Project cost risk analysis using influence diagrams

This paper summarizes new techniques developed in ongoing research of cost risk analysis for the U.S. Department of Energy (DOE) environmental remediation projects. This class of projects has been subject to large cost and schedule overruns due not only to the significant technical uncertainty but also the considerable political/regulatory/social uncertainty as well. Traditionally, Monte Carlo risk analysis methods have been used to evaluate project costs and schedules. However, typical Monte Carlo approaches are not sufficiently powerful to deal with the broad range of risks and uncertainties experienced on environmental remediation projects. This paper is based on the working premise that risks can be classified into two categories: internal or external. Internal risks are those that are inherent to a specific project. They usually affect items in the project cost estimate; internal risks can be evaluated using standard Monte Carlo approaches. External risks can influence the project cost but are not found in the cost estimate. These risks can include regulatory changes, scope changes, and public involvement. To model external risks, variations of Monte Carlo models and Influence Diagramming techniques were tested. This paper describes the various risk analysis formulations that were developed to evaluate both internal and external project risks. We conclude that Influence Diagramming (for external risks) used in conjunction with Monte Carlo methods (for internal risks) were best for evaluating cost risks for these projects.