Final Research Report

Research Project T2695, Task 26

HDM-4 2004

APPLICATION OF HDM-4
IN THE WSDOT HIGHWAY SYSTEM

by

Jianhua Li Stephen T. Muench

Research Assistant Assistant Professor

Joe P. Mahoney George C. White

Professor Research Engineer
Department of Civil and Environmental Engineering
University of Washington

Linda M. Pierce Nadarajah Sivaneswaran

State Pavement Engineer State Pavement Management Engineer
Washington State Department of Transportation

Washington State Transportation Center (TRAC)

University of Washington, Box 354802
1107 NE 45th Street, Suite 535
Seattle, Washington 98105-4631

Prepared for
Washington State Transportation Commission
Departmenst of Transportation
and in cooperation with
U.S. Department of Transportation
Federal Highway Administration

July 2004
1. REPORT NO. 2. GOVERNMENT ACCESSION NO. 3. RECIPIENT’S CATALOG NO.

WA-RD 588.1
4. TITLE AND SUBTITLE 5. REPORT DATE

July 2004
APPLICATION OF HDM-4 IN THE WSDOT HIGHWAY SYSTEM 6. PERFORMING ORGANIZATION
CODE

7. AUTHORS 8. PERFORMING ORGANIZATION

Jianhua Li, Stephen T. Muench, Joe P. Mahoney, George C. White, CODE

Linda M. Pierce, Nadarajah Sivaneswaran

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. WORK UNIT NO.

Washington State Transportation Center

University of Washington, Box 354802 11. CONTRACT OR GRANT NUMBER

University District Building, 1107 NE 45th Street, Suite 535 Research Project
Seattle, Washington (8504-7370 T2695, Task 26
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF REPORT AND PERIOD
COVERED

Research Office
Final Research
Washington State Department of Transportation
Report
Transportation Building, MS 47370 14. SPONSORING AGENCY CODE
Olympia, Washington 98504-7370
Project Manager: Keith Anderson, 360-709-5405
15. SUPPLIMENTARY NOTES

16. ABSTRACT

This project performed extensive study and systematic research on the application of the
Highway Development and Management system (HDM-4) to the Washington State Department
of Transportation (WSDOT) highway system. Data processing, testing, calibration, and analysis
were executed to provide the WSDOT with a new budget planning tool. HDM-4 flexible
pavement deterioration models were calibrated with the assistance of the Washington State
Pavement Management System. A method to calibrate concrete pavement deterioration models
is proposed. This research also explored the application of HDM-4 analysis at the project,
program, and strategic levels. The applications include prediction of pavement conditions
during a defined analysis period, calculation of required budgets for optimal pavement
conditions and maximum economic indicators, establishment of optimized work programs
under varying levels of constrained budgets, and other applications for WSDOT.

17. KEY WORDS 18. DISTRIBUTION STATEMENT

pavement management, pavement performance, models, HDM-4,

vehicle operating costs, pavement preservation, pavement maintenance
19. SECURITY CLASSIF. (of this report) 20. SECURITY CLASSIF. (of this page) 21. NO. OF PAGES 22. PRICE
 DISCLAIMER

The contents of this report reflect the views of the authors, who are responsible

for the facts and the accuracy of the data presented herein. The contents do not

necessarily reflect the official views or policies of the Washington State Transportation

Commission, Washington State Department of Transportation, or Federal Highway

Administration. This report does not constitute a standard, specification, or regulation.

iii
iv
 TABLE OF CONTENTS

1: INTRODUCTION ............................................................................................... 1
1.1: Introduction to the Highway Development and Management System ............... 1
1.1.1: Background................................................................................................ 1
1.1.2: Functions.................................................................................................... 1
1.2: WSDOT Background.......................................................................................... 3
1.3: Research Objectives............................................................................................ 5
1.4: Research Process................................................................................................. 5
1.4.1: Data Preparation ........................................................................................ 5
1.4.2: Calibration ................................................................................................. 6
1.4.3: Output Analysis ......................................................................................... 6
1.5: Report Organization............................................................................................ 7

2: DATA INPUT ...................................................................................................... 8

2.1: Data Sources ....................................................................................................... 8
2.1.1: Introduction to the WSPMS....................................................................... 8
2.1.2: Flexible Pavement Distress Data Conversion............................................ 8
2.1.3: Concrete Pavement Distress Data Conversion .......................................... 13
2.2: Road Networks ................................................................................................... 16
2.3: Vehicle Fleets ..................................................................................................... 18
2.4: Works Standards for Asphalt Concrete-Surfaced Flexible Pavements .............. 19
2.4.1: 45-mm Overlay .......................................................................................... 20
2.4.2: Input Table for 45-mm Mill and Fill ......................................................... 26
2.4.3: Input Table for Pothole Patching ............................................................... 27
2.5: Road Works Standards for Concrete Pavements ................................................ 27
2.5.1: Diamond Grinding ..................................................................................... 28
2.5.2: Reconstruction ........................................................................................... 29
2.6: Road Works Standards for BST-Surfaced Flexible Pavements.......................... 30
2.7: HDM-4 Configuration ........................................................................................ 31
2.7.1: Traffic Flow Patterns ................................................................................. 31
2.7.2: Speed Flow Types...................................................................................... 32
2.7.3: Climate Zones ............................................................................................ 35

3: CALIBRATION OF ROAD DETERIORATION MODELS ......................... 40

3.1: Calibration Classification ................................................................................... 40
3.1.1: Surface Types ............................................................................................ 40
3.1.2: Traffic ........................................................................................................ 41
3.1.3: Climate....................................................................................................... 41
3.2: Calibration Tool: LIMDEP................................................................................. 42
3.3: Calibration of Deterioration Models for Flexible Pavements............................. 43
3.3.1: Primary Modeling Factors ......................................................................... 43
3.3.2: Proposed Calibration Methodology ........................................................... 45
3.3.3: Determination of Calibration Coefficients ................................................ 46
3.3.4: Simplified Deterioration Models for AC-Surfaced Flexible Pavements ... 50
3.3.5: Simplified Deterioration Models for BST-Surfaced Flexible Pavements . 57

v
 3.3.6: Calibration Steps and Results .................................................................... 60
3.3.7: Validation of Modeling Factors................................................................. 64
3.4: Calibration of Deterioration Models for Concrete Pavements ........................... 68
3.4.1: Proposed Calibration Methodology ........................................................... 68
3.4.2: Determination of Calibration Coefficients ................................................ 69
3.4.3: Simplified Pavement Deterioration Models for Concrete Pavements ....... 70
3.4.4: Fatal Errors in HDM-4 Outputs for Concrete Pavements.......................... 71

4: OUTPUT ANALYSIS ......................................................................................... 76

4.1: Project Level Analysis ........................................................................................ 76
4.1.1: Analysis Methods ...................................................................................... 76
4.1.2: Reports Generated...................................................................................... 76
4.1.3: Analysis Results......................................................................................... 77
4.2: Program Level Analysis...................................................................................... 86
4.2.1: Multi-Year Forward Program .................................................................... 87
4.2.2: Life-Cycle Analysis ................................................................................... 89
4.3: Strategic Level Analysis ..................................................................................... 90
4.3.1: Road Network Data ................................................................................... 91
4.3.2: Optimized Road Maintenance and Development Plans ............................ 92
4.3.3: WSDOT Funds .......................................................................................... 92
4.3.4: Road Performance Under Varying Levels of Budgets .............................. 93
4.3.5: Economic Indicators Under Varying Levels of Budgets ........................... 98
4.3.6: Optimized Allocation of Defined Budgets ................................................ 98

5: CONCLUSIONS AND SUMMARY.................................................................. 102

5.1: Conclusions......................................................................................................... 102
5.2: Summary............................................................................................................. 102

REFERENCES......................................................................................................... 104

LIST OF APPENDICES ......................................................................................... 106

vi
 FIGURES

Figure Page

1 Report organization........................................................................................ 7
2 Works effects on roughness in HDM-4 ......................................................... 23
3 Works effects on roughness by WSDOT....................................................... 23
4 Simulation of roughness performance after diamond grinding ..................... 28
5 Speed-flow relationships on multi-lane highways......................................... 33
6 Speed-flow curves with criteria ..................................................................... 33
7 Speed-flow and percentage of time-spent-following flow relationships for
directional segments with base conditions..................................................... 34
8 Six climate zones of Washington state .......................................................... 38
9 Flexible pavement distress interaction........................................................... 43
10 Faulting forecasting by different analysis periods ......................................... 73
11 IRI forecasting by different analysis periods ................................................. 73
12 HDM-4 distress data input interface .............................................................. 74
13 Roughness under two alternatives for concrete pavements ........................... 74
14 Roughness of I-405 MP 13.82-15.17 (D) under four alternatives ................. 78
15 Average roughness of all I-405 flexible pavements ...................................... 79
16 Cracking of SR 21 MP 130.4-153.0 (I) for 40 years ..................................... 80
17 Average roughness of all I-5 flexible pavements for five years .................... 89
18 Average volume capacity ratio of all I-5 flexible pavements for five years.. 89
19 Roughness of all ACPs under varying budgets.............................................. 94
20 Roughness of all BSTs under varying budgets .............................................. 94
21 Surface damage of all ACPs under varying budgets ..................................... 95
22 Surface damage of all BSTs under varying budgets...................................... 95
23 Roughness of all ACPs under WSDOT scenarios ......................................... 97
24 Damaged area of all ACPs under WSDOT scenarios.................................... 97

vii
 TABLES

Table Page

1 Categories of road network input tables ........................................................ 17

2 Values of WP and Elane ................................................................................ 18
3 Maintenance standard for 45-mm overlay ..................................................... 25
4 Maintenance standard for 45-mm mill and fill .............................................. 26
5 Maintenance standard for pothole patching................................................... 27
6 Maintenance standard for diamond grinding ................................................. 29
7 Maintenance standard for reconstruction....................................................... 29
8 Maintenance standard for bituminous surface treatment ............................... 30
9 Volume composition in the commuter traffic flow pattern ........................... 31
10 Volume composition in the inter-urban/rural flow pattern ............................ 31
11 Volume composition in the seasonal flow pattern......................................... 31
12 FFS adjustment for lane width and shoulder width ....................................... 32
13 FFS adjustment for access-point density ....................................................... 32
14 FFS adjustment for lateral clearance.............................................................. 33
15 Speed flow types in Washington state ........................................................... 34
16 Moisture categories........................................................................................ 35
17 Temperature categories.................................................................................. 35
18 Climate zones in HDM-4 ............................................................................... 36
19 WSDOT’s climate zones by county............................................................... 39
20 Climate indices for six climate zones in Washington state............................ 39
21 Climate indices for Western Washington and Eastern Washington .............. 42
22 Climate categories of Washington state......................................................... 42
23 Calibration factors in flexible pavement deterioration models...................... 44
24 Sensitivity classes for RDWE factors ............................................................ 45
25 Values of ICA, ICW, ICT and IRV ............................................................... 49
26 Estimated factors for high ESAL ACPs......................................................... 61
27 Estimated factors for medium ESAL ACPs................................................... 62

viii
28 Estimated factors for low ESAL ACPs.......................................................... 63
29 Estimated factors for all BSTs ....................................................................... 64
30 Climate index for one climate zone of Washington state .............................. 65
31 Maintenance standards for flexible pavements.............................................. 66
32 Traffic composition and annual growth rate for ACPs.................................. 66
33 Validated calibration factors for flexible pavements ..................................... 67
34 Calibration factors used in concrete pavement deterioration models ............ 68
35 Main road network data for concrete pavement tests .................................... 72
36 Current conditions of I-405 MP 13.82-15.17 (D) .......................................... 79
37 Vehicle composition of I-405 MP 13.82-15.17 (D)....................................... 79
38 Pavement conditions of SR 21 MP 130.4-153.0 (I)....................................... 81
39 Traffic composition and growth rates of SR 21 MP 130.4-153.0 (I)............. 81
40 Costs and schedule of 45-mm overlay for SR 405 MP 13.82-15.17 (D)....... 82
41 Costs and schedule of 45-mm mill and fill for I-405 MP 13.82-15.17 (D) ... 82
42 Costs and schedule of pothole patching for I-405 MP 13.82-15.17 (D)........ 83
43 Summary of total annual costs for all I-405 flexible sections ....................... 84
44 Economic indicators for I-405 MP 13.85-15.17 (D)...................................... 85
45 Total economic benefits for all I-405 flexible sections ................................. 86
46 Traffic composition and growth rates of I-5 .................................................. 87
47 Optimized work program for all I-5 flexible sections for five years ............. 88
48 WSDOT historical funds for pavement maintenance .................................... 92
49 WSDOT funding scenarios ............................................................................ 96
50 Economic indicators under varying levels of budgets for all ACPs .............. 98
51 Optimized work program under the current WSDOT budget for ACPs ....... 99

ix
x
 1: INTRODUCTION

1.1: INTRODUCTION TO THE HIGHWAY DEVELOPMENT AND

MANAGEMENT SYSTEM (HDM-4)

1.1.1: Background

The Highway Development and Management System (HDM-4), originally

developed by the World Bank for international use, is a software tool for systematically

addressing these issues. HDM-4 can provide road performance prediction, road

treatment programming, funding estimates, budget allocation, project appraisal, policy

impact studies, and a wide range of special applications. However, its effectiveness is

dependant on its ability to accurately model and predict pavement performance, which is

affected by numerous factors including structural design, materials, construction

variability, traffic, vehicle operating costs, environmental considerations, as well as

maintenance and rehabilitation practices. Therefore, in order to effectively use HDM-4,

its predictive models must be calibrated to local conditions. While Tonga, Thailand,

India, Canada and several other counties have calibrated some HDM-4 models to their

local conditions, there is no evidence that this calibration is valid for Washington State.

To date, there has been no thorough documented calibration and application of HDM-4 in

the U.S. (Kerali et al., 2000b).

1.1.2: Functions

HDM-4 can be used for four basic functions: project analysis, program analysis,

strategic analysis, and research and policy studies. They are briefly described as follows:

1. Project analysis evaluates one or more specific road project or investment options

in short-term planning. By associating costs and benefits, the application analyzes

1
 individual sections with user-selected treatments and determines economic

indicators for different investment options. It provides a detailed economic

appraisal of each option by considering pavement structure performance, life-

cycle predictions of road deterioration and maintenance effects, road-user costs

and benefits, and economic comparisons of project alternatives.

2. Program analysis primarily prioritizes a defined list of candidate road projects

into a one-year or multi-year roadwork and expenditure program under a

constrained budget. A list of candidate road projects is selected as discrete

segments defined by homogeneous physical properties. By using the life-cycle

analysis method or the multi-year analysis method, the optimal combination of

road works options maximizes Net Present Value (NPV) for the whole network,

subject to the sum of treatment costs being less than the available budget.

3. Strategic analysis examines a road network as a whole over the long term. This

typically involves expenditure estimates for road network development and

preservation under various budget and economic scenarios. Typical applications

include fund requirements for specified target road maintenance standards, long-

term forecasts on road network performance under varying levels of budgets,

optimal allocation of funds to sub-networks, and policy studies.

Strategic analysis aggregates individual road segments into various user-defined

categories (e.g., combinations of traffic volume, pavement type and condition,

climate zones). HDM-4 then analyzes each category over a defined time period.

4. Policy studies include funding polices for competing needs, road-user charges for

setting up road funds, impacts of road transport policy changes on energy

2
 consumption, impact of axle load limits, and pavement preservation standards

(Kerali, 2000a).

1.2: WSDOT BACKGROUND

Pavement maintenance is a major and important process for state departments of

transportation. For the Washington State Department of Transportation (WSDOT), this

task covers approximately 18,000 lane-miles of pavements, 3,300 bridges, and 100,000

acres (WSDOT, 2001).

Currently, WSDOT applies pavement preservation on a “least life cycle cost”

basis, but numerous factors such as structural design, materials applied, construction

variability, traffic, environment, and maintenance/rehabilitation types can affect the

prediction capability of pavement performance. Maintenance treatments and pavement

preservation applied on a preventative basis can extend pavement life and lead to more

cost-effective pavement performance, which, in turn, provides a more cost-effective

expenditure of maintenance and preservation funds.

The WSDOT Fiscal Year (FY) 2001-2003 rehabilitation funds were $267.4

million, but the current FY 2004-2005 budget is constrained to $240 million—a reduction

of 10.2%. With constrained budgets, what level of preservation and maintenance will

result in optimized pavement conditions? What are the tradeoffs between

increased/decreased preservation and maintenance budgets?

To address these issues, HDM-4 was introduced to improve the pavement

maintenance/preservation process. The application of HDM-4 to the WSDOT road

network can be beneficial at the project, program, and strategic levels.

3
 At project level, HDM-4 is able to benefit WSDOT by the functions of more

accurate prediction of pavement performance, quantification of pavement

maintenance/preservation benefits, determination of what, when, and where for pavement

maintenance/preservation, and examination of budget tradeoffs and impacts on pavement

performance.

At the program level, HDM-4 can assist WSDOT in planning biennial road works

and expenditure budgets. Once decision makers assign those budgets, HDM-4 can

prioritize desired road works to make optimal use of available funds.

At a strategic level, HDM-4 is beneficial because it can

• forecast long-term road network performance based on proposed levels of

budgets. This can provide decision makers with a clear and convincing picture

of their fund level impacts

• determine budget requirements for a specified set of target road

maintenance/preservation standards, which can provide a baseline optimal

budget level to present to decision makers

• optimize fund allocations to sub-networks (e.g., by functional road

classification or administrative region). This can help WSDOT better allocate

existing funds to achieve the optimal benefits.

These capabilities will be immediately beneficial to WSDOT. At the research and

policy levels, WSDOT and decision makers can see the potential impacts of policy

changes, including the impact of changes to the axle load limit, maintenance/preservation

trigger standards, or pavement design standards (Kerali, 2000a).

4
1.3: RESEARCH OBJECTIVES

Although HDM has been applied in more than 100 countries, most recorded

applications have been in developing countries. In recent years, however, many

industrialized countries have begun to adopt its economic approach and principles, but no

thorough application has been documented in the U.S.

This research used HDM-4 to analyze the WSDOT highway facilities for

systematic maintenance and rehabilitation. This analysis includes developing predictive

strategies for the selection and timing of maintenance/preservation treatments, developing

a detailed examination of road works standards, and developing a process for analyzing

varying budget levels and their pavement impacts for the WSDOT highway system.

Because HDM-4 predicts future road performance from current and/or historical

conditions, the reliability of its results depends upon how well input data represent actual

conditions and how well HDM-4 predictions model actual behavior (Kerali, 2000a).

As the first thorough application of HDM-4 in the U.S., this research mainly

focused on acquisition of data and calibration of deterioration models. The results will be

a view of the effectiveness of various WSDOT preservation funding levels.

1.4: RESEARCH PROCESS

WSDOT implementation of HDM-4 involves four major tasks: 1) data transfer

from the Washington State Pavement Management System (WSPMS) and independent

data gathering, 2) model calibration, and 3) output analyses.

1.4.1: Data Preparation

WSPMS contains a historical archive of WSDOT highway pavement condition

5
data. Most of the required input data for HDM-4 were obtained from the WSPMS either

directly or indirectly. An automated method of data transfer between the WSPMS and

HDM-4 were developed. This method is capable of directly transferring applicable

values from an existing WSPMS database to a new HDM-4 database, and manipulating

existing values from WSPMS to input into HDM-4. Since WSPMS data are dynamic,

automatic data transfer is critical to HDM-4’s current and future usability.

