Integrated Cost / Schedule Risk Analysis

David [Link], Ph.D. Bill Campbell, III

Hulett & Associates, LLC Risk Mitigation Associates
(310) 476-7699 (714) 694-0174
[Link]@[Link] bill@[Link]
[Link] [Link]

Abstract. Project costs often exceed their estimates applied per unit time (“burn rate”) and compensation
because those estimates do not take into consideration per hour for those resources, and uncertainty in the cost
the actual duration of project activities. Cost risk will of activities whose costs do not depend on elapsed time.
also be underestimated if it does not take into
consideration schedule risk. This paper presents a SCHEDULE RISK ANALYSIS
method of incorporating the uncertainty in activities’
durations into the assessment of cost risk. In this
Name Duration Start Finish 4th Quarter
1st Quarter
2nd Quarter
3rd Quarter
4th Quarter
1st Quarter
method, a Monte Carlo simulation of the schedule Start 0d 1/2 1/2 1/2















provides uncertainty in time. Incorporating the Software Subsystem 265 d 1/2 9/23                















schedule risk results into the cost risk model provides Design Software 65 d 1/2 3/7 1/2
               

  
3/7
 


                   


           











the linkage between schedule and cost risk. Then Code Software 150 d 3/8 8/4 3/8      

                    


       








8/4


         











8/5 9/23
equivalence must be established between the schedule Test Software 50 d 8/5 9/23
                 

Hardware Platform 260 d 1/2 9/18

    

and network concepts. Uncertainty in costs is then 

1/2
             

3/2

    

Design Hardware 60 d 1/2 3/2                 
    

represented by uncertainty in “independent costs” (costs Build Hardware 165 d 3/3 8/14 3/3
   




   

8/14











that do not depend on time) and “variable costs” (costs QC Hardware 35 d 8/15 9/18




8/15



9/18








that depend on uncertain time and cost per unit time or Integration 120 d 9/24 1/21


















    





“burn rate” and rate of labor compensation.) Integrate Hardware / Software

80 d 9/24 12/12

9/24  
12/12
 

12/13 1/21

    

Simulation of the cost model combines the results from Test Integrated System40 d 12/13 1/21
        

Deliver System 0d 1/21 1/21

    

1/21
the schedule risk analysis with the uncertainty in the
    

cost assumptions. The results include the probability

distribution of total project costs and sensitivity of that Figure 1. The schedule before risk analysis
distribution to the different inputs. Issues are discussed
and simplified examples are provided. Schedule Risk. The strategy for integration of
schedule and cost risk begins with an analysis of the
INTRODUCTION risk of the schedule. The dates computed in Figure 1
Cost and schedule are related. The estimate at assume all activity durations are known with certainty.
completion (EAC) for a project makes many The scheduling software indicates a finish date for
assumptions. Key among those are the resources that system delivery of January 21. Unfortunately we
will be working on the project activities, their rate of cannot estimate durations with accuracy. Often the
compensation and their durations for the activities that uncertainty in the durations is represented by a 3-point
react to duration. Each of these assumptions contains estimate with optimistic, most likely and pessimistic
potential risk, where risk is defined as an uncertain scenarios defining the three points. These scenarios are
event or condition that, if it occurs has a positive or typically discovered by in-depth interviews of those
negative impact on the cost objective. (PMI 2000) Of who understand the project and its risks. Figure 2
course, even for those activities where cost does not shows a possible set of 3-point estimates that are
respond to duration, there is uncertainty. assumed to represent triangular distributions.

