ANALYSIS OF AQUIFER TEST DATA AND IMPLICATIONS IN AN MTBE-CONTAMINATED FRACTURED BEDROCK AQUIFER, PASCOAG, RHODE ISLAND

An aquifer test was conducted to assess the distribution and magnitude of drawdown in an MTBE-contaminated public drinking water aquifer. The pumping well draws water from highly fractured gneiss with recharge from the overlying glacial sand and gravel overburden. The well was pumped at 240 GPM for 37 days while observations were made at approximately 35 wells, both overburden and bedrock, located throughout the aquifer. Plots of drawdown versus elapsed time for the pumping well and other wells indicate linear drawdown consistent with single fracture transport. One bedrock well located 135 feet from the pumping well shows drawdown similar to the pumping well. This indicates that the well intersects the same fracture and can be considered as an extension of the pumping well. This extended well strikes approximately N40E, which is consistent with the local fracture and lineation orientation. This is relevant because the pumping well and contaminant source locations lie along similar strike. Relatively high levels of MTBE and other gasoline compounds appear to travel along this fracture as a preferred pathway from the site 1500 feet up gradient. MTBE concentrations were 330 ppb in this well which was ten times the concentration obtained from well head sampling. Unexpectedly, the interaction between a stream that runs along the western edge of the aquifer from an up gradient reservoir made deciphering drawdown near the stream difficult. River stage measurements related to precipitation events and reservoir gate adjustments strongly influenced these wells. Late-time apparent drawdown was actually related to the closing of the gate rather than an influence of pumping. This study demonstrates the importance of considering the overall hydrogeology of a site during the construction and interpretation of an aquifer test. The stream and aquifer interaction lead to apparent well drawdown which is actually related to stream stage fluctuation. Well observations in areas close to the stream show that the water table is controlled by the stream stage, while the direct influence of pumping is less significant or obscured by overriding factors. This influence diminishes with distance from the stream as the pumping becomes dominant.