Modeling Contingency in Environmental Remediation Cost Estimation

Cost estimation for hazardous waste cleanup in the U.S. has evolved under the influence of the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (CERCLA). CERCLA mandates cost estimates for cleaning hazardous waste sites, which number in the thousands across the U.S., including mines and wells. Due to the uncertainty surrounding these costs, many federal agencies have issued guidelines for analyzing cost risks and uncertainties. To estimate cost contingency in hazardous waste remediation, estimators should consider: (1) the probability distribution of the expected cost estimate, (2) its standard deviation, (3) the confidence that the final cost will not exceed the stated accuracy range, and (4) stakeholders’ risk tolerances. A method is proposed for estimating contingency that takes these factors into account, using an Extreme Value distribution, and is applied to the largest remediation program in the U.S.: The Hanford Nuclear Site cleanup managed by the U.S. Department of Energy's Office of Environmental Management. The paper concludes that, for most Hanford projects, contingencies (covering “known unknowns”) are reasonably estimated, though they may not be realistic. The proposed cost contingency estimation method can be used to assess whether cost contingency estimates are reasonable for projects in which costs have been estimated using Monte Carlo simulation.