Ground Water Sampling Bias Observed in Shallow, Conventional Wells

A previous field demonstration project on nitrate-based bioremediation of a fuel-contaminated aquifer used short-screened clustered well points in addition to shallow (10 foot), conventional monitoring wells to monitor the progress of remediation during surface application of recharge. These well systems were placed in the center and at one edge of each of two treatment cells. One cell received recharge amended with nitrate (nitrate cell), and the other received unamended recharge (control cell). Data from the clustered well points were averaged to provide a mean estimate for comparison with the associated conventional monitoring well.
Conservative tracer profiles were similar for each of the four systems, with better fits obtained for well systems located at the edge of the treatment cells. However, aromatic hydrocarbon and electron acceptor profiles varied greatly for the two center well systems, with the conventional monitoring well data suggesting that remediation was proceeding at a much more rapid rate than indicated by the cluster well points. Later tests with an electromagnetic borehole flowmeter demonstrated a significant vertical flow through the well-bore of the conventional monitoring well under simulated operating conditions. This created an artifact during sampling, thought to arise from preferential flow of recharge water from the water table to deeper portions of the contaminated zone resulting in several effects, including an actual decreased residence time of water sampled by the conventional well. These data provide additional evidence that conventional monitoring wells may be inadequate for monitoring remediation in the presence of significant vertical hydraulic gradients, even for fairly shallow homogeneous aquifers.