2011 Palisade Risk Conference, Las Vegas, NV Integrated Cost and Schedule Risk Analysis for Engineering Projects Using a Custom @RISK Solution Mark Brumer, MBA 11 November 2011 Integrated Cost and Schedule Risk Analysis for Engineering Projects Using a Custom @RISK Solution Reasons for Integration Design Considerations Consequences of Normality Assembly of Data Set Construction of Model Interpretation of Results Demonstration Questions 2 11 November 2011 Reasons for Integration Limitations of schedule software Relies on additive algorithms Impedes incorporation of capital costs Limits choice of risk functions Restricts introduction of correlated risks Dominance of labor costs Comprise 60 to 80% of phase cost Inversely related under acceleration Nature of potential solutions Comprises a complex surface Risks overlooking behavior at peaks Utilize @RISK v. 6.0 3 11 November 2011 Design Considerations Define purpose and role of the risk analysis Incorporate natural phases of the Project Design, Construction, Systemization, Pilot Testing, et al. Timing of capital expenditures Transitions during the evolution of a phase Recognize inherent bias of estimate Precision of estimate Tendency towards conservatism Establish conditions for risk identification 4 Discretely defined Distributed impacts (vice acute) Relatively small, multiplicative impacts 11 November 2011 Consequences of Normality Project schedules are by definition not normal distributed Consider the consequences of “central limit theorem” Sum of a large number of risks tends to be normally distributed While the product of a large number of risks tends to be log-normally distributed Quantify risks in terms of relative perturbations Calculate product of multiple risks impacting a single activity Include correlations between related risks 5 11 November 2011 Assembly of Data Set: Project Schedules Provide the model foundation Import MS Project files into @RISK v.6.0 directly as *.mpp Establish linkage of risks to activities automatically using customized VBA scripts Require critical analysis before use 6 Let the duration and logic drive milestone dates Verify the validity of critical path Eliminate unnecessary constraints or artificial lags Ensure most activities start “As soon as possible” Avoid dangling activities 11 November 2011 Assembly of Data Set: Cost Estimates Constructed historically using Microsoft Excel Exploit standardized cost breakdown structures Clone existing cost estimates to facilitate data entry 7 11 November 2011 Assembly of Data Set: Risk Quantification Quantify risk events employing an iterative process Develop an aggregated risk function RiskDiscrete({x1,x2,x3}, {p1,p2,p3}) Where pi is the probability of Negative impact Opportunity benefit Status quo Where x1 and x2 are negative and positive consequence distributions bounded at the P10 and P90 Recognize that the consequence at the status quo equals one Generate parameters utilizing a variation of the Delphi method Value creative dissonance 8 11 November 2011 Construction of Model: Overview Integrated Risk Analysis Save Model Project Phases Group Activities Codes Units Managment Update Model Structure Project Cost Project Cost Schedule Schedule Risk Register Risk Register With Weights Periods by Phase Risk Register With Weights Cost Model Creation Cost Categories Cost Elements Risk Events Allocation Risk Correlation Matrix Units Cost and Quantities Cost Categories Period Resources Risk Simulation Settings Cost Subcategories Cost Elements Probability Impact Matrix Definition Risk Definition and Allocation Risk Category PALISADE.COM 9 Risk Events Start Simulation Mitigate Model Correlation Matrix powered by 11 November 2011 Construction of Model: Project Model 10 11 November 2011 Construction of Model: Cost Breakdown Structure 11 11 November 2011 Construction of Model: Risk Definition 12 11 November 2011 Construction of Model: Risk Allocation 13 11 November 2011 Construction of Model: Correlation Matrix 14 11 November 2011 Construction of Model: Integration of Schedule and Cost 15 11 November 2011 Interpretation of Results: Project End Date CL(15 Feb 13) ≈ 30% 1. Availability of Constr. Labor 2. Availability of System. Labor 3. Errors in Design … 16 11 November 2011 Interpretation of Results: Total Project Cost CL($176 mil) ≈ 30% 1. Constr. Labor Costs … 6. Throughput of WRS 8. Throughput of 2nd Rx 17 11 November 2011 Interpretation of Results: Correlation and Phase End Dates Correlation < 30% σ(Con) < σ(Sys) < σ(Pil) 18 11 November 2011 Interpretation of Results: Discussion Allowance for contingency Select confidence level appropriate with risk tolerance Subtract deterministic cost from cost at confidence level Prioritize risk mitigations 19 Evaluate most serious risks Price mitigation measures Rerun mitigated scenario Analyze cost vs. benefit 11 November 2011 Demonstration 20 11 November 2011 Questions 21 11 November 2011 Acknowledgements Richard R. Lunt, PhD † Javier F. Ordóñez, PhD 22 11 November 2011 Contact Mark P. Brumer mark@brumer.org (603) 498-6072 23 11 November 2011