Effect of Biofuels on Biodegradation of Benzene and Toluene at Gasoline Spill Sites

Abstract

The risk that benzene and toluene from spills of gasoline will impact drinking water wells is largely controlled by the natural anaerobic biodegradation of benzene and toluene. Benzene and toluene, as well as ethanol and other biofuels, are degraded under anaerobic conditions to the same pool of degradation products. Biodegradation of biofuels may produce concentrations of degradation products that make the thermodynamics for degradation of benzene and toluene infeasible under methanogenic conditions and produce larger plumes of benzene and toluene. This study evaluated the concentrations of fuel alcohols that are necessary to inhibit the anaerobic degradation of benzene and toluene under methanogenic conditions. At two ethanol spill sites, concentrations of ethanol greater ≥42 mg/L inhibited the anaerobic degradation of toluene. The pH and concentrations of acetate, dissolved inorganic carbon, and molecular hydrogen were used to calculate the Gibbs free energy for the biodegradation of toluene. In general, the anaerobic biodegradation of toluene was not thermodynamically feasible in water with ≥42 mg/L ethanol. In a microcosm study, when the concentrations of ethanol were ≥14 mg/L or the concentrations of n-butanol were ≥16 mg/L, the biodegradation of the alcohols consistently produced concentrations of hydrogen, dissolved inorganic carbon, and acetate that would preclude natural anaerobic biodegradation of benzene and toluene by syntrophic organisms. In contrast, iso-butanol and n-propanol only occasionally produced conditions that would preclude the biodegradation of benzene and toluene.

Introduction

When motor fuel is released into groundwater, the fuel can produce plumes of benzene, toluene, ethylbenzene, and xylenes (BTEX) that may impact drinking water wells. When ethanol is present as a component of the fuel, it can increase the length of the plume of benzene or toluene by inhibiting natural benzene biodegradation (Ma et al. 2013). Ruiz-Aguilar et al. (2003) compared the lengths of the benzene and toluene plumes at 217 spill sites in Iowa, where ethanol was not used in the fuel, to 29 sites in Kansas, where ethanol was a component of the fuel. The mean length of the benzene plume in Iowa was 59 m, compared to 80 m in Kansas. The presence of ethanol increased the length of the benzene plume on average by 36%.

Several computer models are available to predict the impact of ethanol on a benzene plume (Deeb et al. 2002; Molson et al. 2002; Gomez et al. 2008; Gomez and Alvarez 2009; Gomez and Alvarez 2010). These models indicate that whenever ethanol is present in groundwater at concentrations of several hundreds of mg/L, the length of the benzene plume is longer than it would be if ethanol was not present.

The model projections and the empirical survey data on the effect of ethanol on plume lengths have been confirmed by two controlled-release field experiments. Mackay et al. (2006) added approximately 500 mg/L of ethanol to sulfate-reducing groundwater at a motor gasoline spill at Vandenberg AFB, CA. The ethanol inhibited the degradation of benzene and allowed an experimental plume of benzene to expand further than a second experimental plume that was not supplemented with ethanol. Corseuil et al. (2011) released 100 L of Brazilian gasoline containing 24% ethanol into an aquifer at the Ressacada Experimental Farm, Florianopolis, Santa Catarina State, Brazil. They monitored the evolution of the plumes of ethanol and benzene for 10 years. As long as ethanol was present in the groundwater, the plume of benzene continued to expand. The ethanol in the groundwater biologically degraded to acetate and methane, creating conditions that prevented the anaerobic biodegradation of benzene. Once the ethanol was removed, the plume of benzene started to decline through natural anaerobic biodegradation of the benzene.

In the interior of many old and well-acclimated plumes from fuel spills, sulfate and the other soluble electron acceptors, such as oxygen or nitrate, are exhausted. Under these conditions, the initial degradation of benzene or toluene will be carried out by a group of syntrophic microorganisms that ferment the benzene or toluene to acetate or other low-molecular-weight fatty acids and molecular hydrogen. The word syntrophic is a combination of the Greek words syn meaning together and trophe meaning nourishment. Syntrophic bacteria must work with other bacteria to carry out their metabolism. In the syntrophic fermentation reaction, a part of the benzene or toluene molecule is oxidized and a part is reduced. Benzene or toluene serves as both electron donor and electron acceptor. There is no need for an additional electron acceptor.

The fermentation reaction carried out by the syntrophic organisms can only proceed when the concentrations of acetate and hydrogen are low enough to make the reaction thermodynamically feasible. If the degradation products of the fermentation reaction (such as acetate or other low-molecular-weight fatty acids and molecular hydrogen) are not further degraded, they will accumulate and prevent the further degradation of benzene or toluene. Various methanogenic bacteria can consume the acetate and hydrogen and produce methane and carbon dioxide. The syntrophic bacteria carry out the initial metabolism of benzene and toluene, and the methanogenic bacteria complete the degradation of the fermentation products.