A large portion of the required HDM-4 data is essentially independent of WSPMS

data. These data, such as climate zones, traffic-flow schemes and default pavement

conditions, were developed and input into HDM-4.

1.4.2: Calibration

Although HDM-4 models were developed with broad-based applicability in

diverse climates and conditions, the accuracy of the predicted pavement performance still

depends on the extent of calibration, which was applied to adapt the default HDM-4

models to local conditions.

1.4.3: Output Analysis

HDM-4 has three levels of analysis: project level, program level, and strategic

level. Each has different functions for pavement management. The HDM output results

include costs, traffic, pavement deterioration conditions, pavement preservation effects,

road-user effects, environmental effects, and various program and strategic reports

(Kerali et al., 2000b). Further research, such as cost-benefit analyses, was carried out to

confirm the feasibility of the final results.

6
1.5: REPORT ORGANIZATION

Figure 1 shows the organization of this report.

Introduction
Chapter 1

Data Input
Chapter 2

Calibration of Road
Deterioration Models
Chapter 3

Output Analysis
Chapter 4

Conclusions and Summary

Chapter 5

Figure 1: Report organization

Chapter 2 introduces ways to modify HDM-4 configurations and accommodate

WSDOT inputs.

Chapter 3 addresses detailed procedures for deterioration model calibration.

Flexible and concrete pavements were modeled separately in HDM-4. Ways to adapt

WSDOT’s data to the HDM-4 deterioration models are also discussed.

Chapter 4 describes HDM-4 applications at the analysis, program, and strategic

levels for WSDOT.

Chapter 5 draws the main conclusions of this research and summarizes the

report.

7
 2: DATA INPUT

2.1: DATA SOURCES

2.1.1: Introduction to the WSPMS

The WSPMS is both a historical archive of WSDOT highway pavement condition

data and a tool to schedule and predict WSDOT road network preservation demands.

WSDOT collects and analyses the pavement distress conditions and updates the database

every year. The most up-to-date version is WSPMS 2003, which records the road

condition data of 2002.

The WSPMS organizes data into analysis units and project units. An analysis unit

contains homogeneous pavement sections that are structurally uniform (same type of

materials and thicknesses). Project units are established according to similar pavement

performance criteria and made up of one or more analysis units (WSPMS, 2003).

Most calibration variables and road network input data for HDM-4 were derived from

project units. Some specific calibration variables were generated from analysis units.

To convert the WSPMS data into an acceptable form for HDM-4 input, transformation

rules are necessary. A wide range of data types are required for HDM-4, which include

vehicle fleet characteristics, road maintenance and improvement standards, unit costs,

and economic analysis factors. These are stored in the HDM-4 internal database. The

physical attributes of the selected data objects must be exported to a data exchange file

format defined for HDM-4, which is dBase file format [*.DBF (IV)]. Additional details

on inputting data into HDM-4 are discussed in sections 2.1.2 and 2.1.3.

2.1.2. Flexible Pavement Distress Data Conversion

HDM-4 models six types of flexible pavement distress: cracking, rut depth,

8
roughness, potholes, edge-break and raveling. Since WSDOT does not measure and

record edge-break data, this research assumes edge break has a value. WSPMS represents

rut depth, raveling, and roughness in the same way required by HDM-4 so that data can

be used directly. WSDOT and HDM-4 characterize cracking using different methods,

thus a conversion method from WSDOT to HDM-4 was developed. This is explained in

section 2.1.2.1

2.1.2.1:WSDOT Characterization of Cracking

WSDOT uses the pavement structural condition (PSC) to characterize flexible

pavement cracking (Kay et al., 1993). WSDOT considers the following items for PSC

determination. Note that since there are two wheel paths, a 100-foot lane actually has

200 feet of wheel paths for measurement purposes. Therefore, a measurement of 100

percent would mean that both wheel path lengths are completely cracked.

• Alligator (fatigue) cracking (ACEC): the percentage of the wheel path

length that contains alligator cracks. Since there are two wheel paths, a 100-

foot lane has 200 feet of wheel paths for measurement (i.e., a measurement of

100 percent means that both wheel path lengths are completely cracked).

Severity is divided into three levels:

AC1 = hairline

AC2 = spalled

AC3 = spalled and pumping

• Longitudinal cracking (LCEC): the percentage of a 100-foot section length

that contains longitudinal cracking. Severity is divided into three levels:

LC1 = less than ¼ inch wide

9
 LC2 = more than ¼ inch wide

LC3 = spalled

• Transverse cracking (TCEC): the number of transverse cracks per 100-foot

section length. Severity is divided into three levels:

TC1 = less than ¼ inch wide

TC2 = more than ¼ inch wide

TC3 = spalled

• Patching (PTEC): WSDOT considers patching to be a structural defect and

quantifies it using the percentage of the wheel path lengths that contains

patches. The extent of patching is represented as the percentage of both wheel

paths. Severity is divided into three levels:

PT1 = BST patching

PT2 = blade (cold mix) patching

PT3 = full depth patching

The total Equivalent Cracking (EC) is the sum of ACEC, LCEC, TCEC and

PTEC. The calculation of PSC is:

ACEC = AC3 + 0.445(AC2)1.15 + 0.13(AC1)1.35

LCEC = (0.1LC3) + 0.445(0.1LC2)1.15 + 0.13(0.1LC1)1.35

TCEC = (0.8TC3) + 0.445(0. 8TC2)1.15 + 0.13(0. 8TC1)1.35

PTEC = (PT3) + 0.445(0.75PT2)1.15 + 0.13(0.75PT1)1.35

EC = ACEC + LCEC + TCEC + PTEC

PSC = 100 – 15.8(EC)0.5 (2.1)

10
 2.1.2.2: HDM-4 Characterization of Cracking

HDM-4 uses the following factors to characterize flexible pavement cracking:

• Total cracked area of the carriageway (ACA): the percentage of the total

carriageway surface area that is cracked.

• Percentage of total cracking that is wide structural cracking (ACW): the

percentage of ACRA that constitutes wide structural cracking, defined as

cracks greater than 3 mm wide.

• Transverse thermal cracking (ACT): the number of transverse thermal

cracks per kilometer of roadway length. The number of transverse thermal

cracks is capped by setting a time (in years), since thermal crack initiation, to

reach the maximum amount of thermal cracking.

Unlike WSDOT, which uses wheel path length, HDM-4 quantifies cracking by

the percentage of total carriageway area that is cracked.

2.1.2.3: Conversion Method

In order to make HDM-4 results meaningful for WSDOT, HDM-4 must replicate,

as closely as possible, the methods that WSDOT uses to trigger works standards.

Fortunately, rut depth and roughness use the same measurement definitions in both

systems. Unfortunately, this is not the case with cracking. The following presents

conversion methodologies for the various HDM 4 cracking categories:

All Structural Cracking (ACA)

WSMPS categories of alligator (fatigue) cracking and longitudinal cracking will

be added to give a total area of cracking. No distinction between severity levels is made.

The wide structural cracking variable will account for severity throughout the area

11
concerned. In order to convert a linear crack to an area, each wheel path is assumed to be

3 feet wide, and a fully cracked 1.5 x 1.5 ft2 pavement block contains a fully developed

longitudinal and a fully developed transverse crack. Therefore:

100% AC = (3+3) longitudinal cracks + (100/1.5) transverse cracks

= 6 * (100 ft.) + 66.7 * (3 ft. +3 ft.)

= 10 * (100 ft.)

= 10 longitudinal cracks (full 100-foot section length)

= 1000% of longitudinal cracking

And,

100%AC = 1000%LC

AC = 0.1LC (2.2)

Assume that two wheel paths defined by WSDOT constitutes 50% of the total

pavement area (2 wheel paths = 6 feet wide, assume the lane is 12 feet wide). Therefore,

WSDOT percentages are divided by 2 to match HDM-4 percentages.

AC1 + AC 2 + AC 3 LC1 + LC 2 + LC 3
ACA = + 0.1* (2.3)
2 2

Wide Structural Cracking Area (ACW)

WSPMS categories of alligator (fatigue) cracking and longitudinal cracking at

severity level “medium” or “high” are added to give a total area of wide structural

cracking. Since WSPMS “medium” severity level is defined by cracks at least 6 mm

wide, all “medium” severity cracks count as wide structural cracking. Since WSPMS

“high” severity level cracks are more advanced and in poorer condition than “medium”

severity level cracks, all “high” severity cracks count as wide structural cracking.

12
 AC 2 + AC 3 LC 2 + LC 3
ACA = + 0.1* (2.4)
2 2

Transverse Thermal Cracking (ACT)

The WSPMS category of transverse cracking can be converted to an HDM-4

definition for “transverse thermal cracking” by applying a constant to convert from

cracks-per-100 feet to cracks-per-kilometer:

ACT = 32.81(TC1 + TC2 + TC3)

100 ft = 0.03048 km

1 km = 3280.83 ft. (2.5)

Patching

WSPMS uses patching as a contributor to PSC, but HDM-4 models patching

independent of cracking. To avoid confusion, the contribution of patching to PSC will be

ignored. This is a reasonable assumption because the incidence of pavement patching on

WSDOT roads is relatively low (Muench, 2003b).

2.1.3: Concrete Pavement Distress Data Conversion

HDM-4 characterizes concrete pavement deterioration differently than the

WSPMS. Therefore, a method of converting WSPMS distress characterization to HDM-4

characterization must be developed. This section outlines conversion methodologies for

pavement distress.

2.1.3.1: WSDOT Characterization

WSDOT uses PSC to characterize concrete pavement deterioration (Kay et al.,

1993), and the following items are considered in PSC determination:

13
• Slab cracking (CREC): the percentage of slabs that are cracked. Severity is

divided into three levels:

CR1 = percentage of slabs with 1 crack per panel

CR2 = percentage of slabs with 2 or 3 cracks per panel

CR3 = percentage of slabs with 4 or more crack per panel

• Joint and crack spalling (JSEC): the percentage of joints and cracks that

are spalled. Severity is divided into three levels:

JS1 = percentage of joints/cracks with 1/8 – 1 inch spalls

JS2 = percentage of joints/cracks with 1 – 3 inch spalls

JS3 = percentage of joints/cracks with > 3 inch spalls

• Pumping and blowing (PMEC): the percentage of joints and cracks that are

pumping. Severity is divided into three levels:

PM1 = percentage of joints/cracks pumping (slight shoulder depression,

little or no staining)

PM2 = percentage of joints/cracks pumping (moderate shoulder

depression, obvious staining)

PM3 = percentage of joints/cracks pumping (severe shoulder depression,

significant staining)

• Faulting and settlement (FLTEC): the percentage of all panels that exhibit

joints or cracks that are faulting or settling. Three severity levels are:

FLT1 = percentage of panels with 1/8 to ¼ inch faulting or settlement

FLT2 = percentage of panels with ¼ to ½ inch faulting or settlement

FLT3 = percentage of panels with > ½ inch faulting or settlement

14
• Patching (PTEC): the percentage of panels in the travel lane that are

patched. Severity is divided into three levels:

PT1 = percentage of panels patched (1 to 9 percent of surface covered)

PT2 = percentage of panels patched (10 to 24 percent of surface covered)

PT3 = percentage of panels patched (> 25 percent of surface covered)

• Raveling and scaling (RSEC): the percentage of panels that exhibit raveling,

scaling, map cracking, or crazing. Severity is divided into three levels:

RS1 = percentage of pavement surface with slight raveling or scaling

RS2 = percentage of pavement surface with moderate raveling or scaling

RS3 = percentage of pavement surface with severe raveling or scaling

PSC is calculated by Equation 2.6:

CREC = CR3 + 0.24(CR2)1.16 + 0.0054(CR1)1.84

JSEC = 0.075(JS3)1.14 + 0.0061(JS2)1.27 + 0.0034 (JS1)1.03

PMEC = 0.0069(PM1 + PM2 + PM3)1.45

FLTEC = FLT3 + 0.0915(FLT2)1.46 + 0.00115(FLT1)2.32

PTEC = 0.103(PT3)1.19 + 0.0079(PT2)1.55 + 0.00194 (PT1)1.57

RSEC = 0.052(RS3)1.29 + 0.0159(RS2)1.18 + 0.0014 (RS1)1.18

EC = CREC + JSEC + PMEC + FLTEC + PTEC + RSEC

PSC = 100 – 18.6(EC)0.43

(2.6)

2.1.3.2: HDM-4 Characterization

HDM-4 uses the following to characterize concrete pavement deterioration:

• Cracked slabs (CRACKSLABS): the percentage of slabs that are cracked.

15
 • Faulting (FAULT): the average fault height in millimeters.

• Spalled joints (SPALL_JNTS): the percentage of joints that are spalled.

Spalling is assumed to be 75 to 100 mm

• Panel failures per km (FAILURESKM): the number of failed panels per

kilometers of roadway.

2.1.3.3: Proposed Conversion Methods

HDM-4 counts low and medium severity levels of cracking and high severity

level of spalling. For faulting, WSPMS measures it in terms of the percentage of slabs

that are faulted, but HDM-4 measures it in terms of the average height of the faults. Thus,

the conversion from WSPMS to HDM-4 is listed as follows (Muench, 2003c).

CRACKSLABS (%) = (CR1+CR2)

FAULT (mm) = 4.7625*FLT1+9.525*FLT2+12.7*FLT3

SPALL_JNTS (%) = JS3. (2.7)

2.2: ROAD NETWORKS

In HDM-4, road networks store characteristics of road sections, which are the

fundamental unit of the analysis. Each road network contains a detailed listing of each

road section chosen for analysis. Sections, links, and nodes are the three data entities

supported within the road network:

• Sections are lengths of roads over which physical characteristics are

reasonably constant. Sections defined in this research will follow the

definition of project units in WSPMS, where each section has uniform

pavement layers, geometry, and traffic conditions over its entire length.

16
 • Links comprise one or more sections over which traffic is reasonably

constant. Links defined in this research are state route numbers.

• Nodes are used to mark intersections where links have a significant change in

traffic, carriageway characteristics, or administrative boundaries.

Road network data must be in a dBASE IV format to be imported into HDM-4.

Bridge decks were not considered for this research because WSDOT treats bridges in a

separate management system. Because of the data scale limited by HDM-4 computation

power, road network data of all state highways are divided into five categories according

to different pavement types, surface types, and/or 18 kip Equivalent Single Axle Loads

(ESALs). These are listed in Table 1.

Table 1: Categories of road network input tables

ESAL Number of
Pavement Type Surface Material*
(ESAL / year/Elane) Cases
Flexible AC High: (500000+) 374
Flexible AC Medium: (250000,500000] 512
Flexible AC Low: (0, 250000] 1595
Flexible BST -- 412
Concrete Concrete pavement -- 615
*AC denotes asphalt concrete, which is hot mix asphalt, BST indicates bituminous
surface treatments.

Because this research only concerns WSDOT highways, in this report, ACP

represents the AC-surfaced WSDOT flexible highway, BST stands for the BST-surfaced

WSDOT flexible highway, concrete pavements indicates the WSDOT concrete highway.

Elane is the effective number of lanes for the road section:

1
Elane =
WP (2.8)

where WP is the lane distribution factor. WP accounts for the distribution of traffic loads

on the design lane, which is defined by the 1993 AASHTO Guide. Values adopted by this

17
research are listed in Table 2.

Table 2: Values of WP and Elane

Number of Lanes Lane distribution factor (WP)
Elane
in Each Direction 1993 ASSHTO Guide Values adopted
1 100 100 1
2 80-100 90 1.111
3 60-80 70 1.429
4 50-75 65 1.538
5 - - 2
≥6 - - 3

Appendix B1 addresses the data conversion methods from the WSPMS and data

definitions for project level analysis. Appendices B2, B3, B4, B5 and B6 are the road

network input for all high ESAL ACPs, medium ESAL ACPs, low ESAL ACPs, BSTs,

and concrete pavements, correspondingly.

2.3: VEHICLE FLEETS

Vehicle fleets are used to store vehicle characteristics for calculating speeds,

operating costs, travel times, and other vehicle effects. WSDOT uses a simplified version

of the Federal Highway Administration (FHWA) vehicle classification system, which

includes four categories:

1. passenger car
2. single-unit truck. Includes FHWA classes 4, 5, 6 and 7. WSDOT assumes
0.40 ESALs per truck.
3. double-unit truck. Includes FHWA classes 8, 9 and 10. WSDOT assumes
1.00 ESALs per truck.
4. train. Includes FHWA classes 11, 12 and 13. WSDOT assumes 1.75
ESALs per truck (FHWA, 2001).

18
 Appendices C1, C2, C3, and C4 define all vehicle fleet input data for cars, single

units, double units, and trains, respectively. Most of the data are not directly available

from any single data source. Data sources for this research are listed in Appendix C5.

When the data are imported to HDM-4, they can be stored in the HDM-4 internal

database as Appendix C6.

2.4: WORKS STANDARDS FOR ASPHALT CONCRETE (AC) SURFACED

FLEXIBLE PAVEMENTS

Road work standards can be used in all three analysis levels. Two kinds of work

are defined in HDM-4 for flexible pavements. Improvement standards comprise

pavement reconstruction, lane addition, lane upgrading, partial widening, and

realignment. Maintenance standards are applied to meet specific objectives that are

related to functional characteristics of the road network system. Works for crack sealing,

fog seal, edge repair, patching, drainage, edge repair, overlay and flexible pavement

reconstruction are included in maintenance standards. Unlike HDM-4, WSDOT classifies

road works into two categories: 1) preservation and 2) maintenance, which are both

defined as maintenance works in HDM-4. For AC-surfaced flexible pavements, three

basic types of road works are normally adopted by WSDOT (WSDOT, 2002). They are

1. 45-mm overlay

2. 45-mm mill and fill

3. pothole patching

To keep all HDM-4 inputs and output reports consistent, this research names all

three road works standards as maintenance standards.

19
 2.4.1: 45-mm Overlay

Although other thicknesses have been used, WSDOT typically uses a 45-mm

thickness (WSDOT, 2002).

2.4.1.1: Triggers of Rehabilitation Efforts

WSDOT Characterization

WSDOT triggers its rehabilitation efforts (termed “preservation”) on the basis of

any of the following conditions (Kay, 1993):

• pavement structural condition (PSC) ≤ 50

• rut depth ≥10 mm

• roughness ≥ 3.5 m/km

This scheme is heavily weighted toward PSC; therefore, most rehabilitation

efforts are triggered by the PSC value. This was done intentionally because rehabilitation

triggered by PSC will result in roads that are rehabilitated earlier in their life cycles

before an accumulation of defects causes substantial increases in roughness and overall

structural deterioration.

HDM-4 Characterization

HDM-4 uses the following to characterize flexible pavement cracking:

• Total cracked area of the carriageway (ACA): the percentage of the

carriageway surface area that is cracked.

• Percentage of total cracking that is wide structural cracking (ACW): the

percentage of ACRA that constitutes wide structural cracking, which is

defined as cracks > 3 mm wide.

20
 • Transverse thermal cracking (ACT): the number of transverse thermal

cracks per kilometer of roadway length. The number of transverse thermal

cracks is capped by setting a time (in years) since thermal crack initiation, to

reach the maximum number of thermal cracking.

Unlike WSDOT, which uses wheel path area, HDM-4 quantifies cracking by the

percentage of the total carriageway area that is cracked.

