Many of the methods and tools used to model project cost uncertainty have their roots in the time before computers were generally available to support projects. These manual techniques have become embedded in current practice, with all their limitations and shortcomings. They are now regarded as de facto standards simply because they have been in use since before many people working in the field were born but they are inefficient and can be misleading.

1. Creating value from uncertainty
www.Broadleaf.com.au
Cost and schedule risk
assessment
PMI Australia Conference 2014
Presented by Dr Stephen Grey

2. PMI Australia conference 2014
2
Theme
A lot of quantitative risk assessment is
unnecessarily cumbersome and soaks up
effort without adding value to a project.
The approaches used often make it very
hard to think clearly and realistically about
a project.

3. PMI Australia conference 2014
3
Modelling
Reality Belief
Model
Project
Σ

4. PMI Australia conference 2014
4
Describing a model
Assumptions
Sources of
Dependencies uncertainty
Cost

5. The requirement
PMI Australia conference 2014
5

6. PMI Australia conference 2014
Requirement
Inputs $
Project Σ
Target
Duration
Risk of
exceeding
target

7. The candidates
PMI Australia conference 2014
7

8. PMI Australia conference 2014
8
Three methods
Risk events
• List risks
• Probability and impact (range of variation)
• Add them up
Line items
• Line items from estimate (summary)
• Range of variation in line item
• Add them up
Risk factors
• Cost or duration estimating relationships
• Uncertainty in inputs to relationships (ranges)
• Simulate the effect on the base case

9. PMI Australia conference 2014
9
Similarities
Model
RelaƟonships in the model ≡ RelaƟonship in reality?
Is the link between inputs and reality easily understood?
Almost any attention to project risk will be beneficial
Some are more cost‐effective than others

10. Risk event model
Risk Probability Impact P x I
Risk 1 P1 I1 P1 x I1
Risk 2 P2 I2 P2 x I2
Risk 3 P3 I3 P3 x I3
Risk 4 P4 I4 P4 x I4
… … …
Risk n Pn In Pn x In
Total ΣPi x Ii
Monte Carlo simulation
PMI Australia conference 2014
10

11. PMI Australia conference 2014
11
Modelling based on risk events
Σ
Risk Probability Impact P x I
Risk 1 P1 I1 P1 x I1
Risk 2 P2 I2 P2 x I2
Risk 3 P3 I3 P3 x I3
Risk 4 P4 I4 P4 x I4
… … …
Risk n Pn In Pn x In
Total ΣPi x Ii
Target
Risk of
exceeding
target
Monte Carlo simulation

12. Line item modelling
WBS Item Labour Materials Total Contingency
Annnn $ $ $ $
Bnnnn $ $ $ $
Cnnnn $ $ $ $
Dnnnn $ $ $ $
… … … … …
Znnnn $ $ $ $
Total Σ$ Σ$
Monte Carlo simulation
PMI Australia conference 2014
12

13. PMI Australia conference 2014
13
Line item modelling
Σ
WBS Item Labour Materials Total Contingency
Annnn $ $ $ $
Bnnnn $ $ $ $
Cnnnn $ $ $ $
Dnnnn $ $ $ $
… … … … …
Znnnn $ $ $ $
Total Σ$ Σ$
Target
Risk of
exceeding
target
Monte Carlo simulation

14. Risk factors model
Cost estimating relationships (e.g. Cost = Quantity x Unit rate)
Project estimate summary
Labour Facilities Super‐vision
Materials Sub‐contracts
Services Expenses Total
Earthworks 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Concrete 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
… 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Overheads 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Project Total ? ? ? ? ? ? ? ?
Uncertainty
about quantities
of concrete (m3)
Uncertainty about rates for
cost of concrete ($/m3)
PMI Australia conference 2014
14
Simulated cost = Base estimate x (1 + ΔQuantity) x (1 + ΔRate)

15. Simulated cost = Base estimate x (1 + ΔQuantity) x (1 + ΔRate)
Σ
Σ
PMI Australia conference 2014
15
Risk factor modelling
Target
Risk of
exceeding
target
Project estimate summary
Labour Facilities Super‐vision
Materials Sub‐contracts
Services Expenses Total
Earthworks 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Concrete 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
… 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Overheads 1234567 1234567 1234567 1234567 1234567 1234567 1234567 ?
Project Total ? ? ? ? ? ? ? ?
Uncertainty
about quantities
of concrete (m3)
Uncertainty about rates for
cost of concrete ($/m3)
Σ

16. Making it work
PMI Australia conference 2014
16

17. PMI Australia conference 2014
17
Risk event
Variation in cost
if the event occurs
Variation in cost
if the event occurs
No effect
Probability = 33%
Probability = 67%
Min =$2M Likely=$4M Max=$8M
No effect
Probability = 33%
Probability = 67%
Min =$2M Likely=$4M Max=$8M
Inputs Project Σ Outputs

18. PMI Australia conference 2014
18
Risk event model
Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs Inputs Project Σ Outputs
Target
Risk of
exceeding
target
Common factors,
overlaps and interactions

19. PMI Australia conference 2014
19
Line item range
Cost line item
Inputs Σ Outputs

20. PMI Australia conference 2014
20
Line item model
Target
Risk of
exceeding
target
Common dependencies
and correlations

21. Cost
drivers
PMI Australia conference 2014
21
Risk factor model
Risk factors
Inputs Σ Outputs
Risk
factor
Target
Risk of
exceeding
target

22. Risk factor example (partial)
Small IT development project
Professional
services $
Software scale
Team productivity
Professional
services rates
Duration
Overhead rates
($/month)
Overhead $
Market rates for
licenses
Number of users
Design decision
Option 1
Option 2
Installed
Contingency assessment license cost $
methods & trends
22

23. Electrical & instrumentation
Steel, mechanical & piping
Concrete
Earthworks
Risk factor example (partial)
Mineral processing plant construction
Earthworks
direct labour $
Earthworks
quantity
Labour productivity
Labour rates
Option 1
Bulk material
cost $
Bulk material rates
Design decision Lump sum value
Option 2
Major
equipment
item cost $
$/m3
Contingency assessment
methods & trends
23

24. Assessing input parameters
PMI Australia conference 2014
24

25. PMI Australia conference 2014
25
Describing a range
Common practice
1. Start in the middle and work out
2. Explain reason for choosing numbers
Anchoring bias
Confirmation bias
3 1 2 Outcome

26. PMI Australia conference 2014
26
Describing a range
Good practice
1. Establish context
• Assumptions
• Sources of uncertainty
• Pessimistic and optimistic scenarios
Pre‐empt confirmation
2. Start with the extremes and work in
Break anchoring

27. PMI Australia conference 2014
27
Describing a range
Common practice
Assumptions
Sources of
uncertainty
Scenarios
2 3 1 Outcome

28. PMI Australia conference 2014
28
Summary
Modelling using risk factors is generally simpler
and more realistic than using structures based
on risk events or line item ranging
Context setting and methods to limit anchoring
are required to get realistic assessments of
uncertain values’ ranges
It need not be hard work!

29. PMI Australia conference 2014
29
Contact
Dr Stephen Grey
Associate Director
Grey@Broadleaf.com.au
+61 412 223 256
If you would like further information about this topic please contact us.
For further information visit www.Broadleaf.com.au
© 2014 Broadleaf Capital
International Pty Ltd.
All rights reserved.