In Situ Bioremediation of Nitrate and Perchlorate in Vadose Zone Soil for Groundwater Protection Using Gaseous Electron Donor Injection Technology

By Evans, Patrick J; Trute, Mary M

ABSTRACT: When present in the vadose zone, potentially toxic nitrate and perchlorate anions can be persistent sources of groundwater contamination. Gaseous electron donor injection technology (GEDIT), an anaerobic variation of petroleum hydrocarbon bioventing, involves injecting electron donor gases, such as hydrogen or ethyl acetate, into the vadose zone, to stimulate biodegradation of nitrate and perchlorate. Laboratory microcosm studies demonstrated that hydrogen and ethanol promoted nitrate and perchlorate reduction in vadose zone soil and that moisture content was an important factor. Column studies demonstrated that transport of particular electron donors varied significantly; ethyl acetate and butyraldehyde were transported more rapidly than butyl acetate and ethanol. Nitrate removal in the column studies, up to 100%, was best promoted by ethyl acetate. Up to 39% perchlorate removal was achieved with ethanol and was limited by insufficient incubation time. The results demonstrate that GEDIT is a promising remediation technology warranting further validation. Water Environ. Res., 78, 2436 (2006).

KEYWORDS: nitrate, perchlorate, bioremediation, biodegradation, soil, vadose zone, gaseous electron donor injection technology, bioventing, anaerobic bioventing.

doi: 10.2175/106143006X123076

Introduction

Nitrate is a toxic anion known to cause methemoglobinemia, also known as blue baby syndrome, and the U.S. Environmental Protection Agency (U.S. EPA) (Washington, D.C.) has established a maximum contaminant level of 10 mg-N/L in groundwater (U.S. EPA, 1991). Perchlorate is another toxic anion that can inhibit iodide uptake by the thyroid (U.S. EPA, 2005). U.S. EPA has established a reference dose of 0.0007 mg/kg/d, which translates to a drinking water equivalent level of 24.5 g/L (U.S. EPA, 2006). Various technologies have been proven for treatment of nitrate or perchlorate in water (Canter, 1997; Hatzinger et al., 2002; Interstate Technical and Regulatory Council, 2005; Logan, 2001; Min et al., 2004). For example, groundwater can be extracted and treated using anaerobic bioreactors, ion exchange resins, and tailored activated carbon. Groundwater can also be treated in situ using anaerobic bioremediation.

Nitrate and perchlorate can be transported from the vadose zone to groundwater, thus the presence of these contaminants in the vadose zone represents an ongoing source of groundwater contamination (Addiscott, 1996; Burkart and Stoner, 2002; Canter, 1997; Nozawa-Inoue et al., 2005). Therefore, technologies are needed for treatment of nitrate and perchlorate in soil with the goal of groundwater protection. To date, soil treatment technologies have included both ex situ and in situ treatment, but have focused on shallow soil (Interstate Technical and Regulatory Council, 2005; Kastner et al., 2001). The ex situ treatment technologies used successfully have mostly included anaerobic bioremediation; however, thermal treatment and soil washing have also been implemented for perchlorate, to a limited extent (Interstate Technical and Regulatory Council, 2005; Weeks et al., 2003). Excavation for ex situ treatment is limited because it can be expensive or impractical for deep soil. In situ treatment of soil is an alternative to excavation and ex situ treatment, but has seen limited application to date. One example is flushing with aqueous electron donor solutions to stimulate in situ anaerobic bioremediation (Horst et al., 2005). However, in situ flushing with liquids can have limited effectiveness in unsaturated soil, especially when the soil is heterogeneous and the contamination is deep, because of the difficulty of evenly distributing liquids under these conditions.

Bioventing is a proven technology for the in situ aerobic bioremediation of petroleum hydrocarbons in soil (Parsons Engineering Science Inc., 1996). This technology involves addition of oxygen to vadose zone soil, either by injection of air or oxygen or extraction of air, as shown in Figure 1. Oxygen serves as a terminal electron acceptor for aerobic bacteria that can grow while using the petroleum hydrocarbons as an electron donor. Nitrate and perchlorate serve as terminal electron acceptors for certain anaerobic bacteria that are capable of growing on electron donors, including hydrogen, ethanol, and other organic compounds (Logan, 2001; McCarty et al., 1969). Conceptually, gaseous electron donors can be injected to vadose zone soil to stimulate in situ anaerobic bioremediation of nitrate or perchlorate, also shown in Figure 1. Hydrogen is one such gas that may be injected to vadose zone soil for gaseous electron donor injection technology (GEDIT). Hydrogen has been used successfully for in situ sparging of groundwater for the purpose of anaerobic reductive dechlorination of trichloroethene in groundwater (Newell et al., 1997, 2000). Hydrogen has also been injected to the vadose zone for partial reductive dechlorination of tetrachloroethene, 1,1,1 -trichloro-2,2-bis(4-chlorophenyl)ethane (also known as DDT), and dinitrotoluene (Mihopoulos et al, 2001; Shah et al., 2001). The GEDIT can also use vaporized organic liquids, such as ethanol or ethyl acetate, provided that the gaseous concentration is less than the saturation vapor pressure. Such an approach has been evaluated for treatment of high explosives, such as 1,3,5-trirdtrorjerhydro-1,3,5-triazine (also known as ROX) and trinitrobenzene (Rainwateret al., 2001). The purpose of this study was to validate the GEDIT concept in the laboratory and identify engineering parameters important for transitioning to the field.