Conversion of WSDOT Rehabilitation Standards to HDM-4

(a) Cracking

WSDOT triggers rehabilitation when PSC = 50. Therefore, HDM-4 maintenance

trigger levels should simulate this as closely as possible. The only cracking trigger

available in HDM-4 is “total carriageway cracking.” Therefore, several assumptions need

to be made to convert a PSC trigger level to a total cracking area level:

1. Most cracking is at the AC1 and LC1 levels.

2. When AC2/3 or LC2/3 cracking occurs, it is generally at a much lower level.

3. Transverse cracks do not trigger maintenance.

Straightforward calculations show a relationship between PSC and cracking.

WSDOT triggers rehabilitation at varying lengths of cracking, depending upon the

breakdown of severity, so rehabilitation is triggered when:

AC1 + AC2 + AC3 + 0.1(LC1 +LC2 + LC3) = 20% (2.9)

Because HDM-4 total pavement area = 2 x (WSDOT wheel path area), trigger

rehabilitation in HDM-4 when:

ACA ≥ 10% (2.10)

21
 (b) Rutting

This value is set the same as WSDOT. HDM-4 triggers rehabilitation when the

average rut depth ≥ 10 mm.

(c) Roughness

Same as WSDOT, HDM-4 triggers rehabilitation when IRI ≥ 3.5 m/km.

In summary, the triggered distress values to set up road works for WSDOT are

one of the following:

• total carriageway cracked ≥ 10%

• rut depth ≥ 10 mm

• roughness ≥ 3.5 m/km (AL-Omari and Darter, 1992).

2.4.1.2: Work Effects of 45-mm Overlay

The HDM-4 work-effects model for overlays includes several models for the

effects of overlay rehabilitation. These models can be divided into the following:

• User defined: Models that accept user input to determine overlay effects

• Derived: Models that use existing data to determine overlay effects.

HDM-4 models overlay effects on roughness, rutting, surface texture, and skid

resistance. The following discussion focuses on the models for each effect and presents

the selected model for use in WSDOT road network.

(a) Roughness

A generalized bilinear model is used for roughness reduction effects in HDM-4.

Assuming a bilinear relationship (a line with an inflection point), users may enter their

own values for key constants in the relationship.

22
 Reduction in roughness, dRI
a3 dR2

a1 dR1

0 a0 a2
Roughness before overlay, RIb0
Figure 2: Works effects on roughness in HDM-4
where,

dRI: roughness after overlay (m/km)

a0: minimum roughness after overlay (m/km)

a1: slope of the first line

a2: intersection point of the two lines (bilinear mode) (m/km)

a3: slope of the second line. A0 to a3 are user-definable factors.

Figure 3 shows WSDOT data for overlays from 1999 to 2001.

2.50

IRI Reduction = 0.5353(Before IRI) - 0.2807

2.00 2
R = 0.5755
Reduction in IRI (m/km)

1.50

1.00

0.50

0.00

-0.50
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
Before IRI (m/km)

Figure 3: Works effects on roughness by WSDOT

[Source: WSDOT Material Laboratory, 2003]

23
 The model in Figure 3 is linear. Although HDM-4 only allows users to specify a

bilinear model (see Figure 2), a linear one can be made by setting a1 = a3 and a2 = a0.

Therefore, a0 = 0.5244, a1 = 0.5353, a2 = 0.5244, and a3 = 0.5353.

(b) Rutting

A derived method resets rut depth to a user-defined percentage of pre-overlay rut

depth. On the basis of WSDOT experimental results after overlay rehabilitation, the value

of 0 percent is used in HDM-4 rutting reduction models for WSDOT case studies

(Corley-Lay, 1998).

(c) Surface Texture

The HDM-4 derived method resets surface texture to a predetermined value,

depending upon the surface material (= 0.7 mm in almost all cases). HDM-4 surface

texture models for an overlay rehabilitation uses 0.7 mm. (Jayawickrama et al., 1998).

(d) Skid Resistance

HDM-4 resets surface texture to 0.50 mm in all cases. The value is also used in

HDM-4 surface texture models for overlay rehabilitation (Muench, 2003a).

2.4.1.3: Input Table for 45-mm Overlay

A 45-mm overlay without milling is typical for medium or low traffic sections in

Washington State. On the basis of WSDOT’s criteria and the previously stated works-

effect analysis, the HDM-4 input data for the 45-mm overlay maintenance standard is

listed in Table 3.

24
 Table 3: Maintenance standard for 45-mm overlay
General Name: 45 mm Overlay
Short Code: 45 OVER
Intervention Type: Responsive
Design Surface Material: Asphalted Concrete
Thickness: 45 mm
Dry Season a: 0.44
CDS: 1
Intervention Responsive Criteria: Total cracked area ≥ 10%
Rutting ≥ 10 mm
IRI ≥ 3.5 m/km
Min. Interval: 1
Max. Interval: 9999
Last Year: 2099
Max Roughness: 16 m/km
Min ADT: 0
Max ADT: 500,000
Costs Overlay
Economic: 19 dollars/m2 *
Financial: 19 dollars/m2 *
Patching
Economic: 47 dollars/m2 *
Financial: 47 dollars/m2 *
Edge Repair
Economic: 47 dollars/m2
Financial: 47 dollars/m2
Effects Roughness: Use generalized bilinear model
a0 = 0.5244
a1 = 0.5353
a2 = 0.5244
a3 = 0.5353
Rutting: Use rutting reset coefficient = 0
Texture Depth: Use default values (0.7 mm)
Skid Resistance: Use default value (0.5 mm)
[*Costs are derived from data provided by WSDOT]

25
2.4.2: Input Table for 45-mm Mill and Fill

Table 4: Maintenance standard for 45-mm mill and fill

General Name: 45 mm Mill & Fill
Short Code: 45MF
Intervention Type: Responsive
Design Surface Material: Asphalted Concrete
Depth of Mill: 45 mm
Thickness: 45 mm
Dry Season a: 0.44
CDS: 1
Intervention Responsive Criteria: Total cracked area ≥ 10%
Rutting ≥ 10 mm
IRI ≥ 3.5 m/km
Min. Interval: -
Max. Interval: -
Last Year: 2099
Max Roughness: 16 m/km
Min ADT: 0
Max ADT: 500,000
Costs Overlay
Economic: 30 dollars/m2 *
Financial: 30 dollars/m2 *
Patching
Economic: 47 dollars/m2 *
Financial: 47 dollars/m2 *
Edge Repair
Economic: 47 dollars/m2 *
Financial: 47 dollars/m2 *
Effects Roughness: Use generalized bilinear model
a0 = 0.5244
a1 = 0.71
a2 = 0.5244
a3 = 0.71
Rutting: Use rutting reset coefficient = 0
Texture Depth: Use default values (0.7 mm)
Skid Resistance: Use default value (0.5 mm)
[*Costs are derived from data provided by WSDOT]

26
 This type of maintenance is normally used on Interstate routes with high traffic

conditions in Washington State. Unlike the 45-mm overlay, the 45-mm mill and fill has

improved effects on cracking, so the reduction in roughness is greater. Expressed in the

same way as the 45-mm overlay, a1 and a3 are equal to 0.71, instead of 0.5353 based on

work reported by Raymond (2002). Additionally, the work costs of milling and filling are

added to the overlay costs. The inputs to HDM-4 are listed in Table 4.

2.4.3: Input Table for Pothole Patching

WSDOT patches a pothole as soon as it appears. The criteria and related factors

of input to this maintenance standard are listed in Table 5.

Table 5: Maintenance standard for pothole patching

General Name: Pothole Patching
Short Code: POTPAT
Intervention Type: Responsive
Intervention Responsive Criteria: Potholing ≥ 0.01 no./km
Min. Interval: -
Max. Interval: -
Last Year: 2099
Max Roughness: 16 m/km
Max Quantity: 5000 m2/km/year
Min ADT: 0
Max ADT: 500,000
Costs Economic: 47 dollars/m2 *
Financial: 47 dollars/m2 *
Effects Distress Repaired: 100 (%) (potholing only)
[*Costs are derived from data provided by WSDOT]

2.5 ROAD WORKS STANDARDS FOR CONCRETE PAVEMENTS

Road works standards adopted by WSDOT are diamond grinding, dowel bar

retrofit, and reconstruction. Unlike flexible pavements, all road works for concrete

27
pavements are defined as maintenance standards by HDM-4. For the same reason

described in section 2.4, all three of the above standards were considered to be

maintenance standards in this research. In the current HDM-4 version 1.3, dowel bar

retrofit can intervene only as scheduled activity—not as a responsive intervention.

2.5.1: Diamond Grinding

In HDM-4, diamond grinding rehabilitation works are triggered by faulting. The

criterion for triggering diamond grinding is faulting ≥ 6 mm.

HDM-4 measures the works effects of diamond grinding by roughness after

works, but no direct roughness value is available. Figure 4 shows selected concrete

sections in WSPMS 2003 that have been rehabilitated by diamond grinding. Each point

represents one section.

Roughness Performance after Diamond Grinding

6
Roughness (m/km) )

5
4
3
2
1
0
0 5 10 15 20 25 30 35 40
Age (year)

Figure 4: Simulation of roughness performance after diamond grinding

A linear model was taken to regress roughness progression. On the basis of the

data in Figure 4, the linear model is regressed as IRI = 0.0312*Age + 1.4283. This model

represents the road performance after diamond grinding. The intercept is the roughness

measured immediately following the diamond grinding. roughness = 1.4283 m/km.

28
 Table 6: Maintenance standard for diamond grinding
General Name: Diamond Grinding
Short Code: DG
Intervention Type: Responsive
Intervention Responsive Criteria: Faulting ≥ 6 mm
Min. Interval: 1
Max. Interval: 100
Max Roughness: 16 m/km
Max ADT: 500,000
Costs Economic: 1 dollars/m2/mm *
Financial: 1 dollars/ m2/mm *
Effects Grinding thickness: 8 mm
Roughness: 1.4283 m/km
[*Costs are derived from data provided by WSDOT]

2.5.2: Reconstruction

HDM-4 triggers reconstruction works similarly to WSDOT, when the total

carriageway is cracked ≥ 10 percent, and it measures the works effects of reconstruction

by the percentage of the slab replaced after works. For WSDOT, all slabs will be

replaced. Thus, the works standard for reconstruction is shown in Table 7:

Table 7: Maintenance standard for reconstruction

General Name: Reconstruction
Short Code: RECON
Intervention Type: Responsive
Intervention Responsive Criteria: % of cracked slabs ≥ 10%
Min. Interval: 1
Max. Interval: 100
Max Roughness: 16 m/km
Max ADT: 500,000
Costs Economic: 169.9 dollars/m2 *
Financial: 169.9 dollars/m2 *
Effects Slabs replaced: 100%
[*Costs are derived from data provided by WSDOT]

29
2.6: ROAD WORKS STANDARDS FOR BST-SURFACED FLEXIBLE
PAVEMENTS

BST-surfaced flexible pavements are usually used in low traffic locations and

normally rehabilitated every five to ten years in accordance with WSDOT budget

conditions and planning. The only maintenance standard is bituminous surface treatment.

HDM-4 can trigger this work only as a result of cracking. The related input factors are

listed in Table 8.

Table 8: Maintenance standard for bituminous surface treatment

Name: Bituminous Surface Treatment
General Short Code: BSTCRA
Intervention Type: Responsive
Surface Material: Double Bituminous Surface Dressing
Thickness: 12.5mm
Design
Dry Season a: 0.2
CDS: 1
Responsive Criteria: total cracked area ≥ 10%
Min. Interval: 1
Max. Interval: 100
Max Roughness: 16 m/km
Max ADT: 100,000
BST Economic: 2.04 dollars/m2 *
Intervention BST Financial: 2.04 dollars/m2 *
Patching Economic: 47 dollars/m2 *
Patching Financial: 47 dollars/m2 *
Edge Repair Economic: 47 dollars/m2 *
Edge Repair Financial: 47 dollars/m2 *
Crack Seal Economic: 8.5 dollars/m2 *
Crack Seal Financial: 8.5 dollars/m2 *
Roughness: Use user defined method
Roughness: 2 m/km
Effects Mean rut depth: 0 mm
Texture Depth: 0.7mm
Skid Resistance: 0.5mm
[*Costs are derived from data provided by WSDOT]

30
2.7: HDM-4 CONFIGURATION

Configuration data are traffic flow patterns, speed flow types, and climate. These

data are collected to adapt the basic characteristics to WSDOT conditions.

2.7.1: Traffic Flow Patterns

Traffic flow patterns model the temporal variations in traffic. HDM-4 uses traffic

flow pattern data to model congestion effects on vehicle speeds and vehicle operation

costs. The traffic flow types are defined by three steps:

1. Choose typical routes from the WSPMS.

2. Find the routes and mile posts (MPs) in the table.

3. Calculate the volume percentage of different daily time periods.

Table 9: Volume composition in the commuter traffic flow pattern

Period Hours Per Year Hourly Volume % of AADT
Morning Peak (6:30-10:00am) 1277.5 0.093 32.5
Off Peak (10:00am-3: 30pm) 2007.5 0.032 17.5
Evening Peak (3:30-7:30pm) 1460 0.1 40
Night (7:30pm-6:30am) 4015 0.009 10
Total 8760 1.0005 100
[Source: http://www.wsdot.wa.gov/mapsdata/tdo/PDF%20and%20ZIP%20Files/
peak2000.pdf , 2000]

Table 10: Volume composition in the inter-urban/rural flow pattern

Period Hours Per Year Hourly Volume % of AADT
Day Time (7:30am-7:00pm) 4197.5 0.07 80
Night Time (7:00pm-7:30am) 4562.5 0.016 20
Total 8760 1.005 100
[Source: http://www.wsdot.wa.gov/mapsdata/tdo/PDF%20and%20ZIP%20Files/
peak2000.pdf , 2000]

Table 11: Volume composition in the seasonal flow pattern

Periods Hours Per Year Hourly Volume % of AADT
Day Time (7:30am-7:00pm) 4197.5 0.072 82.5
Night Time (7:00pm-7:30am) 4562.5 0.014 17.5
Total 8760 1.003 100
[Source: http://www.wsdot.wa.gov/mapsdata/tdo/PDF%20and%20ZIP%20Files/
peak2000.pdf , 2000]

31
 2.7.2: Speed Flow Types

Speed flow types model the effects of traffic volume on speeds. HDM-4 uses

speed flow data to conduct economic analyses for road capacity improvements. For

individual project segments, the free-flow speeds (FFS) are to be adjusted for the width

of lanes, access point density, and lateral clearance. Lane widths are adjusted for two-lane

roads as illustrated in Table 12. Adjustments for lateral clearance and access point density

for four-lane highways can be made on the basis of tables 13 and 14, respectively. Given

figures 5, 6, and 7 for four-lane and two-lane highways, the changes result in more

accurate values for free-flow speeds and, subsequently, for free-flow capacities and for

ultimate capacities.

Table 12: FFS adjustment for lane width and shoulder width
Reduction in FFS (km/hr)
Lane Width (m)
Shoulder Width (m)
[0.0, 0.6) [0.6, 1.2) [1.2, 1.8) [1.8, + ∞ )
[2.7, 3.0) 10.3 7.7 5.6 3.5
[3.0, 3.3) 8.5 5.9 3.8 1.7
[3.3, 3.6) 7.5 4.9 2.8 0.7
[3.6, + ∞ ) 6.8 4.2 2.1 0
[Source: Highway Capacity Manual 2000]

Table 13: FFS adjustment for access-point density

Access Points per km Reduction in FFS (km/hr)
0 0
6 4
12 8
18 12
≥24 16
[Source: Highway Capacity Manual 2000]

32
 Table 14: FFS adjustment for lateral clearance
Four-lane Highways Six-lane Highways
Total Lateral Clearance Reduction in FFS Total Lateral Clearance Reduction in FFS
(m) (km/h) (m) (km/h)
3.6 0 3.6 0
3 0.6 3 0.6
2.4 1.5 2.4 1.5
1.8 2.1 1.8 2.1
1.2 3 1.2 2.7
0.6 5.8 0.6 4.5
0 8.7 0 6.3
[Source: Highway Capacity Manual 2000]

Figure 5: Speed-flow relationships on multi-lane highways

[Source: Highway Capacity Manual 2000, Exhibit 12-1]

Figure 6: Speed-flow curves with criteria

[Source: Highway Capacity Manual 2000, Exhibit 21-3]

33
Figure 7: Speed-flow and percentage of time-spent-following flow relationships for
directional segments with base conditions
[Source: Highway Capacity Manual 2000, Exhibit 12-7]

For basic conditions, FFS’s are approximately 11km/hr above the 65 and 70

km/hr speed limit, and approximately 8 km/hr above the 80-90 km/hr speed limit. This

can be used to calculate approximate free-flow speeds, the free flow capacities, and the

ultimate capacities. The speed-flow types of studied highways are listed as Table 15.

Table 15: Speed flow types in Washington State

Max
Ultimate Free-Flow Nominal Jam
FFS Acceleration
Speed-Flow Type Capacity Capacity Capacity Speed
(Km/hr) Noise
(PCSE/lane/hr) (PCSE/lane/hr) (Qnom) (km/hr)
(m/s2)

Four-Lane Road 1900-2200 800-1300 0.95 70-110 65-80 0.6

Two-Lane Narrow 850-1600 150 0.7 60-100 30-60 0.7
Two-Lane Standard 1600-1650 150-250 0.9 70-100 40-60 0.65
Two- Lane Wide 1600-1700 150-250 0.9 70-110 40-60 0.6
The above table does not differentiate between urban and rural highways. In

general, the free-flow speed in rural freeways is about 8km/hour higher than those of

urban highways. However, this is the result of varying speed limits (Zhang et al., 2001).

34
 2.7.3: Climate Zones

Climate data is used to model its effects on road performance and user costs.

2.7.3.1: Climate Data Parameter Definitions

HDM-4’s climate data input provides two categories: moisture and temperature.

The default moisture categories include arid, semi-arid, sub-humid, humid, and per-

humid. Temperature categories comprise tropical, subtropical-hot, subtropical-cool,

temperate-cool, and temperate-freezing (Odoki, 2000). Tables 16 and 17 list the

definitions of those categories.

Table 16: Moisture categories

Moisture Thornthwaite Annual
Category Description Moisture Precipitation
Index (mm)
Arid Very low rainfall, high evaporation -100 to –61 <300
Semi-arid Low rainfall -60 to –21 300 to 800
Sub-humid Moderate rainfall or strongly seasonal rainfall -20 to +19 800 to 1600
Humid Moderate warm seasonal rainfall +20 to +100 1500 to 3000
Per-humid High rainfall or very many wet-surface days >100 >2400
[Source: HDM-4 Series, Volumes 5. The World Road Association (PIARC), 1999]

Table 17: Temperature categories

Temperature Temperature
Description
Category Range (°C)
Tropical Warm temperatures in small range 20 to 35
Sub-tropical – hot High day, cool night temperatures, hot-cold seasons -5 to 45
Sub-tropical – cool Moderate day temperatures, cool winters 10 to 30
Temperate – cool Warm summer, shallow winter freeze -20 to 25
Temperate – freeze Cool summer, deep winter freeze -40 to 20
[Source: HDM-4 Series, Volumes 5. The World Road Association (PIARC), 1999]

Among the 25 moisture and temperature combinations, two of them are not

applicable to Washington State. Therefore, a matrix was developed from the

classifications to create 23 unique climate zones, as shown in Table 18.

35
 Table 18: Climate zones in HDM-4
Climate Subtropical Subtropical Temperate Temperate
Zone
Tropical
-hot -cool -cool -freezing
Arid X X X X X
Semi-arid X X X X X
Sub-humid X X X X X
Humid X X X X X
Per-humid X X X N/A N/A

Each temperature and moisture category possesses data parameter definitions. In

HDM-4, the following data parameters are required (Odoki, 2000).