Integrating risk in cost and schedule. Cost estimates

often become disconnected from the facts of the project
that are sometimes different from the assumptions. The
risk that project cost will exceed the estimate is
assessed by taking into account several types of risk,
most notably the risk that project activities’ durations
may differ from those originally assumed (schedule
risk). Other cost risks include the resources to be
 ID Name Duration @RISK: Functions
1 Start 0d Risk Critical
2 Software Subsystem 265 d Finish=RiskOUTPUT() Activity Index
3 Design Software 65 d Duration=RiskTRIANG(50,65,80)
4 Code Software 150 d Duration=RiskTRIANG(140,150,165)
Design Software 58%
5 Test Software 50 d Duration=RiskTRIANG(45,55,70) Code Software 58%
6 Hardware Platform 260 d Finish=RiskOUTPUT() Test Software 58%
7 Design Hardware 60 d Duration=RiskTRIANG(50,60,80)
Design Hardware 42%
8 Build Hardware 165 d Duration=RiskTRIANG(135,165,190)
9 QC Hardware 35 d Duration=RiskTRIANG(30,40,60) Build Hardware 42%
10 Integration 120 d Finish=RiskOUTPUT() QC Hardware 42%
11 Integrate Hardware / Software 80 d Duration=RiskTRIANG(65,80,100) Integrate Hardware / Software 100%
12 Test Integrated System 40 d Duration=RiskTRIANG(30,40,65)
13 Deliver System 0d
Test Integrated System 100%
Deliver System 100%
Figure 2. Add risk to each activity
Figure 4. Risk criticality of activities/paths
Schedule risk results. The risk results provide
possible dates and their probabilities for system INTEGRATION OF SCHEDULE RISK INTO
delivery. This distribution of completion dates is found THE COST ESTIMATE
by simulating the schedule many times. In each
Matching cost and schedule concepts. The cost
iteration a duration is picked at random from the
estimates for each activity should be driven by a
distributions shown in Figure 2 for each activity.
realistic estimate of the duration of the tasks with which
Because we do not know which combination of
they are connected. Similarly, the risk in the cost
durations will actually occur, we tell the computer to
estimate is driven by the risk in the estimate of duration
conduct, say, 3,000 iterations, each with durations
for the same activities. To drive the cost risk for each
selected at random according to the probability
activity by the correct schedule risk, we need to match
distributions for each task’s duration. A probability
the cost estimates to their corresponding schedule
distribution of possible delivery dates results and is
items.
shown in Figure 3.
Finding schedule activities that are the same as the
Distribution for Finish of Integration entries in the cost estimate is not as easy as it should be.
Often the cost estimate is related to the work
1 breakdown structure (WBS) but the schedule is not.
Prob of Value <= X-axis

0.9
0.8 We found that it was easier to take the dollars in the
0.7 CPM date cost estimate and apportion them into the schedule
0.6
is very
Value

0.5 summary tasks than vice versa. The uncertainty in each

0.4
0.3
unlikely summary activity can then drive the appropriate number
0.2 of dollars. This mapping is shown in Figure 5.
0.1
0
12/20/02

