical evaporators

I Heating surface area

(m2)
0.2
Agitator power
(CV)
2.5 to 4
Heating surface area
(m2)
Agitator power
(CV) A2
2 7.5 to 15 7
0.5 3.5 to 7 3 10 to20
1.o 5 to 10 4 10 to20
2.0 7.5 to 15 6 15 to25
3.0 10 to 20 8 15 to30
4.0 10 to 20 12 15 to 30
5.0 15 to 25 18 25 to40
(Source: IFP)
 400 Appendix 2. Pr4-Estime Method

b. Forced-Circulation Evaporators (Table A2.7.6b)

Table A2.7.6b Technical characteristics of forced circulation evaporators

Speed of circulation ( d s ) 1.5 2 2.5 3

Transfer coefficient (kcal/hr.m2."C) 2 800 3 400 3 900 4 200
Heating surface area (m2) 170 155 145 140
Pump power (HP)(I) 26 32 40 50

A2.7.3.3 Evaporator Prices

A. Base Prices

Figure A2.7.3 gives the base prices for various types of evaporators as a hnction of the
exchanger surface, for equipment in the following materials:
Jacketed evaporators: 3 16 L stainless
Thin-film evaporators: 3 16 L stainless
Shell-and-tube evaporators: cast iron shell and copper tubes

10 000 000

1 000 000

100 000
A2
7

loo00
0.1 1 10 100
Exchanger surface (m*)

Figure A2.7.3 Evaporator prices. (Source: IFP)

 Appendix 2. Pr&Estirne Method 401

0 Forced-circulation evaporators: cast iron shell and copper tubes

0 Falling-film evaporators: internal parts in 3 16 L stainless and external parts in carbon
steel.
Corrections must be applied according to the kind of material employed:
corrected price = base price .f,

B. Price Corrections

a. Jacketed Evaporators (Table A2.7.7a)

Table A2.7.7a Price correction factor urn)

for jacketed evaporators

Material
Carbon steel
304 L stainless
3 16 L stainless
Hastelloy G 2.50
(Source: IFP)

b. long-Tube Shell Evaporators (Table A2.7.7b)

Table A2.7.7b Price correction factor Vm)for long shell-and-tube evaporators

Material: shelytubes
Cast iron-copper
Carbon steel
Carbon steel-aluminum 0.7
Nickel-nickel
(Source: IFP)

c. Forced-Circulation Evaporators (Table A2.7.7~)

A2
Table A2.7.7~ Price correction factor urn)for forced circulation evaporators 7

Material: shelytubes fm
Cast iron-copper 1.oo
Monel-cupronickel 1.35
Cast nickel iron-nickel 1.80
 402 Appendix 2. Pr&€stirne Method

A2.7.4 EVA1UATING FILTERS

A2.7.4.1 Filter Selection

Types of filter:
0 Batch filters:
0 Cartridge filters

0 Basket strainers

0 Sand filters (traditional and counter current)

0 Bag filters

0 Sheet filters

Candle or stacked disk filters

0 Vertical leaf filters (self-cleaning or manual cleaning)

0 Level leaf filters (manual cleaning)

Horizontal leaf filters with centrifugal cleaning

0 Frame filters, automatic or not

0 Continuous operation filters:

0 Rotary pressure filters

0 Rotary vacuum filters: string, roll, blade, or belt discharge, with or without pre-coating

0 Belt filters

Horizontal disk vacuum filters.

The choice of a filter essentially depends on the concentration of solids in the feed to be
processed, and the fineness of filtration desired. It also depends on the nature of the operation:
continuous or batch, automatic or not, whether or not the cake is to be recovered, washed, and
dried.

A2.7.4.2 Sizing Filters

Sizing a filter, i.e., determining the surface area required for treating a given feed, depends
on the characteristics of the product to be separated and the properties of the filter. Test mea-
surements are used as references in the consideration of each case.
As a result, it is practically impossible to find even an approximate rule that would be gen-
erally applicable for calculating the surface area.
A2 Nevertheless, Table A2.7.8 presents performance figures for the filters whose prices are
given in this appendix.
7
The power of the motors required to drive rotary or belt filters averages around 4 to 10 HP.
 Appendix 2. Pre-Estirne Method 403

Table A2.7.8 Characteristics of various filters

Concentration Fineness Cake treatment I

Type of filter of solids of filtration
(CL) Recovery 1 1
Washing Drying 1
I

Vertical frame 1-5 to 25-50 X

Horizontal frame <5-10 1-5 to 25-50 x x
Automatic frame filter 1 to 50 1-5 t025-50 x x
Rotary vacuum 5 to 25 5-50 to150-250 X
Belt 5-10 to 50 15-50 to150-250 X
Horizontal disk 5-10 to 50 15-50 to150-250 X

x x : standard operation
x : possible operation
(Source: IFP)

A2.7.4.3 Filter Prices

The base prices of a certain number of filters are given in Fig. A2.7.4, according to the surface
area of filtration.

I 000 000

2
c
m
r
X

6
e
*;
m
c
UI
?
w
0
0

:
E
.-E
0
.-
t
10000 J I I I I I I Ill
1 10 100 1000
Surface area (m2)

Figure A2.7.4 Filter prices. (Source: IFP)

 404 Appendix 2. Pr6Estirne Method

Some corrections need to be made, particularly for the kind of material employed, by mul-
tiplying the base price by&. Tables A2.7.9a and A2.7.9b present the base characteristics and
possible correction factors for each type of filter appearing in Fig. A2.7.4.
The pump unit and the price of filter cloths are included in the base cost of automatic frame
filters.

Table A2.7.9a Effect of kind of material on prices of vertical and horizontal frame filters

I Material 1 Correction factorf, 1

Carbon steel (shell and tubes) 0.4
304 stainless (frames only) 0.6
304 stainless (shell and frames) 1.o
3 16 stainless (frames only) 0.9
3 16 stainless (shell and frames) 1.3
(Source: IFP)

Table A2.7.9b Effect of kind of material on prices of automatic frame filters

Material Correction factorf,

Carbon steel 1.oo
Aluminum 1.75
PVC coating 2.00
Rubber coating 2.50
304 stainless 2.80
3 16 stainless 5.00

For the other filters whose cost is shown in Fig. A2.7.4 the base conditions are as follows:
0 Rotary vacuum filters: 304 stainless

Rotary pressure drum filters: carbon steel

A2 Belt filters: 3 16 stainless, rubber belt
Horizontal disk vacuum filters: 304 stainless
7 Airtight filters: 3 16 stainless.
In each case, the driver and the filtering medium are included in the price.
 Appendix 2. Pr@-€stimeMethod 405

A2.7.5 EVALUATION OF CENTRIFUGES

A2.7.5.1 General
The principal methods of separation using centrifuges, and their applications, are presented
in Table A2.7.10.