Moisture

• Mean monthly precipitation (mm/month)

• Thornthwaite Moisture Index (MI)

100* SWAT − 60* DWAT

MI = I h − 0.6* I a =
NWAT (2.12)

where:

MI: Thornthwaite Moisture Index

Ih: humidity index

Ia: aridity index

SWAT: excess of water (mm)

DWAT: water deficiency (mm)

NWAT: necessary water (mm)

• Dry season duration as a proportion of a year

(365-p)/365 (2.13)

where p is the number of days of precipitation.

36
 Temperature

• Mean annual temperature (°C)

• Temperature range (TRANGE) (°C): the mean monthly ambient temperature

range. Its calculation is based on the temperature ranges for each of the twelve

months of the year, hence the difference between the maximum and minimum

temperature of each month.

• Number of days per year the temperature exceeds 32°C

• Freezing index (FI) (°C - days): the difference between the mean ambient

temperature and 0°C (degrees per day). The freezing index is negative when

the ambient temperature is below 0°C and positive otherwise.

The freezing index is calculated as:

ndays
FI = ∑ ABS[MIN (TEMP, 0)]
i =1 (2.14)

where:

FI: freezing index

TEMP: temperature (°C)

ndays: number of days in one freezing season

• Percentage of driving times on snow-covered roads (%)

• Percentage of driving times on water-covered roads (%)

2.7.3.2: Six Climate Zones in Washington State

WSDOT Regions were used to define six climate zones: Northwest, Olympic,

Southwest, North Central, Eastern, and South Central. The partition of six climate zones

is shown in Figure 8 and follows regional boundaries.

37
 Figure 8: Six climate zones of Washington state
[Source: http://www.wsdot.wa.gov]

The above six climate zone definitions are based on historical weather data from

observation stations. Several stations in each zone were selected so that each station has

at least one of the following characteristics:

• high moisture

• low temperature

• high road density

• loosely based on the six climate-zone classifications.

To look up the climate data for a segment of roadway, the county data listed in the

WSPMS should be cross-referenced. Table 19 lists WSDOT’s climate zones by county.

38
 Table 19: WSDOT’s climate zones by county
North South
Zone Olympic Northwest Eastern Southwest
Central Central
Clallam San Juan Okanogan Ferry Pacific Kittitas
Jefferson Island Chelan Stevens Wahkiakum Yakima
Grays Pend Walla
Whatcom Douglas Lewis
Harbor Oreille Walla
Mason Skagit Grant Lincoln Cowlitz Benton
County
Kitsap Snohomish Spokane Clark Franklin
Thurston King Adams Skamania Klickitat
Pierce Whitman Columbia
Garfield
Asotin

Climate data from each county are listed in Appendix A [Source: http://www.

wrcc.dri.edusummary/mapwa.html]. A uniform moisture and temperature for each county

were developed by averaging the individual climate factors.

Table 20: Climate indices for six climate zones in Washington State
Moisture Temperature
Zone Mean
Duration Mean Temp. Days Freeze
Name Moisture monthly
of dry Temp. Range T>32ºC Index
Index precipitation
season (ºC) (ºC) (89.6ºF) (ºC-days)
(mm)
Olympic 144.77 0.5386 111.98 9.77 9.21 0 11.73
Northwest 75.09 0.5924 65.89 10.60 8.64 0 0.99
North
26.02 0.8296 20.26 10.13 12.30 9.4 465.33
Central
Eastern 58.64 0.7052 41.20 8.57 12.66 2 562.77
Southwest 163.46 0.5036 127.95 10.73 9.92 0 3.13
South
28.24 0.785 22.48 10.21 13.53 11.8 400.03
Central

39
 3: CALIBRATION OF ROAD DETERIORATION MODELS

Road deterioration models simulate future changes from current and/or historical

road conditions, and each model develops and progresses at different rates in different

environments. Therefore, the models should be calibrated to Washington State local

conditions. Coefficients included in the models are used to adjust the deterioration rates

to different types of surface material. Furthermore, the models include a number of user-

definable calibration factors to fit the models to the local conditions. The reliability of

these factors depends on the accuracy of the input data and how well the prediction

models represent local conditions (Odoki et al., 2000).

Thus, calibration of the models includes three primary elements: data preparation,

determination of the optimal calibration factors by regression, and validation.

3.1: CALIBRATION CLASSIFICATION

In order to properly calibrate deterioration models, homogeneous road sections,

(in terms of physical attributes and conditions) must be identified. Therefore, different

pavement groups are defined so that calibration factors for each group can be obtained.

These calibration factors are related to the conditions of climate and environment, traffic,

pavement history, road geometry, pavement structural characteristics, and material

properties. The primary explanatory variables for flexible pavement deterioration were

evaluated to determine the variables needed to define the pavement groups.

3.1.1: Surface Types

HDM uses two general classes of models for pavement performance prediction

based on the road surface classes: incremental models for flexible pavements and

40
absolute models for concrete pavements. Incremental models predict the change in

condition from an initial state, and absolute models predict the condition at a particular

point in time. Furthermore, flexible and concrete pavements manifest various kinds of

distresses, so they must be calibrated independently.

WSDOT flexible pavements use two types of surface material: asphalt concrete

(AC) and 2) bituminous surface treatment (BST). The geometric standards and material

properties for these two flexible pavement types vary considerably, so they were

calibrated in different classes. Therefore, based on the surface materials, WSDOT

highways were divided into three calibration classes:

• AC-surfaced flexible pavements (ACPs)

• BST-surfaced flexible pavements (BSTs)

• Concrete pavements (Odoki et al., 2000).

3.1.2: Traffic

Traffic is measured by ESALs in this research. Specifically, annual quantities of

ESALs divided by Elane. Because concrete pavements are normally used for high traffic

and BSTs for low traffic loads, only ACPs were divided into three classes according to

traffic usage: high, medium, and low as listed in Table 1.

3.1.3: Climate

Climatic conditions for flexible pavements have been classified into six climate

regions according to the moisture and temperature classifications in Tables 16 and 17.

The Cascade mountain range separates Washington State into two significantly

different climate zones, Western Washington (WW) and Eastern Washington (EW). WW

includes the Northwest, Olympic, and Southwest regions. EW includes the North Central,

41
Eastern, and South Central regions. By averaging the climate indices of the three zones of

each class, the corresponding climatic values are shown in Table 21.

Table 21: Climate indices for Western Washington and Eastern Washington
Climate Zone Western Washington Eastern Washington
Moisture Index 127.77 37.64

Moisture Duration of dry season 0.54 0.77

Mean monthly
101.94 27.98
precipitation (mm)
Mean Temperature (ºC) 10.37 9.64
Temp Range (ºC) 9.26 12.83
Temperature
Days T>32ºC (89.6ºF) 0.00 7.73
Freeze Index (ºC-days) 5.29 476.04

On the basis of the climate indices in Table 21, as well as temperatures and

moisture classes defined by HDM-4, WW and EW can be categorized as Table 22.

Table 22: Climate categories of Washington state

Climate Category Moisture Category Temperature Category

Western Washington Per-humid Temperate cool

Eastern Washington Semi-arid Temperate freeze

Table 22 is used for choosing factors to be used in the calibration equations.

Those factors are described in sections 3.3.3 and 3.3.4.

3.2: CALIBRATION TOOL: LIMDEP

LIMDEP is econometric software developed by Econometric Software, Inc., that

can estimate linear and nonlinear regression models, as well as limited and qualitative

dependent variable models for cross-section, time-series, panel-data, and user-specified

models (Greene, 1995). It was used to estimate the optimal calibration factors for large

amounts of local condition data based on models generated by HDM-4.

42
3.3: CALIBRATION OF DETERIORATION MODELS FOR FLEXIBLE
PAVEMENTS

3.3.1: Primary Modeling Factors

Eight types of flexible pavement distresses are modeled in HDM-4. They are 1)

cracking, 2) raveling, 3) potholing, 4) edge-breaking, 5) rutting, 6) surface texture, 7)

skid resistance and 2) roughness. The first five types have high-to-medium sensitivity

level effects on the prediction models; therefore, emphasis is given to the related

calibration factors. Because of limited data and low sensitivity to the prediction models,

skid resistance and surface texture were not calibrated. Rather, the HDM-4 default values

were used. Interactions among the distresses, as defined by HDM-4, are illustrated in

Figure 9. All distresses ultimately have effects on roughness, so roughness can represent

the condition as a whole (Odoki, 2000).

Structural
Crack Crack Deformation
Rutting
Initiation Progression

Roughness
Potholing

Ravelling Ravelling
Enviromental
Initiation Progression
Effects

Figure 9: Flexible pavement distress interaction

[Source: HDM-4 Volume-5]

The key variables used in the road-deterioration models are related to the

conditions of climate and environment, traffic, pavement history, road geometry,

pavement structural characteristics, and material properties. These are described in

Chapter 2. HDM-4 uses 24 factors to represent these conditions, as listed in Table 23.

43
 Table 23: Calibration factors in flexible pavement deterioration models
Deterioration model Calibration factor
Wet/dry season SNP ratio Kf
Drainage deterioration factor Kddf
Drain life factor Kdrain
All structural cracking-initiation Kcia
Wide structural cracking-initiation Kciw
All structural cracking- progression Kcpa
Wide structural cracking- progression Kcpw
Transverse thermal cracking - initiation Kcit
Transverse thermal cracking - progression Kcpt
Rutting-initial densification Krid
Rutting - structural deterioration Krst
Rutting - plastic deformation Krpd
Rutting - surface wear Krsw
Ravelling - initiation Kvi
Ravelling - progression Kvp
Pothole - initiation Kpi
Pothole - progression Kpp
Edge - break Keb
Roughness - enviromental coefficient Kgm
Roughness - SNPK Ksnpk
Roughness - progression Kgp
Texture depth - progression Ktd
Skid resistance Ksfc
Skid resistance - speed effects Ksfcs

Depending upon their different effects on pavement performance, these factors

are categorized into four sensitivity classes by HDM-4, as shown in Table 24.

44
 Table 24: Sensitivity classes for RDWE factors
Sensitivity Impact
Impact Factor
Level Elasticity
Structural number
Modified structural number
S-I High >0.5 Traffic volume
Deflection
Roughness
Annual loading
Age
All cracking area
S-II Moderate 0.2-0.5 Wide cracking area
Roughness-environment factor
Cracking initiation factor
Cracking progression factor
Subgrade CBR (with SN)
Surface thickness (with SN)
Heavy axles volume
Potholing area
S-III Low 0.05-0.2
Rut depth mean
Rut depth standard deviation
Rut depth progression factor
Roughness general factor
Deflection with (SNC)
Subgrade compaction
S-IV Negligible <0.05 Rainfall (with Kge)
Raveling area
Raveling factor
(Note: When SN is given, deflection has negligible impacts.)
[Source: HDM-4, Volume 5]
Calibrating HDM-4 for WSDOT focused first on S-I level factors, with a lesser

priority given to lower sensitivity level factors. Some low, or negligible, sensitivity

factors were not calibrated. Instead, a default value of 1.0 was used, especially when

related data were not available.

3.3.2: Proposed Calibration Methodology

Calibration of a deterioration model is intended to find the adjustment factors that

45
will make model predictions as accurate as possible. The general expression used for the

calibration of deterioration models is:

∆Y = K y *f (Ya , a 0 , a1 , a 2 , ... , a n , X1 , X 2 , ... , X m ) (3.1)

where:

Ky: calibration factor of distress type Y

∆Y : incremental change of distress type Y in the analysis year

Ya: deterioration value at the beginning of the analysis year

a i: default coefficient values given by models and decided by factors of climate and

environment, traffic, pavement history, road geometry, pavement structural

characteristics, or material properties.

Xi: input factors of climate, traffic, pavement history, road geometry, pavement

structural characteristics, and material properties.

f(): non-linear function, which may include various non-linear operations such as

logarithm, exponential and polynomial etc.

If the values of Ya , a 0 , a1 , a 2 , ... , a n , X1 , X 2 , ... , and X m are given, f () acts as one

independent variable of the dependent variable ∆Y . The non-linear models are converted

to simple linear relationship between the dependent variable ∆Y and the independent

variable f () . Only the total roughness deterioration model cannot be transformed to a

linear model. Section 3.3.4 illustrates the specific methodology used to determine the

corresponding factors of roughness. Then, road deterioration models of each calibration

category were regressed independently in LIMDEP.

3.3.3: Determination of Calibration Coefficients

Proposed values for HDM-4 calibration variables are as follows (Odoki, 2000):

46
• CDS (Construction defects indicator for bituminous surfacing): use 1.0, since

the surface condition of the studied sections is supposed to be at an optimal

binder content.

• CDB (Base construction defects indicator): use 0.5 for construction defect

conditions as given values in HDM-4.

• COMP (Relative compaction (%)): use 100 for the assumption of a full

compliance in all layers.

• CRT (Crack retardation time due to maintenance (years)): use 0, since

WSDOT seldom does preventive treatments such as fog seal or joint sealing.

• CRP (Cracking progression retardation due to preventative treatment):

CRP = 1-0.12*CRT = 1 (3.2)

• RRF (Raveling retardation factor due to maintenance): use 1.0 as defaults.

• PASS (Annual number of vehicles with studded tires in one direction

(1000s)): assume 3 %of the annual number of vehicle passes has studded tires

for Western Washington and 8% for Eastern Washington. PASS can be

calculated by Equation (3.3):

365*3% * AADT
WW: PASS = 3
10 * Number Of Direction
365*8% * AADT
EW: PASS = 3
10 * Number Of Direction
(3.3)
• YE4 (Annual number of equivalent standard axles (millions/lane): WSDOT

assumes 0.4 ESALs per single-unit truck, 1.00 ESALs per double-unit truck,

and 1.75 ESALs per train. Then, YE4 can be calculated as (WSPMS, 2003):

47
 0.4* Single-unit + 1.00* Double-unit + 1.75* Train
YE 4 =
106 * ELANE
(3.4)
• YAX (Annual number of axles of all motorized vehicle types in the analysis

year (millions/lane)): assume standard axles of passenger cars (PC), single-

unit trucks, double-unit trucks, and trains are 2, 2, 5, and 6, respectively.

2* PC + 2* Single-unit + 5* Double-unit + 6* Train

YAX = (3.5)
106 * ELANE

• HSNEW (Thickness of the most recent surfacing (mm)): it is the overlay

thickness, which is 45 mm for WSDOT.

• AGE

AGE1 (Preventive treatment age (years)):

= Analysis year – Year of the last prevention

AGE2 (Surfacing age (years))

= Analysis year – Year of the last surfacing

AGE3 (Rehabilitation age(years))

= Analysis year – Year of the last rehabilitation

AGE4 (Base construction age (years))

= Analysis year – Year of the last construction

HDM-4 constrains AGE1/AGE2/AGE3 ≥ AGE4 and condition year ≥

Year of last construction, where condition year equals 2002, since all distress

data of this research were measured in 2002.

• Time to initiation of cracking or raveling:

ICA (Initiation time of all structural cracking)

ICW (Initiation time of wide structural cracking)

ICT (Initiation time of thermal cracking)

48
 IRV (Initiation time of raveling)

Initiation of cracking or raveling is the time when the distress area is 0.5%

of the total carriageway surface area. Ideally, the calculation steps are:

1. Select the specific distress data, surface age, and section length from the

road network input table defined in Chapter 2.

2. Group sections with AC or BST surface.

3. Select AC sections with surface age ≤ 40 years, since longer AC service

lifes are unrealistic. Select BST sections with surface age no greater than

30 years, since BST pavements rarely can serve longer than 30 years.

4. Choose sections whose distress area is 0.5% of the total carriage way area.

5. Weigh surface ages by section length.

A range from 0.4% to 0.6% was chosen instead of 0.5%. Cracking data of

AC sections were grouped by climate zones of WW and EW because cracking

and raveling are highly sensitive to climate. Table 25 illustrates the results.

Table 25: Values of ICA, ICW, ICT and IRV

ICA ICW ICT IRV
Surface Material Climate
(year) (year) (year) (year)
WW 8.95 11.21 10.06
AC 6.33
EW 5.54 7.00 5.70
BST All 3.86 5.67 4.33 4.89

• CCT (Coefficient of thermal cracking): CCT is used to predict ICT for

different climate conditions. CCT of WW equals 100, and CCT of EW is 2.

• NCTeq (Maximum number of thermal cracks (no/km)): use 0 for WW and 20

for EW.

49
 • Teq (Time since initiation to reach NCTeq): use 50 for WW and 7 for EW

(Odoki et al., 2000).

3.3.4: Simplified Deterioration Models for AC-Surfaced Flexible Pavements

HDM-4 classifies AC-surfaced flexible pavements into five categories, while

WSDOT highways take two: 158 sections of AMGB (asphalt mix on granular base), and

1714 sections of AMAP (asphalt mix on asphalt pavements). The coefficient (not the

calibration factors) values proposed by HDM-4 were selected base on surface materials

(Odoki, 2000).

3.3.4.1: Wet/Dry Season SNP Ratio

Use 1.0 as a drainage factor for the good drainage condition of the WSDOT

routes. Use 0.0 as the pothole area at the start of the analysis year because potholes will

be patched as soon as they appear as a practice of WSDOT. Then,

f = K f ⎡⎣1 − 0.075 (1 − e −0.01MMP ) (1 + 0.02 ACRAa ) ⎤⎦

(3.6)

where:

Wet season SNP

f =
Dry season SNP . Use 1.0 since season does not have much of an effect on SNP in

Washington state for unstabilized base and thick, hot mix asphalt layers (normally thicker

than 150 mm).

Kf: Calibration factor for wet/dry season SNP ratio.

ACRAa: Total area of carriageway cracked at the beginning of the analysis year.

ACRAa=ACAa+ACTa

50
MMP: Mean monthly precipitation (mm). Western Washington uses 101.94 and

Eastern Washington uses 27.98, as determined in Chapter 2.

3.3.4.2: Cracking

(a) Initiation of All Structural Cracking (ICA)

In general, WSDOT uses unstabilized base material, so the related road

deterioration model is simplified by using the proposed coefficients.

If HSOLD=0 (that is original surfacing)

⎛ ⎛ YE 4 ⎞
0.14 SNP −17.1⎜ ⎟⎞
⎝ SNP 2 ⎠
ICA = K cia ⎜ 4.21e ⎟ (3.7)
⎜ ⎟
⎝ ⎠
If HSOLD>0 (that is overlays)

⎪⎧ ⎡ ⎛ ⎞ ⎤ ⎪⎫
⎛ YE 4 ⎞
0.14 SNP −17.1⎜ ⎟
⎝ SNP ⎠
ICA = K cia ⎨⎢ MAX ⎜ 4.21e (3.8)
2
,1.125 ⎟ ⎥ ⎬
⎪⎩ ⎣ ⎝ ⎠ ⎦ ⎪⎭
where:

Kcia: calibration factor for initiation of all structural cracking.

(b) Initiation of Wide Structural Cracking (ICW)

Since HDM-4 surface type AMAP (asphalt mix on asphalt pavements) is

representative of more than 90% of WASOT ACPs, coefficients proposed by HDM-4 for

AMAP were used.

ICW = kciw ( 2.04 + 0.98 ICA ) (3.9)

where:

Kciw: calibration factor for initiation of all structural cracking.

