Mar 26, 2000 - FIELD TESTS OF ENHANCED INTRINSIC REMEDIATION OF AN MTBE PLUME. Douglas Mackay. 1. , Ryan Wilson. 1. , Graham Durrant. 1.
Preprint Extended Abstract for Proceedings of ACS National Meeting in San Francisco, CA, March 26-30 2000; to be presented in the Division of Environmental Chemistry Session “Exploring the Environmental Issues of Mobile, Recalcitrant Compounds in Gasoline”
FIELD TESTS OF ENHANCED INTRINSIC REMEDIATION OF AN MTBE PLUME Douglas Mackay1 , Ryan Wilson1 , Graham Durrant 1 , Kate Scow2 , Amanda Smith2 , Murray Einarson1,3, and Bill Fowler3 1 University of Waterloo, Department of Earth Sciences, Waterloo, Ontario, Canada 2 Department of Land, Air and Water Resources, University of California, Davis, California 3 Conor-Pacific/EFW, Palo, Alto, California ABSTRACT We have been studying an MTBE plume at Vandenberg Air Force Base, CA. Microcosm studies with site sediments, conducted at the University of California at Davis and the University of Waterloo, suggest that native aerobic MTBE-degrading bacteria are present in the site sediments and can be stimulated to degrade MTBE solely by adding oxygen. In two separate field tests, dissolved oxygen has been released into the MTBE plume by diffusion through the walls of tubing pressurized with oxygen and in contact with the groundwater flowing through unpumped well screens or permeable walls. In both field tests, significant reductions in MTBE concentrations have been noted downgradient of the diffusive oxygen release systems in repeated sampling events, presumably from in situ biodegradation of the MTBE. Studies are underway to determine if potential breakdown intermediates such as TBA can be detected and if populations of MTBE-degrading bacteria have increased due to oxygenation. INTRODUCTION Laboratory studies have shown that MTBE can be degraded under aerobic conditions by direct metabolism (when the MTBE serves as the carbon and energy source for microbial growth). There is also evidence from field studies that MTBE plumes may be affected by biodegradation during transport through some aquifers (Borden et al., 1997; Schirmer and Barker, 1998), but the rates of degradation are low. When dissolved oxygen is 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. A remedy based on enhancement of in situ aerobic microbial treatment of MTBE must: 1) create steady aerobic conditions, and 2) generate enough microbial biomass to accomplish the treatment at a practically useful rate. Salanitro et al. (1999) injected a non-native MTBE-degrading bacterial culture at Pt. Hueneme, CA, and provided oxygen via a pulsed sparging system, yielding encouraging results. However, it is not yet clear how that method may compare to approaches that rely solely on native MTBE-degrading microorganisms. More generally, there has been little discussion or investigation of whether such in situ bioreactive walls in fact function as they are intended to, i.e. ideally allowing uniform flow of contaminated groundwater through a uniform treatment zone over the long term. At Site 60, Vandenberg Air Force Base, CA, we have found that the existing MTBE plume is weakly anaerobic and located within a relatively shallow and thin aquifer. Microcosm studies with site sediments (UC Davis and U Waterloo) suggested that native aerobic MTBE-degrading bacteria are present in the site sediments and can be stimulated to degrade MTBE solely by adding oxygen (Wilson et al., 1999). Thus, we initiated several field pilot tests of enhancement of aerobic intrinsic remediation near the apparent centerline of the existing MTBE plume and well downgradient of the source area where MTBE is the only VOC at significant concentrations. The main goals of our field tests were 1) to seek field evidence that release of dissolved oxygen into the natural flow of MTBEcontaminated groundwater stimulates and maintains a population of native MTBE-degrading microbes capable of significantly reducing the MTBE flux, and 2) to evaluate the field performance of two configurations of diffusive oxygen-releasing devices, cylindrical and rectilinear. The cylindrical method is described briefly below, while the rest of this paper focuses on the rectilinear “panel” test. Mackay et al., ACS National Meeting, March 2000; p.1
Preprint Extended Abstract for Proceedings of ACS National Meeting in San Francisco, CA, March 26-30 2000; to be presented in the Division of Environmental Chemistry Session “Exploring the Environmental Issues of Mobile, Recalcitrant Compounds in Gasoline”
