LESSONS FOR ENVIRONMENTAL MANAGERS IN CHOOSING AND APPLYING MODELS FOR GROUNDWATER TRANSPORT AND REMEDIATION: A CASE STUDY
SARA SAYLER
WILLIAM F. MCTERNAN
SCOTT A. WARD
DOI: 10.2190/XJDL-WXFG-QALQ-P8TM
Abstract
A case study is presented which highlights the large discrepancy in projected environmental remediation requirements resulting exclusively from the selection and application of alternative modeling approaches for a contaminated groundwater system. Initially, a state of the practice numerical modeling system was employed to determine the final contaminant concentrations and plume configurations which could be expected in future years. These methods and models have become almost standard practice in these types of situations. In this particular application, however, the final model configuration was much more complex than was allowed by the collected data and by the aquifer system being modeled. Internal auto-calibration techniques were needed to calibrate critical field collected data with model results. The resultant model although calibrated projected contaminant plumes in excess of remediation goals well beyond regulatory property lines. Projected costs of remediation were in excess of 4 million USD. The projected concentrations and plume configurations following calibration did not, however, agree with time series data established during an extended monitoring program. These collected data suggested a much slower rate of plume migration. Inspection of the initial transport model suggested that the auto-calibration procedures resulted in parameter values far in excess of those measured by a variety of agencies and individuals. An alternative set of analytical procedures were employed to evaluate the statistically possible range of input parameters possible from aggregate measured data and to employ these data in various modeling configurations to define plume configurations and travel times beneath the study area. Geostatistical analysis and Monte Carlo simulations were employed together with relatively simple analytical transport codes in this effort. The results, statistically more robust than those derived in the initial effort, show a significantly reduced plume footprint in future years, with a corresponding reduction in projected remediation costs. The original 4 million USD estimate was reduced to approximately $155,000 USD. Most significantly, these results and projections are based upon techniques that are consistent with the measured data. This case study illustrates the types of technical concerns confronting environmental managers. By necessity, models frequently must be applied to help amplify complex conditions. If the wrong model is selected or if it is improperly configured, however, the resulting analysis may be woefully incorrect resulting in the expenditure of needless resources that may not provide adequate environmental protection. This article develops one set of methodologies available to address these concerns. This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.