Vogt et al. (2011) reviewed the biochemistry and microbiology of the degradation of benzene to acetate or other fatty acids and hydrogen by syntrophic microorganisms. The pathway for the degradation of benzene can be generalized as Equation 1. A similar pathway for the degradation of toluene is generalized in Equation 2. The equation assumes that the methyl group of toluene is oxidized to bicarbonate. The syntrophic degradation of ethanol is described in Equation 3.

$${\mathrm{C}}_6{\mathrm{H}}_6+6{\mathrm{H}}_2\mathrm{O}\to 3{\mathrm{CH}}_3{\mathrm{COO}}^{-}+3{\mathrm{H}}^{+}+3{\mathrm{H}}_2$$ (1)

$${\mathrm{C}}_7{\mathrm{H}}_8+{\mathrm{H}}_2\mathrm{O}\to 3{\mathrm{CH}}_3{\mathrm{COO}}^{-}+4{\mathrm{H}}^{+}+6.5{\mathrm{H}}_2+{{\mathrm{HCO}}_3}^{-}$$ (2)

$${\mathrm{CH}}_3{\mathrm{CH}}_2\mathrm{O}\mathrm{H}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{CH}}_3{\mathrm{COO}}^{-}+{\mathrm{H}}^{+}+2{\mathrm{H}}_2$$ (3)

As described in Equation 3, ethanol is fermented to the same common pool of intermediates. The anaerobic degradation of ethanol produces acetate, hydrogen, and protons that can inhibit the degradation of benzene and toluene. Following the approach of Corseuil et al. (2011) and Rakoczy et al. (2011), we used thermodynamics to evaluate the possibility of anaerobic biodegradation of benzene and toluene in the presence of the common pool of degradation intermediates.

In groundwater systems that are actively methanogenic, the rate of degradation of acetate and hydrogen to methane is fixed at a maximum value that is characteristic of the aquifer system. Any further increase in the density of the methanogenic organisms is limited by predators or bacterial viruses. There is no such limitation on the rate of degradation of ethanol to acetate and hydrogen. As ethanol is degraded, the acetate and hydrogen will accumulate until the pool of acetate and hydrogen is near the concentrations that make the further degradation of ethanol no longer feasible.

At higher concentrations of ethanol, the particular concentrations of acetate and hydrogen that are necessary to prevent further degradation of the ethanol are also higher. Above a certain concentration of ethanol, the concentrations of acetate and hydrogen that are produced by the ethanol-degrading bacteria should inhibit the syntrophic degradation of benzene and toluene. This concentration of ethanol would be a threshold above which the anaerobic degradation of benzene or toluene should not be expected.

Deeb et al. (2002) offer a simple conceptual model of the impact of ethanol on benzene biodegradation. Whenever the concentration of ethanol is above a threshold concentration, there is no biodegradation of benzene, and any attenuation in the concentrations of benzene is caused by physical processes such as dilution, dispersion, or sorption. As the benzene plume moves down the flow path, after a certain period of time, the degradation of ethanol brings the concentration of ethanol below the threshold. From that point forward in space, the benzene is biologically degraded at a rate that is characteristic of the aquifer.

To help state agencies evaluate the effect of ethanol on benzene plumes, the U.S. Environmental Protection Agency (U.S. EPA) created a simple screening model (called FOOTPRINT) based on the conceptual model of Deeb et al. (2002). The FOOTPRINT model (U.S. EPA 2008) requires a constant rate for the biodegradation of ethanol, a threshold concentration of ethanol above which benzene degradation is inhibited, and a constant rate for the biodegradation of benzene in water whenever the concentration of ethanol is below the threshold. Deeb et al. (2002) offered a default value for the threshold of 3 mg/L ethanol. However, they provided little theoretical justification for the default value. The robust application of the model requires a value for the threshold that is based on the actual constraints on the biodegradation of benzene under anaerobic conditions. This study provides two lines of evidence to determine the threshold.

We sampled groundwater at two ethanol spill sites and determined the pH and concentrations of acetate, dissolved inorganic carbon, and hydrogen that developed during the fermentation of ethanol to methane and carbon dioxide. We also conducted laboratory microcosm studies to determine the pH and the concentrations of acetate, dissolved inorganic carbon, and hydrogen that developed during the fermentation of ethanol.

Other alcohols that are occasionally seen in motor gasoline include n-butanol, iso-butanol, and n-propanol. The fuel component marketed as biobutanol is predominately iso-butanol. In anticipation of the use of these alcohols in gasoline in the future, we included them in the microcosm study.

We calculated the Gibbs free energy available for the anaerobic biodegradation of benzene or toluene under the conditions in the groundwater at the spill sites and in the microcosms. To evaluate the effects of the alcohols on the possibility for natural anaerobic biodegradation, we compared the measured concentrations of the alcohols to the thermodynamic feasibility for anaerobic biodegradation of benzene or toluene.

In their review, Weelink et al. (2010) noted that “All studies regarding anaerobic BTEX degradation have indicated that anaerobic benzene degradation is most difficult and that toluene is one of the aromatic compounds which is relatively easy to degrade anaerobically.” Vogt et al. (2011) noted that “Benzene is considered to be more persistent under anoxic conditions than its alkylated derivatives toluene, ethylbenzene and xylene isomers and the reasons for the recalcitrance of benzene are not yet clear.” One possible explanation is the fact that many alkylbenzenes are degraded under anaerobic conditions through the addition of fumarate to a side chain of the molecule (Weelink et al. 2010; Vogt et al. 2011). This pathway is not available for benzene. Bruce et al. (2010) provide experimental documentation that toluene degrades more rapidly than benzene at a field scale.