Table A2.7.10 Principal types of centrifuge

Particle
Solids
Applications diameter
(%)
tP)
~

Centrifugal dryers
- cyclic:
pusher C 20 to 75 10 to 5000
knife b(?) and c 5 to60 0.7 to 1000
pendular C 5 to60 0.7 to 1000
- spiral screw rotary decanter b(?) and c 2 to40 0.3 to 3000
High-speed disk separators
nozzle b and c 5 to25 0.07 to 70
aperture b and c 1 to10 0.07 to 70
hermetic bowl a, b and c Oto 1 0.07 to 70
a: liquid-liquid separation.
b: liquid-liquid-solid separation.
c: liquid-solid separation.
(Source: IFP)

A2.7.5.2 Sizing Centrifuges

Once again, sizing centrifuges is a matter for specialists. Knowing the diameter of the bowl
and the basket, as well as the characteristic curve for a particular piece of equipment, for the
treatment of a given product, enables the hourly production and the electrical requirement to
be defined. A2
In the absence of up-to-date information on this subject, the curves in Fig. A2.7.5a, which 7
were compiled for specific cases, may be used a rough guide. However, consulting a manu-
facturer is still the best way to obtain reliable, effective advice-ven though this goes
beyond the scope of an evaluation.
 406 Appendix 2. Pr6-Estime Method

0.1 1 10
Bowl or basket diameter (m)

Figure A2.7.5a Sizing of centrifuges. (Source: IFP)

A2.7.5.3 Centrifuge Prices

The prices of centrifugal dryers are shown in Fig. A2.7.5b. The curves were compiled for
A2 3 16 stainless steel, which is the material most commonly used for this type of equipment. For
7 carbon steel apparatus, the values in Fig. A2.7.5b may be divided by two.
For high speed, aperture-type separators, Fig. A2.7.5~should be used; it is also based on
3 16 stainless steel.
 Appendix 2. PrkEstime Method 407

1 000 000

d
g 1ooOoo
a
w
Q9
P.-c
.-8
t
10000
0 1 10
Bowl or basket diameter (m)

Figure A2.7.5b Centrifuge prices. (Source: IFP)

1 000 000

100ow)

A2
loo00
7
1 10 100
Capacity (m3/hr)

Figure A2.7.5~ Prices of high-speeddisk separators. (Source: IFP)

 408 Appendix 2. Pr6-Estirne Method

A2.7.6 EVALUATION OF CRUSHERS AND GRINDERS

A2.7.6.1 General

A very rough classification of crushers and grinders can be made on the basis of the hardness
of the product to be processed, and the initial and final sizes of the solid particles. This is illus-
trated in Table A2.7.11, which also makes possible a quick selection of the type of equipment
to employ.

Table A2.7.11 Classification of crushing and grinding equipment

Product Size in cm
OpCration ‘Qpe of equipment
hardness Feed Product
Crushing:
primary hard 150-30 50-10 Jaw & gyratory crushers
50-10 12-2.5 Roller & bar mills
secondary hard 12-2.5 2.5-0.5 Jaw & gyratory crushers
4-0.6 0.5-0.08 Roller, bar, disk, & hammer mills
weak 50-10 5- 1 Roller, disk, hammer, & rotary knife
mills
Grinding:
coarse hard 0.5-0.08 0.06-0.008 Roller, disk, hammer, & rotary knife
mills
Mills with low and medium
tangential speeds
fine hard 0.12-0.015 0.008-0.001 Mills with low, medium, and high
tangential speeds
Fluid drive mills
Disintegration:
coarse weak 1.2-0.17 0.06-0.008 Mills with high tangential speeds
fine weak 0.4-0.05 0.008-0.001 or fluid drive
(Source: IFP)
A2
7
 Appendix 2. Pr6-Estime Method 409

A2.7.6.2 Price and Performance of Crushers and Grinders

A. Base Prices

Figure A2.7.6 gives the average prices of various types of crushers and grinders as a function
of their processing capacity in t/hr. This graph represents standard steel equipment.
A variety of supplemental information can be supplied.

I00 000000

10 000 000

1 000 000

100 000

Figure A2.7.6 Prices of crushers and grinders. (Source: IFP)

B. Supplemental Technical and Economic Information

a. Bar Mills

Correction for grain size (Table A2.7.12a) A2

Bar mill performance (Table A2.7.12b). 7
Price of bars: 900 f (mid-2000).
410 Appendix 2. Pr6-Estime Method

Table A2.7.12a Effect of product grain size on the cost of rod mills

Particle size Correction factor

Mesh number Opening (side) in mm fgr

number
10 12 1.651 0.6
48 55 0.294 1.o
100 100 0.147 1.7
98% minus 325 mesh - - 2.5

Table A2.7.12b Rod mills: effect of capacity on rod load and motor power

Capacity (t/hr) 3 10 25
Rod load (t/hr) 12 30 70
Power (CV) 125 400 960

b. Hammer Mills (Table A2.7.12~)

Table A2.7.12~ Hammer mills: effect of capacity on motor power

Average capacity ( t h ) 25 40 100 150

Power (CV) 90 150 300 400

Product size (cm)

(min. & max. settings) 2-6 2.5-7 3-8 5-1 1 6-12

Capacity (thr) (min. & max. settings) 4-9 7-17 15-30 30-70 50-90
Motor power (CV) 13 25 35 50 60
 Appendix 2. Pr6-Estime Method 41 1

d. Gyratory Crushers (Table A2.7.12e)

Table A2.7.12e Cone crusher characteristics

I Product size (cm) I 0.6 I 0.9 I 1.0 I 1.2 I

Average capacity (tkr)
Motor power (CV)
(Source: IFP)
I 50
70 I 75
110 I 120
180 1 165
270 1
e. Gyratory Crushers (Table A2.7.120

Table A2.7.12f Gyratory crusher characteristics

1 Product size (cm) 1 0.75 1 1.00 1 1.25 I 1.50 1

Average capacity ( f i r ) 30 90 160 260
Motor power (CV) 60 125 220 270

f. Pulverizers (Table A2.7.12g)