(c) Progression of All Structural Cracking (ACA)

WSDOT measures cracking in wheel paths, but HDM-4 measures it in the total

roadway. Assume that the wheel paths are 50 percent of the total roadway area. Because

51
of WSDOT’s definition, ACA can never be greater than 100 percent, ACA in HDM-4

can never be greater than 50 percent, and then

SCA=ACAa (3.10)
If ACAa>0,
⎡ 1
⎤
(
dACA = kcpa ⎢ 0.2996 + ACAa 0.28 ) 0.28 − ACAa ⎥ (3.11)
⎣ ⎦
Otherwise:
⎡ 1
⎤
(
dACA = kcpa ⎢ 0.2996 * MAX (0, MIN ⎡⎣( AGE 2 − ICA ) ,1⎤⎦ ) + ACAa 0.28 ) 0.28
− ACAa ⎥ (3.12)
⎣ ⎦
where:

ACAa = MAX (ACAa, 0.5) (3.13)

dACA: incremental changes of ACA during the analysis year (%);

ACAa: ACA at the start of the analysis year;

Kcpa: calibration factor for initiation of all structural cracking.

(d) Progression of Wide Structural Cracking (ACW)

As ACA, ACW of Washington State routes can never be greater than 50%, so

SCW =ACWa (3.14)

If ACWa > 0

⎡ ⎛ 1
⎞⎤
dACW = K cpw MIN ⎢( ACAa + dACA − ACWa ) , ⎜ 1.161 + ACWa 0.45 ( ) 0.45 − ACWa ⎟ ⎥ (3.15)
⎣ ⎝ ⎠⎦
Otherwise:

⎡ ) , ⎛⎜ (1.161 * MAX {0, MIN [ AGE 2 − ICW ] ,1} + ACW ⎞⎤

1

dACW = K cpw MIN ( ACA

⎢⎣
+ dACA − ACWa
0.45
) 0.45
− ACWa ⎟
⎝ ⎠⎦⎥
a a

(3.16)
where:

ACWa = MAX (ACWa, 0.5) (3.17)

dACW: incremental change of ACW during the analysis year;

52
ACWa: ACW at the start of the analysis year.

(e) Initiation of Transverse Thermal Cracking (ICT)

If HSOLD=0

ICT = K cit * MAX (1, CCT ) = K cit * CCT (3.18)

Otherwise:

ICT = K cit MAX [1, ( CCT − 0.1)] = K cit ( CCT − 0.1) (3.19)

where:

Kcit: calibration factor for initiation of transverse thermal cracking.

(f) Progression of Transverse Thermal Cracking (NCT)

If HSOLD=0

⎧⎪ ⎡ ⎛ 2 NCTeq ( AGE 3 − ICT − 0.5 ) ⎞ ⎤ ⎫⎪

dNCT = K cpt MAX ⎨0, MIN ⎢( NCTeq − NCTa ) , ⎜ ⎟ ⎥ ⎬ * δ tT (3.20)
⎜ ( eq ) ⎟⎥ ⎪
2

⎩⎪ ⎢⎣ ⎝ T ⎠⎦ ⎭

If HSOLD>0

⎧ ⎡ ⎛ 2 NCT ( AGE 3 − ICT − 0.5 ) ⎞ ⎤⎫

dNCT = K cpt * MIN ⎨( NCT )
− NCTa , MAX ⎢ MIN (0.25 PNCT , ( PNCT − NCT )), ⎜ ⎟ , 0 ⎥ ⎬ * δ tT
eq

⎩ ⎣ ⎝ (T ) 2

⎠ ⎦⎭
eq a

eq

(3.21)

where:

dNCT = 20*dACT

PNCT = 20*PACT (3.22)

{
If ACTa>0, δ tT = 1 Otherwise, δ tT = MAX 0, MIN ⎡⎣( AGE 2 − ICT ) ,1⎤⎦ }
Kcpt: calibration factor for progression of thermal cracking

dNCT: incremental change in number of thermal cracking in analysis year (no/km)

53
dACT: incremental change in ACT in analysis year (%)

PNCT: number of thermal cracks before last overlay (no/km)

NCTa: number of thermal cracks at the start of analysis year.

3.3.4.3: Rut Depth

(a) Initial Densification

The initial densification depends upon the degree of relative compaction of the

base and selected subgrade layers. The construction and compact works are quite good in

Washington state, so Krid = 0.

(b) Structural Deterioration -- Krst

If ACA+ACT=0

( )
∆RDST=K rst 1.1291SNP −1.14YE 40.11 = K rst X 1
(3.23)

Otherwise:

∆RDST=K rst ⎡⎣1.1291SNP −1.14YE 40.11 + 0.0000248SNP −0.84YE 40.14 MMP1.07 ACX a1.11 ⎤⎦ = Krst * X 1
(3.24)
where:

Krst: calibration factor for structural deformation;

∆RDST : incremental change in structural deformation in the analysis year (mm);

ACXa: area of indexed cracking at the beginning of the analysis year (%);

ACXa=0.62ACAa+0.39ACWa (3.25)

MMP: mean monthly precipitation (mm/month). MMP of WW equals 101.94, and

EWequals 27.98.

(c) Plastic Deformation

( )
∆RDPD = K rpd 2.46YE 4 Sh −0.78 HS 0.71 = K rpd X 2 (3.26)

54
where:

Krpd: calibration factor for plastic deformation;

∆RDPD : incremental change in plastic deformation in the analysis year (mm);

Sh: speed of trucks (km/h), assume Sh=SpeedLimit-15;

HS: total thickness of bituminous surface, HS=HSOLD+HSNEW.

(d) Surface Wear

Most roads in Washington are seldom salted, then

∆RDW = K rsw ⎡⎣0.0000248 PASS * W −0.46 S 1.22 ⎤⎦ = K rsw * X 3 (3.27)

where:

Krsw: calibration factor for surface wear;

∆RDW : incremental change in rut depth for studded tires in the analysis year (mm);

W: road width that includes carriageway and shoulder (m);

S: average traffic speed, assuming it equals the speed limit.

(e) Total Rut Depth

If AGE4<=1,

∆RDM = ∆RDPD + ∆RDW = K rpd X 2 + K rsw X 3 (3.28)

Otherwise:

∆RDM = ∆RDSTuc + ∆RDPD + ∆RDW = K rst X 1 + K rpd X 2 + K rsw X 3 (3.29)

where:
∆RDM : incremental change in rut depth in the analysis year (mm).

3.3.4.4: Raveling

(a) Initiation of Raveling (IRV)

(
IRV = K vi 100 e −0.156YAX ) (3.30)

55
where:
Kvi: calibration factor for raveling initiation.
(b) Progression of Raveling (ARV)

Similarly to ACA, Washington state routes can never have sections with ARV

greater than 50 percent, so

⎣( )
dARV = K vp ⎡ ( 0.36 + 1.8YAX ) δ tv + ARVa 0.352 − ARVa ⎤
1/ 0.352
(3.31)
⎦
where:

If ARVa >0, δ tv = 1; (3.32)

Otherwise, δ tv = MAX {0, MIN ⎡⎣( AGE 2 − IRV ) ,1⎤⎦} (3.33)

ARVa = MAX (ARVa, 0.5) (3.34)

Kvp: calibration factor for raveling progression;
DARV: incremental change in ARV during the analysis year (%).

3.3.4.5: Factors Not Calibrated

This research emphasizes high sensitivity level factors. For factors that have low

or negligible sensitivity levels and factors that necessary data are not available, default

values of 1.0 given by HDM-4 were used. These factors are

• pothole initiation and progression - Kpi

• edge-break - Keb

• skid resistance - Ksfc

• skid resistance of speed effects - Ksfcs

• texture depth progression - Ktd

• drainage life factor - Kddf

An exception is Ksnpk, the calibration factor for Structural Number of Pavement

(SNP). It describes the cracking effects on the pavement structure. Because the structural

56
conditions of all Washington state routes are fairly good, this research will use 0 for Ksnpk

instead of the default value 1.0.

3.3.4.6: Roughness

Use Ksnpk = 0, then

∆RI = K gp [ ∆RI s + ∆RI c + ∆RI r + ∆RI t ] + ∆RI e

= K gp [ ∆RI s + ∆RI c + ∆RI r ] + 0.035 * K gm RI a

= K gp ⎡⎣134e (1 + SNP ) YE 4 + 0.0066∆ACRA + 0.088 * ∆RDS ⎤⎦ + 0.035 * K gm RI a

0.035 K gm AGE 3
−5

(3.35)

where:

∆RI s : Roughness increase during the analysis year due to structural deterioration

∆RI c : Roughness increase during the analysis year due to cracking

∆RI r : Roughness increase during the analysis year due to rutting

∆RI t : Roughness increase during the analysis year due to potholing, use 0
∆RI : Roughness increase during the analysis year
K gp : Calibration factor for roughness progression

K gm : Calibration factor for environmental

∆RDS : Increase in standard deviation of rut depth during the analysis year:
∆RDS = RDSb − RDS a
RDSb = ⎡⎣ MAX ( 0.3, 0.9 − 0.04 RDM b ) ⎤⎦ RDM b
RDS a = ⎡⎣ MAX ( 0.3, 0.9 − 0.04 RDM a ) ⎤⎦ RDM a
(3.36)

3.3.5: Simplified Deterioration Models for BST-Surfaced Flexible

Pavements

Some of the default coefficients for BSTs are given differently than those for

ACPs. The related models are listed below, and variables used have the same meaning as

the variables for ACPs (Odoki, 2000).

57
3.3.5.1: Initiation of All Structural Cracking

If HSOLD=0 (that is original surfacing)

⎛ ⎛ YE 4 ⎞
−20.7 ⎜ ⎟⎞
⎝ SNP 2 ⎠
ICA = K cia ⎜13.2e ⎟ (3.37)
⎜ ⎟
⎝ ⎠
If HSOLD>0 (that is overlays)

⎧⎪ ⎡ ⎛ ⎛ YE 4 ⎞
−20.7 ⎜ ⎟ ⎞ ⎤ ⎪⎫
ICA = K cia ⎨⎢ MAX ⎜ 13.2e ⎝ SNP ⎠ , 9.9 ⎟ ⎥ ⎬ (3.38)
2

⎩⎪ ⎣ ⎝ ⎠ ⎦ ⎭⎪

3.3.5.2: Initiation of Wide Structural Cracking

If HSOLD=0

ICW = kciw MAX ⎡⎣( 2.66 + 0.88 ICA ) ,1.16 ICA⎤⎦ (3.39)

If HSOLD>0

ICW = kciw (1.85 + ICA ) (3.40)

3.3.5.3: Progression of All Structural Cracking

The ACA of Washington state routes can never be greater than 50 percent, so

SCA=ACAa (3.41)
If HSOLD=0

If ACAa>0,

⎡ 1
⎤
(
dACA = kcpa ⎢ 0.5632 + ACAa 0.32 ) 0.32 − ACAa ⎥ (3.42)
⎣ ⎦

Otherwise:

⎡ 1
⎤
(
dACA = kcpa ⎢ 0.5632 * MAX (0, MIN ⎡⎣( AGE 2 − ICA ) ,1⎤⎦ ) + ACAa 0.32 ) 0.32
− ACAa ⎥
⎣ ⎦
(3.43)
If HSOLD>0

If ACAa>0,

58
 ⎡ 1
⎤
(
dACA = kcpa ⎢ 0.8194 + ACAa 0.34 ) 0.34 − ACAa ⎥ (3.44)
⎣ ⎦
Otherwise:

⎡ 1
⎤
(
dACA = kcpa ⎢ 0.8194 * MAX (0, MIN ⎡⎣( AGE 2 − ICA ) ,1⎤⎦ ) + ACAa 0.34 ) 0.34
− ACAa ⎥
⎣ ⎦
(3.45)

3.3.5.4: Progression of Wide Structural Cracking

The ACW of Washington state routes can never be greater than 50 percent, so

SCW =ACWa (3.46)

If HSOLD=0

If ACWa > 0

dACW = K cpw MIN ⎡( ACAa + dACA − ACWa ) , 0.625 + ACWa 0.25

⎣ (( )
4
− ACWa ⎤ (3.47) )⎦
Otherwise:

dACW = K cpw MIN ⎡⎣( ACA

a
+ dACA − ACWa ) , ( ( 0.625 * MAX {0, MIN [ AGE 2 − ICW ] ,1} + ACW a
0.25
)
4
− ACWa )⎤⎦
(3.48)

If HSOLD>0

If ACWa > 0

⎡ ⎛ 1
⎞⎤
dACW = K cpw MIN ⎢( ACAa + dACA − ACWa ) , ⎜ 1.19 + ACWa 0.35 ( ) 0.35 − ACWa ⎟ ⎥
⎣ ⎝ ⎠ ⎦ (3.49)

Otherwise:

⎡( ACA ⎛( ⎞⎤
1

dACW = K cpw MIN + dACA − ACWa ) , ⎜ 1.19 * MAX {0, MIN [ AGE 2 − ICW ] ,1} + ACW ) 0.35
0.35
− ACWa ⎟
⎢⎣ a
⎝
a
⎠⎦⎥
(3.50)

3.3.5.5: Initiation of Transverse Thermal Cracking

ICT = K cit * MAX (100, CCT ) = 100 K cit (3.51)

59
 3.3.5.6: Plastic Deformation of Rut Depth
'
∆RDPD = 0 ⇒ K rpd can be any value.

3.3.5.7: Total Rut Depth

If AGE4<=1,

∆RDM = ∆RDW = K rsw X 3 (3.52)

Otherwise:

∆RDM = ∆RDSTuc + ∆RDW = K rst X 1 + K rsw X 3 (3.53)

3.3.5.8: Initiation of Raveling

(
IRV = K vi 10.5 e −0.156YAX ) (3.54)

3.3.6: Calibration Steps and Results

The overall computational logic for regressing deterioration models for each

distress type can be summarized in the following steps:

• Derive the variables from the road network input data.

• Exclude outliers.

• Divide the whole dataset into several sub-tables according to the calibration

category defined in Section 3.1.

• Export the tables to (*.csv) format that can be recognized by LIMDEP.

• Regress the models in different categories in LIMDEP.

Tables 26, 27, 28, and 29 give the values that were regressed in LIMDEP with the

highest R-square based on local variables and models defined by HDM-4 in four

categories. T-statistics refer to the values of the corresponding variables.

60
 Table 26: Estimated factors for high ESAL ACPs
Calibration Factor Class Estimated Factor T-statistic Number of Observations

Kcia 1.74678 36.462 352

Kciw 1.02038 512.290 352
Kcpa 0.69499 9.527 342
Kcpw 0.03395 4.215 345
WW 0.10053 3594.903 248
Kcit
EW 2.51061 64.210 104
Kcpt 0.11289 4.413 212
Krid Age4<=1 0.12434 0.572 28
Krpd Age4<=1 0.04491 0.131 28
Krsw Age4<=1 1.18374 1.557 28
Krst Age4>1 1.76939 3.413 343
Krpd Age4>1 0.00818 0.300 343
Krsw Age4>1 0.31953 3.232 343
Kvi 0.17650 15.488 352
Kvp 0.01731 2.472 345
Kgm 0.49995 5.988 341
Kgp 0.14342 2.718 341

61
 Table 27: Estimated factors for medium ESAL ACPs
Calibration Factor Class Estimated Factor T-statistic Number of Observations

Kcia 1.28748 47.465 461

Kciw 1.01866 559.489 455
Kcpa 0.33471 13.510 418
Kcpw 0.47516 13.736 425
WW 0.10055 1901.345 309
Kcit
EW 1.66277 19.464 146
Kcpt 0.26529 5.254 266
Krid Age4<=1 0.12434 0.572 28
Krpd Age4<=1 0.04491 0.131 28
Krsw Age4<=1 1.18374 1.557 28
Krst Age4>1 0.31945 2.154 255
Krpd Age4>1 0.17583 5.329 255
Krsw Age4>1 0.53993 5.564 255
Kvi 0.12145 37.323 455
Kvp 0.00742 3.672 425
Kgm 0.66950 8.371 328
Kgp 0.15754 1.653 328

62
 Table 28: Estimated factors for low ESAL ACPs
Calibration Factor Class Estimated Factor T-statistic Number of Observations

Kcia 1.26660 107.273 1368

Kciw 1.01951 989.111 1368
Kcpa 0.50630 21.133 1273
Kcpw 0.44102 22.639 1199
WW 0.10074 6751.605 946
Kcit
EW 1.88127 34.588 422
Kcpt 0.37150 7.411 709
Krid Age4<=1 0.12434 0.572 28
Krpd Age4<=1 0.04491 0.131 28
Krsw Age4<=1 1.18374 1.557 28
Krst Age4>1 0.60274 6.791 1268
Krpd Age4>1 0.29016 5.199 1268
Krsw Age4>1 1.03010 9.028 1268
Kvi 0.08604 95.111 1368
Kvp 0.02824 6.964 1213
Kgm 0.76619 12.757 1281
Kgp 0.29695 2.600 1281

63
 Table 29: Estimated factors for all BSTs
Calibration Factor Class Estimated Factor T-statistic Number of Observations

Kcia 0.36716 49.634 376

Kciw 0.99222 -- 376
Kcpa 0.49920 9.452 353
Kcpw 0.50040 9.452 353
WW 0.04338 6499.995 257
Kcit
EW 3.29640 77.269 118
Kcpt 0.62233 6.206 353
Krid Age4<=1 -- -- 0
Krpd Age4<=1 -- -- 0
Krsw Age4<=1 -- -- 0
Krst Age4>1 0.22147 4.756 352
Krpd Age4>1 0.00000 0.000 352
Krsw Age4>1 2.05351 9.469 352
Kvi 0.89950 231.770 376
Kvp 0.10260 2.046 353
Kgm 0.98860 7.692 353
Kgp 0.11181 1.196 353

3.3.7: Validation of Modeling Factors

Calibration factors were chosen by maximizing R-squared in LIMDEP. All

deterioration model factors were further accessed by the process of validation. Validation

was used to determine how well the models represent the real system and to adjust model

factors with real observations.

3.3.7.1: Validation Data Preparation

At the project unit, WSPMS 2003 consists of 3,508 sections (excluding bridges),

including 2,893 flexible sections and 615 concrete sections. To simplify the validation

process, these sections were merged into 42 typical sections (24 flexible sections and 18

concrete sections) according to the following classification:

64
 • Pavement types (flexible, concrete)

• Traffic volume (high, medium, low)

o High: when ESAL/Elane/year > 500,000

o Low: when ESAL/Elane/year < 250,000

o Medium: when 250,000 ≤ ESAL/Elane/year ≤ 500,000

• Surface material (AC, BST)

• Pavement condition (good, fair, poor)

o Good: when rut depth < 5mm and ACA < 1.00%

o Poor: when rut depth >10mm or ACA > 5.00%

o Fair: between good and poor

• Road class (interstate, non-interstate).

Climate zones are merged into one, as shown in Table 30.

Table 30: Climate index for one climate zone of Washington state
Climate
Moisture Index 109.75
Duration of dry season 0.59 (as a fraction of a year)
Mean monthly precipitation 97.15 mm
Mean temperature 10.22 ºC
Average Temperature range 9.97 ºC
Days T>32 ºC 1.55 days
Freeze Index 99.44 ºC-days
Percentage of Time Driven
On snow covered roads 0
On water covered roads 20
[Source: WSPMS 2003]

According to the classification, all road network variables for the 42 sections were

set as the medians of corresponding project level sections. Therefore, these 42 sections

can represent the main characteristics of all WSDOT highways. Results based on these

65
typical sections are thereby valid and can be used in all project level sections.

