PNNL

Abstract

As part of its environmental remediation program at the Hanford Site, the U.S. Department of Energy needed a method to es-timate ground contaminant source term inventories and uncertainties. To address this need, a stochastic model, the Hanford Soil Inventory Model (SIM), was created to provide inventories and uncertainties for 75 analytes to 377 different locations encompassing 57 years of operations using historical production and surveillance data. Open Crystal Ball (OCB) and Crystal Ball® software from Decisioneering® was chosen for SIM’s foundation because of its unique capabilities to deal with such a complex simulation given the project requirements. Using OCB, comprehensive simulations were performed, producing 450,000 forecasts from 2.6 million assumptions. Reference values for selected locations and analytes fall within the SIM predicted 0.5 to 99.5 percentile uncertainties 68% of the time. Numerous waste management contractors and regulatory or-ganizations at the Hanford Site have used, and continue to use, the results.