Table A2.7.12g Pulverizer characteristics

[ Average capacity ( f i r ) I 1.0 I 1.6 I 2.2 I 2.7 I 3.2 I 3.9 I

Motor power (CV) 25 30 40 50 60 75

A2.7.7 EVALUATION OF GAS CYCLONES

A:
Gas cyclones are mainly used for removing dust particles of between 10 p and 1 mm from a 7
gas. Below that size, small diameter, high load loss cyclones (multiple cyclones) are used.
When the gas to be purified or the particles themselves are corrosive, equipment designed
for severe operating conditions is used.
Figure A2.7.7a, which shows the efficiency of cyclones as a function of the size of particle
to be removed, enables a decision to be made about whether or not to apply such equipment,
and the quantity of material that can be recovered.
Figure A2.7.7b shows the variation of average prices with capacity.
412 Appendix 2. Pr6-Estime Method

Particle diameter (p)

1 10 100
100
99
98

95

90

50

10

Figure A2.7.7a Gas cyclone efficiency. (Source: IFP)

100 000

10000

1 000
1 000 10000 100000 1 000 000
Capacity (m3hr)

Figure A2.7.7b Gas cyclone prices. (Source: IFP)

 Appendix 2. Pr6-Estime Method 413

A2.7.8 EVALUATION OF VIBRATING SCREENS

A2.7.8.1 General

Price data are more readily available for vibrating screens than for other equipment used in
separating grains of varying size.
The price of the equipment essentially depends on the screen surface area, and this param-
eter is dependent on the granulometry of the feed to be processed and the products sought, as
well as the type of screen utilized.
Although precise standards do exist, it appears difficult to develop a quick calculation
method: it is preferable to consult the manufacturers directly.

A2.7.8.2 Prices of Vibrating Screens

As a rough guide, Fig. A2.7.8 shows the variation of simple (one stage only) vibrating screens
as a function of their surface area.
Corrections may be applied for the number of stages, by multiplying the base price byf,,
(Table A2.7.13).
The most common sizes and the power required are summarized in Table A2.7.14.

a
E
w
Qv
0
E
.-C
A:
7

Figure A2.7.8 Prices of vibrating screens. (Source: IFP)

 414 Appendix 2. Pr6-Estime Method

Table A2.7.13 Effect of number of stages on price of vibrating screens

I Number of stages I f", I

I
1
2
3
II 1.oo
1.20
1.35
II
(Source: IFP)

Table A2.7.14 Size and energy requirement of vibrating screens

Size Number Power

(m) of stages (CV)
0.30-0.90 1 0.50
2 0.75
0.45-0.90 1 0.50
2 0.75
0.6- 1.20 1 0.75
2 1.o
0.9- 1.80 1 2.0
2 3.0
1.20-2.40 1 3.0
2 5.0
1.50-3.00 1 5.0
2 7.5
(Source: IFP)

A2.7.9 EVALUATION OF CONVEYORS

A2.7.9.1 General
A2 The main types of conveyor are the following:
7 0 Belt conveyors: flat or raised edge. These are horizontal or inclinable, characterized by the

width of the transporter belt, the distance between roller axles, the linear speed of the belt,
and the motor power, as required by the flow of product to be carried and its density.
Screw conveyors. These too may or may not be inclined, and are characterized by the
diameter of the Archimedes screw, the distance between its ends, the speed of the rotor, and
the power of the motor, considering the required output. The kind of product to be transported
also affects the choice of the material from which the equipment is made.
 Appendix 2. Pr&Estime Method 415

0 Bucket elevators. The most important parameters are the unit capacity of the buckets, the

height of the lift, the speed of the chain that carries the buckets, and the power of the driver:
these are a function of the flow of product to be moved, and its density.

A2.7.9.2 Conveyor Sizes and Prices

Figure A2.7.9 provides information on the price of this equipment according to the distance
of transport (length or height), the other dimensions being fixed in advance, and taken as ref-
erences.
Correction factors must be applied to obtain prices for equipment having different charac-
teristics.

1 000 000

100 000

a
e
w
0
8 10 000
7
E
E
.-C
.-8
t
1 000

Figure A2.7.9 Conveyor prices. (Source: IFP)

A. Belt Conveyors A2
7
a. Technical Parameters
Capacity (t/hr) (Table A2.7.15a)
Motor power (CV) (Table A2.7.15b).
 416 Appendix 2. Pre-Estime Method

Table A2.7.15a Belt conveyors: effect of belt width and speed on quantity carried (d = 1)

1.0 1.5 2.0

0.500 80 100 150 200

0.650 150 180 270 360
0.800 200 250 380 500

(Source: IFP)

Table A2.7.15b Belt conveyors: effect of belt width on driver power, for a given speed (1 d s )
and a particular product tonnage (d c- 1)

Motor power (CV)

Belt width (m) Distance (m)
10 100
0.500 3.0 7.5
0.650 4.5 10.0
0.800 5.5 12.5

b. Price Corrections
0 Base price. The prices in Fig. A2.7.9 are for flat or raised edge belt conveyors with the
following characteristics:
- Width: 0.50 m
- Speed: 1 m l s .
0 Correction factor (Table A2.7.16).

Table A2.7.16 Effect of belt width on conveyor price

A2
Width (m) Aarg
7
0.500 1.oo
0.650 1.20
0.800 1.30
1.000 1.50
1.200 1.70
 Appendix 2. PrtLEstirneMethod 417

B. Screw Conveyors

a. Parameters (Table A2.7.17)

Table A2.7.17 Capacity (t/hr) and motor power (CV) of screw conveyors

Screw diameter Distance Solid carried (t/hr)

(m) Grains
0.300
0.400
0.400
0.400 110 250 5.0
0.400 15 110 250 12.0
(Source: IFP)

b. Price Corrections
0 Base price. The curve in Fig. A2.7.9 for screw conveyors in carbon steel is based on the
following conditions:
- Diameter of Archimedes screw: 0.40 m
- Rotor speed: 60 rpm.
0 Correction factors (Table A2.7.18).
For screw conveyors in 304 L stainless steel the price curve may be obtained from the pre-
vious one (for carbon steel) by using a multiplying factor of about 1.7.