We have developed and field-tested diffusive oxygen releasing devices or “emitters” (Wilson and Mackay, 1995). The early emitters were cylindrical, with LDPE tubing coiled around a PVC frame, and could be placed in an umpumped well screen. One end of the LDPE was connected to a gas cylinder and the other end to a venting valve. To date, the gas cylinder has contained either oxygen or a mixture of oxygen and sulfur hexafloride (SF6 ), the latter as a conservative tracer. The tubing was pressurized and the imposed chemical gradient causes oxygen (or oxygen and SF6 ) to diffuse through the walls of the tubing and then dissolve into the water flowing through the well screen. If the total pressure and gas partial pressures are held constant within the tubing, the rate of solute flux from the release device should be constant and uniform (in the absence of expressed demand). Previous field tests with these cylindrical emitters have confirmed this steady performance. FIELD TEST OF OXYGEN RELEASE PANEL One of the key goals of the field test discussed herein was the evaluation of a flat panel approach for housing the diffusive emitter tubing. Flat panels are appealing because they offer the potential for relatively uniform solute release over considerable cross-sectional areas (e.g. entire plumes). Figures 1 and 2 present a plan view and vertical section of our small-scale release panel test. The panel was constructed of three layers of prefabricated plastic stripdrain material. Continuous lengths of ¼” LDPE tubing were woven around the internal supports of the stripdrain and the panel covered with a high permeability geotextile. Based on our prior work, the total length of LDPE tubing per cross-sectional area of panel was anticipated to achieve DO levels on the order of 25 mg/L downgradient of the panel in the absence of oxygen demand. The panel and associated devices were emplaced in a trench dug by a backhoe, which was then backfilled with pea gravel to above top of panel and with native materials to surface. Three 1” PVC injection wells were installed to allow release of bromide tracer to investigate water flow patterns, panel clogging, etc. Multilevel monitoring wells were constructed from 1/8” OD Teflon tubing attached to PVC centerstocks. PVC monitoring wells were placed along both sides and the bottom of the panel. One end of each LDPE tube within the panel was connected to a gas cylinder containing a mixture of oxygen and SF6 ; the other end was connected to a venting valve. The gas was supplied to the panel starting in mid-August 1999. We have been monitoring for over two months at the time of writing. A bromide tracer test was conducted (data not included herein). Figure 3 presents selected DO, MTBE and TBA monitoring data from October 1999. The figure contains three plan views of the test, each illustrating the location of the panel and selected monitoring points within the backfill (outer rectangular boundary). Only the middle and lower sampling points on the multilevel samplers are presented (the upper points could not be sampled due to the low water table) and from the two wells on either side of the panel. In the top frame of Figure 3, it is clear that oxygen has emanated downgradient from the panel, with DO on the order of 10 mg/L except for the monitoring location on the right side (approx. 4 mg/L). We have also observed SF6 tracer roughly proportional to the oxygen observed. Thus the emitter appears to be working as expected, especially given that there is a substantial oxygen demand in the groundwater. The oxygen data, especially when considered along with data from the bromide tracer test (not included herein) and the MTBE data in the middle frame of Figure 3, suggest that the flow through the panel excavation is not perfectly orthogonal to the panel. Rather, it appears that the groundwater flow direction in mid October 1999 angles to the left (Figure 3). Thus, we believe the right multilevel downgradient of the panel (with 4.5/3.5 mg/L DO in its middle/lower points) is sampling a zone impacted by water flowing through the panel and low DO water skirting the right edge of the panel. Note that the MTBE concentrations in the groundwater approaching the panel from upgradient ranged from about 100 to 400 µg/L on that day. In marked contrast, the MTBE concentrations measured downgradient of the panel range from below detection limit (
