Illustration of sampling-based methods for uncertainty and sensitivity analysis.

A sequence of linear, monotonic, and nonmonotonic test problems is used to illustrate sampling-based uncertainty and sensitivity analysis procedures. Uncertainty results obtained with replicated random and Latin hypercube samples are compared, with the Latin hypercube samples tending to produce more stable results than the random samples. Sensitivity results obtained with the following procedures and/or measures are illustrated and compared: correlation coefficients (CCs), rank correlation coefficients (RCCs), common means (CMNs), common locations (CLs), common medians (CMDs), statistical independence (SI), standardized regression coefficients (SRCs), partial correlation coefficients (PCCs), standardized rank regression coefficients (SRRCs), partial rank correlation coefficients (PRCCs), stepwise regression analysis with raw and rank-transformed data, and examination of scatter plots. The effectiveness of a given procedure and/or measure depends on the characteristics of the individual test problems, with (1) linear measures (i.e., CCs, PCCs, SRCs) performing well on the linear test problems, (2) measures based on rank transforms (i.e., RCCs, PRCCs, SRRCs) performing well on the monotonic test problems, and (3) measures predicated on searches for nonrandom patterns (i.e., CMNs, CLs, CMDs, SI) performing well on the nonmonotonic test problems.

Performance assessment in support of the 1996 compliance certification application for the Waste Isolation Pilot Plant.

The conceptual and computational structure of a performance assessment (PA) for the Waste Isolation Pilot Plant (WIPP) is described. Important parts of this structure are (1) maintenance of a separation between stochastic (i.e., aleatory) and subjective (i.e., epistemic) uncertainty, with stochastic uncertainty arising from the many possible disruptions that could occur over the 10,000-year regulatory period that applies to the WIPP, and subjective uncertainty arising from the imprecision with which many of the quantities required in the analysis are known, (2) use of Latin hypercube sampling to incorporate the effects of subjective uncertainty, (3) use of Monte Carlo (i.e., random) sampling to incorporate the effects of stochastic uncertainty, and (4) efficient use of the necessarily limited number of mechanistic calculations that can be performed to support the analysis. The WIPP is under development by the U.S. Department of Energy (DOE) for the geologic (i.e., deep underground) disposal of transuranic (TRU) waste, with the indicated PA supporting a Compliance Certification Application (CCA) by the DOE to the U.S. Environmental Protection Agency (EPA) in October 1996 for the necessary certifications for the WIPP to begin operation. The EPA certified the WIPP for the disposal of TRU waste in May 1998, with the result that the WIPP will be the first operational facility in the United States for the geologic disposal of radioactive waste.

A Monte Carlo procedure for the construction of complementary cumulative distribution functions for comparison with the EPA release limits for radioactive waste disposal.

A Monte Carlo procedure for the construction of complementary cumulative distribution functions (CCDFs) for comparison with the U.S. Environmental Protection Agency (EPA) release limits for radioactive waste disposal (40 CFR 191, Subpart B) is described and illustrated with results from a recent performance assessment (PA) for the Waste Isolation Pilot Plant (WIPP). The Monte Carlo procedure produces CCDF estimates similar to those obtained with importance sampling in several recent PAs for the WIPP. The advantages of the Monte Carlo procedure over importance sampling include increased resolution in the calculation of probabilities for complex scenarios involving drilling intrusions and better use of the necessarily limited number of mechanistic calculations that underlie CCDF construction.

Contamination of surface-water bodies after reactor accidents by the erosion of atmospherically deposited radionuclides.

Reactor safety analyses usually do not consider the population risk which might result from the contamination of surface-water bodies after reactor accidents by the erosion of atmospherically deposited radionuclides. This paper is intended to provide perspective on the reasonableness of this omission. Data are presented which are suggestive of the rates at which atmospherically deposited radionuclides might erode into surface-water bodies. These rates are used in the calculation of potential health effects resulting from surface-water contamination due to such erosion. These health effects are compared with predicted health effects due to atmospheric and terrestrial pathways after reactor accidents. The presented results support the belief that the contamination of surface-water bodies after reactor accidents by the erosion of atmospherically deposited radionuclides is not a major contributor to the risk associated with such accidents.

Sensitivity analysis of a model for the environmental movement of radionuclides.

Results are presented from a sensitivity analysis study of a model developed to represent the environmental movement of radionuclides. This model is designated the Environmental Transport Model. The study has three purposes: (1) to develop sensitivity analysis techniques applicable to the Environmental Transport Model, (2) to provide insight and experience with respect to the performance of a sensitivity analysis of this model and (3) to develop understanding of the overall operation of the model and the variables which influence this operation. Two variations of a hypothetical river receiving a radionuclide discharge containing 99Tc, 245Cm, 241Pu, 234U, 230Th and 226Ra are defined. Independent variables of the following types are introduced: variables which define physical properties of the river system (e.g. soil depth, river discharge and sediment resuspension) and variables which summarize radionuclide, the following dependent variables are investigated: (1) radionuclide concentration in soil, (2) dissolved radionuclide concentration in surface-water and (3) total radionuclide concentration in surface-water. The investigation employs sensitivity analysis techniques based on Latin hypercube sampling, rank transformations and stepwise regression. Among the important variables indicated in the analysis are distribution coefficients, river discharge and suspended sediment concentration.