Methodology

Chemicals. All chemicals were purchased from Sigma Aldrich (St. Louis, Missouri).

Soil. The soil used for this study was a silty sand from the greater Los Angeles, California, vicinity, which was contaminated with perchlorate (23 mg/kg) and nitrate (14 mg-N/kg). The initial moisture content of the soil ranged from 7.2 to 7.5%, and the initial organic carbon content was 0.34%.

Microcosm Studies. An initial microcosm study was conducted to evaluate two different electron donors (i.e., hydrogen and ethanol) under ambient soil conditions and under conditions where moisture, nutrients, or both moisture and nutrients were amended. Microcosms were set up in 26-mL anaerobic pressure tubes sealed with thick butyl rubber stoppers and aluminum crimp seals (Table 1 ). Approximately 20 g of soil were added to each tube; tubes were then sealed and flushed with nitrogen to remove oxygen. Amendments and electron donors were added using nitrogenflushed syringes through the butyl rubber stopper, and the microcosms were shaken to distribute amendments evenly. Nitrogen and phosphate were added to the soil as aqueous solutions of ammonium phosphate [(NH^sub 4^)^sub 2^HPO^sub 4^]. Hydrogen equal to 120 mol electron equivalents and ethanol equal to 210 mol electron equivalents were added initially. Carbon dioxide was added with hydrogen at a final vapor concentration of 4.5% by volume to provide a carbon source for the hydrogen microcosms. Electron donor gas-phase concentrations were monitored twice per week for the duration of the experiment. Additional hydrogen or ethanol was added when gas concentrations approached zero. The times when these additions were made to specific microcosm bottles are shown in Table 1. Microcosms were incubated in the dark for 34 days. A second microcosm study evaluated various electron donor concentrations and was conducted similarly, with the following exceptions. The moisture content was maintained at 9.5%, and nutrients were not added. These microcosms were incubated in tightly sealed canning jars with septum fittings to allow initial flushing with nitrogen and weekly headspace sampling. These microcosms were incubated for 105 days. All microcosms were conducted in duplicate.

Column Studies. Column studies were conducted to evaluate electron donor transport and subsequent biodegradation of nitrate and perchlorate. Electron donor vapor transport studies were conducted using columns constructed from transparent Schedule 40 PVC pipe. Although soil for the column studies came from same area as microcosm soil, nitrate and perchlorate concentrations were lower than expected, so aqueous solutions of sodium nitrate and sodium perchlorate were added to soil to obtain final concentrations of 72 mg-N/kg and 41 mg ClO^sub 4^/Rg, respectively. The amended soil was homogenized and loaded into the columns to a density of 1.6 g/mL. Soil was added in 485-g batches per 15-cm column length, for a total soil length of 90 to 110 cm. After each batch addition, the columns were tapped along the side to settle the soil, while minimizing further compaction of the underlying soil. Soil was supported on plastic screens 5 cm above the gas inlet to evenly distribute inlet gas. Argon was used as the carrier gas, and inlet gas flow was monitored using a gas flow meter (Aalborg, Orangeburg, New York). Outlet gas flowrates were measured using an ADM1000 mass flow meter (J&\W Scientific, Folsom, California). Electron donor was delivered using a multichannel Syringe Pump (Cole-Palmer, Niles, Illinois) and 200-mL gas sampling bulb. The liquid electron donor was injected by the syringe pump to the sample port on the gas sampling bulb, as argon carrier gas flowed through the bulb. The electron donor vaporized into the gas stream and was routed using tubing to the column inlet where the electron donor gas concentration was monitored. Septa were located every 15 cm along the column for use as gas sampling ports. Electron donor concentrations along the column were monitored daily or every other day, while gas was flowing through the columns. Gas samples were collected in duplicate using a gas-tight sample lock syringe.

An initial transport study was conducted using ethanol. Ethanol flow to the gas sampling bulb was kept constant at 0.2 L/min, and the initial argon flowrate ranged from 3.5 to 4.0 mL/min. The measured influent ethanol concentration averaged 25 000 ppmv. On day 37, the argon flowrate was increased to 50 mL/min, while the ethanol flowrate was maintained at 0.2 L/min, and the average measured influent ethanol concentration after day 38 was 1300 ppmv. Gas samples collected periodically from the column sample ports were analyzed for ethanol to determine the distance the ethanol vapor front had traveled through the soil column over time.