Because toluene degrades faster than benzene, the ratio of the concentration of toluene to benzene decreases as they are degraded. To further evaluate the effects of ethanol on the anaerobic biodegradation of toluene, we compared the ratio of toluene to benzene at two fuel spill sites to the concentration of ethanol in the groundwater.

Materials and Methods

Analyses for Alcohols

The samples were analyzed for ethanol, n-butanol, iso-butanol, and n-propanol by a modification of the EPA Method 8260 (U.S. EPA 2013). In the field samples, the method detection limit was 0.025 mg/L for ethanol, 0.014 mg/L for n-propanol, 0.015 mg/L for iso-butanol, and 0.016 mg/L for n-butanol.

Analyses for Organic Acids

Samples were analyzed for acetate, n-butyrate, iso-butyrate, and n-propionate by high-performance liquid chromatography using a Dionex IonPac ICE-AS1 ion exclusion column and a conductivity detector with an anion micro membrane. The samples were frozen to preserve them until analysis. The method detection limits were less than 0.1 mg/L for acetate, n-propionate, iso-butyrate, and n-butyrate.

Analyses for Hydrogen and Methane

The concentration of molecular hydrogen and methane was determined in the gas headspace of the microcosms using an Agilent Micro 3000 gas chromatograph. The procedure followed that of Feeney and Larson (2002) with modifications. The method detection limit and quantitation limit for hydrogen were 4.6 and 15 ppm v/v, respectively. The method detection limit and quantitation limit for methane were 3 and 15 ppm v/v, respectively.

Groundwater samples for methane and dissolved hydrogen were collected without headspace in a 50-mL serum vial. The samples were preserved with trisodium phosphate to pH >10.5. A 10% headspace was created in the vial, and the vial was shaken to equilibrate gases into the headspace (Kampbell and Vandegrift 1998). Then, the headspace gas was analyzed as described above. The appropriate Henry’s Law constants and the ratio of gas phase to water phase were used to calculate the concentration of gases originally dissolved in the water sample. The method detection limit and quantitation limit of hydrogen was 0.02 and 0.7 μg/L, respectively. The method detection limit and quantitation limit of methane was 0.2 and 3 μg/L, respectively.

Analyses of Benzene and Toluene

The samples were analyzed for benzene and toluene by a modification of the EPA Method 8260 (U.S. EPA 2013). The method detection limit for benzene was 0.06 μg/L, and the reporting limit was 0.5 μg/L. The method detection limit for toluene was 0.03 μg/L, and the reporting limit was 0.5 μg/L. The samples from the microcosms were diluted 30-fold before analysis.

Analyses of pH

The pH of the water in microcosms was determined using an Accumet Excel XL 25 Dual Channel pH/Ion Meter. On day 76 and day 976 of incubation, 2.0 mL of standing water from the microcosms that had been amended with ethanol was collected and analyzed for pH. On day 270 and day 1170 of incubation, 2.0 mL of standing water was collected from the microcosms that were amended with n-butanol, iso-butanol, and n-propanol and analyzed for pH. The pH of the water at the site at Berryville, VA was determined using a Yellow Springs Instrument multiparameter meter model #YSI-556 MPS. The pH of the water at the site at Montvale, VA was determined with a Horiba multiparameter meter. Both meters monitored well water in a flow-through cell that was supplied by a peristaltic pump.

Analyses of Dissolved Inorganic Carbon

The concentration of dissolved inorganic carbon was determined by a Dohrman DC-80 Carbon Analyzer. The method detection limit is 0.12 mg/L. The reporting limit is 0.5 mg/L.

Construction of Laboratory Microcosms

Laboratory microcosms were constructed using sediment from an old spill of JP-4 jet fuel at the U.S. Coast Guard Support Center in Elizabeth City, NC. This sediment was well acclimated for the anaerobic biodegradation of benzene under methanogenic conditions (Wilson et al. 2000).

The microcosms were constructed in 160-mL serum bottles. They contained 40 mL of wet sediment (approximately 96 grams dry weight). The sediment contained 16 mL of pore water. An additional 5 mL of water was added for a total content of 21 mL of water. The gas headspace occupied 115 mL. The microcosms were spiked with selected alcohols and BTEX compounds by adding solutions of the compounds in water. The intended final concentration of ethanol, n-propanol, n-butanol, and iso-butanol was 2,000 mg/L. These concentrations are plausible concentrations of fuel alcohols in groundwater in contact with a fresh spill of gasoline.

The intended final concentration of benzene, toluene, ethylbenzene, o-xylene, m-xylene, and p-xylene was 1 mg/L of each compound. Microcosms were spiked with alcohols and BTEX compounds in July 2009.