Table A2.7.18 Effect of screw diameter on conveyor price

Diameter fdiam
(m) Conveyor: carbon steel Conveyor: 304 L stainless
0.150 0.65 0.60
0.300 0.85 0.80
0.400 1.oo 1.oo
0.500 1.15 1.20

C. Bucket Elevators
Base price. The curve in Fig. A2.7.9 for bucket elevators was based on the following
conditions:
- Capacity: 80 tfhr
- 304 L stainless steel buckets.
 418 Appendix 2. Pr6-Estime Method

0 Correction factor (Table A2.7.19).

Table A2.7.19 Effect of capacity on bucket elevator prices

Capacity(') (tlhr)

1.oo
120 1.20
(1) Density near 1.
(Source: IFP)

A2.7.10 EVALUATION OF INSTRUMENTATION

The factorfg, which is applied to determine the erected cost of each category of equipment
and which is given in Tables 3.21 and 3.22, covers standard instrumentation.
However, it does not cover automatic controls, whose cost depends not only on the number
of lines, but also on just how complex each one is. Moreover, such controls are installed only
in the final stages of plant construction, by specialist companies whose fees cover both the
equipment and its installation on the unit. In other words, in this case it is not possible to sep-
arate the equipment part of the cost from the other items, in particular the cost of erection. In
practice, then, the price of the automatic control system is added to the production battery
limits investment, as calculated without including this item. In any event, a rough estimate of
this amount may be obtained from the cost of the principal equipment or its component
pieces. Accordingly, there are two possible options:
(a) For standard units, add 10% of the overall price of the principal equipment
(b) Or break it down by categories of equipment, taking:
15% for the reactor and column set-up
0 5%fordmms

10% for compressors

and add the results together.

A2 IMPORTANT COMMENT
7 It should be emphasized that the values calculated by either method are assumed to
already include the costs of erection or other related expenses, and are to be added to
the erected costs of the principal equipment.
 Appendix 2. Pr6-Estime Method 419

A2.8 EVALUATION OF UNITS FOR PROVIDING UTILITIES

The investment costs cover the following items:

Production of utilities: these are for the complete, installed units, in operating condition,
which have to be connected to the various distribution systems,
Distribution of the utilities to the units that use them: this covers the cost of setting up
the systems for handling the utilities.

A2.8.1 EVALUATION OF UTILITY-PRODUCING UNITS

These units generally include the following equipment:

0 Steam production

0 Electricity generation

0 Obtaining boiler water, process water, and cooling water (including recycling, i.e.,

returning the water to its initial state)

Refrigeration.
It should be remembered that facilities for producing utilities rarely serve a single unit:
they supply the needs of all or part of the industrial complex concerned, and the capacity
selected for determining the investment must be adjusted appropriately.
In economic calculations utilities are often reckoned as if they were purchased: their inter-
nal transaction prices are determined by the methods discussed earlier.

A2.8.1.1 Steam Production

This can be provided either:

By packaged boiler units, which are boilers delivered fully assembled and complete.
Their capacity is limited (100 to 150 f i r ) , simply because they have to be transportable.
Or by boilers that are delivered as components and assembled onsite.
Figure A2.8.1 gives the investment costs for these two types of units (complete facilities
powered by heating fuel oil), as a function of production capacity.
Depending on the type of boiler (fuel, desulphurization, or de-nitrogenization), the boiler
pressure, and the required superheating of the steam, we apply correction coefficients:
price of facility = base price . (1 +f,+ft ffp)
A2
8
 420 Appendix 2. Pr6-Estime Method

A. Base Price

-36
%
(Y
O0 Oo0 Oo0

$
-7
8
c
10000000
m

W
8
:
0

.z 1000000
.-C
.-i3
B
jj
e
w 100000
1 10 100 1000
Capacity (t/hr of steam)

Figure A2.8.1 Prices of boilers powered by heating fuel oil.

(Source: IFP)

B. Correction Coefficients
Table A2.8.1 Correction factors for boiler base prices

a. Effect of type

I Type of boiler I fc
Natural gas boiler -0.3
Heating fuel oil boiler 0.0
Boiler with anti-pollution system 0.7
(desulphurization, de-nitrogenization)
(Source: IFP)

1 1
b. Effect of superheating

I I f t
;:;:
Temperature of superheating ("C)
A2 Saturation
l;;
8
150 0.20
200 0.25
(Source: IFP)
 Appendix 2. Pr@-EstirneMethod 42 1

c. Effect of pressure

I Pressure (bar) I fP I
10 0.00
40 0.65
100 0.75

A2.8.1.2 Electricity Generation

We may require either:
0 The production of electricity alone

Or a combined production of electricity and steam.

For the production of electricity alone, natural gas binary cycle methods are increasingly
preferred: a gas turbine coupled to a boiler (with or without postcombustion), whose steam
expands in a condensation turbine. The gas turbine can be used by itself, but its efficiency is
low. Lastly, we may also use a condensing steam turbine, fed, for example, by a fuel oil-bum-
ing boiler.
The combined production of steam amounts to using steam either directly from the boiler,
or as it leaves a back-pressured turbine.
We may also mention integrated binary cycle systems, powered by a coal gasifier.
There are a great number of possible configurations, each one adapted to a particular
energy need. We will therefore limit ourselves here to presenting the investments for certain
type systems (Fig. A2.8.2) and their corresponding energy efficiencies (Table A2.8.2).

Table A2.8.2 Electricity generation: overall energy efficiencies for various alternatives

'Qpe of facility

Combined cycle (integrated gasification) using coal

I Energy efficiency
("/I
41
1
Combined cycle using natural gas
Ratio of therms (steam) to therms (electricity): 0.0 52
0.3 55
3.5 82
Gas turbine only 25
Boiler + steam turbine 39
I A2
(Source: IFP) 8
 422 Appendix 2. Pre'-Estime Method

V
r
K
1000000000
t
5
I

* 100000000
2
w
0
0
10 000 000

1 000 000

100 000
1 10 100 1 000
Power (MW)

Figure A2.8.2 Prices of electrical generating facilities.