Individual column transport studies were then conducted in duplicate with ethanol, ethyl acetate, butyl acetate, and butyraldehyde. Liquid electron donor flowrates were maintained at 0.5 L/min, and argon carrier gas flowrates were maintained at 42 mL/ min. One column from each set of duplicates was sealed following electron donor breakthrough and stored at room temperature for the purpose of measuring nitrate and perchlorate biodegradation. For this set of duplicate columns, the ethyl acetate and butyraldehyde electron donors broke through the soil columns after 3 days and were immediately sealed. Ethanol and butyl acetate traveled 46 and 61 cm, respectively, after 10 days, and the columns and were sealed at this time. All four columns were cut transversely into 30- to 37-cm sections 34 days following initiation of electron donor flow. The soil from each column section was homogenized, sampled, and analyzed for nitrate, perchlorate, pH, and percent moisture. Thus, the ethyl acetate and butyraldehyde columns were sealed for a total of 31 days, and the ethanol and butyl acetate columns were sealed for a total of 24 days.

Analytical Chemistry. Microcosm study soil samples were submitted to Columbia Analytical Services (Kelso, Washington) for analysis of perchlorate by U.S. EPA method 314.0 (U.S. EPA, 2000), nitrogen as nitrate plus nitrite by U.S. EPA method 353.2 (U.S. EPA, 1983), and percent solids by U.S. EPA method 160.3 (U.S. EPA, 1983). Column study soil samples were submitted to Severn Trent Laboratories (Los Angeles, California) for analysis of nitrate by U.S. EPA method 300.0 (U.S. EPA, 1983), perchlorate by U.S. EPA method 314.0, and percent solids by U.S. EPA method 160.3. Soil pH for microcosms was determined by measuring the pH of a slurry of soil and distilled water (1:1 by weight) using a standard pH probe. Soil pH for the column studies was measured by Severn Trent Laboratories using U.S. EPA method 9045C (U.S. EPA, 1995b). Hydrogen and carbon dioxide were measured using a Hewlett Packard 5890 gas Chromatograph (Hewlett Packard, Palo Alto, California) equipped with a 4.7-m, 60/80 mesh, Carboxen 1000 column (Supelco, Bellefonte, Pennsylvania) and a thermal conductivity detector. Ethanol, butyraldehyde, ethyl acetate, and butanol were analyzed on a Hewlett Packard 5890 gas Chromatograph equipped with a 75-m DB-624 column (Supelco) and flame ionization detector. Nitrogen was used as the carrier gas in both cases, and all gas samples were analyzed in duplicate.

Results

Soil Microcosms. Soil microcosms were prepared to characterize the effects of electron donor, moisture, and nutrients on nitrate and perchlorate reduction. Final nitrate and perchlorate concentrations after 34 days of incubation in the first microcosm study are shown in Figures 2 and 3, respectively. A slight reduction in nitrate was observed in the absence of an electron donor on supplementation with moisture, nutrients, or both moisture and nutrients (Figure 2). Similar results for perchlorate were observed, but only in the case of supplementation with nutrients or moisture and nutrients (Figure 3). In no case was nitrate or perchlorate completely reduced in the absence of an electron donor. The electron donors hydrogen and ethanol promoted complete nitrate reduction on supplementation with moisture or moisture and nutrients (Figure 2). Ethanol, but not hydrogen, promoted complete nitrate reduction in the unsupplemented microcosms (i.e., no moisture or nutrients) and in the microcosms supplemented only with nutrients. Hydrogen promoted partial nitrate removal in the microcosms supplemented only with nutrients. Ethanol promoted complete perchlorate reduction, and hydrogen promoted partial perchlorate reduction, in the microcosms supplemented only with moisture (Figure 3). Perchlorate reductions in the microcosms supplemented with moisture and nutrients were less than in the microcosms supplemented only with moisture. A possible reason is that nutrients may have promoted greater carbon assimilation into cellular material relative to oxidation to carbon dioxide that, in turn, would yield fewer reducing electron equivalents per mole of electron donor consumed. Supplementation with nutrients only did not promote perchlorate reduction. These data demonstrate that, in this particular soil, nutrient supplementation was not necessary. This result is consistent with studies on hydrocarbon venting, demonstrating that nutrient supplementation is typically unnecessary (Parsons Engineering Science Inc., 1996).

The final pH values in the microcosms are shown in Figure 4. Elevated pH values were observed in the microcosms supplemented with moisture and nutrients relative to the other microcosms. A pH value between 9.3 and 9.8 did not prevent complete reduction of nitrate. On the other hand, the elevated pH in the microcosms supplemented with moisture and nutrients may have led to partial inhibition of perchlorate reduction.

