Validation, Proof-of-Concept, and Postaudit of the Groundwater Flow and Transport Model of the Project Shoal Area

Abstract

EXECUTIVE SUMMARYThe groundwater flow and radionuclide transport model characterizing the Shoal underground nuclear test has been accepted by the State of Nevada Division of Environmental Protection. According to the Federal Facility Agreement and Consent Order (FFACO) between the U.S. Department of Energy and the State of Nevada, the next steps in the closure process for the site are the model validation (or postaudit), the proof-of-concept, and the long-term monitoring stage. This report addresses the development of the validation strategy for the Shoal model, needed for preparing the subsurface Corrective Action Decision Document-Corrective Action Plan (CADD/CAP) and the development of the proof-of-concept tools needed during the five-year monitoring/validation period. The proposed approach builds on the previous model validation work conducted for the Central Nevada Test Area (CNTA). The CNTA validation approach is adapted and modified to the site-specific conditions and challenges of the Shoal area, which are significantly different from those at CNTA (faster transport rates are predicted for Shoal).Groundwater model validation definitions and literature debates are briefly discussed. Most of the literature debate is on the terminology and not on the process itself. No one disputes the importance and necessity of the model validation or evaluation process and no one disagrees on the concept of using an independent data set to test the model. The disagreement is on what the process should be called and what the implications are for the term used. This has resulted in a lack of focus on development of model validation processes or tools. Although there is rough agreement on the goal of the process, there is no agreement on or development of a uniform methodology for executing model validation. We postulate that there is an urgent need for a validation process that can be used to evaluate model performance and build confidence in its suitability for making decisions. This confidence building is a long-term iterative process and it is this process that should be termed model validation.The FFACO's 10-step model validation process and the 12-step modeling protocol established in the hydrogeologic literature are discussed and contrasted. It is noted that one of the eliminated steps in the conversion from the hydrogeologic literature process to the FFACO process is the "model redesign" step. This model redesign step is a necessary complement to the model postaudit step, which appears in both processes. It is only meaningful to have a postaudit step together with the "possibility" for a model redesign step. Otherwise, there is no need for the model postaudit step.Previous model postaudit studies are reviewed and reasons for poor model performance are discussed. Based on these reviews and on the validation process proposed for CNTA, a similar validation approach for Shoal is proposed. The details of the proposed approach are not repeated here, but rather more explanation and discussion of some of the challenging asp...