12/31/02

1/11/03

1/22/03

2/13/03

2/24/03

3/18/03

3/29/03
2/2/03

3/7/03

Map to Schedule
Cost Element ($ 000) Schedule Element ($ 000)
Summary Activity:
A S/W 4,594 A S/W 5,359
Date
B H/W 3,804 B H/W 4,569
C Integ. 2,077 C Integ. 2,842
Figure 3. Distribution of delivery dates A/B/C Mat. 1,976
A/B/C P.M & Support 320
These results indicate that the date of 2/21 is very Total Cost per EAC 12,770 Total Cost per Sch. 12,770
unlikely. There are other results from the schedule risk
analysis. The software shows which activities and Figure 5. Map Costs to Schedule Activities
paths are most likely to delay the project to assist risk Schedule element input to the cost model. Total cost
response planning. Figure 4 shows that Integration is is the sum of variable and independent costs. Each of
always on the critical path and that software is the next these has uncertainty. Variable costs are the product of
most likely to delay the project. This information is three items and each is uncertain.
quite helpful to the project manager.
 schedule simulation easily because we used the same
maker’s software for each of the simulations. Being
Input Item Uncertain Source of Data specific, the schedule was simulated using @RISK for
Schedule Risk (Microsoft) Project that produced a histogram as an
Duration Yes
Anaylsis output. We used a 30-point distribution in this analysis
Histogram for accuracy. An example of that histogram might be
“Burn Rate” Yes Interviews the following: Duration=RiskHISTOGRM(242,383,
Compensation {0.001,0.004,0.009,0.017,0.0255,0.032,0.053,0.0715,0.
Yes Interviews
rate 0655,0.0895,0.088,0.09,0.0955,0.0885,0.056,0.0545,0.0
485,0.032,0.027,0.019,0.013,0.0115,0.0015,0.0045,0.00
Figure 6. Components of Variable Cost 15,0,0.001,0,0,0}).
The first problem is to find a common denominator for The cost risk model was programmed in Microsoft
these three items. Hours can be used, but integration (MS) Excel. The simulation program for the cost risk
with the schedule is easier if they are translated to days. model was @RISK for Excel, which accepts the
The second problem is to determine the uncertain histograms from its cousin in Project. If the software
duration of the schedule summary elements. Each of that simluates the Excel model does not accept the
these inputs is the probability distribution of the outputs of the schedule risk model, for instance if Risk+
appropriate summary activity, such as those listed in is used to simulate MS Project and Crystal Ball
Figure 5, above calculated from the simulation of the simulates the Excel model, an approximation of the
schedule. This distribution or “histogram” shows the schedule risk outputs must be made and input into the
different duration that the summary activity could take cost risk model, introducing some error that may be
and their relative likelihood. large or small. A larger error might occur if there are
probabilistic branches in the schedule and the histogram
Calculation of the number of days the Software from the schedule risk analysis has two maxima. It
Subsystem and Hardware Platform take can be found would be difficult to estimate a single distribution or to
directly from the simulation of the schedule because make a custom distribution in Excel for such a shape.
these two summary activities start on a fixed day, the
first day of the project. (We have used 7-day weeks in Collecting variable cost risk data. The data collection
the schedule because the days will be translated to process starts by defining the components of cost that
Excel that does not distinguish weekends from are needed to simulate the cost risk analysis. These
weekdays in computing the date mathematics. Thus the items are all uncertain unless they are in the past. In
schedule date will be wrong, unless it is a plant fact, if the analysis is done in the middle of the project,
turnaround with 7-day schedules but the duration in there should be a history of “actuals” that would be
days should be accurate.) certain and added onto any uncertain estimate of costs
The histogram from the schedule simulation for the yet to go.
Integration summary task will not represent faithfully The first element to be collected from the project team
the number of days of integration work because its leaders is the “burn rate” or number of resources (hours,
starting date is not fixed but depends on the completion heads) applied to the task per unit time (day). The burn
of the longer of its two uncertain predecessor paths, rate (e.g., heads per day) will be multiplied by the
software or hardware. We simulated the Integration duration (e.g., days) to compute the resources (e.g.,
tasks separately, starting from an artificial fixed date, to total hours) for the schedule summary task.
develop the histogram of uncertain integration days that
abstracts from the uncertain start date for those tasks. The burn rate is a partly-new concept for the teams
because they tend to think in time-phased terms. We
Integration in the Original Schedule
Integration Summary 120 d 9/24 1/21 Finish=RiskOUTPUT()
could not incorporate any time phasing of staffing in
Integrate H/W & S/W 80 d 9/24 12/12 Duration=RiskTRIANG(65,80,100) the simulation model, so the concept had to be the
Test System 40 d 12/13 1/21 Duration=RiskTRIANG(30,40,65) equivalent based on an assumption of level staffing. It
Deliver 0d 1/21 1/21
turned out that this concept was easy enough to
Integration Simulated Separately to Derive a Histogram in Elapsed Days estimate once it was explained and the team was given
Integration Summary 120 d 1/2 5/1 Finish=RiskOUTPUT() encouragement.
Integrate H/W & S/W 80 d 1/2 3/22 Duration=RiskTRIANG(65,80,110)
Test System 40 d 3/23 5/1 Duration=RiskTRIANG(30,40,65)
Deliver 0d 5/1 5/1
The team leader was expected to meet with the team on
all of these data collection tasks and develop 3-point
Figure 7. Simulate a downstream path estimates that could be turned into probability
The histogram is transferred to the cost model from the distributions for simulation. For burn rate we provided
data on the estimate of burn rate computed from the
 cost estimate. This is found by computing the total estimate was quite different from that elicited from the
hours for the tasks in the base estimate and dividing by interviews. The table below is a typical entry for the
the number of days assumed in that estimate. The uncertain compensation rate for variable costs.
result is either in hours or headcount. The team fills out
the table with their first estimate of optimistic (low), Uncertainty in Labor Rate
most likely and pessimistic (high) number of hours or Labor Rate
heads per day.
Dollars/Hour (Using current data) $122.12 $110.00 $130.00 $160.00
Note that the number of resources that emerges from
this exercise need not be the number in the base Figure 9. Interview data on labor rate
estimate. Sometimes the base estimate is out of date. Collecting risk data on independent costs. The term
Sometimes it was “squeezed” to fit some target handed “independent cost” refers to the cost of the project
down by management or the customer and was never a elements that do not rely on the duration of the activity.
true estimate of cost. Given these factors, it is not Cost elements in this category might include software
surprising that the “most likely” estimate of resources releases that are provided by a vendor, a component or
per day is often not that implied by the original piece of equipment provided by a subcontractor, or a
estimate. Infrequently, but sometimes, the optimistic piece of purchased equipment. The basic characteristic
(low) estimate of resources is above that implied by the of these could be that their cost, while uncertain, does
base estimates. not depend on the date of arrival. Thus, while we might
The team is then interviewed on their first guess at the not know how much a piece of equipment may cost at
three-point estimates. The most common problem with an early point in the project before we get a firm
these estimates is that they are too narrow. It is well- quotation, its cost will not necessarily be higher if it
established that people who have no or little experience arrives later rather than earlier in the project.
in providing data on risk will initially underestimate the There were really two types of independent costs
true uncertainty in the numbers. For this reason, an in- elements in the analysis. One, the simpler, would be
depth interview is needed to check both their represented by a piece of equipment or purchased item
understanding of the concepts and also their that had no time dimension at all. The 3-point estimate
representation of the optimistic, pessimistic and most for this element of cost has the same concept as is
likely estimate. The following table shows the result of usually found in direct estimation of cost risk. The
one of those interviews. The hours and equivalent interviewee prepares three scenarios representing the
headcount are both provided and teams may differ in low, most likely and high estimates of cost. These
the way they want to provide the data. The translation scenarios are discussed and sometimes challenged
can be made in the simulation. during the risk interview and three estimates of cost are
Uncertainty in Variable Hours developed that correspond to the scenarios. These three
estimates are used to develop a probability distribution,
Nominal Low Most Likely High
often a triangular distribution, of the risk that the
Duration Dependent Labor
element will cost more or less than estimated. The cost
EAC Burn Rate (hours/day) 118.5 115.0 120.0 165.0
model simulation uses samples directly from this
(Equivalent headcount) 14.8 14.4 15.0 20.6
distribution.
Figure 8. Interview data on “burn rate”
Three-point estimates for other so-called independent
To review, the total variable cost is found by costs were really derived from assumptions about how
multiplying (1) the number of days’ duration by (2) the long activities would take. Some of these independent
hours per day and then by (3) the compensation per cost items were actually estimated with assumptions of
hour. Each input comes from a specific source as hours, but not necessarily calendar duration. This
shown in Figure 6, above. category included items that did not have any clear
linkage to an available schedule concept. (Risk
The compensation per hour really represents the skill
analysis is often the art of the possible, and sometimes
mix of the people presumed to be on the job. If skill
even the best concepts cannot be implemented.)
levels 2, 3 and 4 are used in the original estimate, an
average compensation can be computed from the Correlation of uncertain cost concepts. Correlation
weighted average of those to be assigned in the baseline often occurs in projects when the elements’ costs might
plan. During the teams’ deliberations they often noticed be expected to move together in real projects.
that there were a predominance of levels 4 and 5 on the Elements’ costs could move together if a common
job. This finding, or projection, implied that the rate outside influence is driving them.
per hour in the estimate was not accurate. In several
One example of correlation between elements’ costs
instances, the skill mix assumed in the initial baseline
can be expected when a particular task or group of tasks
is more difficult than expected in the original estimate.
For instance, suppose software-coding activities are Sample Correlation
taking more than 265 days in the current schedule. This Correlation Pairs Correlation
might occur if the software turns out to be more
difficult than originally expected, or if the estimates of Duration: Burn Rate 0.5
software coding durations were unwarrantedly short as
Duration: Labor Rate 0.5
in a “success-oriented” schedule. Perhaps the original
assumptions, both for cost and schedule, included reuse Burn Rate: Labor Rate 0.5
of existing software, but that turns out not to be
feasible. Figure 10. Example Correlations
If software (or hardware or integration) tasks are taking In the real project the independent costs could be
longer several things might happen: (1) the schedule is correlated with these as well. Also, in the real project
in the high end of its estimated distribution, (2) project the teams were queried about the correlations. This
management puts more people on the activity concept was unfamiliar to them and had to be explained
increasing the per/time unit burn rate, and (3) the at some length. Some interviewees did not expect to
people added are more skilled than the average assumed see any correlation and others did.
in the estimate, which drives up the compensation or
RESULTS
labor rate.
Main findings. The cost estimation model was
In this scenario, it would be possible and even expected
simulated using uncertainty in the schedule duration in
that the three items in question would all be above their
days represented by the histogram from the schedule
level in the original cost estimate. That is, the time risk analysis, variable cost uncertainties represented by
would be longer, the hours per day more and the labor
burn rate and labor rate uncertainties, and independent
rate higher, and these items would all be expected
cost uncertainties. Correlations were inserted. The
together in the same project. The effect could also be
results show that it is very important to represent
expected to operate in reverse. That is, if the software uncertain durations along with the traditional cost risk
(hardware, integration) tasks turn out to be easier than
analysis variables.
originally estimated, then duration would be shorter, the
head count lower and the skill level (hence the labor The model was set up so that different combinations of
rate) lower together. assumption uncertainties could be turned on and off.
This allowed us to identify sensitivities and compare
It would therefore be said that these three items are
the impact of cost risk variables against the schedule-
positively correlated. That is, if many projects were
related uncertainty. For instance, in the first sensitivity,
done and some had harder software (hardware,
one of the scenarios, the “Schedule Risk Only”
integration) tasks than originally estimated, they would scenario, looked like this:
find that all three types of cost elements were moving
against them. These simultaneous adverse events would
reinforce each other when multiplied for variable costs
to produce a sort of magnified high cost of the project.
The same effect would work in reverse to produce
significantly low cost results. This is the working of
correlation.
In this demonstration cost risk model we installed
correlations between the three pairs of elements as
follows.
 is addressed in Figure 13 below.
Risk-Type Variables Turned On
Distribution of Total Program Cost vs. EAC
Schedule Variation Yes Cost Risk Only, Sche dule Risk Only, Cos t & Schedule Ris k

