Achieving Predictable Projects

1
Achieving Predictable Projects
In a World of Black Swan Risks
2
SESSION 9
Achieving Predictable Projects
In a World of Black Swan Risks
Gary Berman, PE, FCMAAPresident / CEOGREYHAWK
3
SESSION 9
Achieving Predictable Projects
In a World of Black Swan Risks
Paul McNuttManager, Project Risks
Reviews ConocoPhillips
4
SESSION 9
Achieving Predictable Projects
In a World of Black Swan Risks
TBDTBDTBD
5
Agenda

• What risks are you including?
• Risk Spectrum
• Discussion
• How do you assess and manage those risks?
• Levels of Analysis
• Risk Management Techniques
• Discussion
• How will you management risk in the future?
• Discussion

6
How Deep is Your Risk Pond
7
A Spectrum of Risk
Tactical Quantity accuracy Labor
rates Productivity Weather Margin of error Design
allowance
Black Swan Market crash War Extreme
weather Major scope change
Future Foreign exchange Hyper-inflation Changing
regulation
Strategic Basis of design unc. Partner
alignment Permitting
8
Labor Availability
9
Design Allowance
10
Material Availability
11
Weather Delay
12
Scope Definition
13
Governance Model
14
Global Hyper-inflation
15
Geopolitcal
16
Considerations When Modeling Risk

• Definition of risk in dollars or days
• Frequency
• Probability Rating Scale
• Optimistic / Likely / Pessimistic
• Combinations
• Lone wolf or hunting in packs
• Iterative Analysis to determine ranges
• Determine cost effectiveness of risk mitigation

17
Risk Translated into Real Impact
18
Risk Spectrum
19
The Four Levels of Risk Assessment

• Level 1 Sensitivity Analysis
• Use of plus/minus percents to test a projects
  weaknesses
• Generally uses symmetric ranges (e.g., 10 and
  -10)
• Level 2 Three Point Ranging
• Use of continuous distributions on cost estimate
  line items
• Generally ignores dependency between similar
  items
• Level 3 Three Point Ranging with Dependencies
• Use of continuous distributions with correlation
• Generally ignores risk events and schedule slips
• Level 4 Full Cause and Effect Modeling
• Use of three point ranges on cost estimate
  variables coupled with dependencies and risk
  events including a linked schedule risk

Ranges generated from Levels 1-3 are typically
narrow and unskewed Ranges generated from Level 4
are wider and skewed reflecting observed data
20
Any more comments or questions
Achieving Predictable Projects
In a World of Black Swan Risks

• Gary Berman GREYHAWK
• Paul McNutt ConocoPhillips
• TBD TBD