(Source: IFP)

A2.8.1.3 Production of Cooling, Boiler, and Process Waters

In view of the large flows of cooling water that circulate in industrial chemical complexes,
such water must be recycled, once its temperature has been brought down to a level accept-
able for plant operations. In general the temperature of water leaving the production units is
found to reach 35"C, whereas it usually enters at about 25"C, and sometimes 20°C.
The cooling is done in towers, using atmospheric air in counter-current circulation with the
water to be cooled in the form of a fine spray.
There are three major types of coolers:
Natural aeration: spray-pipe coolers, stepped coolers
Natural draft: wood shaft and drip system, concrete shaft
Forced ventilation.
The boiler water is simply softened; the process water is demineralized.
For cooling and boiler water, the make-up amount is reckoned to be 5% of the total circu-
lating flow. The investment costs (Fig. A2.8.3) include the various required treatments.
A2 In addition to the various chemicals, whose cost must be evaluated in each particular case,
8 there is a consumption of electricity to be considered. Its requirement in kW/hr per m3/hr of
water used amounts to about:
0.17 for cooling water
0.06 for boiler water
0.14 for process water.
 Appendix 2. Pr4-fstirne Method 423

p 100000000
0
c)

3
P
0
g 10000000
z- in,,,,, , I , I I I U

d
L

g 1000000
0
B
E
E
.-c
p) 100000
.-0
k
x
c)

F
w 10000
1 10 100 1000 10 000 100 000
Capacity (m3/hr)

Figure A2.8.3 Prices of water production facilities. (Source: IFP)

Finally, in some plants the temperature of the cooling water is allowed to rise by more than
10°C: the basic investment then has to be corrected by means of a coefficient whose value is
given in Table A2.8.3.

Table A2.8.3 Effect of inleb‘outlet temperature difference

on cost of water cooling

3
1
(Source: IFP)
At (“C)

1.8

A2.8.1.4 Refrigeration A2
8
In chemical plants it is often necessary to have production cycles for various levels of cold.
The systems employed vary considerably depending on the temperature required (propane,
ethane, or methane cycles): the investment cost depends on the temperature and the maximum
unit production capacity.
 424 Appendix 2. Pr6-Estirne Method

Figure A2.8.4 shows the unit price (mid-2000 Mf per Mkcal) as a function of temperature
for the corresponding maximum unit capacity: it refers to installed units, including centrifu-
gal compressors, evaporators, condensers, etc., and connection and erection costs.
The investment cost is calculated from the maximum capacity selected, by applying an
extrapolation exponent of 0.7 for the smaller sizes; otherwise we determine the number and
capacity of the modules required to service the perceived overall needs.

c.
3
X
6 2000000
?!
zI
m^ 1500000
e
W
0
0
1000000
.-
E
.-E
.-0 500000

-
h
P
?!
W 0
-100 -90 -80 -70 -60 -50 40 -30 -20 -10 0
Temperature (“C)

Figure A2.8.4 Price of refrigeration facilities. (Source: IFP)

The production cost includes the considerable electricity consumption, which rise in pro-
portion to the temperature level sought (Fig. A2.8.5).
Lastly, we can consider the consumption ofwater, which can be obtained from Fig. A2.8.6.

A2.8.2 DISTRIBUTION O F UTILITIES TO THE USER UNITS

A2 The facilities that consume utilities are often close to their point of production. For this rea-
son distribution costs are sometimes negligible in comparison to the cost of the production
8 units.
However, in some cases the investment required by distribution is on the same order of
magnitude as that for production: this may be due to distance, or to the particular nature of
the utility concerned (insulation of steam pipes, for example).
 Appendix 2. Pr6-Estime Method 425

2 500

2 000

500

0
-1 00 -80 40 40 -20 0
Temperature (“C)

Figure A2.8.5 Electricity requirement for refrigeration.

(Source: IFP)

300

250

200
-m
$
n
150

E
100

50

0
-1 00 -80 -60 40 -20 0
Temperature (“C) A2
8
Figure A2.8.6 Cooling water requirement for refrigeration.
(Source: IFP)
 426 Appendix 2. Pr6-Estime Method

The utilities to be considered here are the following:

0 Steam

0 Cooling water

0 Electricity.

A2.8.2.1 Steam
Depending on the distance between the producing and consuming facilities, and the size of
the steam production unit, the amount of the investment assigned to distribution may vary
from:
3 500 to 13000 mid-2000 € per t/hr of steam produced.
If data concerning distance are lacking, it is wise to increase the fixed costs in the steam
transaction price by 10%.

A2.8.2.2 Cooling Water

Distribution of cooling water, when the water is obtained from cooling towers, carries an
investment cost that may vary from 150 to 600 mid-2000 f per m3/hr required.
When the water is taken directly from a river, the investment ranges from 80 to 350 mid-
2000 € per m3/hr required.

A2.8.2.3 Electricity
There are several different cases, depending on the area of investment:
0 Electricity for general use (lighting circuits): 340 to 480 mid-2000 €/kW

0 Primary transformer: 150 to 240 mid-2000 €/kVA; it is necessary to correct for capacity

(see Table A2.8.4)

Capacity (LVA) fcap

3 000 1 .o
5 000 0.6
10000 0.4
20 000 0.3
A2 (Source: IFP)
8

0 Secondary transformer: 140to 190 mid-2000 €/kVA; it is necessary to correct for capac-
ity (see Table A2.8.5).
 Appendix 2. Pre-Estime Method 427

Table A2.8.5 Effect of size of secondary transformers on their price

Capacity (kVA) fdim

1500
2 000 0.5
(Source: IFP)

A2.8.3 SPECIAL CASES

Some types of utility are produced using standard equipment, whose price can be found in
Appendices A2.1 to A2.7. This provides the investment for principal equipment.
We can also calculate the investment costs for distribution.
Utilities whose use involves the common types of equipment include:
Compressed air (compressor, filter, dryer)
Treated water (filter, reactor, distillation column)
0 Fuel (pump, storage tank, etc.).

The investments pertaining to distribution can then be obtained as follows:

Compressed air: 60 to 240 mid-2000 €/lo3 € of principal equipment cost, depending on
the complexity of the system
Treated water: 130 to 240 mid-2000 f/m3 of filtered and softened water
Fuel: 45 to 300 mid-2000 €/lo3 f of principal equipment cost for liquid fuel (heating
oil), 75 to 360 mid-2000 €/lo3 f of principal equipment cost for gaseous fuel (fuel-gas).

A2
8
 Appendix 2. Pr6-Estime Method 429

A2.9 EVALUATION OF STORAGE

Storage tanks include:

0 Vessels used at atmospheric pressure

0 Pressurized vessels:

- Cylindrical
- Spherical.
Figures A2.9. la, A2.9.lb, A2.9.2, and A2.9.3 give prices for this equipment as a function
of capacity, for very specific types of storage and particular conditions of utilization. Correc-
tion factors must be applied for other cases.

