2009 Portland GSA Annual Meeting (18-21 October 2009)

Paper No. 12
Presentation Time: 9:00 AM-6:00 PM

MODELING A CONVERGENT-FLOW TRACER TEST IN FRACTURED SEDIMENTARY ROCKS


HSIEH, Paul A., U.S. Geological Survey, Menlo Park, CA 94025, SHAPIRO, Allen M., Water Resources Division, U.S. Geological Survey, 12201 Sunrise Valley Drive, 431 National Center, Reston, VA 20192 and TIEDEMAN, Claire R., U.S. Geological Survey, 345 Middlefield Road, Mail Stop 496, Menlo Park, CA 94025, pahsieh@usgs.gov

The U.S. Geological Survey conducted a convergent-flow tracer test to evaluate groundwater flow and solute transport properties in fractured sedimentary rocks at the former Naval Air Warfare Center in West Trenton, New Jersey. At the site, fractured mudstones are contaminated with trichloroethene (TCE). At the start of the test, 194 liters of bromide solution having a bromide concentration of 3760 mg/liter was injected over 30 minutes into a well located 41 m from a well that was pumped at a constant rate of 32 liters/min. Tracer injection was followed by injecting 362 liters of bromide-free water in 60 minutes to flush the bromide from the injection well into the bedrock. The bromide breakthrough curve monitored at the pumped well showed first detection about 30 hours after injection. Peak concentration (0.097 mg/liter) was reached in 169 hours. The breakthrough curve showed a very long tail. Bromide concentration at the pumped well remained above the background concentration (about 0.036 mg/liter) for approximately 6 months. After 6 months only 20% of the bromide injected into the bedrock was recovered at the pumped well. Results indicate that a small portion of the injected tracer moved quickly to the pumped well, while the majority of the injected tracer remained in the subsurface for a long time.

The tracer test was modeled using MODFLOW to simulate groundwater flow and MT3DMS to simulate solute transport. Spatial variability of hydraulic conductivity was incorporated into the flow model in order to reproduce the observed large range of tracer residence times in the subsurface. The model simulation suggested that a small portion of the tracer was injected into high velocity flowpaths that enabled rapid transport to the pumped well. However, a large portion of the tracer was injected into regions that became relatively stagnant once the injection stopped. This portion of the tracer remained in the vicinity of the injection well and slowly diffused into higher velocity flowpaths to the pumped well, resulting in long residence times. The analysis suggested that both the heterogeneity in hydraulic conductivity and method of tracer injection exert a strong control on the outcome of tracer tests in fractured rock.