Environ. Sci. Technol. 2002, 36, 190-199
In Situ MTBE Biodegradation Supported by Diffusive Oxygen Release
limited or absent, as is often the case in groundwater impacted by petroleum hydrocarbon spills, natural attenuation of MTBE plumes may be insufficient for risk management. Thus, there is considerable interest in developing approaches for in situ treatment of MTBE, with biotreatment approaches receiving the most attention.
R Y A N D . W I L S O N , * ,† D O U G L A S M . M A C K A Y , †,‡ A N D KATE M. SCOW‡ Department of Earth Sciences, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada, and Department of Land, Air and Water Resources, University of California, Davis, California 95616
MTBE has been shown in microcosm studies to be biodegraded by a number of pure (7-10) and mixed (6, 1113) microbial cultures under aerobic conditions and, in some cases, demonstrated to serve as the sole carbon and energy source for these organisms. Tert-butyl alcohol (TBA) is noted as an intermediate of aerobic degradation of MTBE, although it can be aerobically degraded by some microorganisms (6, 9, 13, 14). Because degradation rates under aerobic conditions are believed to be more rapid than those under anaerobic conditions (9) and given the variety of ways to increase the oxygen concentration of contaminated groundwater, there have been many attempts to increase the rate of in situ aerobic degradation of MTBE. Most attention has been directed to the permeable reactive barrier (PRB) approach, i.e., the creation of an in situ aerobic biotreatment zone through which the plume migrates under the natural gradient and within which MTBE is degraded. A PRB effective at enhancing in situ aerobic microbial treatment of MTBE must (i) create steady aerobic conditions, (ii) generate and/or sustain enough microbial biomass to accomplish the treatment at a practically useful rate, and (iii) ensure that the contaminated groundwater continues to flow through the aerobic treatment zone.
Microcosm studies with sediments from Vandenberg Air Force Base, CA, suggest that native aerobic methyl tertbutyl ether (MTBE)-degrading microorganisms can be stimulated to degrade MTBE. In a series of field experiments, dissolved oxygen has been released into the anaerobic MTBE plume by diffusion through the walls of oxygenpressurized polymeric tubing placed in contact with the flowing groundwater. MTBE concentrations were decreased from several hundred to less than 10 µg/L during passage through the induced aerobic zone, due apparently to in situ biodegradation: abiotic MTBE loss mechanisms were insignificant. Lag time for initiation of degradation was less than 2 months, and the apparent pseudo-first-order degradation rate was 5.3 day-1. Additional MTBE was added in steps to raise the influent concentration to a maximum of 2.1 mg/L. With each step, MTBE was degraded within the preestablished aerobic treatment zone at rates ranging from 4.4 to 8.6 day-1. Excess dissolved oxygen suggested that even higher MTBE concentrations could have been treated. Continued flow through the treatment zone was repeatedly confirmed through tracer and other tests. These and others’ results suggest that it is possible to create permeable in situ treatment zones solely by releasing oxygen to support native microbial degradation of MTBE.
Introduction Methyl tert-butyl ether (MTBE), primarily because of its widespread use as a gasoline additive, has been inadvertently released to the subsurface environment at thousands of sites in the United States (1), in Europe, and perhaps elsewhere. In some cases, these releases have impacted or posed a potentially significant threat to water supply wells (1, 2). While field evidence suggests that considerable anaerobic transformation of MTBE may occur within or near source zones under some conditions (3), the flux of MTBE out of those zones is often high enough to generate plumes of concern. Other evidence suggests that the plumes are bioattenuated during transport through some subsurface environments, but biodegradation rates are apparently low (4, 5) except where sufficient dissolved oxygen (DO) is present such as at groundwater-surface water interfaces (6). When DO is * Corresponding author phone: (519)888-4567, ext 5372; fax: (519)746-7484; e-mail:
[email protected]. † University of Waterloo. ‡ University of California. 190
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 2, 2002
The field research on in situ MTBE treatment reported to date has generally succeeded in demonstrating the first two requirements but has provided little evidence that the third requirement is met. Salanitro et al. (9, 15) injected a nonnative MTBE-degrading bacterial culture into an existing MTBE plume at Port Hueneme, CA, and, providing oxygen via a pulsed sparging system, showed evidence of treatment over a sustained period of time. Interestingly, in a nonbioaugmented comparison plot also amended with oxygen, microorganisms native to Port Hueneme groundwater were also observed to degrade MTBE in situ after a lag period of approximately 173-230 days. Salanitro et al. (9) noted that under the conditions of their study, the bioaugmented plot apparently had an initially higher rate of degradation than the non-bioaugmented plot. In addition, they reported that TBA appeared to emanate untreated or only partially treated from the zone of oxygen stimulation of the native microorganisms in the non-bioaugmented plot. These findings led them to suggest that in many cases bioaugmentation may be preferred even in the presence of native MTBEdegrading microorganisms. However, their work did not address a possible side effect of employing bioaugmentation and injecting oxygen gas, namely, a reduction in the permeability of the aquifer within the intended treatment zone. Such a permeability reduction might lead to reduced groundwater flow through the treatment zone and thus result in partial bypass of contaminated groundwater around it. Since the groundwater flow within the treatment zone was somewhat uncertain, the field data do not yield reliable estimates of the rate of in situ MTBE degradation. However, in microcosm studies using sediments and groundwater from the site (not bioaugmented but spiked with MTBE to approximately 11 mg/L), Salanitro et al. (9) reported a lag period of 14-21 days, after which time the MTBE was degraded by native microbial populations at an apparently zero-order rate (254 µg/L day-1) to nondetect within 63 days. No results 10.1021/es015562c CCC: $22.00
2002 American Chemical Society Published on Web 12/13/2001