Ethanol degraded more rapidly than we expected. The microcosms that were amended with ethanol were amended a second time to produce an intended final concentration of 5000 mg/L in January 2010. Only the data from the second amendment are presented. Microcosms were incubated on a roller in the dark at room temperature. The microcosms turned over once every 15 s. The roller mixed the sediment in each microcosm and maintained uniform conditions within each microcosm.

There was no clear evidence that toluene and benzene degraded in the microcosms even after 693 days of incubation. Due to an error, the microcosms were constructed with sediment that had residual NAPL. The residual NAPL was highly weathered and depleted in toluene and benzene. The initial concentrations of toluene and benzene in the pore water were more than 10-fold lower than the expected concentrations because the toluene and benzene that were added to the microcosms partitioned into the NAPL. As a result, the initial concentrations were near the method-reporting limit. It was not possible to distinguish removals in the living microcosms from removals in the sterile autoclaved controls. See Figure 1 for example data.

Figure 1.

Reduction of concentrations of benzene in a living microcosm that was amended with BTEX (#27), a microcosm amended with ethanol and BTEX (#30), and a sterilized microcosm amended with BTEX (#37).

Corrections for Dilution in Microcosms

The total water content of the microcosms was 21 mL. At each sampling interval, up to 4.0 mL of water was taken from the microcosms. After all the samples were taken, the standing water that was taken for analysis was replaced with up to 4.0 mL of sterile, deaired RO water. The reported concentrations were corrected for the dilution caused by the volume of water that was added to replace the water taken for analysis.

Theoretical Framework

Thermodynamic Feasibility of Benzene and Toluene Degradation The Nernst Equation was used to evaluate the thermodynamic feasibility of the biodegradation of benzene by syntrophic bacteria under the conditions that prevailed in the microcosm study. We assumed that benzene was degraded to acetate, hydrogen ions, and molecular hydrogen following Equation 1 and that toluene was degraded following Equation 2. The value of DG′ for benzene was calculated following Equation 4, and the value for toluene was calculated by Equation 5:

$$\mathrm{\Delta }G\prime =\mathrm{\Delta }{G}^{o\prime }+RT*\mathrm{Ln}\left[\frac{{\left[{\mathrm{CH}}_3{\mathrm{COO}}^{-}\right]}^3{\left[{\mathrm{H}}^{+}\right]}^3{\left[{\mathrm{H}}_2\right]}^3}{{\left[{\mathrm{C}}_6{\mathrm{H}}_6\right]}^1}\right]$$ (4)

$$\mathrm{\Delta }G\prime =\mathrm{\Delta }{G}^{o\prime }+RT*\mathrm{Ln}\left[\frac{{\left[{\mathrm{CH}}_3{\mathrm{COO}}^{-}\right]}^3{\left[{\mathrm{H}}^{+}\right]}^4{\left[{\mathrm{H}}_2\right]}^{6.5}{\left[\mathrm{H}\mathrm{C}{\mathrm{O}}_3^{-}\right]}^1}{{\left[{\mathrm{C}}_7{\mathrm{H}}_8\right]}^1}\right]$$ (5)

where DGo ′ is the change in free energy for Equation 1 or Equation 2 as calculated from the heats of formation of the reactants and products; where R = 0.008314 kJ/(mol K); and where T is the temperature in K. Calculations and heats of formation follow Thauer et al. (1977). The value of DGo ′ for benzene was taken to be 181 kJ/mol, and the value of DGo ′ for toluene was considered to be 313 kJ/mol. The heat of the formation of benzene and toluene was corrected for the heat of the solution following the procedure applied to naphthalene in Dolfing et al. (2009).

The concentrations are molar concentrations, with the exception of molecular hydrogen. Following the usage of Jackson and McInerney (2002), the concentration of hydrogen is the concentration in the atmosphere in the headspace above the sediment and porewater. Data on pH in the microcosm studies were only available for two sampling dates. The calculations were performed with the value of pH for the date that was closest in time to the sampling date.

Results

Conditions at the Montvale, Virginia Site

A release of fuel-grade ethanol (E95) was reported at a gasoline service station in Montvale, VA in February 1998. The release was discovered when an underground storage tank was excavated. Samples from eight wells were taken in June, September, and December 2011 and March 2012 (Figure 2a, Table S1, Supporting Information). The most contaminated wells were MW-2 and MW-11. At the time the samples were collected, the spill was being remediated by a pump and treat system. All the monitoring wells were contained within the capture zone of the system. The concentration of ethanol was as high as 4280 mg/L. The concentration of acetate was as high as 557 mg/L. The highest concentration of dissolved hydrogen was equivalent to 64,634 ppm in the gas phase. Despite the fact that the spill was 13 years old, the concentrations of benzene were as high as 6.1 mg/L.

Figure 2.

Locations of monitoring wells at the (a) Montvale and (b) Berryville field sites.