Final average moisture concentrations in the microcosms supplemented with moisture or moisture and nutrients ranged from 10.0 to 12.6%. Final average moisture in the remaining microcosms ranged from 7.65 to 8.35%. Moisture clearly had a positive effect on nitrate and perchlorate reduction, as shown in Figures 2 and 3. Significant nitrate reduction was observed with hydrogen only when moisture was supplemented. Complete nitrate reduction was observed for ethanol without moisture supplementation, proving that nitrate reduction was possible at the lower moisture content. Therefore, the 34-day incubation period may have been insufficient for complete nitrate reduction in the presence of hydrogen at the low moisture content. Significant, albeit incomplete, perchlorate reduction was only observed in the microcosms supplemented with moisture. Perchlorate reduction in the unsupplemented microcosms (i.e., containing hydrogen or ethanol, but no moisture or nutrients) was not significant, but one of the duplicate bottles for each electron donor contained less perchlorate than the corresponding control microcosms without an electron donor. Therefore, similar to the observation for nitrate, insufficient time (i.e., 34 days) may have been allowed for perchlorate reduction. The hypothesis that insufficient time was allowed for complete perchlorate reduction in this first microcosm study was validated by the second microcosm study, as discussed later in this section.

Sufficient electron donor must also be provided to promote complete nitrate and perchlorate reduction. Nitrate present in the 20-g soil aliquots used in the microcosms required 93 moles of electron equivalents (eq) for reduction of nitrate to dinitrogen gas and an additional 46 eq for reduction of perchlorate to chloride ion (Table 2). Hydrogen added to the microcosms provided 120 to 200 eq, which was sufficient for nitrate reduction, but ranged from being insufficient to being slightly in excess for perchlorate reduction, because a total of 138 eq was required for reduction of both nitrate and perchlorate. Nitrate is reduced first, before perchlorate (Coates and Achenbach, 2004). Ethanol added to the microcosms was in excess of that required.

The second microcosm study was conducted with varying amounts of hydrogen or ethanol, a constant moisture content of 9.5%, and no nutrients. The control microcosms received no electron donor, and the final nitrate and perchlorate concentrations after 105 days of incubation were 10.3 0.13 mg-N/kg and 24 3.6 mg/kg, respectively. Microcosms supplemented with 100, 200, 300, or 400% of stoichiometric ethanol contained <0.04 0 mg/kg perchlorate after 105 days. Microcosms supplemented with 100% stoichiometric hydrogen contained 19 0.07 mg/kg perchlorate, and microcosms supplemented with 200, 300, or 400% stoichiometric hydrogen contained <0.04 0 mg/ kg perchlorate. All supplemented microcosms contained <0.5 O mg-N/ kg nitrate. These data support the hypothesis that insufficient time and hydrogen were present in the first microcosm study for complete perchlorate reduction.

Electron Donor Transport. In addition to promoting nitrate and perchlorate biodegradation, electron donors must be transported through vadose zone soil to make GEDIT effective. Column studies were conducted to characterize transport of various electron donors through moist soil. The transport of 25 000-ppmv ethanol in an argon carrier gas through a column containing soil with 10% moisture is shown in Figure 5. The argon carrier gas flow was initially 3.5 to 4.\0 mL/min and then increased to 50 mL/min on day 37. These flowrates translate to an increase in the bulk gas velocity, from approximately 0.01 to 0.15 cm/s, based on an assumed void fraction of 0.35. Recommended minimum design bulk gas velocities for soil vapor extraction and bioventing range from 0.001 to 0.01 cm/s (U.S. Army Corps of Engineers, 2002). After 37 days of column operation, the influent ethanol concentration was reduced, from 25 000 to 1300 ppmv, when the bulk gas velocity was increased to maintain a constant ethanol mass loading rate to the column. The ethanol transport data in Figure 5 show that ethanol was transported slowly at the initial bulk gas velocity and progressed approximately 30 cm in 1 month. Increasing the bulk gas velocity to 0.15 cm/s increased the ethanol transport rate significantly.

The transport of additional electron donors was investigated, and results are presented in Figure 6. These data demonstrate that ethyl acetate and butyraldehyde were transported quickly, and ethanol and butyl acetate were transported more slowly. These data and physical properties for these electron donors are presented in Table 3. The Henry’s constant describes equilibrium partitioning between gas and water (i.e., soil moisture) phases and appears partially to describe the observed difference in gaseous electron donor transport. However, butyl acetate was transported slower than expected, based on comparison of its Henry’s constant value with that of ethyl acetate. Another factor is the organic carbon-water partitioning coefficient, K^sub OC^. The value of K^sub OC^ was greatest for butyl acetate and might have contributed to retardation of butyl acetate transport. However, the total organic carbon concentration of the soil was only 0.34%, and significant partitioning was not expected. Esters are subject to hydrolysis under alkaline conditions (Mabey and Mill, 1978), and this transformation may have played a role in limiting transport of butyl acetate. However, ethyl acetate was transported relatively quickly and has a slightly greater hydrolysis rate constant. The relatively low vapor pressure of butyl acetate may have played a role in its retardation, with respect to an increased potential for condensation, because its influent concentration was 21% of the saturation vapor pressure. Finally, biodegradation of electron donors may have occurred during the column study, which would also limit observed transport. Therefore, no single parameter has been identified that controlled electron transport through moist soil. Nevertheless, hydrogen, ethyl acetate, and butyraldehyde have good transport characteristics and merit further investigation. Ethanol and butyl acetate are not transported effectively through this soil and are unlikely to be practical for GEDIT. Henry’s constant, saturation vapor pressure, potential for hydrolysis, and K^sub OC^ should be considered when evaluating suitability of other candidate electron donors.