Independent Hour Variation No 100.0%

Burn Rate Variation No 80.0%

Labor Rate Variation No

60.0% Schedule Risk Only

Percentile
Material Variation No Cost Risk Only
Cost & Schedule Risk
40.0%
EAC
Figure 11. Specifying partial scenarios:
Schedule-risk-only- scenario 20.0%

The first question was whether the traditional cost risk 0.0%
11.0 13.0 15.0 17.0 19.0

items such as independent cost, burn rate, labor rate or $Millions

material variations were more important in determining

the overall cost risk than schedule risk (histograms from Figure 13. Simulating schedule and cost
the schedule risk analysis) alone. For this sensitivity, risk elements together
we specified the scenario in Figure 11 above, than
reversed the items for the cost risk analysis alone. The results in Figure 13 indicate that the total cost risk
These results are shown in Figure 12. of this project would be incompletely specified if only
cost risk elements or only schedule risk elements were
Distribution of Total Program Cost vs. EAC
Schedule Risk Only, Cost Risk Only
used to represent project risk. The cumulative
distribution of project cost uncertainty is further to the
100.0% right than either cost or schedule elements alone would
indicate. Of course it is very unlikely that a cost risk
80.0%
analyst would incorporate only the uncertain schedule
60.0%
data in a cost risk, but it has been known to happen.
Percentile

Schedule Risk Only

Cost Risk Only
Usually the cost risk is represented by the uncertain
40.0% EAC cost risk elements alone. If schedule risk is not taken
into account the cost risk will be underestimated, in
20.0% some cases by a substantial margin.