The assigned standards follow the routine works of WSDOT as Table 31:

Table 31: Maintenance standards for flexible pavements

Asphalt Concrete-Surfaced Flexible Pavements
-Base Alternative (Do Nothing)
-Pothole Patching
-45-mm Overlay
Bituminous Treatment-Surfaced Flexible Pavements
-Base Alternative (Do Nothing)
-BST

Vehicle composition and annual growth rate also need to be confirmed on the

selected road sections in the given year. Medians for specific surface types are applied as

shown in Table 32:

Table 32: Traffic composition and annual growth rate for ACPs
Traffic Composition Annual Growth Rate
Vehicle Type
(%) (%)

ACP
Car 87.11 1.66
Single Unit 6.11 1.66
Double Unit 5.37 1.66
Train 1.41 1.66
BST
Car 83.14 1.72
Single Unit 9.05 1.72
Double Unit 5.72 1.72
Train 2.09 1.72

3.3.7.2: Validation Criteria

The following heuristic values from WSDOT were set as the validation criteria.

• For validation of AC surfaced flexible pavements models, roads should be

overlaid about every 12 years when cracking ≥10 percent. At that time

66
 roughness is smaller than 3.5m/km and rut depth is close to 10 mm.

• For validation of BST-surfaced flexible pavements models, roads should be

resealed about every 6 years, when cracking ≥10 percent. At that time

roughness is lower than 3.5m/km and rut depth is close to 10mm. But the

average road conditions are worse than that of ACPs.

3.3.7.3: Validation Results

The validation results are listed in Table 33. These factors will be used for future

analysis in Chapter 4. Note that the results in Table 33 differ from the calibration factors

in Tables 26, 27, 28 and 29. The factors in these Tables are not validated.

Table 33: Validated calibration factors for flexible pavements

ACP
Calibration
Traffic BST
Factors
High Med Low
Kcia 1.00 0.84 0.76 0.20
Kciw 0.40 0.40 0.40 0.30
Kcpa 0.71 0.78 0.82 0.50
Kcpw 0.11 0.30 0.45 0.50
Kcit 0.10 0.10 0.10 0.04
Kcpt 0.20 0.20 0.20 0.62
Krid 0.12 0.01
Krst 0.15 0.22
Krpd 0.01 0.02
Krsw 0.32 2.05
Kgm 0.70 1.00
Kgp 1.62 0.70
Kvi 1.00 1.00
Kvp 0.04 1.00
Kpi 1.00 1.10 3.00 1.00
Kpp 0.10 0.08 0.40 1.00
Ksnpk 0.00 0.00

67
3.4: CALIBRATION OF DETERIORATION MODELS FOR CONCRETE
PAVEMENTS

The deterioration models for concrete pavements in HDM-4 are basically absolute

models. Absolute models make predictions based on conditions at a particular point in

time (Odoki, 2000). HDM-4 sets this point of time as the construction year for concrete

pavement deterioration models.

WSDOT uses jointed plain concrete pavement. Table 34 lists related calibration

factors used in those deterioration models.

Table 34: Calibration factors used in concrete pavement deterioration models

Calibration Factor Deterioration model
Kjpc Cracking calibration factor
Kjpnf Faulting calibration factor in pavements without dowels
Kjpdf Faulting calibration factor in pavements with dowels
Kjps Joint spalling calibration factor
Kjpr Roughness progression calibration factor
[Source: HDM-4 Volume 4]

3.4.1: Proposed Calibration Methodology

The general expression used for the deterioration models of concrete pavements

in HDM-4 is as following:

Predicted Distress: Y ' = K 'y *f (Ya , a 0 , a1 , a 2 , ... , a n , X1 , X 2 , ... , X m ) (3.55)

where:

K 'y : default calibration factor of distress type Y given by HDM-4 (1.0);

Y' : predicted value of distress types Y by HDM-4;

ai: default coefficient values given by models and decided by factors of climate and

environment, traffic, pavement history, road geometry, pavement structural

characteristics or material properties;

68
Xi: input factors of climate, traffic, road geometry, pavement history, pavement

structural characteristics and material properties.

The optimal calibration factors were obtained by following the steps below:

• Use default value 1.0 as the calibration factors.

• Run HDM-4 in Project Level for one-year forecasting from 2002. Export

• Export the predicted distress values of 2003 from the output reports.

• Exclude outliers.

• Regress the pavement deterioration models in LIMDEP for Ky.

Y = K y *Y ' (3.56)

where:

Y: Value of distress type Y in WSPMS 2003;

Y ': HDM-4 predicted value for distress type Y by using default calibration

factors;

Ky: Calibration factor of distress type Y.

3.4.2: Determination of Calibration Coefficients

The proposed values for calibration coefficients are listed as follows:

• Subgrade k static modulus of reaction: 54 mpa/m.

• Modulus of elasticity of concrete (Ec): 27500 Mpa (4,000,000 psi).

• Modulus of rupture (flexural strength) of concrete: 5 Mpa (725 psi).

• Thermal coefficient of concrete: 0.000006 for gravel aggregate type.

• Shrinkage coefficient: 0.00045 in/in.

• Dowels diameter: 38 mm.

69
 • Joint seal material: asphalt because WSDOT PCC joints are usually hot pour

liquid sealant – asphalt.

• Use corrosion coated or not: yes because WSDOT dowel bars are epoxy

coated or stainless steel.

3.4.3: Simplified Pavement Deterioration Models for Concrete Pavements

WSDOT concrete pavements are generally jointed plain without load transfer

dowels, and the corresponding pavement deterioration models are listed as follows:

3.4.3.1: Thermal Cracking

CRACKINGt =Kjpc * Function (Slab thickness, Joint spacing, Concrete flexural strength,

Climate)

where:

CRACKINGt: Cracking at time t (% of total carriageway);

Kjpc: Calibration factor of cracking (default value 1.0).

3.4.3.2: Faulting

FAULTt = Kjpnf * Function (Drainage condition, Base type, Shoulder/lane width, Joint

spacing, Slab thickness)

where:

FAULTt: Average faulting at time t;

Kjpnf: Calibration factor of faulting (default value 1.0).

3.4.3.3: Spalling

SPALLt = Kjrs * Function (Age, Seal type, Material characteristics, Joint spacing)

where:

70
SPALLt: Spalling at time t (%);

Age: Age since pavement construction (years);

Kjrs: Calibration factor of spalling (default value 1.0).

3.4.3.4: Roughness

ROUGHNESSt = Kjpr * Function (ROUGHNESS0, Fault, Spall, Crack, Joint Spacing)

where:

ROUGHNESSt: Roughness at time t;

ROUGHNESS0: Roughness at the timne of pavement construction;

Kjpr: Calibration factor of roughness (default value 1.0).

3.4.4: Fatal Errors in HDM-4 Outputs for Concrete Pavements

The HDM-4 concrete pavement deterioration models do not function properly,

which essentially renders the whole concrete pavement portion of HDM-4 non-

functional. This conclusion is drawn from a large number of tests based on various

pavement conditions, traffic conditions, and calibration factors. The errors are

summarized as follows:

1. Regardless of the selected calibration factors, values of cracking and spalling

never change. They always equal the values input for the construction year during

the whole service life. Even when calibration factors are set at the maximum

value of 20, the results do not change.

2. HDM-4 reports do not output values of deteriorated cracks and failures, even

though the related deterioration models are given by HDM-4.

71
3. Prediction of faulting and roughness is not reliable. As an example, a typical

concrete pavement section is input into HDM-4 for testing. Table 40 shows the

main characteristics. Three different analysis periods were tested on this section:

o TEST 1: 20 years from 2004 to 2023

o TEST 2: 21 years from 2023 to 2043

o TEST 3: 40 years from 2004 to 2043

Because the analysis was performed on the same section, and traffic

growth rates were set to 0, all tests should have had the same distress conditions

in 2023. However, faulting and roughness in TEST 2 were significantly lower

than those of TEST 1 from 2023 to 2043. Faulting and roughness estimated by

HDM-4 are illustrated in Figures 10 and 11, respectively. Unreasonably low

faulting and roughness values are always predicted by HDM-4 at the beginning of

any analysis period. Therefore, the faulting and roughness performance forecasted

by HDM-4 is not stable and reliable.

Table 35: Main road network data for concrete pavement tests
Speed Flow Type Four Lane Wide
Traffic Flow Type Urban
Road Class Interstate
Number of Lanes 2
AADT 60000
Speed Limit (km/hr) 95
Subgrade Type 1 (Granular)
Base Type 0 (Asphalt Treated)
Base Thickness (mm) 225
Construction Year 1962
Calibration Factors 1 (default)
Traffic Growth Rate 0.00

72
 Average Faulting by Different Analysis Periods

4
TEST3
TEST2
Faulting (m/km) )

3
TEST1
2

0
2004 2009 2014 2019 2024 2029 2034 2039

Year

Figure 10: Faulting forecasting by different analysis periods

Average Roughness by Different Analysis Periods

3
TEST3
TEST2
Roughness(m/km) )

TEST1
2

0
2004 2009 2014 2019 2024 2029 2034 2039

Year
Figure 11: IRI forecasting by different analysis periods

73
4. Two preservation standards were input into HDM-4 and analyzed on the same

section: reconstruction and do nothing. If “reconstruction” is never triggered in a

specific analysis period, the pavement distress conditions should be the same as

“do nothing.” But as seen in Figure 12, produced by HDM-4, the roughness

values of the two alternatives are totally different, especially in the beginning of

the analysis period. This is unreasonable.

Average Roughness by Different Works

3
Roughness (m/km)

Reconstrction
Do Nothing
1

0
2004 2008 2012 2016 2020

Year

Figure 12: Roughness under two alternatives for concrete pavements

5. In the input interface shown in Figure 13, although the distress data in a specific

year can be added manually in the second column, only the first column is used,

which is restricted by HDM-4. Therefore, the interface is futile.

74
 Figure 13: HDM-4 distress data input interface

6. Some maintenance standards, such as reconstruction (or slab replacement), can

only be triggered by cracking, but the estimated cracking stay at the same value as

the construction year, so this maintenance standard is invalid.

Therefore, it is concluded that the HDM-4’s concrete module is not yet functional.

It cannot be applied until improvements and careful tests have taken place. Thus, the

concrete module is NOT included in the later research.

75
 4: OUTPUT ANALYSIS

As stated in Chapter 1, HDM-4 output analyses include three levels: 1) project

level, 2) program level, and 3) strategic level.

4.1: PROJECT LEVEL ANALYSIS

A “Project” is defined as several road works, or more than one road section,

grouped together in one contract. Project title, road network, or vehicle fleet information

can create a project. Work standards, general traffic composition and growth rate, extra

benefits, and costs must be specified to start a project analysis. Project analysis predicts

pavement conditions and costs during a user-specified time period. The costs include

capital investment, road maintenance works, and vehicle operation costs. Accident costs

and emission costs may also be included (Kerali, 2000). All road agency costs in this

research are financial, which exclude the costs of overhead, taxes, project engineering,

and safety from the total road work economic costs.

4.1.1: Analysis Methods

Options for analyzing investment are provided by section or project. In analysis

by section, economic indicators (such as NPV and IRR) are calculated individually for

each section alternative; analysis by project will group alternatives to perform an

economic analysis, and a base alternative is compared for each project alternative to give

annual total economic indicators.

4.1.2: Reports Generated

The goal of project level analysis is to identify the most cost-effective solutions

by comparing several project alternatives. HDM-4 project level analysis can generate the

76
following types of reports:

• traffic condition forecasting, which includes AADT, traffic volume, and

volume/capacity ratio

• road deterioration condition prediction, which includes roughness, pavement

distress conditions, and roadwork lists

• road-user cost estimation, which includes an accident-rate summary, cargo

holding hours, crew hours, fuel consumption, labor hours, lubricant hours,

overhead costs, parts consumption, vehicle speed

• environmental effects

• comparison of the project alternatives

• input data, which can be used for checking and review.

4.1.3: Analysis Results

All ACP sections on I-405 and all BST sections on SR 21 are used as examples of

project level analyses. Pavement preservation alternatives used for ACPs are: 1) base

alternative (do nothing), 2) pothole patching, 3) 45-mm overlay, and 4) 45-mm mill and

fill. The alternatives for BSTs are: 1) base alternative (do nothing), and 2) bituminous

surface treatment. Functions are performed as described in the following groups.

4.1.3.1: Pavement Distress Forecasting

In a defined analysis period, distress conditions analyzed at the project level under

different maintenance alternatives are generated.

(a) AC-Surfaced Flexible Pavements

As an example, Figure 14 shows the roughness changes during a 40-year analysis

period of section I-405 MP 13.82 – 15.17 (D), which was constructed in 1957 and

77
overlaid in 1994. Other related information about this 1.35-mile-long section is listed in

Tables 36 and 37.

Annual Average Roughness

16
Base Alternative
Roughness (m/km)

12
Pothole Patching
8
45mm Overlay
4
45mm Mill & Fill
0
2004 2009 2014 2019 2024 2029 2034 2039

Year

Figure 14: Roughness of I-405 MP 13.82 – 15.17 (D) under four alternatives

The figure illustrates the following conclusions for this specific section:

• Without any maintenance to the road (the base alternative), the roughness will

increase to 16 m/km (the maximum allowed by HDM-4) in year 2018, which

is 24 years from the last overlay.

• By only patching potholes, the roughness will increase to 4 m/km in year 2021

and 10 m/km in year 2041.

• 45-mm mill and fill has the best works effects on roughness, and it lasts

longer than 45-mm overlays.

78
 Table 36: Current conditions of I-405 MP 13.82 – 15.17 (D)
Current Surface Thickness 46 mm
Base Thickness 152 mm
IRI 1.54 m/km
ACRA 0.04%
Pothole 0
Rut Depth 5 mm
Skid resistance 1
MT AADT 97,813
Number of Lanes 3
[Source: WSPMS 2003]

With an annual traffic growth rate of 1.48 %, traffic compositions are as follows:

Table 37: Vehicle composition of I-405 MP 13.82 – 15.17 (D)

Vehicle Composition (%)
Car 92.7
Single Unit 3.4
Double Unit 3.5
Train 0.4
[Source: WSPMS 2003]

A project level analysis “by project” considers all I-405 flexible sections as one

project. Figure 15 shows the result of average roughness (weighted by each section

length) for 40 years.

Average Roughness (IRIav) for each Project

(weighted by section length)
16
45 mm Overlay
14 45mm Mill with
45mm Overlay
12 Base Alternative
Pothole Patching
IRIav (m/km)

10

0
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043

Year

Figure 15: Average roughness of all I-405 flexible pavements (from HDM-4 screen)

79
 (b) BST-Surfaced Flexible Pavements

A BST section, SR 21 MP 130.4 – 153.0 (I), was analyzed in HDM-4 during a 40-

year analysis period. This section was constructed in 1955 and resurfaced in 1975, 1984,

1992, and 1999, and other related data are listed in Tables 38 and 39. According to theses

input information, HDM-4 estimated the pavement conditions, such as roughness,

raveling, cracking, rut depth, and number of potholes. Figure 16 shows the cracking

(percentage of total carriageway area).

Annual Average Cracking

100

80 Base Alternative
Cracking

60
Bituminous
Surface Treatment
40

20

0
2004 2009 2014 2019 2024 2029 2034 2039

Year

Figure 16: Cracking of SR 21 MP 130.4 – 153.0 (I) for 40 years

Note that the “base alternative” is “do nothing,” and “bituminous surface

treatment” is defined in Chapter 2.

80
 Table 38: Pavement conditions of SR 21 MP 130.4 – 153.0 (I)
Current Surface Thickness 15 mm
Previous Surface Thickness 85 mm
IRI 2.42 m/km
ACRA 0.82%
Raveling 0.62%
Pothole 0 NPT
Rut Depth 4 mm
Skid resistance 1
MT AADT 537
Number of Lanes 2 (Both direction)
[Source: WSPMS 2003]

Table 39: Traffic composition and growth rates of SR 21 MP 130.4 – 153.0 (I)

Vehicle Type Traffic Composition(%) Annual Growthrate(%)

Car 75.17 3.86
Single Unit 2.45 3.86
Double Unit 13.29 3.86
Train 9.09 3.86

[Source: WSPMS 2003]

4.1.3.2: Road Agency Costs of Maintenance

(1) By Section

By choosing the “By Section” option for project analysis, HDM-4 can illustrate

the direct maintenance costs and work schedule for each section and for each specific

road maintenance alternative.

Tables 40, 41, and 42 show schedules for the optimal works for I-405 MP 13.82 –

15.17 (D), with economic indicators spanning 40 years. The distress types following the

maintenance alternatives in tables 40 and 41 point out the types that trigger road works.

Costs are road agency costs without consideration for discount rates.

During this 40-year analysis period, maintenance of 45-mm overlay and 45-mm

81
mill & fill are triggered three times, but the latter serves one to three years longer than the

former in each case.

Table 40: Costs and schedule of 45-mm overlay for I-405 MP 13.82 – 15.17 (D)
Road Agency Cost Work Quantity
Year Description
($) (m2)
2010 45mm Overlay - Cracking 452,293 23,805
2021 45 mm Overlay - Rutting 452,293 23,805
2032 45 mm Overlay - Rutting 452,293 23,805
Total cost for the section: 1,356,879

Table 41: Costs and schedule of 45-mm mill and fill for I-405 MP 13.82 – 15.17 (D)
Road Agency Cost Work Quantity
Year Description
($) (m2)
2010 45mm Mill & Fill - Cracking 714,147 23,805
2027 45mm Mill & Fill - Rutting 714,147 23,805
2040 45mm Mill & Fill - Rutting 714,147 23,805
Total cost for the section: 2,142,441

82
 Table 42: Costs and schedule of pothole patching for I-405 MP 13.82 – 15.17 (D)
Road Agency Cost Work Quantity
Year Description
($) (m2)
2015 Pothole Patching 236 5.0
2016 Pothole Patching 306 6.5
2017 Pothole Patching 385 8.2
2018 Pothole Patching 469 10.0
2019 Pothole Patching 550 11.7
2020 Pothole Patching 625 13.3
2021 Pothole Patching 693 14.7
2022 Pothole Patching 753 16.0
2023 Pothole Patching 805 17.1
2024 Pothole Patching 849 18.1
2025 Pothole Patching 887 18.9
2026 Pothole Patching 919 19.6
2027 Pothole Patching 945 20.1
2028 Pothole Patching 966 20.6
2029 Pothole Patching 983 20.9
2030 Pothole Patching 997 21.2
2031 Pothole Patching 1,007 21.4
2032 Pothole Patching 1,015 21.6
2033 Pothole Patching 1,021 21.7
2034 Pothole Patching 1,026 21.8
2035 Pothole Patching 1,029 21.9
2036 Pothole Patching 1,031 21.9
2037 Pothole Patching 1,032 22.0
2038 Pothole Patching 1,033 22.0
2039 Pothole Patching 1,034 22.0
2040 Pothole Patching 1,034 22.0
2041 Pothole Patching 1,035 22.0
2042 Pothole Patching 1,035 22.0
2043 Pothole Patching 1,035 22.0
Total cost for the section: 24,733

(2) By Project

If using the “By Project” option, HDM-4 will treat all given sections as one

project and calculate annual road agency costs for each alternative. Under each

alternative, work schedules are optimized to gain the greatest economic advantage. Table

43 lists all related costs for all I-405 flexible pavements over a period of 40 years.

