LOST TIME ANALYSIS - A KEY

TO PRODUCTION OPTIMIZATION

Determining equipment potential and accurately tracking determine if the losses were due to reasons
associated with the operating department or
production losses can help mill personnel maintain high due to external reasons.
operating profitability, states Dr. Ron Sanclemente in the Aurell and Isacson (1) described a
simplified version of the Swedish SSG
following article. Dr. Sanclemente is Manager of Pulp Standard 2000, for lost time analysis. They
Projects at Jaakko Poyry Fluor Daniel, Raleigh, North outlined the modified Swedish process and
described a basic system for collecting data
Carolina. and analyzing the information. They point
out the difficulty in defining maximum
If the above statement sounds familiar production potential (production at
Optimizing production, operating costs
then a program should be implemented capacity) for a department.
or product quality in a pulp and paper
mill is dependent on maintaining a high which tracks production losses due to
operating efficiency in all departments equipment slowback and equipment Production Potential
of the mill. This requires an operational, downtime. Such analyses lead to in- If the impact of individual departments on
maintenance and capital program which creased equipment availability which will lost production is not important, then ar-
responds to all of the issues contributing improve productivity reduce operating bitrarily assigning a value for production
to lost production. costs and improve product quality. potential is all that is necessary. If com-
Most mills track downtime, in par- A key element in effectively tracking parison of departments and their relative
ticular extended downtime, and most of lost production is to maintain simplicity impact on lost production is important then
these same mills react to the issues in collecting, compiling, analyzing and a method which defines the rate at which a
which contributed to the downtime. It presenting the data. Reasons for loss in department can operate is required. The
isn't unusual, however to hear the lament production should he identified and their. method must be applicable to all
"We worked on the obvious impact quantified in a m a n n e r w h i c h departments and must compare them on a
maintenance problems and fixed those i s easy to understand and interpret. This similar basis.
things which caused downtime, but should include not only the obvious
"downtime" but should also include pro- Difficulty in defining the maximum
equipment availability and productivity
duction lost due to "slowback" of the sustainable production level can lead to
are still low, costs are still high and
equipment. In addition it is important to arbitrarily assigning a value which esti-
product quality is still variable.”
mates the potential. Typically production
potential, for an operating unit, is defined by
nameplate or design criteria, by the
Figure 1 operating superintendent drawing on process
experience, or by any number of other
methods. Each of the methods is ambiguous
since it does not utilize actual production
data.
1,2 Lacking a good definition of the pro-
1,1 Maximum sustainable production duction capacity for individual departments,
1 most mills track downtime, but do not track
last production associated with slowback. A
Bleached air dry tpd (000)

0,9
0,8
majority of the lost production capability is
Average production
0,7
often a result of slowback, rather than
downtime. This means that many of the
0,6
issues associated with lost production are
0,5
not being addressed.
0,4
In order to remove the ambiguity as-
0,3
sociated with defining departmental pro-
0,2
duction potential, actual operating data
0,1
should be used. The duration curve, a
0
method routinely utilized by Jaakko Poyry,
0 10 20 30 40 50 60 70 80 90 100 %
removes this ambiguity.
Production for a period of time, typically
the daily average for the previous year, is
sorted from the day of greatest production to
the day of lowest production. In order to
clearly understand how
production has varied over the year a will provide early insight into future mation should be entered daily to allow
curve showing sorted production versus maintenance issues. easy access, on an as needed basis. Once a
percentage of operating time is prepared A simple form for collection of the month the data should be sorted by reason
as shown in Figure 1. The maximum required information, by production per- for lost production and further analyzed to
sustainable production level is defined as sonnel, is shown in Figure 2. Input of provide direction for the maintenance.
that production level which reported pro- information begins on a new sheet at the engineering and production departments.
duction exceeds 10% of the time. If data beginning of the day shift, commonly at It is important when generating this
for a 365 day period is compiled and 07:00 hours. The rate at this time is en- information that slowback of the equip-
sorted from the day of greatest production tered, and the reason for being at that rate ment is included and that the slowback is
to the day of lowest production, the avera- is carried over from the previous shift. It defined in equivalent downtime units. For
ge production of the 36th and 37th day is is evident in Figure 2, that there were example, if a piece of equipment is rated at
the maximum sustainable rate. This Tech- several reasons for lost production and 100 tons per day and is operated at 50 tons
nique can be applied to any operating not all of the reasons were associated with per day then the slowback, in equivalent
department or any piece of equipment in shutting down the equipment. downtime, would be 12 hours. The same
the mill. Lost production in the pulp mill Lost time should be collected by production could have been attained by
is equivalent to lost production on the every major production unit in the mill. operating the equipment at a rate of
paper machines, or any other part of the such as, Kamyr Digester # 1, Machine #2,
mill, if this method is used to define Reco-
production potential.
A close look at Figure 1 clearly Figure 2:
shows that the maximum sustainable rate
for the pulp mill is 1000 BADT/d. It can
be seen that the average rate for the
period of time included is 920 BADT/d, PULP MIL LOST PRODUCTION REPORT DATE 1/1/89
and that this average rate was exceeded
69% of the time. Another feature of the TIME RATE I/E REASON EXPLANATION
duration curve, which becomes apparent 700 15 2 3 Evaporators down for water wash
after a large quantity of production 800 15 2 3 Ramping up after evaporator wash
information has been gathered from many 850 16 1 6
different mills, is the ratio between the 1000 0 1 5 Kicked out
maximum sustainable rate and the average 1020 15 1 5 Ramping up after HPF KO'd
rate. This ratio has b e e n defined by 1100 16 1 6
Jaakko Poyry as the Capacity Efficiency 2020 0 2 7 Power failure
(CE). The ratio can act as an indicator as
2310 14 2 7 Ramping up after power failure
to how well the mill is running and when
110 16 1 6
compared to similar mills, can provide an
indication of the production level that can 700 16 1 6
be realistically expected.