Perchlorate and Nitrate Biodegradation in Columns. Column studies were also used to evaluate whether electron donors transported as gases through moist soil could promote biodegradation of nitrate or perchlorate. Nitrate removals in columns with butyl acetate, ethanol, ethyl acetate, or butyraldehyde electron donors are shown in Figure 7. Only ethyl acetate promoted consistent and high percentages of nitrate reduction up to 100% throughout the soil column. Butyl acetate promoted nitrate reduction in the middle of the soil column and near the outlet, but not near the inlet. Because butyl acetate was not transported well (Figure 6), it may have accumulated near the inlet at high and potentially inhibitory concentrations. Ethanol promoted nitrate reduction near the inlet and in the middle, but not near the outlet. Lack of ethanol transport through the soil column (Figure 6) likely explained the lack of nitrate reduction near the outlet. Butyraldehyde was not effective in promoting nitrate reduction, possibly because aldehydes are especially toxic.

Figure 8 presents the perchlorate removal data for these same columns. Complete perchlorate removal was not observed in any of the columns, and the ethanol column promoted the greatest perchlorate removal, up to 39%. No perchlorate reduction was observed in the final third of the ethanol column, as expected, because of the poor transport characteristics of this electron donor. The pattern of perchlorate reduction in the ethanol column was consistent with the observed pattern of nitrate reduction. Thus nitrate removal was a necessary predecessor to perchlorate removal, as has been observed previously in aqueous systems (Coates and Achenbach, 2004; Hatzinger et al., 2002).

Electron equivalents required for complete reduction of nitrate and perchlorate in the soil in each column are shown in Table 4. Also shown are the electron equivalents for each electron donor injected to each column. The electron donor equivalents based on injected donors were greater than the demand of both nitrate and perchlorate, but the relative amounts of ethanol and butyl acetate delivered to the columns were greater than those of ethyl acetate and butyraldehyde. Greater electron equivalents added to the soil did not translate into greater nitrate biodegradation, but may have facilitated greater perchlorate biodegradation. Additionally, the final moisture in the soil ranged from 7.9 to 8.1%. The microcosm results indicated that perchlorate reduction did not occur within 34 days at this moisture content, but did occur in 105 days at a moisture content of 9.5%. Thus, greater time was likely required for perchlorate reduction in the soil columns. An alternative explanation is that perchlorate-reducing bacteria are completely inhibited at this moisture content. However, nitrate-reducing bacteria were not inhibited at this moisture content. Therefore, the lack of incubation time is a more reasonable explanation.

These data demonstrate that an electron donor, in addition to being able to support anaerobic biodegradation of nitrate or perchlorate, must be capable of being transported through soil. Sufficient electron donor must be delivered to these bacteria, and this delivery will be affected, not only by the mass injected, but also by the partitioning from the gas to soil and soil moisture containing the bacteria. This partitioning will be controlled, not only by soil moisture and the Henry’s constant, but also by soil organic carbon and K^sub OC^. The soil moisture is important, because it will affect electron donor partitioning and, as shown in the microcosm studies, the biological reduction of nitrate and perchlorate. Finally, sufficient time must be allowed for the bacteria to consume the electron donor and reduce the nitrate or perchlorate. The data in Figure 7 indicate that sufficient ethyl acetate was distributed through the soil, and sufficient time was allowed for nitrate reduction. The data in Figure 8 and Table 4 indicate that, while sufficient ethyl acetate was distributed to the soil, perchlorate was not reduced. Sufficient time apparently was not allowed for complete perchlorate reduction. The relatively lesser amount of ethyl acetate delivered to the soil column may also have contributed to the limited extent of perchlorate biodegradation.

Engineering Considerations. The GEDIT can be implemented by direct injection of a gaseous electron donor that is a gas under normal soil temperatures (i.e., hydrogen) or of a mixture of a vaporized liquid (i.e., ethyl acetate) and a carrier gas, such as nitrogen (Figure 9). Alternatively, soil vapor can be extracted, mixed with a gaseous electron donor, and reinjected to the vadose zone. Engineering design of GEDIT must consider gaseous electron donor demand and transport. In addition, soil lithology and permeability must be considered, as for a standard bioventing system (U.S. EPA, 1995a). Electron donor transport will be a function of achievable bulk gas velocity in the specific formation, soil lithology, soil moisture, and organic carbon, and electron donor physical properties, including Henry’s constant, organic-carbon partitioning coefficient, and saturation vapor pressure. Treatability studies similar to the column studies conducted herein will be necessary until accurate mathematical models describing gaseous electron donor transport can be developed for GEDIT. In the direct gaseous electron donor injection configuration, the primary electron acceptors will be nitrate and perchlorate, but may include oxygencontaining soil gas that is not displaced by the injected gaseous electron donor. In the soil vapor extraction and reinjection configuration, oxygen will likely be present and must be included in stoichiometric demand calculations.