A2.9.1 ATMOSPHERIC PRESSURE TANKS

A2.9.1.1 Capacity Less than 100 m3 (Table A2.9.1)

corrected price = base price .f,

where f, is the correction factor for the material employed.

Table A2.9.1 Effect of kind of material, for tanks of less

than 10 m3 capacity (atmospheric pressure)

Material fm
Carbon steel 1.oo
Aluminum 1.40
Rubber coated 1.50
Lead coated 1.55
Stainless steel 3.20
Enameled steel 3.25

A2
9
 430 Appendix 2. Pr6-Estime Method

pi
r
F!
ic
+
u)
e
W 10000
0
0

2
.-
E
.-0
C

.-
h
P
L

F! 1 000
W
0 1 10 100
Capacity (m3)

Figure A2.9.1 a Price of cylindrical atmospheric pressure tanks (smaller

then 100 m3). (Source: IFP)

I 000 000

100 000

10 000
100 1 000 10000 100 000 1 000000
Capacity (m3)

A2 Figure A2.9.1 b Price of cylindrical atmospheric pressure tanks (larger

then 100 m3). (Source: IFP)
9
 Appendix 2. Pr6-Estirne Method 43 1

5 1000OOO
0
4-

100OOO

1oOOO

1000
1 10 100 1000
Capacity (m3)

Figure A2.9.2 Price of horizontal cylindrical pressure tanks.

(Source: IFP)

100OOO

1 10 100 1 000
Capacity (m3)

Figure A2.9.3 Price of spherical pressure tanks. A2

(Source: IFP)
9
 432 Appendix 2. Pr6-Estime Method

A2.9.1.2 Capacity Greater than 100 m3 (Tables A2.9.2a and A2.9.2b)

corrected price = base price .fd f,

where
fd is the correction factor for type
f, is the correction factor for material.

Table A2.9.2a Effect of type on tanks of capacity greater

than 100 m3 (atmospheric pressure)

TYPe fd
Cone roof 1.o
Floating roof 1.1
Elevator-type roof 1.3
(Source: IFP)

Table A2.9.2b Effect of kind of material on tanks of capacity greater

than 100 m3 (atmospheric pressure)

Material
Carbon steel
Rubber coated
Lead coated
304 stainless steel 2.8
(Source: IFP)

A2.9.2 PRESSURIZED STORAGE TANKS

A2.9.2.1 Cylindrical Tanks (Table A2.9.3)

corrected price = base price .fp
where fp is the pressure correction factor.

A2
9
 Appendix 2. Pr6-Estime Method 433

Table A2.9.3 Effect of pressure on cylindrical pressure tanks

Pressure (bar) fP
<lo 1.oo
10 to 15 1.20
15 to 20 1.so

A2.9.2.2 Spherical Tanks (Table A2.9.4)

corrected price = base price .fp
wherefp is the pressure correction factor.

Table A2.9.4 Effect of pressure on spherical pressure tanks

Pressure (bar) fP
3 1.o
5 1.3
7 1.5
10 1.8
15 2.3

A2
9
INDEX

Index Terms Links

Accessories (see Vessels, pressure)

Acetylene 58
Actors in a project 230
Added value
long-term 69
short-term 69
Agitators
power 336
price 335
Air-coolers 349
price 355
principle 349
sizing 354
exchange area 355
power requirement 354
width 354
Allen D.H. and Page R.C. (method) 254
Allowable stress 320
Alumina (see Dryers)
American association of cost engineers 274
Annual discounted cash flow 55 86
general expression 88
principal type applications 89
financing from the general treasury 89
income, expenses, depreciation constant 97
known terms 90
no taxation 92
This page has been reformatted by Knovel to provide easier navigation.
Index Terms Links

ASTM designation 320

Audit
external 36
internal 36

Bach N.G. (method) 266

Battery Limits Investments (see Investments)
Bottoms (see Vessels, pressure)
Boulitrop R. (index) 195
Boyd’s indices 213
Break-even point 77
Bridgewater A.V. 211
Buildings (see Secondary material)
By-products 118

Callagher J.T. 274

Capital
borrowed 79
company (see Equity funds)
depreciable 165 178
fixed 165 172 271
working 182
Capitalized discounted cash flow 93
financing from company treasury 93
financing method known 94
income, expenses, depreciation constant 97
no taxation 94

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Cash flow 72
annual discounted calculation 86
capital repayment 111
capitalized discounted 93
financing capacity 111
growth 111
maintenance 111
practical significance 111
Cash from operations (see Cash flow)
Catalysts 179
heterogeneous phase 118
homogeneous phase 120
mass 135
supported 135
CE (chemical engineering) plant cost index 200
Centrifuges 405
Charges
financial (interests) 72 163
fixed 151
capital 154
operating 151
indirect 271
variable 116
Chilton C.H. 270
Civil engineering 264
Cladding (see Vessels, pressure) 66
Clerk J. 274
Coefficient
α 96
β 98
overall gross heat transfer (tube-type heat
exchangers) 337

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Coefficient (Cont.)
overall gross transfer coefficient (tube-type
exchangers) 338 339
overall heat transfer (evaporated) 398
Columns (see Vessels, pressure)
Commodity material 264
Compressed air 427
Compressor 368
prices 370
types 368
Construction (see Erection)
Contingencies 264
Contractor fees 175
Conveyors 414
belt conveyors 415
bucket elevators 417
general 414
price 416
screw conveyors 417
sizes 415
Co-products 117
Correlation (mathematical with precursor index) 26
Cost
definition 264
direct
average structure 270
Guthrie method 279 280 284
solids handling or special equipment 284
standard chemical equipment 280
economic cost of price 107
operating expenses invariant 108
operating expenses variable 109

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Cost (Cont.)
indices (see Indices)
examples 195
use 194
indirect, Guthrie method 280 284
labor 149
location effect (see Investments)
of price 112
operating 112 182 183
breakdown 114
components 115
production (see Price, cost of)
shadow - of equity fund 90
transport costs and risks standardized
language 128
utilities, definition 137
Cran J. 208
Criteria of profitability 71
standardization 71
typological classification 71
criteria based on the concept of present
value 71
empirical criteria 71
Crushers (see Grinders) 408
performance 409
price 409
Crystallizers 394
crystallization efficiency 394
prices 396
selection 395
Cumulative discounted cash flow (see
Capitalized discounted cash flow)