Biodegradation of Toluene at Field Sites

At these two sites, it was difficult to directly document the degradation of toluene by comparing changes in the concentration of toluene over time or with distance from the source of the spill. Groundwater was pumped from both sites as part of the remedy. As water table elevations changed, more or less of the NAPL source was in contact with the groundwater. The concentration of toluene in wells downgradient of the source was diluted with clean water that was produced along with the contaminated water. To remove these interferences, the ratio of the concentration of toluene to the concentration of benzene (μg/L/μg/L) will be used as an index for the extent of the degradation of toluene. Under anaerobic conditions in groundwater, toluene degrades more rapidly than benzene (Bruce et al. 2010). As a result, the ratio of the concentration of toluene to benzene decreases as toluene and benzene are degraded.

In wells in the source area of groundwater contamination at the two sites, the ratio of concentrations of toluene to benzene varied from 10:1 to 1:1, regardless of the concentration of toluene in the water (Figure 3). In these wells, partitioning from NAPL controlled the concentrations of toluene and benzene and masked any contribution of the biodegradation of toluene.

Figure 3.

Effect of proximity to NAPL in the source area of the release on the ratio of concentrations of toluene to benzene in groundwater. If the ratio of the concentration of toluene to benzene is less than 1.0, that is considered evidence for the anaerobic degradation of toluene.

In addition to the preferential biodegradation of toluene, there are two additional interactions that can perturb the ratio of toluene to benzene in groundwater. At high concentrations, ethanol increases the concentration of benzene and toluene dissolved in groundwater in contact with the gasoline present as an NAPL (Corseuil et al. 2004). This cosolvency effect of ethanol is stronger for toluene than for benzene, and the presence of ethanol in groundwater tends to increase the ratio of toluene to benzene. The flow of groundwater over the NAPL, from either the natural flow of groundwater or the pump and treat remedies, can be expected to physically weather the NAPL. Benzene is more soluble in water than toluene, and a greater proportion of the benzene would be removed from the NAPL by the flow of water. As the NAPL weathers over time, the ratio of toluene to benzene in groundwater in contact with the NAPL should increase. Because these two interactions act contrary to a selective biodegradation of toluene, they cannot provide a false indication that toluene is being selectively degraded.

Figure 4 compares the ratio of toluene to benzene in groundwater from the two field sites to the concentration of ethanol. The figure only includes data that was “downgradient of the release” in Figure 3. At the Montvale site, when concentrations of ethanol were greater than 18 mg/L, the ratio was closer 4:1 and was consistently above 1:1. There was no evidence for toluene degradation at concentrations of ethanol above 18 mg/L. At concentrations of ethanol less than 1 mg/L, the ratio is often much lower (Figure 4a). The shift in the ratio when the concentration of ethanol was less than 1 mg/L indicates that toluene was being degraded.

Figure 4.

Relationship between the anaerobic biodegradation of toluene in groundwater at two ethanol spill sites and the concentration of ethanol in the groundwater, the thermodynamic feasibility for the anaerobic biodegradation of toluene, and the feasibility of benzene degradation. A decrease in the ratio of the concentration of toluene to benzene is considered an indication of toluene biodegradation. Solid symbols in panels (d) and (f) are exceptions; see text for description. (a) Ratio of toluene to benzene at Montvale site, (b) ratio of toluene to benzene at Berryville site, (c) toluene at Montvale site, (d) toluene at Berryville Site, (e) benzene at Montvale site, (f) benzene at Berryville site.

Similar to the the case at the Montvale site, when the concentration of ethanol was greater than 42 mg/L at the Berryville Site, the ratio was close to 4:1 and was consistently above 1:1 (Figure 4b). At one well, at one sampling time, the concentration of ethanol was 3.3 mg/L, and the ratio was 0.3, indicating the degradation of toluene at this concentration of ethanol (Figure 4b). At other wells and other sampling times, there was no indication of toluene degradation until the concentration of ethanol was at or below the analytical detection limit of 0.025 mg/L.

At both the Montvale site and the Berryville site, there was no evidence for toluene degradation when the concentration of ethanol was greater than 42 mg/L.

Thermodynamic Feasibility of Toluene and Benzene Degradation in Groundwater at Spill Sites

Figure 4 compares the thermodynamic feasibility for the anaerobic biodegradation of toluene to the concentration of ethanol (Panels [c] and [d]). For each data point, the concentrations of toluene, acetate, bicarbonate ion, hydrogen ion, and molecular hydrogen were entered into Equation 5 to calculate the value of DG′ for the degradation of toluene that pertained to conditions in that well at that time of sampling. Figure 4 also compares the thermodynamic feasibility for the anaerobic biodegradation of benzene to the concentration of ethanol (Panels [e] and [f]). The concentrations of benzene, acetate, hydrogen ion, and molecular hydrogen were entered into Equation 4 to calculate the value of DG′ for the degradation of benzene. The concentration data and associated calculated values of DG′ are presented in Tables S1 and S2.

Degradation is theoretically feasible if the value of DG′ is negative. Schink (1997) noted that the minimum amount of energy that can be converted to make ATP is in the range of −20 kJ per reaction. He reasoned that DG′ must be more negative than −20 kJ per reaction for biodegradation to proceed. However, Jackson and McInerney (2002) showed that certain syntrophic bacteria can carry out reactions at somewhat less negative values for DG′ than −20 kJ per reaction. In Figure 4, the shaded bar bounds values of DG′ between 0 and −20 kJ/mole. If the value of DG′ is above the bar, then the syntrophic biodegradation of toluene should be feasible. If the value is below the bar, then degradation should not be feasible. If the value is within the bar, biodegradation may or may not be feasible.

