POTENTIAL FOR MISINTERPRETING GROUND WATER CONTAMINATION DATA FROM WELLS IN FRACTURED CRYSTALLINE BEDROCK OWING TO CONCENTRATION AVERAGING AND FLOW RECIRCULATION

It has long been known that vertical concentration averaging in shallow overburden wells can provide data that can be misleading with respect to contaminant severity, the extent and orientation of a problem, and the source of contamination. Although largely ignored, concentration averaging in wells in fractured crystalline bedrock can also be misinterpreted. Contamination in fractured rock is generally assessed by sampling domestic bedrock wells. If domestic wells are sampled at the discharge of usage, they are useful for evaluating potability in consideration of drinking water standards. However, contaminant concentrations determined in sampling these wells can be very misleading in evaluating sources and migration of contaminants. It can be shown that concentration averaging in bedrock wells that are intersected by several water contributing fractures makes interpretation of such data dubious. Concentrations can vary depending on sampling rate, owing to differences in fracture transmissivities and hydraulic heads, the number of contributing fractures, and which fractures are contaminated. Conceptual models will be presented that illustrate alternative interpretations of bedrock contamination. Pesticide contamination in Stamford, Connecticut will be used to illustrate the situation. The presence of a high density of domestic bedrock wells and associated septic fields presents a potential complex situation where ground water pumped from the bedrock can be recirculated back to the bedrock from septic discharge in the overburden. Contaminant concentrations appear clustered in some places and spotty in others. Contamination may be detected at one home but not next door. There is a lack of plume definition. No definitive source has yet to be identified. Identifying and sampling the contributing fractures that intersect wells are needed to improve site assessments. But this is hampered by costs and logistical issues. Even then to evaluate flow continuity requires larger scale hydraulic and tracer tests. The results of these tests, however, can also be influenced by averaging. More detail needs to be paid to the three-dimensional nature of a flow field which requires well profiling and good characterization during drilling. More cost effective methods are also needed for downhole fracture and contaminant characterization.