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Cumulative net present value (see Capitalized

discounted cash flow)
Currency
constant 39
current 39
Customers (intax oriented methods,
forecasting) 35
Cyclones 411

Depreciation
background 155
calculating depreciation 158
declining balance 159
economic 109
of currency (rate of) 84
operating period and depreciation period 156
provision, allocation for 72 155
salvage value of facilities 158
sinking fund 162
straight-line 159
sum of the years’ digits 162
Direct analogy (see Estimation)
Distribution of utilities 424
cooling water 426
electricity 426
steam 426
Driver, pump 363
Drums (see Vessels, pressure)
Dryers 385
flash dryers 388
multipurpose dryers 391

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Dryers (Cont.)
prices 391
rotary drum dryers 390
selection 385
sizing 385
alumina or molecular sieves 387
surface exchange drying 387
spray dryers 388
steam-tube rotary dryers 391
vacuum rotary dryers 389

Ejectors, steam
calculations 377
consumption of utilities 380
maximum suction flow 377
number of stages 377
prices 383
steam 377
Elasticity 6
anticipation elasticity 7
coefficient 7 18
cross elasticity of demand 8
empirical or casual elasticity 7
price-demand elasticity 7
substitution price elasticity 9
theoretical elasticity 7
Electric motor 364
Electrical systems (see Secondary material)
EMIP (Equivalent maximum investment
period) 78
Engineering, fees 172

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

ENR (Engineering news-record)

construction cost index 196
Equipment
consumable 120
principal 263
special 385
Equity funds 79 84
Erection 264 268 283
303
Estimation 229
choosing an estimation method 231
classification of estimation methods 235
activity or analytical approach 235
product or global approach 237
procedure 223
respective contributions of evaluation and
estimation 229
selecting an estimation method 244
selection criteria 232
Evaluation (see Estimation)
commercial evaluation software 290
Evaporators 397
prices 400
selection 397
sizing 398
technical characteristics 399
Exchangers
conventional nomenclature 343
double pipe 342
plate 358
price 360
sizing 359

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Exchangers (Cont.)
prices of exchanger tubes 349
prices of tube-type 342
double pipe exchangers 347
exchangers with tube bundles 342
selection 341
TEMA standard 342
tube-type heat 337
Experts
Delphi method 35
Exponential methods (see Investments,
methods for determining)
“extrapolation factor” procedure 248
flow methods 253
Allen D.H. and Page R.C. 254
Stallworthy E.A. 253
functional modules method 263
notional operating unit methods 248
Hill R.D. 248
Wilson G.T. 251
Zevnik F.C. and Buchanan R.L. 249
significant steps methods 256
Taylor J.H. 256
Viola J.L. 260
External (see Vessels, pressure)
Extrapolation
exponent 189
factor 186

Facilities for the production and

distribution of utilities 168

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Factor, stream 112

effective 113
theoretical 113
Factorial methods (see Investments, methods
for determining)
methods employing constant multiplying
factors 264
methods employing variable multiplying
factors 269
methods that consider the effect of capacity 274
methods that consider the effects of materials 274
methods that consider the particular nature
of each project 271
cost structure of principal equipment 273
identifying the most relevant criteria of
variation 271
severity of operating conditions 273
type of equipment 273
unit production capacity 272
Fans 369
Filters 402
prices 403
selection 402
sizing 402
Financial charges (see Interest)
Forecasts
application methods 39
areas of application 19
dominantly future-oriented methods 34
external or exogenous methods 26
frequency 19
internal or endogenous methods 21

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Forecasts (Cont.)
intuitive procedures 20
limits of methods 38
long-term 20
medium-term 20
methods 18
objective procedures 20
processing and extrapolation of historical
data 20
short-term 20
types 19
very long-term 20
very short-term 20
Foundations (see Civil engineering and also
Secondary material)
Functional modules method (FMM) 292 309
application 313
auxiliary modules 313
description of the method 311
distillation module 310
functional modules 312
principle 292
Furnaces 373
energy cost 375
general observations 373
price 373

General services and common facilities 170

Giffen effect or paradox 8
GNP Deflator 41
GNP per capita 27

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Graphical adjustment and extrapolation of

data 21
Grassroots investments 183
Grignard (see Reactor agitated)
Grinders 408
performance 409
price 409
Gross domestic product 27
Gross National Product 27
Guthrie K.M.
location factors 209
method to calculate investments 279

Hand W.E. 267 274

Heads (see Vessels, pressure)
Hill R.D. 248
Hirsch J.H. and Glazier E.M. (method) 275

IFP (methods developed by) 291

“functional modules” method 292
background 291
pre-estime method 286 297
description 302
development and application 298
state of development 297
validity and application limits 305
Indices
deflation 39
implicit GDP price 41

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Indices (Cont.)
industrial production (IPI) 27
location (Boyd) 214
tightness 55
updating or cost index 190
Boulitrop’s 195
chemical engineering 200
concept of 190
ENR (Engineering news-record)
construction cost index 196
equipment cost 196
IFP Eval and bieval 220
Laspeyres index 192
M&S 196
Nelson 197
Paasche (De) index 192
principal features and consistency 201
Vandamme 195
Von Kölbel and Schulze 201
WEBCI (Werkgroep Begrotingsproblemen
in der Chemische Industrie) 201
Initial loads 178
Instrumentation 264 268 418
Insulation (see Secondary material)
Insurance 153
Interest
compound 79
continuous compound 80
on construction loans 179
simple 79
Internals (see Trays, packings)

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Investments
adjustment of capacity 185
battery limits
definition 166
determination of 223
structure 263
charges 164
location effect 185 206
location effect 206
background 206
first Attempt 207
systematic methods 211
methods for determining 247
total grass roots 166
updating 190
concept 190
examples 195
problem 190
use 194
IRP (Interest recovery period) 78
IRR 55

Johnson R.J. 208

Klumpar I.V. and Slavsky S.T. (method) 286

Labor 146
concept of the shift worker 146
costs 149

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Land costs 184

Lang H.J. 265
Laspeyres index 192
Latent heat of vaporization of water 340
Life operating (lifetime of facilities) 87
Line of equivalence 41
Lining (see Vessels, pressure)
Loans 84
borrowed capital 79
Location 185 206
Boyd’s ndices 213
IFP technical currency 218
Logarithmic mean temperature difference
(LMTD) 351