Biodegradation may not occur even though the conditions for biodegradation are thermodynamically feasible. The appropriate organisms may not be present, or the organisms may be present at such s low population density that they do not impact the concentrations of benzene or toluene. The energy available to the organisms may not be adequate to support growth of the organisms, which would lead to an increase in their numbers and result in the observable consumption of toluene or benzene. In this case, although values of DG′ are negative, or even more negative than −20 kJ/mole, observable biodegradation may not occur. On the other hand, if biodegradation is not thermodynamically feasible, it cannot occur. If values of DG′ are positive, degradation cannot occur at the field sites.

There is a relationship between the concentrations of ethanol and conditions that make the anaerobic biodegradation of toluene thermodynamically unfeasible. When the concentration of ethanol at the Montvale site was  ≥18 mg/L, the values of DG′ for degradation of toluene were positive (Figure 4c), and as expected, there is no evidence of the degradation of toluene (Figure 4a). At concentrations of ethanol  <18 mg/L, the values of DG′ for the degradation of toluene are negative in most, but not all, of the water samples, and toluene is degraded in most, but not all, of the samples.

As was the case at the Montvale site, when concentrations of ethanol in groundwater from the Berryville site were  ≥42 mg/L, there was no evidence of the degradation of toluene (Figure 4b). At the Montvale Site, all but one of the water samples with concentrations of ethanol <18 mg/l showed evidence of toluene biodegradation. In contrast, several of the samples at the Berryville site with ethanol concentrations <42 mg/L did not show evidence of toluene degradation. The threshold for the inhibition of toluene degradation by ethanol may have been lower in these samples, or due to the shorter exposure time, the microbial community may not have been acclimated to the anaerobic degradation of toluene. In either case, none of the samples from the Berryville site with ethanol concentrations ≥42 mg/L showed evidence of the degradation of toluene.

When concentrations of ethanol in water from the Berryville site were ≥42 mg/L, values for DG′ were either positive or were in the range between 0 and −20 kJ/mole where biodegradation might not be feasible (Figure 4d). There are three exceptions. The filled circle in Panel (d) is data from well W-6 taken in June 2012, the filled diamond is data from well W-8 taken in April 2013, and the filled square is data from well W-7 taken in December 2010. In the case of MW-7, the water was collected in December 2010 after the spill was reported in October 2010. There may have not been adequate time for bacterial acclimation to the anaerobic degradation of ethanol. Although the concentration of ethanol was high, the concentration of acetate was relatively low, and the concentration of hydrogen was very low (Table S2). With respect to the other two wells, there is no obvious explanation why values DG′ would be less than −20 kJ/mole in these wells at these sampling periods.

When conditions are thermodynamically unfavorable for the anaerobic biodegradation of toluene, they are generally unfavorable for the biodegradation of benzene (compare Panels [c] an [d] of Figure 4 to Panels [e] and [f]). When concentrations of ethanol in water were ≥42 mg/L, values of DG′ for the anaerobic degradation of benzene were either positive or were in the range between 0 and −20 kJ/mole. There was one exception. The filled square in Panel (f) is data from well W-7 taken in December 2010. As discussed above, there may have not been adequate time for acclimation to the anaerobic biodegradation of ethanol.

The data from the field sites indicate that 42 mg/L is a reasonable boundary for the concentration of ethanol that can lead to the inhibition of the anaerobic degradation of toluene and benzene. At concentrations of ethanol above 42 mg/L, there is no evidence of the biodegradation of toluene in the ratio of concentrations of toluene to benzene. With a few exceptions as discussed above, at concentrations of ethanol above 42 mg/L, the calculated value of DG′ indicates that the syntrophic biodegradation of toluene or benzene is not thermodynamically feasible.

Effect of the Age of the Spill on Anaerobic Degradation of Toluene

The release at the Montvale site is older than the release at the Berryville site. The field data used in this study from the Montvale site were collected 12.6 years to 14.1 years after the release was reported. Data were collected from the Berryville site 0.2 to 2.6 years after the report of the release. At the Montvale site, wells MW-3 and M-11 were directly impacted by the release, and wells MW-1, MW-4, MW-5, MW-7, MW-12, and RW-6 were downgradient of the release (Figure 2 and Table S1). At the Berryville site, wells W-1, W-2, W-5, W-6, W-7, and W-8 were directly impacted, and wells W-9, W-10, and W-11 were downgradient of the release (Figure 2 and Table S2).

Figure 3 compares the evidence of the anaerobic biodegradation of toluene at the two sites. At both sites, there was no evidence of anaerobic degradation of toluene in wells that were situated near the release. The concentrations of toluene and benzene and the ratio of toluene to benzene were probably controlled by dissolution from residual gasoline. At the Berryville site 3 years after the release, five of ten samples from downgradient wells showed evidence of the anaerobic degradation of toluene. At the Montvale site 13 years after the release, 15 of 17 samples showed evidence of degradation.

