Risk Based Estimating Self Modeling

1
Risk Based EstimatingSelf Modeling
Ovidiu Cretu, Ph.D., P.E. Terry Berends,
P.E. David Smelser
2
(No Transcript)
3
Traditional Estimating
Risk Based Estimate
Threat 1
Base Estimate
Base Estimate
Contingency
Opportunity
Threat 2

• All known and unknown risks are equally weighted
• Allows little control over the project
  cost/schedule
• Reactive

• Clear recognition of projects threats and
  opportunities
• Allows a reasonable control over the project
  cost/schedule
• Proactive

4
Risk Based Estimate
Engineers Estimate
Monte Carlo Method
Likelihood of Occurrence
Identify Quantify Risks
Impact ,Mo
5
Base Cost and Schedule Validation

• Review the project assumptions
• Review the project cost and schedule based on the
  information available
• Update unit price
• Update quantities
• Capture the cost of unknown cost of miscellaneous
  items
• Remove some contingencies

6
Variability of the Base Cost and Schedule

• The entire construction cost/duration
• A major group of pay items
• An individual pay item
• Symmetrical distribution
• Beta3 Distribution

7
Cost Duration Mo Variability 2 to 10
Validate Base Cost Duration
Risk Based Estimate
Engineers Estimate
Monte Carlo Method
8
Risks Identification and Quantification

• Focus is on
• Identify the key risky events
• Estimate how likely it is that the risky event
  will materialize
• Estimate why and by how much events may turn out
  differently from the base estimate

9
Probability of Risk Occurrence

• Lowest value 0
• Highest value 1
• Middle value 0.5

10
Probability of Risk Occurrence

• Very Low 5
• Low 25
• Medium (As Likely As Not) 50
• High 75
• Very High 95
• It is important to be approximately right. Do
  not waste time being precisely wrong.

11
Define Range and Shape

• Three Point Estimate about as much information
  an expert can provide.
• MIN the first point
• MAX the second point
• The Best-guess

Range
Shape
12
Shape

• The Best-guess This will be the experts
  median guess
• Median Actual outcomes evenly distributed over
  the median guess
• The Best-guess cant be too close to the max or
  the min.

13
ELICIT VALUES
MIN 100 MAX 700
Best Guess 400
Most Likely400
Entire range (100 to 700) includes 100 of the
possibilities
14
ELICIT VALUES
MIN 100 MAX 700
Best Guess 200
Expert Costs are more likely to be at the lower
end of the range
Most Likely 130
Entire range (100 to 700) includes 100 of the
possibilities
15
ELICIT VALUES
MIN 100 MAX 700
Best Guess 600
Expert Costs are more likely to be at the
higher end of the range
Most Likely670
Entire range (100 to 700) includes 100 of the
possibilities
16
Cost Duration Mo Variability 2 to 10
Validate Base Cost Duration
Risk Based Estimate
Engineers Estimate
Monte Carlo Method
Likelihood of Occurrence
Identify Quantify Risks
Impact ,Mo
17
RESULTS
Risk Based Estimate
End CN
Schedule
Ad Date
18
INPUT
OUTPUT
RBE
Base Cost Duration Variability Estimating
Date Escalation Factor Risks Cost,
Duration Status Project Phase Probability Range
and Shape Critical Path Info Markups
MCM
Cost CY YOE Diagram Table Schedule AD Date End
CN Diagram Table Sensitivity Analysis
The Model
10,000 Plausible Cases
19

• MCS -- DEMO

20
Conclusions

• Better understanding of the projects challenges
• Crafts the project risk management plan with
  clear target on how to enhance the project value
• Helps in maximizing the projects opportunities
  and reducing or eliminating the projects threats

21
The RBE Self-modeling

• Two Major Functions
• Estimating Function
• Risk Management Function

22
Conclusions Self-modeling

• The model allows registration of meaningful
  information and it produces valuable results that
  may be used by decision makers.
• The model does not require any special software
  or specialized skills.
• WSDOT - Self-modeling Spread Sheet

23
Any Questions??