Conceptual gaseous electron donor injection requirements are shown in Table 5, for the purpose of comparing potential gaseous electron donors. The calculations show that the number of pore volumes of injected gaseous electron donor varies greatly among the different compounds, and this variation is a function of the saturation vapor pressure, reducing power, and molecular weight. The saturation vapor pressure determines how much electron donor can be practically vaporized without a significant risk of condensation upon injection. Hydrogen and propane are gases at ambient soil temperatures; thus, they can be injected as pure gases. The other compounds are organic liquids at ambient temperature and pressure and must be vaporized before injection; a practical maximum concentration in the carrier gas of 20% by volume was selected for this exercise. However, greater concentrations up to the saturation vapor pressure are possible. Greater reducing power (i.e., number of reducing equivalents per mole) of an electron donor is preferable, because it is one factor determining the mass of electron donor that must be injected. The other factor is the mol\ecular weight. These two factors determine the minimum mass of electron donor that must be injected to promote reduction of nitrate and perchlorate. The mass of each electron donor per reducing equivalent varies as a function of these two factors (Table 5). The minimum mass of electron donor required per unit soil volume varies similarly and was used in combination with unit chemical prices to determine the minimum chemical cost per unit soil volume. The costs listed in Table 5 are similar and are anticipated to be small compared with other major technology implementation costs, such as injection well installation. Additionally, unit chemical costs (i.e., hydrogen) may be reduced through use of greater quantities or technical grades of chemicals. Typical costs for aerobic bioventing range from $10 to $70 per cubic meter (Federal Remediation Technologies Roundtable, 2006), and subsurface infrastructure requirements for GEDIT are anticipated to be similar to aerobic bioventing.

The quantity of electron donor that must be injected, in combination with the maximum practical concentration of electron donor that is achievable in the injected gas, determines the minimum volume of gaseous electron donor and any carrier gas that must be injected to promote reduction of nitrate and perchlorate. This volume is presented in Table 5 in terms of the unit volume occupied per reducing equivalent and in terms of equivalent soil pore volumes. These values span three orders of magnitude and demonstrate that hydrogen, propane, ethyl acetate, and hexene are preferable gaseous electron donors with respect to minimization of the number of pore volumes that must be injected. Hydrogen requires approximately 10 times more volume per equivalent and thus more soil pore volumes than propane, because propane liberates approximately 10 times more electron equivalents per mole, and 1 mole of an ideal gas occupies 22.4 L at standard temperature and pressure. Ethyl acetate requires 40 times more volume than propane, because its saturation vapor pressure limits the vapor concentration that can be injected.

Discussion

Suitable electron donors for GEDIT must (1) be capable of supporting biological reduction of nitrate or perchlorate, (2) be capable of being transported through moist soil, (3) be added in a sufficient quantity to promote anaerobic reduction of nitrate and/ or perchlorate, and (4) have properties that minimize gas volume that must be injected to meet the stoichiometric demand. This study demonstrated that various gaseous electron donors, including hydrogen, ethanol, ethyl acetate, butyl acetate, and butyraldehyde, are capable of partially, and, in some cases, completely promoting biological reduction of nitrate or perchlorate. Other electron donors, such as propane and 1-hexene, also have promise based on their physical properties (i.e., high Henry’s constant and vapor pressure) and potential for supporting biological perchlorate and nitrate reduction. Propane supported perchlorate reduction in one of three sites tested, indicating that it has the potential to promote perchlorate biodegradation at some, but not all sites (Envirogen, 2002). Hexene has not been evaluated to our knowledge, but hexane supports the biological reduction of nitrate (Ehrenreich et al., 2000). Therefore, it is possible that hexene also can support biological nitrate reduction because of a double bond. Alkanes are considered to be one of the least chemically reactive compound classes, and the addition of a double bond increases its reactivity (Ehrenreich et al., 2000). The moisture content of the soil is an important parameter and affected biological reduction of nitrate and perchlorate. Nutrient addition was not beneficial, as has been observed with the aerobic bioventing of petroleum hydrocarbons (Parsons Engineering Science Inc., 1996). Transport rates of electron donors through moist soil varied greatly, and this variation was partially attributed to differences in Henry’s constants. Other factors, including organic carbon partitioning, saturation vapor pressure, and ester hydrolysis, also may have contributed to the observed differences in transport. No single parameter completely explained the variations, thus emphasizing the need for site-specific treatability or pilot studies. Substantial nitrate reduction was observed in column studies, and ethyl acetate promoted consistent nitrate reduction throughout the soil, demonstrating that it has properties conducive to both nitrate reduction and electron donor transport. Other electron donors promoted nitrate reduction only in portions of the soil column because of poor transport or toxicity. Minor perchlorate reduction was observed in the columns, even though sufficient electron donor for its reduction was injected. Based on results of the second microcosm study, it is likely that insufficient time was allowed for perchlorate reduction in the column study. Engineering calculations indicated that the volumes of different gaseous electron donors needed to satisfy the stoichiometric demand for nitrate and perchlorate reduction can vary significantly. This variation is a function of the saturation vapor pressure and reducing power. The injection volume is an important parameter because it will directly affect GEDIT remediation system operating time and operating cost.

