This recommended practice (RP) defines general practices and considerations for inherent cost risk analysis and estimating contingency using a cost estimate ranging analysis with the Monte Carlo simulation (MCS) method. The method is only recommended for quantifying inherent cost uncertainty when the project scope is well-defined (i.e., Class 3 or better) and when the project has no new technology and minimal complexity. In all cases, the method is not to be used alone when there are significant project-specific risks (e.g., contingent risks or events). The method is applicable to two situations:

• As the sole method for cost contingency determination (e.g., a small project for an owner or a contractor bid estimate for a simple project when the owner retains responsibility for most scope and event risks).
• As part of a hybrid approach that uses this RP for inherent cost uncertainty and the expected value method to quantify project-specific cost risks (i.e., contingent risks or risk events).

A key principle for recommended quantitative risk analysis (QRA) methods is that they be risk-driven; i.e., clearly link the risk to the impacts. Estimate ranging using subjective team inputs from those involved in the project design, planning, and estimating can be expected to reliably capture the inherent cost uncertainties.

Estimate ranging does not meet the risk-driven principle when systemic and project-specific risks are significant. In particular, it is not recommended when the project scope is not well-defined (i.e., Class 10, 5, or 4), or the project has technology uncertainty, or any significant level of physical and/or execution complexity. Additionally, when there are significant project-specific risks (the main source of schedule delays as well), estimate ranging cannot be said to clearly link these defined risks to their impacts.

In both situations, estimate ranging tends to underestimate contingency, often by significant amounts. Alternative methods for these situations of greater uncertainty and risks are described in Professional Guidance Document PGD-02, Guide to Quantitative Risk Analysis. RP 41R-08, Understanding Estimate Ranging broadly describes the class of QRA methods called estimate ranging.

In general, estimate ranging methods with MCS replace fixed values in a cost estimate with 3-point or uniform distributions, establish correlations between the distributions, and running MCS to obtain an overall cost outcome distribution. There are many variations possible within this general description. For example, the fixed values replaced could be for individual estimate item costs or for various estimate subtotals. The values replaced could also be at the estimate element level such as quantities, productivity, item pricing, and so on, rather than for estimate item totals. This RP describes typical estimate breakdown variations with specific consideration of the need to address MCS correlation which becomes more challenging as more distributions are assigned. In addition to the topic of correlation, the RP also discusses MCS considerations for distribution assignment and determination of 3-point values; however, this RP is not a detailed treatment of MCS.

Contributors:

John K. Hollmann, PE CCP CEP DRMP FAACE Hon. Life (Primary Contributor)
Peter R. Bredehoeft, Jr. CEP FAACE
Colin H. Cropley
Francisco Cruz Moreno, PE
Larry R. Dysert, CCP CEP DRMP FAACE Hon. Life
Dr. David Hulett, FAACE
Sagar B. Khadka, CCP DRMP PSP FAACE
Luis Henrique Martinez
Martin Snyman, DRMP
Robert F. Wells, CEP

About John: