DIFFERENTIATING BETWEEN DIFFUSION AND ADVECTION OF CHEMICAL CONSTITUENTS IN FRACTURED ROCK AQUIFERS

In fractured rock aquifers, tracer tests are often conducted under hydraulically stressed conditions, where water tagged with chemical constituents is injected into the formation and withdrawn at nearby pumping boreholes. The tracer breakthrough curves at the pumped boreholes are used to interpret the magnitude of physical and chemical processes and estimate formation properties. Usually elongated tails of breakthrough curves are attributed to diffusive processes, which in fractured rock may be the diffusion into and out of the primary porosity of the rock. Elongated tails of breakthrough curves, however, are not always indicative of diffusion; heterogeneous hydraulic properties can lead to the advection of chemical constituents and breakthrough curves with characteristics similar to that of diffusion.

Tracer tests were conducted in bedding plane fractures of the sedimentary rocks of the Newark Basin between boreholes separated by about 35 m. The breakthrough curve showed the first detection and peak concentration at 2 and 7 days, respectively, after the start of the test, and the tracer continued to be detected in the pumped borehole for approximately 6 months. The immediate interpretation may point to the significance of diffusion; however, the tracer mass collected from the pumped borehole accounted for only 20% of the injected mass. Diffusion alone is not consistent with the mass recovery. Subsequent sampling at monitoring locations between the injection and pumping boreholes revealed high concentrations of the tracer in the groundwater in the bedding plane fractures. The elevated tracer concentrations at intermediate locations between the injection and pumping boreholes point to the significance of heterogeneous hydraulic properties that leads to areas of slow advection and stagnant groundwater. Models of chemical transport that incorporate the heterogeneous hydraulic properties within bedding plane fractures can reproduce the general characteristics of the breakthrough curves; however, the injuection of the tracer solution into the formation needs to be explicitly modeled. These results illustrate that monitoring only at an abstraction point in a hydraulically stressed tracer test may not provide a unique interpretation of the physical and chemical processes that affect chemical transport.