Degradation of Alcohols in the Microcosm Study

The time course of the degradation of ethanol is presented in Figure 5. The courses of degradation of n-butanol, n-propanol, and iso-butanol are presented in Figures S1, S2, and S3, respectively. Table 1 presents degradation rate constants that are fit to the concentrations of the alcohols. A zero-order and a first-order rate constant are provided for each alcohol. The coefficient of variation was used to evaluate the fit of the data to the two different rate laws. The results were mixed; a zero-order law was a better fit in eight cases, and a first-order law was a better fit in 3 cases. The rate of degradation of n-butanol, n-propanol, and iso-butanol were remarkably consistent, varying from a zero-order rate constant of 2.7 to 1.8 mg/(L day). The rate constants for the degradation of ethanol were larger, varying from 7.2 to 29 mg/(L day).

Figure 5.

Time period of the degradation of ethanol and production of hydrogen, acetate, n-butyrate, and methane in the microcosms.

Table 1.

Rate Constants for the Degradation of Fuel Alcohols in the Microcosm Experiment

Alcohol	Microcosm Number	Microcosm Condition	First-Order Rate Constant ±95% Confidence Interval (per year)	Coefficient of Variation	Zero-Order Rate Constant ±95% Confidence Interval (mg/(L day))	Coefficient of Variation
Ethanol	14	Live	4.2 ± 1.9	0.17	7.2 ± 3.3	0.18
Ethanol	15	Live	7.4 ± 4.5	0.14	29 ± 26	0.21
Ethanol	16	Live	5.1 ± 4.2	0.19	24 ± 4.3	0.04
Ethanol	35	Sterile	0.018 ± 0.24		0.12 ± 0.77
iso-Butanol	20	Live	2.7 ± 1.1	0.16	2.7 ± 0.5	0.08
iso-Butanol	21	Live	2.0 ± 0.27	0.06	2.0 ± 0.5	0.11
iso-Butanol	22	Live	1.9 ± 1.1	0.23	2.7 ± 0.8	0.13
iso-Butanol	35	Sterile	0.11 ± 0.33		0.36 ± 1.0
n-Butanol	17	Live	2.8 ± 1.3	0.19	2.4 ± 0.3	0.06
n-Butanol	18	Live	2.6 ± 1.0	0.16	2.3 ± 0.4	0.07
n-Butanol	19	Live	1.8 ± 0.5	0.12	2.7 ± 0.8	0.11
n-Butanol	35	Sterile	0.30 ± 0.72		0.80 ± 2.0
n-Propanol	23	Live	2.5 ± 0.9	0.15	2.6 ± 0.4	0.07
n-Propanol	25	Live	2.0 ± 0.6	0.14	1.8 ± 0.3	0.08
n-Propanol	35	Sterile	0.21 ± 0.51		0.64 ± 1.5

Figure 5 presents the time course of the biodegradation of ethanol in three replicate microcosms. The microcosms are identified for convenience by arbitrary numbers. Ethanol, at an initial concentration near 4000 mg/L, was degraded within 400 days of incubation. While ethanol was available and degrading, the concentration of hydrogen was often above 5000 ppm (v/v) in the headspace of the microcosms.

While ethanol was being degraded, the concentrations of acetate and butyrate were as high as 1000 to 2000 mg/L. Ma et al. (2011) reported the production of butyrate from ethanol. Presumably, the butyrate was produced from hydrogen and acetate following Equation 6.

$$2{\mathrm{CH}}_3{\mathrm{COO}}^{-}+{\mathrm{H}}^{+}+2{\mathrm{H}}_2\to {\mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{CH}}_2{\mathrm{COO}}^{-}+2{\mathrm{H}}_2\mathrm{O}$$ (6)

As the fatty acids were depleted, methane was produced. The data reported in Figure 5 are the cumulative production of methane, not the pool of methane measured at a particular sampling interval. To facilitate comparisons to the concentrations of alcohols and fatty acids, the data on methane production, as measured in the headspace of the microcosms, are expressed as if the methane remained in the pore water. After 776 days of incubation, microcosm 15 and 16 achieved 89 and 91%, respectively, of the concentration expected from the stoichiometric conversion of ethanol to methane. For some unknown reason, the conversion of acetate and butyrate to methane stalled in microcosm 14.

As a point of comparison, Corseuil et al. (2011) saw concentrations of ethanol as high as 2,000 mg/L and concentrations of acetate near 100 mg/L in their field study.

The degradation of n-butanol was slower; 1200 mg/L was degraded within 460 days of incubation (Figure S1). The n-butanol was first transformed to n-butyrate, which built up until the n-butanol was depleted. There were lower concentrations of acetate, never more than 75 mg/L. While n-butanol was available and was being degraded, the concentrations of hydrogen varied from 1820 ppm to 5750 ppm. Over the time period of the experiment, the yield of methane was 67, 65, and 62% of the expected stoichiometric yield.

