Environ. Sci. Technol. 2006, 40, 6123-6130
Impact of Ethanol on the Natural Attenuation of Benzene, Toluene, and o-Xylene in a Normally Sulfate-Reducing Aquifer D O U G L A S M . M A C K A Y , * ,† NICHOLAS R. DE SIEYES,† MURRAY D. EINARSON,‡ KEVIN P. FERIS,† ALEXANDER A. PAPPAS,† ISAAC A. WOOD,† LISA JACOBSON,† LARRY G. JUSTICE,† MARK N. NOSKE,† KATE M. SCOW,† AND JOHN T. WILSON§ Department of Land, Air, and Water Resources, University of California at Davis, Davis, California 95616, Geomatrix Consultants, Oakland, California 94612, and U. S. Environmental Protection Agency, Ada, Oklahoma 74821
Side-by-side experiments were conducted in a sulfatereducing aquifer at a former fuel station to evaluate the effect of ethanol on biodegradation of other gasoline constituents. On one side, for ∼9 months we injected groundwater amended with 1-3 mg/L benzene, toluene, and o-xylene (BToX). On the other side, we injected the same, adding ∼500 mg/L ethanol. Initially the BToX plumes on both sides (“lanes”) extended approximately the same distance. Thereafter, the plumes in the “No Ethanol Lane” retracted significantly, which we hypothesize to be due to an initial acclimation period followed by improvement in efficiency of biodegradation under sulfate-reducing conditions. In the “With Ethanol Lane”, the BToX plumes also retracted, but more slowly and not as far. The preferential biodegradation of ethanol depleted dissolved sulfate, leading to methanogenic/acetogenic conditions. We hypothesize that BToX in the ethanol-impacted lane were biodegraded in part within the methanogenic/acetogenic zone and, in part, within sulfate-reducing zones developing along the plume fringes due to mixing with sulfatecontaining groundwater surrounding the plumes due to dispersion and/or shifts in flow direction. Overall, this research confirms that ethanol may reduce rates of biodegradation of aromatic fuel components in the subsurface, in both transient and near steady-state conditions.
Introduction Ethanol is an increasingly common component of fuels, in part due to phasing out of methyl-tert-butyl ether (MTBE) as a fuel oxygenate. It is important to determine if ethanol poses significant risks to groundwater, including through impacts on the biodegradation of other fuel components. Ethanol should degrade rapidly and without an acclimation period under most redox conditions unless present at high * Corresponding author phone: (650)324-2809; fax (650)618-1571; e-mail:
[email protected]. † University of California at Davis. ‡ Geomatrix Consultants. § U. S. Environmental Protection Agency. 10.1021/es060505a CCC: $33.50 Published on Web 09/01/2006
2006 American Chemical Society
concentrations, i.e., near large and/or recent spills of ethanolblended fuels (1). Ethanol would be expected, however, to influence the natural biodegradation of fuel components, notably the BTEX compounds (benzene, toluene, ethylbenzene, and xylene isomers), due to rapid preferential biodegradation of ethanol causing (1) depletion of readily available electron acceptors, thus slowing or stopping BTEX biodegradation by native microbes, due to slower kinetics compared to more energetically favorable metabolic pathways, and/or (2) reduction of the fraction of the native microbial community able to biodegrade BTEX (1-8). Laboratory batch studies have shown that BTEX compounds persist in the presence of ethanol and sometimes for considerable periods of time after ethanol is exhausted (23, 9). However, batch studies do not capture the dynamics of biodegradation in the subsurface, where flowing groundwater delivers more dissolved species and microorganisms to contaminated zones and where the variation of redox conditions along flowpaths allows spatially variable and distinct microbial communities to develop. Computer simulations of ethanol impacts on BTEX may capture more field complexity but often make limiting assumptions, e.g., that only aerobic biodegradation is of significance (10); under that assumption, simulated benzene plumes are longer in the presence than absence of ethanol but by less than a factor of 3. Similar apparent impacts on plume length were noted in a study of BTEX plumes in the U. S. Midwest: In aquifers previously exposed to ethanol-amended gasoline, the median length of benzene plumes was 80 m, compared to 48 m at sites where ethanol-amended fuel was never used (7). These observations are generally consistent with results from a controlled field experiment in a previously uncontaminated aerobic aquifer (11) in which reduced BTEX degradation rates were hypothesized to result from depletion of electron acceptors by preferential degradation of ethanol. The overall goal of this research was to determine the effects of ethanol on the natural attenuation of other gasoline constituents through a controlled field experiment at a typical fuel station. Our goal was not to examine the worst conditions that might occur for a large and catastrophic gasohol spill but rather to examine the probably more common case of long-term, small-volume releases of gasohol to the subsurface at an operating fuel service station with no ongoing active remediation of past spills. Most fuel-contaminated sites in the U. S. overlie aquifers in which the dominant dissolved electron acceptor is sulfate (12), so we selected such a site for this work. This paper addresses impacts on selected BTEX species, while a subsequent paper addresses impacts on MTBE.
Site and Pre-Experimental Conditions Site 60, Vandenberg Air Force Base (VAFB), CA, has been described previously (13), but we summarize and/or update this information below and in the Supporting Information. A gasoline leak was noted in 1994 at the fuel service station; tanks and piping were excavated in 1995, and the excavation was backfilled with relatively permeable media (Figure 1). Around the backfill, the surficial media are relatively impermeable and in some areas covered with pavement. Downgradient of the backfill, several thin, horizontal, sandy layers exist within 8 m of the ground surface (Supporting Information). Layers of less permeable silt or clayey silt separate these sandy layers. The layers are not as distinct or internally homogeneous as conceptualized, yet within the study area the S3 aquifer (∼1 m thick) is the primary pathway for downgradient advection of water and contaminants from VOL. 40, NO. 19, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6123
FIGURE 1. Map of the experimental area immediately around and downgradient of the former gas station. EUG to EK are transects of monitoring wells. The arc of background monitoring wells contains two (circled) used to supply water for injection. Injection occurred in two sets of 3 wells in the ER transect to create two “lanes” of injected water (“Without Ethanol Lane” and “With Ethanol Lane” depicted as blue and red outlined arrows, respectively). the backfill area. Outside of the backfilled excavation, vertical movement of water is impeded by low permeability surficial and deeper layers. For years, there has been very little BTEX or other petroleum hydrocarbon contamination remaining in the source area and almost no detectable downgradient migration of BTEX or other petroleum hydrocarbons in the S3 aquifer, apparently due to excavation of the spill and past/ ongoing biodegradation under predominantly sulfate-reducing conditions. For example, gasoline-range total petroleum hydrocarbon (TPHg) was