0.0%
Figure 14 below indicates the degree to which the
11.0 12.0 13.0 14.0 15.0 16.0 17.0 18.0 19.0 integrating schedule risk makes a difference.
$Millions

Compare Cost, Schedule and Integrated Risk

Cost &
Cost Risk Schedule
Figure 12. Scenarios of cost risk and Schedule
Only Risk Only
Risk
schedule risk separately ($millions)
Minimum 11.5 11.6 12.3
The curves show that the cost risk scenario has the
Maximum 16.2 15.1 19.1
greater impact on the uncertainty in project costs, with Mean 13.9 13.4 15.8
some iterations being somewhat below and others being 10% 12.8 12.5 14.0
quite a bit higher than the EAC, which is shown as a 20% 13.0 12.7 14.4
triangle. This shows that even if only cost risk items 30% 13.2 12.9 14.6
40% 13.4 13.0 14.9
were considered and schedule assumptions were held
50% 13.6 13.1 15.1
fixed at their baseline levels, costs could be two million 60% 13.7 13.2 15.4
higher at least 5% of the time. The curve for the 70% 14.0 13.4 15.6
schedule risk (histograms) only is much steeper than 80% 14.2 13.5 16.0
that for cost risk elements only. 90% 14.6 13.7 16.5

The next question is whether the interaction between Figure 14. Components of project cost risk
cost and schedule risk adds much to the results when
One measure is the likelihood of overrunning the EAC,
using cost risk. The answer to this question will
which was $12.8 million in this example. The table
demonstrate whether the integration of cost and
shows that this value could possibly be achieved with
schedule risk elements into one model and simulation
either cost risk elements or schedule risk elements
has been worth the efforts described above. This issue
alone, but it is not very likely with both cost and and the ratio benefit / cost of burn rate control might be
schedule risk elements taken into account. very low.
Another measure of the contribution of integrating cost The cost risk element that contributes the greatest total
and schedule risk elements into the cost risk estimate is uncertainty from positive (under runs) to negative (over
the mean of the distributions. The EAC of $12.8 runs) is the labor rate in compensation per hour.
million increases to $13.6 million when all cost risk Uncertainty about the skill mix of people on the job can
elements are considered, but it increases to $14.3 cause the cost to under run or over run the estimate as
million when schedule risk elements are added. Thus, shown in Figure 16.
the overrun at the mean is only $0.6 million with cost
risk elements alone, but it is $1.5 million, more than Distribution of Cost Risk Elements and Total Cost Risk vs. EAC

twice that, when schedule risk elements are added. 100%

90%
The same story is told at a greater level of safety, taking
80%
the 80% level. Here cost risk elements alone have an 70% EAC
estimate of $14.2 million but a complete analysis with 60% Fixed Hour

Percentile
schedule risk elements shows $15.0 million. 50%
Burn Rate
Labor Rate
40%
Clearly the inclusion of schedule risk in the cost risk Material
All Cost Risk
30%
analysis is meaningful and necessary. Without the 20%
schedule uncertainty, the cost risk analysis would 10%
underestimate the risk of meeting cost objectives and 0%

not provide the project manager complete information. $11.0 $12.0 $13.0 $14.0 $15.0 $16.0 $17.0 $18.0 $19.0
$Millions

Using the results to assist risk response planning.

The cost risk model was configured to allow each type Figure 16. Labor rate contributes the most
of risk to be simulated separately or to be subtracted
separately from the total cost risk. The first example of total uncertainty in the cost risk
this use of the model is to see which type of cost risk Another way to examine the benefit of risk mitigation is
variable provides the most risk. There are two ways to to subtract each cost risk type from the total risk
look at risk in this context: (1) the greatest increase in representing all cost risk types and total schedule risk.
cost risk from the EAC representing overruns, and (2) In this way the project manager may see what risk
the greatest spread of risk from lowest to highest responses might reduce total project cost risk and
representing uncertainty or imprecision in any prioritize risk mitigation actions. The information in
particular estimate of cost. Figure 17 shows that the greatest improvement in total
. cost risk might come from improving the burn rate. It
Distribution of Cost Risk Elements and Total Cost Risk vs. EAC shows what would happen to total cost risk if the
100%
project manager might hold the burn rate to baseline
90%
levels. Whether this is possible or not is another story.
80% Still, controlling the number of resources applied per
70% EAC unit time would seem to offer significant cost risk
60% Fixed Hour reduction.
Percentile

