Quantitative Assessment of Risk Sensitivity in Construction Cost Estimating

Abstract

Realistic construction estimating requires optimal decisions that not only minimize construction time and cost, but also address other important factors such as project risks. Probabilistic estimating is generally used to incorporate risk into estimated costs, the most common measure of which is the expected monetary value (EMV). Yet, EMV does not capture the way most people make decisions under uncertainty because it cannot reflect the decision maker’s risk sensitivity. This paper presents a probabilistic estimating model that can quantify and incorporate the contractor’s risk aversion. A highway tunneling project is presented to illustrate the application of the proposed model. Tunnel cost estimating is formulated as a risk-sensitive Markov decision process (MDP) where the contractor’s risk sensitivity is modeled by an exponential utility function. The contractor’s decisions depend on his risk aversion coefficient and the variability in tunneling costs, which in turn depend on geologic uncertainty. The risk sensitive MDP is solved by maximizing the expected utility value (EUV) of tunneling costs. The final results include the optimal excavation and support sequence, and the risk-adjusted tunneling costs for the project, as functions of the contractor’s risk sensitivity.