83
Table 43: Summary of total annual costs for all I-405 flexible sections
Road Agency Cost ($)
Year
45mm Mill & Fill 45mm Overlay Pothole Patching
2006 513,396 325,151 0
2008 2,043,711 1,294,350 0
2009 566,052 358,500 0
2010 9,350,097 5,921,728 512
2011 4,067,676 2,576,195 714
2012 0 0 907
2013 542,412 343,528 1,172
2014 0 0 1,503
2015 573,156 362,999 2,962
2016 0 0 4,198
2017 0 0 5,023
2018 0 0 7,436
2019 0 0 8,966
2020 2,043,711 1,294,350 10,314
2021 1,007,046 637,796 11,624
2022 513,396 325,151 12,798
2023 5,072,928 3,212,854 14,241
2024 955,260 604,998 15,258
2025 3,775,371 2,391,068 16,162
2026 1,873,188 1,186,352 16,966
2027 0 0 17,649
2028 1,300,032 823,354 18,223
2029 542,412 343,528 18,694
2030 0 0 19,074
2031 0 0 19,379
2032 3,623,913 2,295,145 19,621
2033 0 0 19,811
2034 4,067,676 2,576,195 19,957
2035 0 0 20,067
2036 566,052 358,500 20,147
2037 1,468,656 930,149 20,204
2038 439,200 278,160 20,243
2039 1,785,795 1,131,004 20,270
2040 1,989,576 1,260,065 20,288
2041 0 0 20,300
2042 1,873,188 1,186,352 20,307
2043 2,563,689 1,623,670 20,313
Total 53,117,589 33,641,140 465,300

84
 4.1.3.3: Optimum Maintenance Standards

(1) By Section

A project analysis “by section” estimates works effects and economic indicators

for each section. The most effective alternative is the one with the highest indicators. An

example using section I-405 MP 13.82 - 15.17 (D) is illustrated in Table 44.

The Economic Indicators Summary Report gives a summary of costs, discounted

Net Present Value (NPV), and Internal Rate of Return (IRR) by project alternatives. The

discount rate is 4 percent.

Table 44: Economic indicators for I-405 MP 13.85 – 15.17 (D)

Increase in Decrease in Net Present Internal Rate of
Alternative
Agency Cost User Cost Value (NPV) Return (IRR)
Base Alternative 0.00 0.00 0.00 0.00
Pothole Patching 0.01 206.77 206.76 No Solution
45mm Overlay 0.67 249.13 248.46 93.2
45mm Mill & Fill 1.06 249.13 248.07 82.3
(All costs are in millions of dollars.)

For this case, the of 45-mm overlay alternative has the highest NPV and IRR.

Therefore, it is the best alternative for this section in 40 years.

(2) By Project

By comparing the economic indicators of each alternative with a base alternative,

the most cost-effective alternative for the whole project is determined.

Table 45 illustrates the costs and economic indicators in 40 years for all I-405

flexible sections with a discount rate of 4 percent. The base alternative is “Do Nothing.”

Since 45-mm overlay has the highest NPV and IRR, it is the optimal maintenance

alternative for all I-405 flexible sections.

85
 Table 45: Total economic benefits for all I-405 flexible sections
Increase in Decrease in Net Present Internal Rate of
Alternative
Agency Cost User Cost Value (NPV) Return (IRR)
Pothole Patching 0.17 4,773.23 4,773.07 --
45mm Overlay 17.00 5169.676 5,152.68 96.7
45mm Mill & Fill 26.84 5169.676 5,142.84 80.1
(All costs are in millions of dollars.)

4.2: PROGRAM LEVEL ANALYSIS

The program level analysis selects a combination of treatments for sections by

optimizing an objective function under budget constraints. It generates the same type of

reports as project level analysis, such as traffic reports, pavement distress conditions,

works effects, road-user effects, and costs. Furthermore, it compares outputs by different

categories under constrained budgets or optimized conditions. The reports generated in

program level analysis are (Kerali, 2000a):

• optimum section alternatives for varying levels of budgets

• a pavement surface condition summary by link ID (state route number) or

road class (interstate, major collector, minor arterial, or principal arterial)

• average roughness (weighted by section length) by link ID, pavement surface,

or road class

• average speed by link ID, pavement surface, or road class

• volume/capacity ratio by link ID, pavement surface, or road class

• work programs optimized or unconstrained by section or by year.

To demonstrate the functions, all I-5 flexible pavements, 141 sections and 872.73

lane miles were analyzed in program level. Table 46 lists the traffic composition and

growth rates.

86
 Table 46: Traffic composition and growth rates of I-5
Vehicle Type Traffic Composition(%) Annual Growthrate(%)
Car 90.1 1.48
Single Unit 3.6 1.48
Double Unit 5.9 1.48
Train 0.4 1.48

4.2.1: Multi-Year Forward Program

The multi-year Forward Program method requires one preservation treatment or

one treatment after the previous treatment that has been assigned to each road section.

The treatments are triggered based on reaching a distress threshold.

4.2.1.1: Maintenance Triggered and Related Costs

Details of the works activity and condition responsive criteria are summarized in

Table 47. In this case, a 45-mm overlay is assigned to each I-5 flexible section during an

analysis period of five years. No discount rate is applied to the road agency costs.

Thus, in five years, 17 sections out of 141 sections need to be maintained to

obtain the optimal road condition, which means that each time a 45-mm overlay is

triggered, the section will be overlaid. The total agency financial cost is $8.57 million.

87
 Table 47: Optimized work program for all I-5 flexible sections for five years
Length Financial Cumulative
Year Section AADT Work Description
(km) Cost Cost

005 MP182.67 - 183.96 (D) 2.08 90906 45 mm Overlay Rutting 0.43 0.43
005 MP181.51 - 182.67 (D) 1.87 87669 45mm Overlay Cracking 0.39 0.82
2004
005 MP182.86 - 183.96 (I) 1.77 92292 45 mm Overlay Rutting 0.37 1.19
005 MP188.70 - 189.30 (I) 0.97 73985 45 mm Overlay Rutting 0.20 1.39
005 MP6.07 - 7.00 (D) 1.50 34656 45 mm Overlay Rutting 0.21 1.60
2005
005 MP19.83 - 20.08 (D) 0.40 32089 45 mm Overlay Rutting 0.11 1.71
005 MP198.19 - 198.51 (D) 0.51 60772 45 mm Overlay Rutting 0.11 1.82
2006
005 MP228.25 - 229.33 (D) 1.74 38504 45mm Overlay Cracking 0.24 2.06
005 MP190.21 - 191.60 (I) 2.24 93678 45 mm Overlay Rutting 0.62 2.69
005 MP181.51 - 182.86 (I) 2.17 90916 45 mm Overlay Rutting 0.45 3.14
2007
005 MP183.96 - 189.30 (D) 8.59 89293 45mm Overlay Cracking 1.79 4.93
005 MP135.52 - 139.50 (D) 6.41 88521 45 mm Overlay Rutting 1.78 6.71
005 MP190.12 - 191.60 (D) 2.38 94930 45 mm Overlay Rutting 0.66 7.37
005 MP276.20 - 276.56 (I) 0.58 8860 45mm Overlay IRI 0.08 7.45
2008 005 MP180.10 - 180.75 (D) 1.05 94931 45 mm Overlay Rutting 0.22 7.67
005 MP195.11 - 197.18 (D) 3.33 62584 45 mm Overlay Rutting 0.69 8.36
005 MP8.10 - 8.70 (I) 0.97 48638 45 mm Overlay Rutting 0.20 8.57
(All costs are in millions of dollars)

4.2.1.2: Average Roughness Distress by Road Class

The annual average roughness, weighted by length for road classes (interstate and

non-interstate), of the optimized work program is shown in Figure 17.

88
 1.04

1.02 Interstate
Roughness (m/km)

0.98

0.96

0.94

0.92

0.9
2004

2005

2006

2007

2008

2009
Year

Figure 17: Average roughness of all I-5 flexible pavements for five years (from
HDM-4 screen)

4.2.1.3: Volume/Capacity Ratio by Pavement Surface

The annual average volume capacity ratio weighted by length under the optimized

work program is shown in Figure 18.

0.412

0.408 Interstate
Volume Capacity Ratio

0.404

0.4

0.396

0.392

0.388

0.384
2004

2005

2006

2007

2008

2009

Year

Figure 18: Average volume capacity ratio of all I-5 flexible pavements for five years
(HDM-4 screen)

4.2.2: Life-Cycle Analysis

Different than the multi-year program analysis, the life-cycle analysis requires at

least two standards for each section to compare the defined works alternatives with the

base alternative during the specific analysis period. It selects optimal alternatives for each

89
section that maximize the economic benefits for the whole network while constraining

the financial costs to less than the available budget.

4.3: STRATEGIC LEVEL ANALYSIS

For medium to long-term planning, the strategic level analysis groups all road

segments with similar characteristics into the road network matrix. As with the program

level analysis, it provides a combination of works alternatives, which optimizes objective

functions under constrained budgets. The objective functions can be

• maximizing Net Present Value (NPV) for the best benefits

• maximizing the improvement of roughness to obtain the best road condition

• minimizing cost for target roughness.

As explained in Chapter 3, Section 3.3.1, improvement of IRI signifies an

improvement in road conditions. Whenever a roadway drops beneath the target

roughness, maintenance is triggered. Thus, road distress will never be worse than the

targeted level of roughness. The system will select the maintenance scenario that

provides the most cost-beneficial schedule for WSDOT highways. This strategic level

analysis can benefit WSDOT by:

• determining the required budget levels for specified preservation treatments

• demonstrating the effects of varying budget level on the long-term road

performance.

• optimizing the allocation of funds

• presenting economic indicators for varying budget levels (Odoki, 2000).

90
 4.3.1: Road Network Data

Strategic level analysis adopts all 24 typical WSDOT flexible sections prepared

in Chapter 3, Section 3.3.7. Appendix D1 describes the definitions of the 18 ACP and 6

BST sections.

These sections represent all WSDOT flexible pavements, so most of the sections

are very long. A large amount of funding is needed when any one of these long sections

needs repairing. When that time comes, if the given budget is less than required, the

sections will not be maintained at that time. The budget cannot be applied to subsequent

budget periods. In other words, a budget defined in a specific budget period can only be

used during that specific time period; only sections requiring less than the given budget,

and with higher economic returns, can be repaired. To entirely spend the allocated

budget, the long sections must be divided into shorter sections (keep in mind that all of

this is due to HDM-4 software constraints, but this needs to be stated). But how short

should they be? If the sections were as small as the project level, there would be 2481

ACP sections and 412 BST sections. Can the strategic analysis handle so many sections?

Although HDM-4 dose not specify any limitations in the number of sections,

experiments ascertained that 49 is the maximum number of sections that can be handled

by the strategic level analysis. Therefore, the 18 ACPs were divided into 49 sections, and

the 6 BSTs were divided into 20 sections. In this configuration, each section is shorter

than 300 km. The related road network inputs are listed in appendices D2 and D3 for

ACPs and BSTs, respectively.

Because of the limited number of input sections permitted by the system, the

strategic level road network cannot simulate the WSDOT highway system as effectively

as project level analyses can. For long-term forecasting, the results are reliable, but they

91
can be considered solely as a reference for an analysis detailed in a specific year.

4.3.2: Optimized Road Maintenance and Development Plans

The economic analysis produces a work program that offers alternatives with the

minimum costs for a target road distress condition. A target road condition defines the

conditions under which the road will be maintained. The road works assigned to each

section represents the optimal maintenance standards for each specific section.

The road agency costs in this report exclude overhead, taxes, project engineering,

and safety. All costs and funds in this report are financial. For all ACPs, HDM-4

estimates an optimal work program with the total financial cost of $4,224.1 million for 40

years; for all BSTs, the financial cost is $339.9 million. The relative annual costs are

$105.6 million and $8.5 million, which represent constant annual purchasing power

without considering the discount rate.

4.3.3: WSDOT Funds

WSDOT obtains the pavement maintenance funds every two years. The historical

budgets are listed in Table 48.

Table 48: WSDOT historical funds for pavement maintenance

Financial Funding
Biennium Total Economic Funding per Biennium (and per Year) (Biennium)
AC BST PCCP
1995-1997 $258.3 ($129.2/ year) $198.5 $13.1 $42.4
1997-1999 $319.1 ($159.6/ year) $204.9 $16.6 $32.9
1999-2001 $275.6 ($137.8/ year) $189.0 $20.6 $3.7
2001-2003 $267.4 ($133.7/ year) $174.1 $13.8 $14.5
(All costs are in millions of dollars) (Refer to Chapter 2, sections 2.4 and 2.5)

92
 The WSDOT FY 2001-2003 preservation funding (again, HDM-4 refers to this

with the generic term of “maintenance”) is $174.1millon for ACPs and $13.8 million for

BSTs. They were used to simulate the current WSDOT’s budget scenario for 40 years:

$3,482 million for ACPs and $276 million for BSTs, which represent the same constant

biennium purchasing power without considering the discount rate.

4.3.4: Road Performance under Varying Levels of Budgets

HDM-4 strategic analysis can calculate minimum costs for a target IRI (smaller

than 3.5 m/km) with optimal maintenance. Under varying levels of budgets, it provides

an optimal work program.

HDM-4 shows that the minimum cost for target IRI (MCTI) of all ACPs is $4,224

million for 40 years without considering the discount rate. In the following simulation,

five budget levels starting in 2004 for 40 years are assumed:

• MCTI ($4224 million)

• WSDOT ($3482 million)

• 75% of MCTI ($3168 million)

• 50% of MCTI ($2112 million)

• 0 (no budget)

For all BSTs, the MCTI is $339.9 million for 40 years without considering the

discount rate. Five budget levels starting in 2004 for 40 years are assumed:

• MCTI ($340 M)

• WSDOT($276M)

• 75% of MCTI ($255M)

• 50% of MCTI ($170M)

93
 • 0 (no budget)

Figures 19 and 20 illustrate roughness (weighted by length) performance for all

ACPs and BSTs, respectively.

Annual Average Roughness of ACPs

16
Roughness (m/km)

12 0

50% MCTI
8
WSDOT

4 75% MCTI

MCTI
0
2004 2009 2014 2019 2024 2029 2034 2039
Year
Figure 19: Roughness of all ACPs under varying budgets

Annual Average Roughness of BSTs

16
0
Roughness (m/k m

12 50% MCTI

8 75% MCTI

WSDOT
4

MCTI
0
2004 2009 2014 2019 2024 2029 2034 2039

Year
Figure 20: Roughness of all BSTs under varying budgets

Figures 21 and 22 show surface damage conditions. The condition is presented by

94
the total surface damage area (percentage of total carriageway area), which is composed

of edge-breaks, potholes, cracking, and raveling areas.

Annual Average Damaged Area of ACPs

100

0
80
Damaged Area (%)

50% MCTI
60

75% MCTI
40

WSDOT
20

MCTI
0
2004 2009 2014 2019 2024 2029 2034 2039
Year

Figure 21: Surface damage of all ACPs under varying budgets

Annual Average Damaged Area of BSTs

100
0

80
Damaged Area (%)

50% MCTI

60

75% MCTI
40

WSDOT
20

0 MCTI
2004 2009 2014 2019 2024 2029 2034 2039
Year
Figure 22: Surface damage of all BSTs under varying budgets

The five budget levels for ACPs can be expressed by the constant annual

purchasing power for 40 years beginning in 2004, without considering the discount rate.

The amounts are as follows:

95
 • MCTI ($105.6 million)

• WSDOT ($87 million)

• 75% of MCTI ($79.2 million)

• 50% of MCTI ($52.8 million)

• 0 (no budget)

In FY 2004-2005, if $30 million is cut from the current WSDOT budget for

ACPs, and the funding will revert back to the current level in subsequent bienniums, what

is the impact of this cut? What is the added user cost due to this cut in 40 years? How

much will it cost over the next 10 years to bring the network back to the same condition

had there been no cut? To answer these questions, all related funding scenarios are listed

in the Table 49.

Table 49: WSDOT funding scenarios

Budget Distribution
Scenario Description
2004-2005 2006-2015 2016-2043
Cut $30M from the current WSDOT
budget for ACPs in FY 2004-2005, then $144.1 $4012.8
A
apply the MCTI budget for the ($72/year)* ($105.6/year)*
remaining 38 years
Cut $30M from the current WSDOT
budget for ACPs in FY 2004-2005, then $144.1 $870.5 $2437.5
B
back to the current WSDOT budget level ($72/year)* ($87/year)* ($87/year)*
for the remaining 38 years
Cut $30M from the current WSDOT
budget for ACPs in FY 2004-2005, then $144.1 $1050 $2437.5
C
bring the network back to the same ($72 /year)* ($105/year)* ($87/year)*
condition as the scenario D in 10 years
The WSDOT current level of budget for $3482
D
ACPs ($87/year)*
(All costs are in millions of dollars)
(* the constant annual purchasing power in the defined FYs)

Figure 23 illustrates roughness performance curves for all ACPs under the four

funding scenarios above. Figure 24 shows surface damage conditions. The differences

96
between the road performance curves of scenarios B and D indicate the effects of the

budget cut. The total road user cost of Scenario B in 40 years is $200.2 billion.

Comparing that with Scenario D, $37.5 billion is added in 40 years. Moreover, to bring

the road condition back in 10 years to the same condition it would have been without the

budget cut would cost an additional $ 179.5 million.

Annual Average Roughness of ACPs

8

6 Senario B
Roughness (m/km)

4 Senario C/D

2
Senario A

0
2004 2009 2014 2019 2024 2029 2034 2039
Year

Figure 23: Roughness of all ACPs under WSDOT scenarios

Annual Average Damaged Area of ACPs

50

40
Damaged Area (%)

Senario B
30

20
Senario C/D

10

Senario A
0
2004 2009 2014 2019 2024 2029 2034 2039

Year

Figure 24: Damaged area of all ACPs under WSDOT scenarios

97
 4.3.5: Economic Indicators Under Varying Levels of Budgets

Reports produced by Optimum Section Alternatives provide an economic analysis

using a defined budget level. Table 50 collects costs and maximum benefits using

different budget levels. In this table, the discount rate is 4 percent.

Table 50: Economic indicators under varying levels of budgets for all ACPs
Length 40-year Agency Cost Average
Annual
Scenario Repaired Roughness NPV
Budget
(Lane-km) Undiscounted Discounted (m/km)
50% MCTI 52.8 8,642 2,112 1,193 4.63 146,619
75% MCTI 79.2 13,094 3,168 1,785 2.93 159,682
WSDOT 87 14,294 3,482 1,951 2.55 162,461
MCTI 105.6 17,420 4,224 2,372 1.33 198,052
(All costs are in millions of dollars)

4.3.6: Optimized Allocation of Defined Budgets

An optimal work program is produced for a defined budget. The program has a

total cost that falls within the specified budget. Table 51 lists optimal works, time to

apply, and road agency costs under the current WSDOT budget for ACPs in 40 years.

The actual total cost is $ 3,449 million.

