Iterative approach to groundwater flow modeling of the Martinsville Alternative Site, under consideration for low-level radioactive waste storage in Clark County, Illinois, U.S.A.

Abstract

The regulatory requirements for characterization of the Martinsville Alternative Site (MAS) were fulfilled by applying an iterative approach to the groundwater flow modeling of the site and surrounding area. The approach consisted of field data collection and development of an initial conceptual model. The numerical model was then constructed to be consistent with the data and conceptual model. Next, the calibration results were evaluated statistically, and visually by a groundwater modeling review committee, to determine if the model accurately represented groundwater flow at the site. Initial results failed acceptance criteria because the values of numerical model input parameters had to be varied beyond observed data ranges to calibrate the results, and therefore the model was inconsistent with the initial conceptual model. This led to additional field data collection in areas where the numerical model deviated most from field-determined data. The new data provided sufficient information to revise the conceptual model and calibrate the numerical model successfully. Model calibration was followed by validation. Validation of the numerical model provided additional assurance that the model correctly simulated the observed system. No additional data were found to be necessary during validation of the MAS numerical model. The iterative approach proved to be successful for calibrating and validating this groundwater flow model and should be implemented from the onset of characterization planning in other applications.