470-481 Gw J-a 03

The release of nonaqueous phase liquids (NAPLs) to the subsurface environment can create long-term sources of ground water contamination as the NAPL slowly dissolves into ground water (Mercer and Cohen 1990; Cohen and Mercer 1993). Effective remediation of subsurface NAPL contamination requires that NAPL be accurately located and saturations quantified. This is particularly important for dense nonaqueous phase liquids (DNAPLs) because their high density causes them to migrate below the water table and move along pathways distinct from water flow (Schwille 1988; Nelson and Brusseau 1996). Laboratory and field studies have shown that partitioning tracers can be used to locate and quantify NAPL contamination (Jin et al. 1995; Wilson and Mackay 1995; Nelson and Brusseau 1996; Annable et al. 1998; Nelson et al. 1999; Young et al. 1999). Partitioning tracers have the advantage of interrogating larger aquifer volumes compared to traditional coring techniques. These studies have typically involved the injection of a suite of conservative and partitioning tracers at one well, followed by the measurement of the tracers at one or more monitoring wells (i.e., an interwell tracer test). An alternative approach involves the use of single well “push-pull tests” in which the tracers are injected and extracted from the same well (Schroth et al. 2000; Davis et al. 2002). Retardation factors for injected partitioning tracers are determined from concentration breakthrough curves and, assuming linear equilibrium partitioning, NAPL saturations are calculated (see as follows). Naturally occurring radon-222 (hereafter referred to as radon) can be used in lieu of injected partitioning tracers for locating and quantifying NAPL contamination. Radon is a naturally occurring, radioactive, inert isotope that occurs in ground water as a dissolved gas. A part of the uranium-238 decay series, radon has a half-life of 3.83 days and is continuously produced through the α-decay of radium-226 Abstract Naturally occurring radon in ground water can potentially be used as an in situ partitioning tracer to characterize dense nonaqueous phase liquid (DNAPL) saturations. The static method involves comparing radon concentrations in water samples from DNAPL-contaminated and noncontaminated portions of an aquifer, while the push-pull method involves the injection (push) and extraction (pull) of a radon-free test solution from a single well. In the presence of DNAPL, radon concentrations during the pull phase are retarded, with retardation manifested in greater dispersion of radon concentrations relative to a conservative tracer. The utility of these methods was investigated in the laboratory using a physical aquifer model (PAM). Static and push-pull tests were performed before and after contamination of the PAM sediment pack with trichloroethene (TCE), and after alcohol cosolvent flushing and pump-and-treat remediation. Numerical simulations were used to estimate the retardation factor for radon in push-pull tests. Radon partitioning was observed in static and push-pull tests conducted after TCE contamination. Calculated TCE saturations ranged up to 1.4% (static test) and 14.1% (push-pull test). Post-remediation tests showed decreases in TCE saturations. The results show that radon is sensitive to changes in DNAPL saturation in space and time. However, the methods are sensitive to DNAPL saturation heterogeneity, test location, sample size, and test design. The influence of these factors on test results, as well as the apparent overestimation of the retardation factor in push-pull tests, warrant further investigation.