were reported for respikes of the microcosms with additional MTBE. Landmeyer et al. (6) recently reported a field trial of oxygen release within an MTBE plume in South Carolina and noted the rapid onset of degradation of MTBE without accumulation of intermediates such as TBA. In their work, a fine particulate oxygen-releasing metal peroxide compound (ORC; Regenesis Bioremediation Products, CA) was injected as a slurry into the aquifer from a transect of wells. Their results suggest that, at their study site at least, in situ MTBE remediation by native microorganisms can be stimulated solely by release of oxygen, with no need for bioaugmentation. They did not investigate whether the presence of the injected particles of ORC in the pore spaces of the aquifer reduced the permeability of and flow through the intended treatment zone, as might be expected. Thus, since the groundwater flow rate through the treatment zone was not directly measured, it is not possible to use the field data to make a reliable estimate of the rate of in situ degradation of MTBE. They did, however, report results of microcosm studies of biofilm material collected at a groundwater discharge point that mineralized approximately 2 mg/L MTBE within 80 days; no results were reported for respikes of the microcosms with additional MTBE. Mackay et al. (16, 17) reported the results of a field test in which oxygen release was found to support in situ aerobic degradation of MTBE by native microorganisms at Vandenberg Air Force Base, CA. In that work, DO was released to groundwater as it flowed through a pea gravel-filled trench oriented transverse to flow. Within the trench had been emplaced a rectilinear, permeable panel containing lowdensity polyethylene (LDPE) tubing pressurized with a gas mixture of 95% oxygen and 5% sulfur hexafluoride (SF6), the latter serving as a nonreactive tracer. The results indicated that oxygen and SF6 levels were elevated in the groundwater downgradient of the panel while MTBE concentrations were dramatically decreased, apparently due to in situ degradation by native organisms migrating into and populating the panel or pea gravel downgradient of the oxygen release. TBA was detected only transiently after oxygen release began, apparently being treated to nondetectable levels thereafter by the native microorganisms. While this field test confirmed that in situ degradation of MTBE and TBA could be sustained solely by oxygen release, monitoring suggested that the permeability of the panel may have decreased over time, leading to reduced flow through the panel and partial bypass of groundwater around it. Thus, it was not possible to make a reliable estimate of the rate of in situ MTBE degradation from that work. In summary, there appears to be evidence that in situ degradation of MTBE can be supported and maintained by oxygen release to the subsurface, with or without bioaugmentation. However, the field tests reported to date have not been conducted in a way that allows clear confirmation of groundwater flow rate through the intended treatment zone over extended periods of time. In this paper, we describe a series of field experiments on in situ aerobic degradation of MTBE by native microorganisms. The field design allowed for considerable monitoring detail along the flow path into and through the treatment zone, including assessment of MTBE removal rate and continuation of groundwater flow. It is therefore possible to make more reliable estimates of rates of in situ degradation than were possible in prior field tests.
Field Site Since early 1998, we have been conducting field and related lab research focused on an existing MTBE plume and its source, a former gas station (site 60), at Vandenberg Air Force Base (VAFB) in central California (Figure 1). Einarson et al.
FIGURE 1. Map of the vicinity of site 60, Vandenberg Air Force Base (VAFB), CA. The fine dashed line encloses all MTBE detections greater than 2 µg/L (19). The area enclosed by the heavier dashed line is the estimated extent of detectable BTEX species. Noted on the figure are the locations of some of the monitoring conducted as part of our project, notably transects A-C. The closeup reveals orientation and scale of two field experiments: (a) the Longitudinal Trench Facility (LTF) and (b) the Panel Test. (18) describe the results of our site characterization activities in detail, including detailed monitoring along transects A-C (Figure 1). Here we present only a brief overview of the site. Release History and Plume Formation. Leaking underground storage facilities at a service station resulted in the generation of a plume of petroleum hydrocarbons and MTBE. The station was taken out of service in 1994, and the underground storage tanks and piping (located in the source zone noted in Figure 1) were removed in 1995 via two excavations, which were backfilled with pea gravel and/or sand. Lee and Ro (19) estimated the areal extent of MTBE contamination (> 2 µg/L) as roughly 75-90 m wide and approximately 520 m long (Figure 1). A BTEX plume is also present at the site but apparently attenuates to below detection limits within 15-30 m of the source (Figure 1). Hydrogeologic Setting. Site 60 is situated in a small northtrending canyon that feeds into the broad, west-trending Santa Ynez River Valley. A heterogeneous mixture of sand, silt, and clay alluvium fills the canyon to a thickness of approximately 12 m beneath the former service station. Near the leading edge of the plume, the alluvium thickens up to 24 m. Groundwater within the canyon alluvium occurs at depths ranging from 1.5 to 2.5 m below ground surface (bgs) and flows northeast from the former service station. On the basis of slug test, pump test, and water level data, we estimate the average linear groundwater velocity to be on the order of 0.3-0.6 m/day. Figure 2 presents a cross section along transect B (see ref 18 for details), illustrating the layered nature VOL. 36, NO. 2, 2002 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
191
FIGURE 2. Cross section along transect B, illustrating the layered nature of the alluvium and the approximate location of the MTBE plume in March 1999. Permeable, sandy units are depicted in white. Less permeable, silty/clay units are depicted in gray. of the media and the fact that the MTBE plume is almost entirely migrating within one thin aquifer, i.e., within a vertical interval of 2.7-3.6 m bgs. Groundwater at the site, both upgradient of the source area and within the plume, is weakly anaerobic (DO
