CASE STUDIES IN ESTABLISHING GROUNDWATER BASELINE VALUES FOR URANIUM ISR SITES

To evaluate the efficacy of monitored natural attenuation in restoring a uranium in situ recovery (ISR) site to premining conditions, it is necessary to establish a baseline or baseline water quality of those premining conditions. The current industry standard is to calculate groundwater baseline concentrations using data from the mining zone and to restore to the mean value. This restoration goal is very conservative and water-intensive. Furthermore, restoring to the mean involves cleaning up a portion of groundwater uranium that would have been present premining. For this reason, a statistical protocol was established for determining baseline values for different parts of the system. The same criteria and methodology was used for establishing baseline values as were used for determining groundwater baseline for Los Alamos National Laboratory, an approach accepted by the New Mexico state regulator.

Baseline was established at two locations, Crownpoint, New Mexico, a site in premining conditions, and Rosita, Texas, where mining and site restoration have occurred. Upper tolerance limits (UTLs) were calculated for 19 of 33 constituents at the Mobil Unit 1 at Crownpoint, where nondetects represented less than 50% of the data. A UTL is the upper 95% confidence limit on the 95^th percentile. For large data sets following normal distributions, a UTL can be roughly estimated as the mean plus two standard deviations. A minimum of eight analyses is required for the statistical analysis, but for the approach to produce robust results, a much larger data set is desirable. Two sets of baseline calculations were determined for Rosita Production Areas 1 and 2. Protocol recommends that the potential existence of outliers be discussed with the regulator. Results of the baseline determinations produced higher UTLs for many constituents than restoration values used by industry. The specific results are interpreted with caution but the approach is as reliable as the data set, which points to the need for increased premining characterization of proposed ISR sites.