Quantifying groundwater flow variability in a poorly cemented fractured sandstone aquifer to inform in situ remediation.

This study applies innovative methods to characterize and quantify the magnitude of groundwater flow in a fractured and variably cemented sandstone aquifer to inform an in-situ remediation strategy for trichloroethene (TCE) contamination. A modified active-distributed temperature sensing (A-DTS) approach in which fiber optic cables were permanently grouted in the borehole was used to quantify groundwater flow rates. Two additional tracer tests were conducted: 1) fluorescein tracer injection followed by rock coring and sampling for visual mapping and porewater analysis, and 2) deployment of passive flux meters in conventional monitoring wells to evaluate groundwater velocity and mass flux distributions. Forced gradient injection of fluorescein tracer suggests a dual porosity flow system wherein higher rates of groundwater flow occur within discrete features including highly permeable bedding planes and fractures, with slower flow occurring within the rock matrix. Tracer was observed and detected in the unfractured matrix porewater >1.5 m away from the injection well. Beyond this distance, >6 m radially away from the injection hole, tracer was primarily detected within and adjacent to high transmissivity fractures serving as preferential flow paths. The Darcy flux calculated using active distributed temperature sensing (A-DTS) shows depth-discrete values ranging from 7 to 60 cm/day, with average and median values of 23 and 17 cm/day, respectively. Passive Flux Meters (PFMs) deployed in three conventional monitoring wells with slotted screens and sand filter packs showed groundwater flux values ranging from 2 to 11 cm/day, with an overall average of 4 cm/day and are likely biased low due to spreading in the sand pack. The study results were used to inform an in-situ remediation system design including the proposed injection well spacing and the amendment delivery approach. In addition, the results were used to build confidence in the viability of delivering an oxidant to the rock matrix via advective processes. This is important because 1) the matrix is where the majority of the TCE mass occurs, and 2) it provides insights on processes that directly affect remedial performance expectations given advective delivery to preferential pathways and the matrix overcomes diffusion only conditions.

Improving estimates of groundwater velocity in a fractured rock borehole using hydraulic and tracer dilution methods.

A straddle-packer system for use in boreholes in fractured rock was modified to investigate the average linear groundwater velocity (v¯f) in fractures under ambient flow conditions. This packer system allows two different tests to be conducted in the same interval between packers without redeploying the system: (1) forced gradient hydraulic tests to determine the interval transmissivity (T), and (2) borehole dilution experiments to determine the groundwater flow rate (Qt) across the test interval. The constant head step test method provides assurance that flow is Darcian when determining T for each interval and identifies the flow rate at the onset of non-Darcian flow. The critical Reynolds number method uses this flow rate to provide the number of hydraulically active fractures (N) in each interval, the average hydraulic aperture for the test interval and the effective bulk fracture porosity. The borehole dilution method provides Qt values for the interval at the time of the test, and v¯f can be estimated from Qt using the flow area derived from the hydraulic tests. The method was assessed by application to seven test intervals in a borehole 73 m deep in a densely fractured dolostone aquifer used for municipal water supply. The critical Reynolds number method identified one or two fractures in each test interval (1.1 m long), which provided v¯f values in the range of 10 to 8000 m/day. This velocity range is consistent with values reported in the literature for ambient flow in this aquifer. However, when hydraulically active fractures in each interval is identified and measured from acoustic and optical televiewer logs, the calculated v¯f values are unreasonably low as are the calculated values of the hydraulic gradient needed to provide the Qt value for each tested interval. The combination of hydraulic and dilution tests in the same interval is an improved method to obtain values of groundwater velocity in fractured rock aquifers.