Burn Rate
50%
Labor Rate
40% Distribution of Total Program Cost vs. EAC
Material
Risk Mitigation -- Control the Burn Rate Uncertainty to Baseline
30% All Cost Risk

20% 100%

10% 90%

80%
0%
$11.0 $12.0 $13.0 $14.0 $15.0 $16.0 $17.0 $18.0 $19.0 70%

$Millions 60%
Percentile

EAC
50% Cost & Schedule

40% No Burn Rate

Figure 15. Burn rate contributes the most 30%

cost risk difference from the EAC 20%

10%

Figure 15 above shows that the burn rate of all cost risk 0%
$11.0 $12.0 $13.0 $14.0 $15.0 $16.0 $17.0 $18.0 $19.0
elements contributes the most to increases of the risk $Millions
over the EAC. Control of the rate of resources per unit
time may help to manage the risk of cost overruns the
most. Of course, controlling this factor may be difficult
Figure 17. Controlling the burn rate could first time and a different way in a later interview.
contribute the most to reducing cost risk Usually the second time more risk will be reported.
The mature organization will understand that people
OBSERVATIONS become more comfortable discussing risk as they learn
how to do it.
Methodology. There are several issues with the
methodology of integrating cost and schedule risk CONCLUSIONS
analysis.
This paper has demonstrated that a good cost risk
♦ Cost and schedule structures can become disjoint, analysis needs to incorporate schedule risk as well as
making linkage of the two difficult. Serious effort was the traditional cost risk elements. Integrating the results
required to link the structures so the costs could be from the schedule risk into the cost risk model provides
allocated to schedule concepts. The recommendation is a more complete picture of cost risk than if it were
to structure both at a common starting point, probably excluded. The picture of cost risk that emerges will
the WBS. depend significantly on the degree of schedule risk that
is found in the project.
♦ Modelling issues were serious as well.
Decomposing the baseline estimate into its components Often cost risk analysis looks at individual cost
and discovering what must have been the baseline elements and tries to figure out what drives their
assumptions was tricky. Keeping the estimates clean uncertainty. Sometimes participants in a traditional
was not possible, and many approximations were cost risk analysis reveal uncertainty in activity
needed. Making the model work when many different durations, but those concerns are usually ad hoc and
cost and schedule elements are present, well beyond the incomplete. The disciplined approach of this paper
simple example described here, implies a very makes the consideration of schedule risk explicit and
complicated spreadsheet model. Including the facility unavoidable.
to simulate one or another type of variable alone was
Often a “troubled project” is having as much difficulty
complex modelling as well.
keeping on schedule as it is on budget. These two
Data collection issues. Collecting high-quality data problems reinforce each other, magnifying the
that are as accurate as possible is the most important problems. This magnification is represented in the
part of the analysis. present model by two things: (1) multiplication of time,
and therefore time uncertainty by burn rate and
♦ There are new concepts for the interviewees, such therefore burn rate uncertainty, and (2) correlation
as the burn rate, average compensation and correlation between burn rate and duration.
between elements. We simplified the burn rate to a
level-loading concept, which further complicated it for The cost estimates often lose track of schedule realities,
the interviewees. Still, the teams were able to causing the estimates of cost to be unrealistically low in
understand and respond. many cases. Estimators feel that they did not have a
full kit of information when the facts of the schedule
♦ The explicit use of burn rate and labor rate are made clear to them, making them look bad at
highlights the basic forces of the cost risk elements of estimating and causing them to lose credibility with the
variable costs. It was found that the participants could project executives. Insisting on a close communication
look at both elements separately, which we believe between the scheduling and the cost estimation
improved the representation of variable cost risk. functions will help to reduce this problem. Insisting on
♦ Data collection is most difficult when the looking at the schedule risk when evaluating the cost
participants have never been part of a risk analysis, as is risk will improve the accuracy of the estimates of
often the case. Expert facilitation is often necessary. project cost and of cost risk.
The good news is that this is not brain surgery, and
participants can learn while doing.
♦ Individual and organizational maturity comes from
treating data collection seriously. It includes preparing
the teams and giving them a sense of importance for the
effort. It includes providing enough time for the
exercise and making it part of the regular work, not an
extra. It also includes recognizing that the team leaders
and members may not provide perfect data the first time
they are interviewed. They may answer one way the