Collecting the Data STANDARD REASONS FOR LOST PRODUCTION

The method used for collection of the data 1. Scheduled 5. High Pressure Feeder
should be simple and easy to use. The 2. Pulp Inventory 6. Operating at design rate
data should be entered by the operators 3. WBL Inventory 7. Other
since they can most easily define the 4. WL Inventory
reason for the lost time. If data collection
is too complex daily time constraints, in Note: If lost production is due to internal reasons put a “1” in the I/E column, if the
operations, will result in lack of interest reason is extern to the pulp mill put a “2” in the I/E column.
or lack of time to provide the necessary
information.
The rules for determining if an entry
is to be made on the downtime report are very Boiler #1. etc, This will allow the 100 tons per day for 12 hours and then
very simple. If the operator makes a potential for the various departments to be taking the equipment down for the re-
change in the rate of the process. then an compared and will provide information as maining 12 hours.
entry is made. The time of the change, to where resources can best be applied.
new process rate, and reason for the rate Analyzing the Data
change are noted on the downtime report. Compiling the Data
The reason is further clarified by deter-
mining if it was internal or external to the Data is of little value unless it is analyzed
process. In order to make the job easier, a Spreadsheets. combined with personal and distributed to the proper people in a
list of standard reasons for slowback or computers make the job of compiling and form which can help them do their job
downtime can be provided with a catchall manipulating the data easy. A spreadsheet better. After sorting the data by reason, for
"other" category. If the "other" category is (Figure 3) can be prepared which will lost production it is easy to develop a
used it is important that the reason for lost separate the lost production by Internal Pareto chart. This chart will clearly show
time be clearly defined. Analysis of the DownTime (IDT), Internal SlowBack the reasons for lost time and provide
"other" category will provide additions to (ISB), External DownTime (EDT) and visual information about their relative
t h e list of standard r e a s o n s and External SlowBack (ESB). The spread- significance (Figure 4).
sheet should be compatible with the Equipment problems which contribute
downtime report on which the daily infor- to significant lost rime, such as the power
mation is manually entered. The infor- failure shown in Figure 3, are normally
investigated and a plan is quickly deve-
loped to solve the problem. The trouble Figure 4: A Pareto chart can be used to show the reasons for
with looking at the large lost time items lost time and provide relative significance far each reason.
is that those production problems which
recur, such as weak black liquor (WBL)
inventory, are ignored. The Pareto
Chart, illustrated in Figure # 4, indicates 4 ,5
that, while the power failure did 4 ,0
contribute to lost production. WBL

Min. Lost Production (000)

inventory had a much more significant 3 ,5
impact on production loss. 3 ,0
Another useful visual representation,
2 ,5
that can be developed from the compiled
data, is a bar chart comparing equipment 2 ,0
availability and equipment utilization. 1 ,5
Aurell and Isacson (1) defined, and ex-
plained, Availability and Utilization as 1 ,0
lost time parameters. Availability is very 0 ,5
simply the percentage of time a
particular piece of equipment is 0 ,0
available to produce product. Utilization
000

!*

*!
is the percentage of time the equipment

(
'
actually produces.

&
,#

1
/

%
.

*
A bar chart comparing Availability

&!