M&S (Marshall and Swift, formerly

Marshall and Stevens)
equipment cost index 196
Maintenance 151
Margins 54
disappearing 49
reconstructing 52
Market
analysis 5
raw materials and products available 10
raw materials available, limited for
products 15
raw materials, very big for products 11
Maturity
curve 60
degree 58

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Methanol
example 46 47
Methods
Allen D.H. and Pages R.C. 254
Bach N.G. 266
exponential 248
factorial 263
functional modules 309
Guthrie K.M. 279
Hand W.E. 267 274
Hill R.D. 248
Hirsch J.H. and Glazier E.M. 275
IFP 291
Klumpar I.V. and Slavsky S.T. 286
Lang 266
least square (linear regressions) 22
Miller C.A. 211 276
pré-estimation 286
pre-estime method 319
return on investment (ROI) 75
SRI (Stanford research institute) 290
Stallworthy E.A. 253
Taylor J.H. 256
variable multiplying factors 269
Viola J.L. 260
Wilson G.T. 251
Zevnik F.C. and Buchanan R.L. 249
Miller C.A. 211 276
Models
econometric 30
behavioral 31
economic analysis 31

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Models (Cont.)
several explanatory variables 30
inter-industry exchange 31
Molecular sieves (see Dryers)
Moving average method 22
MTBE
example 11

Nelson (now Nelson-Farrar) refinery

construction cost indices 197
Nelson refinery construction cost indices
chemical product index 136
Nelson W.L.
location effect 207 209
Net present value (see Annual discounted
cash flow)

Operating cost 182 183

Operating Manual 178
Overheads
factory 154
head office 154

Paasche (De) index 192

Packings 333
price 333
Painting (see Secondary material)

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Payback period with discount (discounted

POT) 103
“Payout time” or simplified payback period
method 73
Penetration, market saturation 60
Peters M.S. 270
Piping (see Secondary material)
Polymethyl-4 pentene-1 example 16
Present value 78 81
background 78
borrowed capital 79
Price 39
% of raw material cost 134
average 12
calculated 127
ceiling 57
consumer 41
contract 123
cost of
breakdown 114
floor 48 57
increase 50
for evaluation purposes 129
forward 125
minimum profitable sales 49 74 75
102
official selling or producer 122
optimum 18
posted or list 122
price/volume relationships 5
products and raw materials prices
relationships 41

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Price (Cont.)
reference 122
spot 122
threshold 127
Process Data Book 177
Production of utilities 419
cooling, boiler, and process waters 422
electricity 421
refrigeration 423
steam 419
Products
evolution 48
price structure 48
raw materials/products duality 5
Profitability
project 67
Profits
gross 72
real 72
Pumps 363
centrifugal pumps 363
driver 363
reciprocating pumps 366
rotary and metering pumps 368

Rate
currency depreciation 84
discount (see Present value) 81
constant money 83
current money 83

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Rate (Cont.)
growth
average 63
constant real 64
real 63
variable real 65
Internal rate of return (IRR) 99
load 112
psychological discount 84
taxation 69
utilization and load 113
Raw materials 41 116
products duality 5 6
Reactors (see Vessels, pressure)
agitated 335
comparable to columns or drums 335
Reagents 118
Relationships prices raw materials and line
of equivalence 41
Relative capital enrichment (RCE) 106
Repayment 91
Representatives (future-oriented method,
forecasting) 34
ROI 55 71 75
Royalties
paid-up 175
running 176

Sales
annual 72
normalization of evolution curves 58
volume 58
This page has been reformatted by Knovel to provide easier navigation.
Index Terms Links

Sales volume 39 58
Salvage value (see also Depreciation
provision) 87
“Scaling factor” or “six tenths factor” 185
Secondary material 264 268 283
287 289 303
Sensitivity (estimation) 232
Shell (see Vessels, pressure)
exchangers 343
Shift worker 146
Site preparation (see Civil Engineering)
Skirt (see Vessels, pressure)
Smoothing, exponential 24
Spare parts 174
SRI (Stanford Research Institute) method 290
Stages in the progress of a project 224
implementation phases 228
construction proper 229
studies and supplies 228
study or pre-study stages 225
feasibility studies 226
identification 226
transition period 227
Stallworthy E.A. 253
Startup expenses 180
Steel structures (see Secondary material)
Stevenson D. 61
Storage 169 429
atmospheric pressure 429
3
capacity greater than 100 m 432
capacity less than 100 m3 429
definition 169

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Storage (Cont.)
pressurized storage tanks 432
cylindrical tanks 432
spherical tanks 433
Study (market)
general procedure 36
Styrene-butadiene rubber (example) 62
Superchargers 369
Supervision 150

Taxes 68 153
Taylor J.H. 256
Technical dollar exchange rate (IFP) (see
Investment cost, location effect)
Terris P. 209
Thermal insulation (see Secondary material)
Thickness 319
application 321
basic formula 319
maximum allowable stress 320
welding coefficient 320
Timmerhaus K.D. 270
Traceability 232
Trays 330
price 330
Treated water 427
Trend, strong 49
Tube bundles 342
Turbines 364
Turbo-blowers 369
Turnover ratio 76

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Unit capacity 185

Updating (see Indices)
Utilities 137
definition 137
distribution 424
prices 137
cooling, process, and boiler water 145
electricity 143
fuels 139
refrigeration 145
steam 143
production of 419
cooling, boiler, process waters 422
electricity 421
refrigeration 423
steam 419
updating (see Investment, cost of)

Valves (see Secondary material)

Vandamme 195
Veblen effect 8
Vessels, pressure 319
price of accessories 329
price of shell and bottoms 323
carbon steel with cladding 325
carbon steel with lining 327
homogeneous material 323
price of skirt 329
weight of the bottoms 322

This page has been reformatted by Knovel to provide easier navigation.

Index Terms Links

Vessels, pressure (Cont.)

weight of the shell 322
weight of the skirt 322
Vibrating screens 413
general 413
prices 413
Viola J.L. 260
Volumes 39
Von Kölbel and Schulze 201

WEBCI (Werkgroep Begrotingsproblemen

in der Chemische Industrie) 201
Wilson G.T. 251
Working capital 182

Yard improvements 184

Zevnik F.C. and Buchanan R.L. 249

This page has been reformatted by Knovel to provide easier navigation.