A Comparison of Geostatistically-Based Inverse Techniques for Use in Performance Assessment Analyses at the WIPP Site—Results from Test Case No. 1

by D. A. Zimmerman, GRAM, Inc, Albuquerque, United States,
David P. Gallegos, GRAM, Inc, Albuquerque, United States,

Document Type: Proceeding Paper

Part of: High Level Radioactive Waste Management 1993


The groundwater flow pathway in the Culebra Dolomite aquifer at the Waste Isolation Pilot Plant (WIPP) has been identified as a potentially important pathway for radionuclide migration to the accessible environment. Consequently, uncertainties in the models used to describe flow and transport in the Culebra and to be addressed. A 'Geostatistics Test Problem' is being developed to evaluate a number of inverse techniques that may be used for flow calculations in the WIPP performance assessment (PA). The Test Problem is actually a series of test cases, each being developed as a highly complex synthetic data set; the intent is for the ensemble of these data sets span the range of possible conceptual models of groundwater flow at the WIPP site. The Test Program analysis approach is to use a comparison of the probabilistic groundwater travel time (GWTT) estimates produced by each technique as the basis for the evaluation. Participants are given observations of head and transmmissivity (possibly including measurement error) or other information such as drawdowns from pumpings wells, and are asked to develop stochastic models of groundwater flow for the synthetic system. Cumulative distribution functions (CDFs) of groundwater flow (computed via particle tracking) are constructed using the head and transmissivity data generated through the application of each technique; one semi-analytical method generates the CDFs of groundwater flow directly. This paper describes the results from Test Case No 1. Of the five techniques compared, those based on the linearized form of the groundwater flow equation exhibited less bias and less spread in their GWTT distribution functions; the semi-analytical method had the least bias. While the results are not sufficient to make generalizations about which techniques may be better suited for the WIPP PA (only one test case has been exercised), analyses of the data from this test case provides some indication about the relative importance of other aspects of the flow modeling (besides inverse method or geostatistical approach) in PA. These ancillary analyses examine the effect of gridding an the effect of boundary conditions on the groundwater travel time estimates.

Subject Headings: Groundwater flow | Model tests | Data processing | Travel time | Hydrologic models | Flow distribution | Comparative studies | Groundwater | United States | North America | Pennsylvania | Oregon

Services: Buy this book/Buy this article


Return to search