Characterization and reduction of uncertainty in remedial activities and performance assessments in support of rational decision making for short- and long-term stewardship

This project applies advanced uncertainty analysis and data fusion techniques that aid rational decision making for short- and long-term stewardship. The outcomes from this project facilitate a transition from deterministic to probabilistic modeling at DOE site as well as the maintenance of PAs by providing systematic procedures and tools for the regular reevaluation of PAs with data collected over the duration of the compliance period. As a matter of fact, regular reevaluation of the PAs can be an important mechanism for generating early warnings of remediation failures.

This project overcomes the technological and scientific barriers mentioned in the introductory discussion:

• The availability of methods developed through this project for generating computationally efficient reduced models that are equivalent to the full complex models currently used in the PA process make it technologically feasible to conduct probabilistic PAs.
• The availability of methods developed through this project ensure that the probability distributions of uncertain inputs are consistent with existing data.

Specific efforts to be pursued under this study include:

• Development of computationally efficient equivalent representations of vadose zone models based on the PORFLOW system (which is used to conduct PAs and CAs at SRS), and use of these representations to perform uncertainty and sensitivity analyses.
• The enhancement of conducting and maintaining probabilistic PAs and CAs by using Bayesian updating to improve the characterization of uncertainties in inputs to groundwater contaminant transport models; this is achieved by incorporating extant and future contaminant concentration data in the modeling application.
• Utilization of Stochastic Response Surface Methods for model calibration and prediction using FACT flow model. Application of flexibility analysis and FACT detailed flow modeling to determine the radius of influence of remediation wells.
• Application of HDMR (High Dimensional Model Representation) methods for complex model reduction.
• Participation in efforts of the Remediation Center.

Complex mechanistic environmental and biological models, required for comprehensive exposure assessments, are often very computationally demanding. In performing distribution-based exposure assessments, large numbers of simulations employing these mechanistic models are often required to obtain the population exposure estimates for the pollutant of concern. Therefore, the computing time and resources can be prohibitively expensive or infeasible. For this reason, exposure assessments often utilize simpler, empirical formulations, and many (over-) simplifying assumptions. A desirable alternative, however, consists of deriving simple, computationally efficient substitutes to the comprehensive process models ("Fast Equivalent Models") through a formal, systematic process of mathematical model reduction that considers and retains essential characteristics of input-output relationships in the original model.