TREATMENT CELL SUPPLEMENT TO A PERMEABLE REACTIVE BARRIER: PERFORMANCE ASSESSMENT MONITORING

A permeable reactive barrier (PRB) containing 300 tons of zero valent iron (ZVI) filings was installed in June 1999 at a former uranium milling site near Monticello, Utah. The PRB has three zones consisting of (from upgradient): 2 ft of gravel/ZVI mixture, 4 ft of ZVI, and 2 ft of gravel. Ground water samples and water level data have been collected at regular intervals from more than 50 wells in the PRB.

Satisfactory performance of the PRB requires that contaminant concentrations are lowered to regulated levels, a high proportion of the ambient ground water flow is treated and, there are no adverse affect to land use. Contaminant concentrations remain low in most wells in the ZVI zone to satisfy the first criterion. Attainment of the second and third criteria is less certain however considering a continuously rising ground water mound behind the PRB. This occurrence is likely coupled to an order of magnitude loss of permeability as determined by serial gas-injection slug tests, presumably due to observed mineral precipitation in the ZVI zone. As a result, a very shallow water table now threatens crop production and the proportion of treated ground water may have decreased significantly from an initial mass-balance estimate of 4 gallons per minute. Flux estimates based on Darcian principles are limited by high spatial and temporal variability.

To alleviate the ground water mounding and provide additional treatment capacity a supplemental treatment cell was installed in June 2005. The cell is currently (July 2005) treating about 2.4 gpm. Performance data (flows, pressures, and pH), collected and transmitted daily to operations personnel, indicate that the cell is operating normally. Effluent concentration of uranium remains less than the ground water standard. While there are some operational and maintenance costs for the treatment cell, the costs are relatively minor. The amount of data required to confirm the performance of the treatment cell is much less than the amount required to confirm the performance of the PRB at a similar level of confidence.