&
-

and Utilization defines the bottleneck(s) in

the process (Figure 5). When Availability
and Utilization are the same, that part of
utilization of the evaporators is almost the historical level of availability.
the process is a bottleneck. If Availability
same. indicating that the evaporators are a
is low, the bottleneck can be eased by The information collected by this
bottleneck. Closer inspection of the lost
focusing on the internal reasons for process provides real data to justify capi-
production information collected for the
slowback or downtime. If Utilization is tal expenditures and can clearly define
evaporators indicates that the evaporators
low the bottleneck can be eased by the impact that capital expenditure will
are having to be water washed every 2 days,
focusing on the external reasons. If neither have on the process. This information
A review of historical information shows
is low, the bottleneck can only be eased by will aid in the development of realistic
that in the past, the evaporators were boiled
increasing the production potential for the ROI information for the capital request.
once every two weeks. In the last year, the
equipment. This very often requires the
evaporators, have progressively required
expenditure of capital. Distribution of the Analysis Each mill
more frequent water washouts. Inspection
It could have been concluded from the needs to decide the most appropriate
of the evaporators showed that most of the
Pareto charts shown in Figure 4 that pro- individuals to receive the results of the
robes were plugged and that they should be
duction could have been increased by analysis. It is important to determine the
hydro-blasted or acid washed. After
merely installing more WBL storage ca- level of information required by each
cleaning, the evaporators came back to a
pacity. Figure 5 indicates that this is not person. Some individuals will require a
necessarily the case. The availability and
Figure 3:
!
"# ! $ % ! $ %"#

PULP MIL LOST PRODUCTION REPORT

Month - January
Date Time Time Rate Rate I/E Reas Explanation of Variance ISB IDT ESB EDT
from to from to on

" #$ !
! %&' "()

& *
!
! *!
++
++

,* #$
very sophisticated and complete Figure 5: & '
analysis of the data, while others will
require only the summary.
Production personnel can use the
data to determine which issues should
100 %
be addressed and which areas should be
90 %
monitored in order to improve 80 %
operations. Engineering personnel can 70 %
use the material to determine future 60 %
capital expenditures and to provide 50 %
justification for those expenditures. 40 %
Maintenance personnel can use the 30 %
information to assure that they are 20 %
addressing the "critical few" 10 %
maintenance issues and not just greasing 0 %
the "squeaky wheel". A close review of

*
#3
$

"
the "other" category in the reasons for

#
#
2

4
!
slowback or downtime will provide

2
insight into future maintenance issues.
Management can use the analysis to
assess the operation of different depart-
ments, on an equal basis. Knowing what
the bottlenecks are will help management
decide how to spend capital and allocate
resources to increase production,
improve quality and reduce operating
costs.
Definitions: Calculations:
Conclusions
The goal of any production unit is to 1. Internal Down Time (IDT) Slow Back
operate safely, at maximum rate at the
lowest cost and at the same time to provide The time during which no product is
a product whose quality meets the produced due to reasons internal to the TRR * ( MSR-AR)/MSR
requirements of the customer. process.
Analysis of the potential for produc- TRR = Time at Reduced Rate (min.)
tion improvement and recognition of the 2. External Down Time (EDT)
MSR = Maximum Sustainable Rate
impact that slowback has on production The time during which no product is
is one of the tools which will assist produced due to reasons external to the AR =Actual Rate
operations in meeting all of the above process. Availability - %
goals.
A difficulty which arises when 3. Internal Slow Back (ISB)
trying to determine production losses, (TT - IDT - ISB)/TT
due to equipment slowback, is defining The time during which product is made,
IDT = Internal Down Time
the potential for a piece of equipment or at a reduced rate due to reasons internal
an area. The duration curve provides a to the process, expressed as equivalent ISB = Internal Slow Back
method by which production potential downtime.
TT = Total Time
can be determined, This method is
applicable to all areas of the mill. 4. External Slow Back (ESB) Utilization - %
Collection and compilation of the
The time during which product is made,
data should be kept simple. Personal
at a reduced rate due to reasons external
computers have been shown to be very (TT - IDT - ISB- EDT - ESB)/TT
to the process, expressed as equivalent
valuable, for compiling and EDT = External Down Time
downtime.
manipulating the data which has been
collected. Once the data has been ESB = External Slow Back
collected and compiled it is important 5. Availability
that it be presented in a form which is The percent of time a process is
easy to understand. The Pareto Chart available to produce product.
and a bar chart comparing equipment
availability and utilization are two ways 6. Utilization
which can be used to present the data in
a form which is easy to understand and The percent of time a process actually
interpret. produces a product
Finally, it is important that the
information be distributed to those 7. Maximum Sustainable Rate (MSR)
individuals in the mill who can best The production rate, as defined by the
utilize it. The distribution should duration curve, which could be sustained
include those actually doing the at 100% Availability.
maintenance as well as members of
management who impact how time and
capital are allocated.