98
 Table 51: Optimized work program under the current WSDOT budget for ACPs
Length Financial Cumulative
Year Section (km) Work Description Cost Cost
Interstate Flexible Pavement High Traffic Poor Condition 14 45mm Mill & Fill Rutting 3 3
Other Flexible Pavement Medium Traffic Poor Condition1 164 45 mm Overlay Rutting 25 28
Other Flexible Pavement High Traffic Poor Condition 23 45mm Mill & Fill Rutting 9 37
2004
Other Flexible Pavement Low Traffic Poor Condition1 213 45 mm Overlay Rutting 32 70
Other Flexible Pavement Low Traffic Poor Condition2 200 45 mm Overlay Rutting 30 100
Interstate Flexible Pavement Low Traffic Poor Condition 3 45mm Mill & Fill Rutting 1 101
Interstate Flexible Pavement High Traffic Fair Condition 119 45mm Overlay Cracking 18 119
Interstate Flexible Pavement Medium Traffic Fair Condition 205 45mm Overlay Cracking 31 150
Other Flexible Pavement Medium Traffic Fair Condition1 130 45mm Overlay Cracking 20 170
Other Flexible Pavement Medium Traffic Fair Condition2 200 45mm Overlay Cracking 30 200
Other Flexible Pavement Medium Traffic Fair Condition3 200 45mm Overlay Cracking 30 231
Other Flexible Pavement Medium Traffic Fair Condition4 200 45mm Overlay Cracking 30 261
Other Flexible Pavement Medium Traffic Fair Condition5 200 45mm Overlay Cracking 30 291
2007 Other Flexible Pavement Medium Traffic Fair Condition6 200 45mm Overlay Cracking 30 322
Other Flexible Pavement Medium Traffic Fair Condition7 200 45mm Overlay Cracking 30 352
Other Flexible Pavement High Traffic Fair Condition 64 45mm Mill & Fill Cracking 25 377
Other Flexible Pavement Low Traffic Fair Condition1 213 45mm Overlay Cracking 32 409
Other Flexible Pavement Low Traffic Fair Condition2 200 45mm Overlay Cracking 30 440
Other Flexible Pavement Low Traffic Fair Condition3 200 45mm Overlay Cracking 30 470
Other Flexible Pavement Low Traffic Fair Condition4 200 45mm Overlay Cracking 30 501
Interstate Flexible Pavement Low Traffic Fair Condition 7 45mm Mill & Fill Cracking 2 502
Interstate Flexible Pavement Medium Traffic Good Condition1 180 45mm Overlay Cracking 27 530
Interstate Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 560
Interstate Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 591
Other Flexible Pavement Medium Traffic Good Condition10 200 45mm Overlay Cracking 30 621
Other Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 651
Other Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 682
Other Flexible Pavement Medium Traffic Good Condition4 200 45mm Overlay Cracking 30 712
Other Flexible Pavement Medium Traffic Good Condition5 200 45mm Overlay Cracking 30 743
Other Flexible Pavement Medium Traffic Good Condition6 200 45mm Overlay Cracking 30 773
Other Flexible Pavement Medium Traffic Good Condition7 200 45mm Overlay Cracking 30 803
2009 Other Flexible Pavement Low Traffic Good Condition10 200 45mm Overlay Cracking 30 834
Other Flexible Pavement Low Traffic Good Condition11 200 45mm Overlay Cracking 30 864
Other Flexible Pavement Low Traffic Good Condition2 200 45mm Overlay Cracking 30 895
Other Flexible Pavement Low Traffic Good Condition3 200 45mm Overlay Cracking 30 925
Other Flexible Pavement Low Traffic Good Condition4 200 45mm Overlay Cracking 30 955
Other Flexible Pavement Low Traffic Good Condition5 200 45mm Overlay Cracking 30 986
Other Flexible Pavement Low Traffic Good Condition6 200 45mm Overlay Cracking 30 1,016
Other Flexible Pavement Low Traffic Good Condition7 200 45mm Overlay Cracking 30 1,047
Other Flexible Pavement Low Traffic Good Condition8 200 45mm Overlay Cracking 30 1,077
Other Flexible Pavement Low Traffic Good Condition9 200 45mm Overlay Cracking 30 1,107
Interstate Flexible Pavement Low Traffic Good Condition 12 45mm Mill & Fill Cracking 3 1,110
Interstate Flexible Pavement High Traffic Poor Condition 14 45mm Mill & Fill Cracking 3 1,114
2015 Other Flexible Pavement Medium Traffic Poor Condition1 164 45mm Overlay Cracking 25 1,139
Other Flexible Pavement High Traffic Poor Condition 23 45mm Mill & Fill Cracking 9 1,147
Other Flexible Pavement Low Traffic Poor Condition1 213 45mm Overlay Cracking 32 1,180
2016
Other Flexible Pavement Low Traffic Poor Condition2 200 45mm Overlay Cracking 30 1,210

99
 Table 51 (Continued)
Length Financial Cumulative
Year Section Work Description
(km) Cost Cost
2016 Interstate Flexible Pavement Low Traffic Poor Condition 3 45mm Mill & Fill Cracking 1 1,211
2018 Interstate Flexible Pavement High Traffic Fair Condition 119 45mm Overlay Cracking 18 1,229
Interstate Flexible Pavement Medium Traffic Fair Condition 205 45mm Overlay Cracking 31 1,260
Other Flexible Pavement Medium Traffic Fair Condition1 130 45mm Overlay Cracking 20 1,280
Other Flexible Pavement Medium Traffic Fair Condition2 200 45mm Overlay Cracking 30 1,310
Other Flexible Pavement Medium Traffic Fair Condition3 200 45mm Overlay Cracking 30 1,341
Other Flexible Pavement Medium Traffic Fair Condition4 200 45mm Overlay Cracking 30 1,371
Other Flexible Pavement Medium Traffic Fair Condition5 200 45mm Overlay Cracking 30 1,402
2019 Other Flexible Pavement Medium Traffic Fair Condition6 200 45mm Overlay Cracking 30 1,432
Other Flexible Pavement Medium Traffic Fair Condition7 200 45mm Overlay Cracking 30 1,462
Other Flexible Pavement High Traffic Fair Condition 64 45mm Mill & Fill Cracking 25 1,487
Other Flexible Pavement Low Traffic Fair Condition1 213 45mm Overlay Cracking 32 1,520
Other Flexible Pavement Low Traffic Fair Condition2 200 45mm Overlay Cracking 30 1,550
Other Flexible Pavement Low Traffic Fair Condition3 200 45mm Overlay Cracking 30 1,581
Other Flexible Pavement Low Traffic Fair Condition4 200 45mm Overlay Cracking 30 1,611
2020 Interstate Flexible Pavement Low Traffic Fair Condition 7 45mm Mill & Fill Cracking 2 1,613
Interstate Flexible Pavement Medium Traffic Good Condition1 180 45mm Overlay Cracking 27 1,640
Interstate Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 1,670
Interstate Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 1,701
Other Flexible Pavement Medium Traffic Good Condition10 200 45mm Overlay Cracking 30 1,731
Other Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 1,762
Other Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 1,792
Other Flexible Pavement Medium Traffic Good Condition4 200 45mm Overlay Cracking 30 1,822
Other Flexible Pavement Medium Traffic Good Condition5 200 45mm Overlay Cracking 30 1,853
Other Flexible Pavement Medium Traffic Good Condition6 200 45mm Overlay Cracking 30 1,883
Other Flexible Pavement Medium Traffic Good Condition7 200 45mm Overlay Cracking 30 1,914
2021
Other Flexible Pavement Low Traffic Good Condition10 200 45mm Overlay Cracking 30 1,944
Other Flexible Pavement Low Traffic Good Condition11 200 45mm Overlay Cracking 30 1,974
Other Flexible Pavement Low Traffic Good Condition2 200 45mm Overlay Cracking 30 2,005
Other Flexible Pavement Low Traffic Good Condition3 200 45mm Overlay Cracking 30 2,035
Other Flexible Pavement Low Traffic Good Condition4 200 45mm Overlay Cracking 30 2,066
Other Flexible Pavement Low Traffic Good Condition5 200 45mm Overlay Cracking 30 2,096
Other Flexible Pavement Low Traffic Good Condition6 200 45mm Overlay Cracking 30 2,126
Other Flexible Pavement Low Traffic Good Condition7 200 45mm Overlay Cracking 30 2,157
Other Flexible Pavement Low Traffic Good Condition8 200 45mm Overlay Cracking 30 2,187
Other Flexible Pavement Low Traffic Good Condition9 200 45mm Overlay Cracking 30 2,218
2022 Interstate Flexible Pavement Low Traffic Good Condition 12 45mm Mill & Fill Cracking 3 2,220
Interstate Flexible Pavement High Traffic Poor Condition 14 45mm Mill & Fill Cracking 3 2,224
2027
Other Flexible Pavement Medium Traffic Poor Condition1 164 45mm Overlay Cracking 25 2,249
2028 Other Flexible Pavement High Traffic Poor Condition 23 45mm Mill & Fill Cracking 9 2,258
Other Flexible Pavement Low Traffic Poor Condition1 213 45mm Overlay Cracking 32 2,290
2029
Other Flexible Pavement Low Traffic Poor Condition2 200 45mm Overlay Cracking 30 2,320
Interstate Flexible Pavement High Traffic Fair Condition 119 45mm Overlay Cracking 18 2,339
2030
Interstate Flexible Pavement Low Traffic Poor Condition 3 45mm Mill & Fill Cracking 1 2,339
Interstate Flexible Pavement Medium Traffic Fair Condition 205 45mm Overlay Cracking 31 2,370
2031
Other Flexible Pavement Medium Traffic Fair Condition1 130 45mm Overlay Cracking 20 2,390

100
 Table 51 (Continued)
Length Financial Cumulative
Year Section Work Description
(km) Cost Cost
Other Flexible Pavement Medium Traffic Fair Condition2 200 45mm Overlay Cracking 30 2,421
Other Flexible Pavement Medium Traffic Fair Condition3 200 45mm Overlay Cracking 30 2,451
Other Flexible Pavement Medium Traffic Fair Condition4 200 45mm Overlay Cracking 30 2,481
2031
Other Flexible Pavement Medium Traffic Fair Condition5 200 45mm Overlay Cracking 30 2,512
Other Flexible Pavement Medium Traffic Fair Condition6 200 45mm Overlay Cracking 30 2,542
Other Flexible Pavement Medium Traffic Fair Condition7 200 45mm Overlay Cracking 30 2,573
Other Flexible Pavement High Traffic Fair Condition 64 45mm Mill & Fill Cracking 25 2,598
Other Flexible Pavement Low Traffic Fair Condition1 213 45mm Overlay Cracking 32 2,630
2032 Other Flexible Pavement Low Traffic Fair Condition2 200 45mm Overlay Cracking 30 2,660
Other Flexible Pavement Low Traffic Fair Condition3 200 45mm Overlay Cracking 30 2,691
Other Flexible Pavement Low Traffic Fair Condition4 200 45mm Overlay Cracking 30 2,721
Interstate Flexible Pavement Medium Traffic Good Condition1 180 45mm Overlay Cracking 27 2,748
Interstate Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 2,779
Interstate Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 2,809
Other Flexible Pavement Medium Traffic Good Condition10 200 45mm Overlay Cracking 30 2,840
Other Flexible Pavement Medium Traffic Good Condition2 200 45mm Overlay Cracking 30 2,870
2033
Other Flexible Pavement Medium Traffic Good Condition3 200 45mm Overlay Cracking 30 2,900
Other Flexible Pavement Medium Traffic Good Condition4 200 45mm Overlay Cracking 30 2,931
Other Flexible Pavement Medium Traffic Good Condition5 200 45mm Overlay Cracking 30 2,961
Other Flexible Pavement Medium Traffic Good Condition6 200 45mm Overlay Cracking 30 2,992
Other Flexible Pavement Medium Traffic Good Condition7 200 45mm Overlay Cracking 30 3,022
Other Flexible Pavement Low Traffic Good Condition10 200 45mm Overlay Cracking 30 3,052
Other Flexible Pavement Low Traffic Good Condition11 200 45mm Overlay Cracking 30 3,083
Other Flexible Pavement Low Traffic Good Condition2 200 45mm Overlay Cracking 30 3,113
Other Flexible Pavement Low Traffic Good Condition3 200 45mm Overlay Cracking 30 3,144
Other Flexible Pavement Low Traffic Good Condition4 200 45mm Overlay Cracking 30 3,174
2034 Other Flexible Pavement Low Traffic Good Condition5 200 45mm Overlay Cracking 30 3,204
Other Flexible Pavement Low Traffic Good Condition6 200 45mm Overlay Cracking 30 3,235
Other Flexible Pavement Low Traffic Good Condition7 200 45mm Overlay Cracking 30 3,265
Other Flexible Pavement Low Traffic Good Condition8 200 45mm Overlay Cracking 30 3,296
Other Flexible Pavement Low Traffic Good Condition9 200 45mm Overlay Cracking 30 3,326
Interstate Flexible Pavement Low Traffic Fair Condition 7 45mm Mill & Fill Cracking 2 3,328
2036 Interstate Flexible Pavement Low Traffic Good Condition 12 45mm Mill & Fill Cracking 3 3,331
2039 Interstate Flexible Pavement High Traffic Poor Condition 14 45mm Mill & Fill Cracking 3 3,334
2040 Other Flexible Pavement Medium Traffic Poor Condition1 164 45mm Overlay Cracking 25 3,359
Interstate Flexible Pavement High Traffic Fair Condition 119 45mm Overlay Cracking 18 3,377
2042
Other Flexible Pavement High Traffic Poor Condition 23 45mm Mill & Fill Cracking 9 3,386
Other Flexible Pavement Low Traffic Poor Condition1 213 45mm Overlay Cracking 32 3,418
2043
Other Flexible Pavement Low Traffic Poor Condition2 200 45mm Overlay Cracking 30 3,449
(All costs are in millions of dollars)

101
 5: CONCLUSIONS AND SUMMARY

5.1 CONCLUSIONS

This study collected pavement related data, calibrated the HDM-4 models for

flexible pavements, and applied them to the WSDOT road network. The calibration

factors were shown to be reliable. Significant findings are:

1. HDM-4 can be used for the WSDOT road network.

2. Road deterioration models can be calibrated for flexible pavements. For the WSDOT

network, it is impossible to apply one set of calibration factors to the entire network.

The WSDOT road network requires calibration factors significantly different than

default HDM-4 values.

3. The current version of HDM-4 (version 1.3) does not provide meaningful analysis

output for WSDOT concrete pavements. Therefore, HDM-4 cannot be used to

analyze concrete pavements.

4. WSPMS data are essential to this research. The annually updated data make the

calibration procedure possible.

5. Current research mainly focuses on road deterioration models and road works effects

models. Further studies on traffic and environmental effects would benefit WSDOT.

5.2 SUMMARY

The research has contributed the following:

1. HDM-4 input data were collected and processed from existing WSPMS data and

other available sources. The data included traffic, climate and environment,

102
 vehicle mechanical characteristics, and costs and benefits of road works.

2. An integrated method of deriving and formatting pavement condition data from

WSPMS to HDM-4 was created and applied.

3. Most pavement deterioration models were systematically simplified, except in

several models where factors had negligible impacts on road performance.

4. A method of calibrating concrete pavement deterioration models is proposed.

Although HDM-4 cannot currently execute those models, the method can be used

for future applications.

5. An econometric software-LIMDEP was successfully applied to estimate

calibration factors based on local condition data and HDM-4 models.

6. All calibration factors of flexible pavement deterioration models (based on HDM-

4 models) were proved to be reliable by the validation process. The estimated

performance of road conditions matched the actual local conditions reasonably

well and proved that these factors can be used for WSDOT’s future applications.

7. Four major functions were derived on the basis of the three levels of analysis:

• Predict the required budget based on selected target road conditions.

• Produce strategies for selecting and timing of road works under varying

budget levels.

• Assist WSDOT and policy makers in determining the long-term effects of

different funding scenarios, including the effects of near-term funding cuts on

(1) long-term pavement condition, (2) long-term user costs and (3) future

funding required to restore the network.

103
 REFERENCES

Al-Omari, B. and Darter, M.I. (1992), “Relationship Between IRI and PSR”, Interim
Report for 1991-1992, Report No. UILU-ENG-92-2013, Illinois Department of
Transportation, Springfield, IL.

Corley-Lay, J.B. (1998), “Friction and Surface Texture Characterization of 14 Pavement

Test Sections in Greenville, North Carolina”, Transportation Research Record
1639, National Research Council, Washington, D.C. pp. 155-161.

Greene, William H. (1995), “LIMDEP user's manual: version 7.0”, Bellport, N.Y.,
Econometric Software, 1995.

Jayawickrama, P.W. and Thomas, B. (1998), “Correction of Field Skid Measurements

for Seasonal Variations in Texas”, Transportation Research Record 1639,
Transportation Research Board, National Research Council, Washington, D.C.
pp. 147-154.

Kay, R.K.; Mahoney, J.P. and Jackson, N.C. (1993), “The WSDOT Pavement
Management System - A 1993 Update”, Washington State Transportation Center
(TRAC), WSDOT, OlympiaKerali, H.G.R. (2000a), “Highway Development and
Management (HDM-4) Volume 1: Overview of HDM-4”, the World Road
Association (PIARC), Paris and the World Bank, Washington, D.C.

Kerali, H.G.R., McMullen Derek, Odoki, J.B. (2000b), “Highway Development and
Management (HDM-4) Volume 2: Application Guide”, the World Road
Association (PIARC), Paris and the World Bank, Washington, D.C.

Muench, Stephen T. (2003a), “HDM-4 Works Effects:”45 mm Overlay without

Milling”, Technical Report (unpublished).

Muench, Stephen T. (2003b), “Conversion of WSPMS Flexible Pavement PSC to HDM-

4 Variables”, Technical Report (unpublished).

Muench, Stephen T. (2003c), “Conversion of WSPMS Rigid Pavement PSC to HDM-4

Variables”, Technical Report (unpublished).

Odoki, J.B. and Kerali, H.G.R. (2000), “Highway Development and Management (HDM-
4) Volume 4: Analytical Framework and Model Descriptions”, the World Road
Association (PIARC), Paris and the World Bank, Washington, D.C.

104
Raymond, Christopher M. (2002), “Analysis of Influences on As-Built Pavement
Roughness in Asphalt Overlays,” International Contest on LTPP Data Analysis,
2002.

Washington State Department of Transportation (2001), “Maintenance Accountability

Process,” Manual, Maintenance Office, Washington State Department of
Transportation, Olympia, Washington, Feb. 2001.

Washington State Department of Transportation (2002), “Measures, Markers and

Mileposts—Gray Notebook—Quarter Ending March 31, 2002,” Fifth Edition,
Washington State Department of Transportation, Olympia, Washington, Mar.
2002.

Washington State Department of Transportation (2003), “Washington State Pavement

Management System (WSPMS)”, 2003.

Xiaoping Zhang and Naveen Kumar Juwa (2001), “Vehicle Fleet and Traffic Flow Types
in Washington State” (unpublished).

105
 LIST OF APPENDICES

Because many of these appendices would be too large and unwieldy to print, they

are available electronically.

Appendix A: Climate Data of Washington State

Appendix B1: Road Network Definitions for Project Analysis

Appendix B2: Road Network Input Table for Project Analysis—High ESAL ACPs

Appendix B3: Road Network Input Table for Project Analysis—Medium ESAL ACPs

Appendix B4: Road Network Input Table for Project Analysis—Low ESAL ACPs

Appendix B5: Road Network Input Table for Project Analysis—BSTs

Appendix B6: Road Network Input Table for Project Analysis—PCCPs

Appendix C1: Vehicle Fleet Definitions for Cars

Appendix C2: Vehicle Fleet Definitions for Single Units

Appendix C3: Vehicle Fleet Definitions for Double Units

Appendix C4: Vehicle Fleet Definitions for Trains

Appendix C5: Data Sources of Vehicle Fleet Input

Appendix C6: Vehicle Fleet Input Table

Appendix D1: Road Network Definitions for Strategic Analysis

Appendix D2: Road Network Input Table for Strategic Analysis—ACPs

Appendix D3: Road Network Input Table for Strategic Analysis—BSTs

106