Towards optimal sampling schedules for integral pumping tests

Ethanol use as a gasoline additive is increasing, as are the chances of groundwater contamination caused by gasoline releases involving ethanol. To evaluate the impact of ethanol on dissolved hydrocarbon plumes, a field test was performed in which three gasoline residual sources with different ethanol fractions (E0: no ethanol, E10: 10% ethanol and E95: 95% ethanol) were emplaced below the water table. Using the numerical model BIONAPL/3D, the mass discharge rates of benzene, toluene, ethylbenzene, xylenes, trimethylbenzenes and naphthalene were simulated and results compared to those obtained from sampling transects of multilevel samplers. It was shown that ethanol dissolved rapidly and migrated downgradient as a short slug. Mass discharge of the hydrocarbons from the E0 and E10 sources suggested similar first-order hydrocarbon decay rates, indicating that ethanol from E10 had no impact on hydrocarbon degradation. In contrast, the estimated hydrocarbon decay rates were significantly lower when the source was E95. For the E0 and E10 cases, the aquifer did not have enough oxygen to support complete mineralization of the hydrocarbon compounds to the extent suggested by the field-based mass discharge. Introducing a heterogeneous distribution of hydraulic conductivity did little to overcome this discrepancy. A better match between the numerical model and the field data was obtained assuming partial degradation of the hydrocarbons to intermediate compounds. Besides depending on the ethanol concentration, the impact of ethanol on hydrocarbon degradation appears to be highly dependent on the availability of electron acceptors.