Degradation of n-propanol resulted in a stoichiometric accumulation of n-propionate (Figure S2). Very little acetate was produced while n-propanol was available and was being degraded. The maximum concentration of hydrogen was 2530 ppm, and the maximum concentration of acetate was 1042 mg/L. The maximum concentration of acetate occurred after the n-propanol and propionate was degraded.

Degradation of iso-butanol also resulted in a stoichiometric accumulation of iso-butyrate (Figure S3). Once the iso-butanol was degraded, the iso-butyrate degraded. The maximum concentration of hydrogen was 1340 ppm, and the maximum concentration of acetate was 67 mg/L. The yield of methane was 42, 58 and 69% of the expected stoichiometric yield.

Effect of Alcohols on Thermodynamic Feasibility in Microcosms

In groundwater in contact with unweathered residual gasoline, the concentration of toluene should be near 46 mg/L (Falta 2004). This value will be considered a plausible upper boundary of the concentration of toluene in groundwater at a gasoline spill. The MCL for toluene is 1.0 mg/L. This will be considered a lower concentration of regulatory concern. Figure 6 compares the relationship between the concentration of the fuel alcohols in the microcosm study and the thermodynamic feasibility of anaerobic toluene degradation when the toluene is present at 46 and 1.0 mg/L. If the calculated value for DG′ falls above the grey bar in each panel of Figure 6, degradation is thermodynamically feasible. If the value falls below the grey bar, degradation is not feasible. If the value falls within the grey bar, degradation may or may not be feasible.

Figure 6.

Relationship between the concentration of ethanol (a), n-butanol (b), iso-butanol (c), or n-propanol (d) in water in a microcosm and the Gibbs free energy for the anaerobic biodegradation of toluene. (X) is the value of DG′ when the toluene concentration is 46 mg/L. (O) is the value of DG′ when the toluene concentration is 1.0 mg/L. (a) Ethanol biodegradation, (b) hydrogen production, (c) acetate production, (d) n-butyrate production, (e) methane production.

Concentrations of ethanol above 14 mg/L or butanol above 16 mg/L produced conditions that precluded the anaerobic biodegradation of toluene (panels [a] and [b] of Figure 6). The impact of iso-butanol and propanol was not as strong. These alcohols occasionally produced concentrations that precluded the anaerobic biodegradation of toluene (panels [c] and [d]).

The concentration of benzene in groundwater in contact with gasoline should be near 37 mg/L (Falta 2004), and the MCL for benzene is 0.005 mg/L. Figure 7 compares the relationship between the concentration of the fuel alcohols in the microcosm study and the thermodynamic feasibility of anaerobic benzene degradation when benzene is present at 37 and 0.005 mg/L.

Figure 7.

Relationship between the concentration of ethanol (a), n-butanol (b), iso-butanol (c), or n-propanol (d) in water in a microcosm and the Gibbs free energy for the anaerobic biodegradation of benzene. (X) is the value of DG′ when the benzene concentration is 37 mg\/L. (O) is the value of DG′ when the benzene concentration is 0.005 mg/L.

When the concentrations of ethanol were ≥14 mg/L or concentrations of n-butanol were ≥16 mg/L, biodegradation of the alcohols generally produced conditions that would preclude the natural anaerobic biodegradation of benzene by syntrophic organisms.

In general, concentrations of iso-butanol and n-propanol up to 1000 mg/L did not produce conditions that would preclude the biodegradation of benzene if the benzene were present at 37 mg/L. If the concentrations of benzene were near the MCL, iso-butanol and n-propanol may preclude the biodegradation of benzene.

In a microcosm study conducted with sediment from an old gasoline spill at Site 60 at Vandenberg Air Force Base, CA, Schaefer et al. (2010) compared the effects of ethanol and iso-butanol on the anaerobic degradation of benzene, toluene, ethylbenzene, and total xylenes. Unless nitrate or sulfate was added, benzene did not degrade in their sediment, even when the alcohols were not present. However, when no electron acceptor was added (methanogenic conditions), both ethanol and iso-butanol slowed the rate of degradation of toluene, ethylbenzene, and total xylenes.

Relevance to Gasoline Spill Sites

O’Reilly et al. (2016) extracted information on the distribution of ethanol at gasoline spill sites in California from the GeoTracker database (State Water Resources Control Board [SWRCB] 2015). They queried data from seven counties in northern and southern California. Data were available from 495,761 samples from 48,681 wells at 3493 sites. Ethanol was detected in 0.6% of the samples, 4.5% of the wells, and 22% of the sites. At sites where ethanol was detected at least once, the median concentration was 0.3 mg/L, and the maximum concentration was 13,000 mg/L. Of the 3493 sites, 6.7% reported an ethanol concentration >1 mg/L, 2.6% reported a concentration >10 mg/L, and 0.8% reported a concentration >100 mg/L.

Our microcosm results suggest that the inhibition of the degradation of benzene and toluene can be expected whenever the concentration of ethanol is above 14 mg/L. Data from the two field studies indicate that the inhibition of toluene degradation is seen at ethanol concentrations above 42 mg/L. Based on the distribution of ethanol concentrations at gasoline spill sites in California, ethanol will inhibit the natural anaerobic degradation of benzene and toluene at a small fraction of sites.