Conclusion

The results of this laboratory study demonstrate that GEDIT has the potential to be an effective remediation technology for nitrate and perchlorate in vadose zone soil. Hydrogen and ethyl acetate are two electron donors that show promise based on the results of this study. Propane, in the form of liquefied petroleum gas, and 1- hexene may have promise, but further data supporting their ability to support biological reduction of nitrate and perchlorate are needed. Additional laboratory research focusing on complete perchlorate reduction and the effects of different site conditions is needed. Ultimately, a field pilot test of GEDIT is a necessary next step for its validation as a practical soil remediation and groundwater protection technology.

References

Addiscott, T. M. (1996) Fertilizers and Nitrate Leaching. Issues Environ. Sci. Technol., 5(1), 1-26.

Burkart, M. R.; Stoner, J. D. (2002) Nitrate in Aquifers Beneath Agricultural Systems. Water Sd. Technol., 45 (9), 19.

Canter, L. W. (1997) Nitrates in Groundwater. CRC Press: Boca Raton, Florida.

Coates, J. D.; Achenbach, L. A. (2004) Microbial Perchlorate Reduction: Rocket-Fuelled Metabolism. Nature Rev. Micmbiol., 1 (7), 569-580.

Ehrenreich, P.; Behrends, A.; Harder, J.; Widdel, F. (2000) Anaerobic Oxidation of Alkanes by Newly Isolated Denitrifying Bacteria. Arch. Micmbiol., 173 (1), 58-64.

Envirogen Inc. (2002) In Situ Bioremediation of Perchlorate, Strategic Environmental Research and Development Program Project CU- 1163 Final Report; Envirogen Inc.: Lawrenceville, New Jersey.

Federal Remediation Technologies Roundtable (2006) Remedial Technologies Screening Matrix and Reference Guide, Version 4.0. section 4.1 -Bioventing, http://www.frtr.gov/matrix2/section4/ 4_1.html.

Hatzinger, P. B.; Whittier, M. C.; Arkins, M. D.; Bryan, C. W.; Guarini, W. J. (2002) In-Situ and Ex-SiW Bioremediation Options for Treating Perchlorate in Groundwater. Remediation, 12 (2), 69.

Horst, J. F.; Hempleman, D.; Wuerl, B. ).; Lenzo, F. (2005) Successful In Situ Bioremediation of Perchlorate-Impacted Vadose Zone Soil. Proceedings of the Eighth International In Situ and On- Site Bioremediation Symposium, Baltimore, Maryland, June 6-9; Battelle Press: Columbus, Ohio.

Interstate Technical and Regulatory Council (2005) Technology Overview. Perchlorate: Overview of Issues, Status, and Remedial Options; Interstate Technical and Regulatory Council: Washington, D.C., September.

Kastner, J. R.; Das, K. C.; Nzengung, V. A.; Dowd, J.; Fields, J. (2001) In-Situ Bioremediation of Perchlorate-Contaminated Soils. Bioremediation of Inorganic Compounds: The Sixth International In Situ and On-Site Bioremediation Symposium, San Diego, California, June 4-7; Battelle Press: Columbus, Ohio, 289.

Logan, B. E. (2001) Assessing the Outlook for Perchlorate Remediation. Environ. Sd. Technol., 35 (23), 483A.

Mabey, W.; Mill, T. (1978) Critical Review of Hydrolysis of Organic Compounds in Water Under Environmental Conditions. J. Chem. Phys. Ref. Data., 7 (2), 383.

McCarty, P. L.; Beck, L.; St. Amant, P. (1969) Biological Denitrification of Wastewaters by Addition of Organic Materials. Proceedings of the 24th Industrial Waste Conference, May 6-8; Purdue University: Lafayette, Indiana.

Mihopoulos, P. G.; Suidan, M. T.; Sayles, G. D. (2001) Complete Remediation of PCE Contaminated Unsaturated Soils by Sequential Anaerobic-Aerobic Bioventing. Water Sd. Technol., 43 (5), 365.

Min, B.; Evans, P. J.; Chu, A. K.; Logan, B. E. (2004) Perchlorate Removal in Sand and Plastic Media Bioreactors. Water Res., 38 (1), 47-60.

