Observations from borehole dilution logging experiments in fractured crystalline rock under variable hydraulic conditions

Identifying hydraulically active fractures in low permeability, crystalline-bedrock aquifers requires a variety of geophysical and hydrogeophysical borehole tools and approaches. One such approach is Single Borehole Dilution Tests (SBDT), which in some low flow cases have been shown to provide greater resolution of borehole flow than other logging procedures, such as vertical differential Heat Pulse Flowmeter (HPFM) logging. Because the tools used in SBDT collect continuous profiles of water quality or dye changes, they can identify horizontal flow zones and vertical flow. We used SBDT with a food grade blue dye as a tracer and dual photometer-nephelometer measurements to identify low flow zones.

SBDT were conducted at seven wells with open boreholes (exceeding 300 ft). At most of the wells HPFM logs were also collected. The seven wells are set in low-permeability, fractured granite and gneiss rocks underlying a former tetrachloroeythylene (PCE) source area at the Savage Municipal Well Superfund site in Milford, NH. Time series SBDT logs were collected at each of the seven wells under three distinct hydraulic conditions: (1) ambient conditions prior to a pump test at an adjacent well, (2) mid test, after 2-3 days of the start of the pump test, and (3) at the end of the test, after 8-9 days of the pump test. None of the SBDT were conducted under pumping conditions in the logged well. For each condition, wells were initially passively spiked with blue dye once and subsequent time series measurements were made.

Measurement accuracy and precision of the photometer tool is important in SBDT when attempting to detect low rates of borehole flow. Tests indicate that under ambient conditions, none of the wells had detectable flow as measured with HPFM logging. With SBDT, 4 of the 7 showed the presence of some very low flow. None of 5 (2 of the 7 wells initially logged with HPFM under ambient conditions were not re-logged) wells logged with the HPFM during the pump test had detectable flow. However, 3 of the 5 wells showed the patterns of very low flow with SBDT during the pump test including pumping induced changes of inflow and outflow patterns at one well.