Newell, C. J.; Fisher, R. T.; Hughes, J. (1997) Direct Hydrogen Addition for the In-Situ Biodegradation of Chlorinated Solvents. Proceedings of the Petroleum Hydrocarbons and Organic Chemicals in Ground Water: Prevention, Detection, and Remediation Conference, Houston, Texas, Nov. 12- 14;NationalGroundwaterAssociation:Westerville,Ohio, 791.

Newell, C. J.; Haas, P. E.; Hughes, J. B.; Khan, T. (2000) Results from Two Direct Hydrogen Delivery Field Tests for Enhanced Dechlorination. In Bioremediation and Phytoremediation of Chlorinated and Recalcitrant Compounds, Wickramanayake, G. B., Gavaskar, A. R., Alleman, B. C., Magar, V. S. (Eds.); Battelle Press: Columbus, Ohio, 21.

Nozawa-Inoue, M.; Scow, K. M.; Rolston, D. E. (2005) Reduction of Perchlorate and Nitrate by Microbial Communities in Vadose Soil. Appl. Environ. Micmbiol., 71 (7), 3928.

Parsons Engineering Science Inc. (1996) Bioventing Performance and Cost Results from Multiple Air Force Test Sites, USAF-F33615-90- D-4014, Technical Demonstra\tion and Final Technical Memorandum Prepared for the Air Force Center for Environmental Excellence; Parsons Engineering Science Inc.: Denver, Colorado.

Perry, R. H.; Chilton, C. H. (1973) Chemical Engineers’ Handbook, 5th ed.; McGraw-Hill: New York, 3-97.

Rainwater, K.; Brown, J.; Heintz, C.; Mollhagen, T.; Hansen, L. D. (2001) Design, Construction, and Operation of a Field Demonstration for In Situ Biodegradation of Vadose Zone Soils Contaminated with High Explosives. U.S. Army Corps of Engineers: Washington, D.C.

Shah, J. K.; Sayles, G. D.; Suidan, M. T.; Mihopoulos, P.; Kaskassian, S. (2001) Anaerobic Bioventing of Unsaturated Zone Contaminated with DDT and DNT. Water Sd. Technol., 43 (2), 35.

Syracuse Research Corporation (2006). Physical Properties Database, PHYSPROP; Syracuse Research Corporation: Syracuse, New York, http://www.syrres.com/esc/physdemo.htm.

U.S. Army Corps of Engineers (2002) Engineering and Design. Soil Vapor Extraction and Bioventing, Engineer Manual No. 1110-1-4001; U.S. Army Corps of Engineers: Washington, D.C., June 3.

U.S. Environmental Protection Agency (2006) Assessment Guidance for Perchlorate, memorandum from Susan Parker Bodine to Regional Adminstrators; U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response: Washington, D.C.

U.S. Environmental Protection Agency (1991) Nitrate, CASRN 14797- 55-8. Integrated Risk Information System (IRIS); U.S. Environmental Protection Agency: Washington, D.C., http://www.epa.gov/iris/subst/ 0076.htm, revised Oct. 10.

U.S. Environmental Protection Agency (1995a) Manual: Bioventing Principles and Practice, EPA-540/R-95-534a; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (1983) Methods for Chemical Analysis of Water and Wastes. EPA-600/4-79-020; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (2000) Methods for the Determination of Organic and Inorganic Compounds in Drinking Water, Vol. 1, EPA-8 15/R-OO-014; U.S. Environmental Protection Agency: Washington, D.C.

U.S. Environmental Protection Agency (2005) Perchlorate (ClO4″) and Perchlorate Salts. Integrated Risk Information System (IRIS); U.S. Environmental Protection Agency: Washington, D.C., http://www. epa.gov/iris/subst/1007.htm, revised Feb. 18.

U.S. Environmental Protection Agency ( 1995b) Test Methods for Evaluating Solid Waste, Physical/Chemical Methods (SW846), Update HB; U.S. Environmental Protection Agency: Washington, D.C.

Weeks, K. R.; Veenstra, S. C.; Hill, D. L.; Gregson, B. P. (2003) A Study of Treatment Options to Remediate Explosives and Perchlorate in Soils and Groundwaterat Camp Edwards, Massachusetts. Remediation, 13 (2), 131.

Acknowledgments

Authors. Patrick J. Evans is currently an associate at CDM, Bellevue, Washington. Mary M. Trute is currently a graduate student at the University of Washington. At the time of this work, she was a research assistant at CDM. Correspondence should be addressed to Patrick J. Evans, CDM, 11811 N.B. First Street, Suite 201, Bellevue, Washington 98005; e-mail: evanspj@cdm.com.

Submitted for publication March 21, 2006; revised manuscript submitted May 18, 2006; accepted for publication May 23, 2006.

The deadline to submit Discussions of this paper is March 15, 2007.

Copyright Water Environment Federation Dec 2006

(c) 2006 Water Environment Research. Provided by ProQuest Information and Learning. All rights Reserved.