GREEN REMEDIATION OF cvoc DNAPL IN BEDROCK

GREEN REMEDIATION OF cvoc DNAPL IN BEDROCK I. Richard Schaffner, Jr., P.G., C.G.W.P. (richard.schaffner@gza.com ) and Steven R. Lamb, P.G., C.G.W.P. (GZA GeoEnvironmental, Inc.; Manchester, New Hampshire) Eric C. Lindhult, P.E. (eric.lindhult@gza.com) (GZA GeoEnvironmental, Inc.; Fort Washington, Pennsylvania) Bernard Fenelon, P.G. (GZA GeoEnvironmental, Inc.; Waukesha, Wisconsin) Michael Reed (Massachusetts Department of Environmental Protection; Springfield, MA) Abstract Natural bioremediation of chlorinated volatile organic compounds (cvocs) via a reductive dechlorination pathway is occurring at many sites, but typically has limited effectiveness due to limited electron donor availability. Enhanced reductive dechlorination (ERD) is a cost-effective treatment technology that stimulates the natural biodegradation of cvocs to expedite site clean-up and reduces the cost of remediation in comparison to conventional remedies. ERD has also been successfully employed at sites where the cvocs exist in fractured bedrock. Three case studies are presented in which a novel bioremediation additive remediated residual, dense non-aqueous phase liquids (DNAPLs) consisting of cvocs in bedrock. Case Study No. 1 Groundwater cvoc contamination was identified at a former manufacturing facility in North Adams, Massachusetts. Historical trichloroethene (TCE) concentrations were as high as 120,000 microgram per liter (μg/l), which indicated the probable presence of residual DNAPL. Complicating the project, the highest TCE concentrations were detected in fractured phyllite-muscovite-quartz schist bedrock. Case Study No. 2 In 1989, an investigation was performed at a Wisconsin farm that reportedly received ~550,000 pounds (250,000 kilograms [kg]) of spent solvent in the early 1970s. The spent solvent was illegally discharged, resulting in contamination of the site s drinking water well. Groundwater analytical results indicated the presence of DNAPL migrating through the glacial till and into the sandstone aquifer, with TCE concentrations as high as 990,000 micrograms per liter (µg/l) in the till groundwater system and up to 75,000 µg/l in the sandstone aquifer. Case Study No. 3 A wastewater treatment plant (WWTP) in New Hampshire was closed in 1986; however, industrial waste haulers apparently entered the facility and illegally discharged industrial septage into the former sludge pits. Groundwater impacted with cvocs extended through the overburden and into the moderately fractured and slightly weathered granofels and schist bedrock. The cvoc concentrations in the bedrock were as high as 11,000 µg/l for TCE and 49,200 µg/l for cis-1,2-dichloroethene (cis-1,2-dce). Case Study No. 1 North Adams, Massachusetts Background and Hydrogeology - Groundwater at a former electrical capacitor manufacturing facility in North Adams, Massachusetts was impacted by a TCE release. The resulting plume extended to a primary recharge area for a municipal well. Historical TCE concentrations were as high as 120,000 μg/l, which is approximately 11% of the aqueous solubility and indicative of the presence of source mass likely in the form of residual DNAPL. The project remedial

investigation and design was conducted in accordance with the Massachusetts Contingency Plan (MCP) with regulatory oversight by the Massachusetts Department of Environmental Protection (MADEP). Site geology consists of fine to coarse sand and gravel, with cobbles and boulders from ground surface to a depth of approximately 10 feet (3 meters [m]) in the source area. Underlying the granular subsurface is moderately weathered fractured phyllite-muscovite-quartz schist bedrock, with minor limestone inclusions, with areas of saprolite observed in the study area. There are numerous transmissive fractures in the bedrock in the source area vicinity. Figure 1 presents a general cross-section of the source area and monitoring well locations. Figure 1 Generalized cross-section of North Adams, Massachusetts Site Groundwater Quality Historical TCE concentrations (up to 120,000 μg/l) were detected in groundwater collected from two bedrock monitoring wells, MS-7 and MS-28 in the source area. Groundwater collected from well MS-3R installed in the overburden contained TCE at a concentration of 6,700 μg/l. The maximum concentration of TCE in groundwater allowable by the MCP in drinking water well recharge areas is 5 μg/l. Baseline groundwater sampling performed prior to commencement of remedial activities (in June 2007) detected TCE concentrations at 4,500 μg/l in overburden well MS-3R, and 40,000 and 58,000 μg/l in bedrock wells MS-28 and MS-7, respectively. Green Remediation Program Implementation ERD was proposed to address cvocs in both the bedrock and overburden groundwater systems and to reduce the source mass and size of the cvoc plume. After an evaluation of the existing biogeochemical conditions, an injection program was designed and implemented that consisted of 1,600 pounds (725 kg) of a beta

version of ERD ENHANCED from BioStryke Remediation, which was gravity injected into monitoring wells MS-3R, MS-7, and MS-28 in October 2007. Due to permeability differences, the majority of the slurry (approximately 98%) was introduced into the overburden according to the MADEP. A second similar injection was performed in August 2008. Approximately one year after the initial injection, the TCE concentrations in groundwater samples collected from the test wells indicated a decrease of 73% in the overburden well and approximately 98% in both fractured bedrock wells. The data are shown in Table 1. During this time, the breakdown products cis-1,2-dce and vinyl chloride, which had not previously been detected at the site at significant concentrations, were detected in each of the three wells, with cis-1,2-dce concentrations increasing nearly 2,000% in samples collected from bedrock well MS-7, thus confirming TCE dechlorination. The order of magnitude increase in ethene concentrations in samples collected from well MS-7, from less than 3 μg/l to 81 μg/l, verified the reduction of cis-1,2-dce to vinyl chloride and ultimately to ethene. Table 1 cvoc concentrations (μg/l) in bedrock groundwater well in Massachusetts Well cvoc Baseline Post-ERD ENHANCED Injection Reductions from Peak* Jun-07 Jan-08 Jul-08 Nov-08 Jun-09 Jul-10 Jul-11 MS-28 TCE 40,000 20,000 700 370 <25 <25 <1 ~100% Bedrock cis-1,2-dce 340 170 470 1,800 2,200 <25 23 99.0% MS-7 TCE 58,000 19,000 1,400 420 16,000 3,900 <10 ~100% Bedrock cis-1,2-dce <1,000 13,000 19,000 2,800 3,100 1,900 68 99.6% MS-3R TCE 4,500 N/S 1,200 N/S 27 88 67 98.5% Overburden cis-1,2-dce 380 N/S 770 N/S 47 770 4 99.0% *Reduction from peak TCE and cis-1,2-dce concentrations. Groundwater samples collected in July 2011, less than 4 years after the initial injection, indicated the mass of cvocs (based on mole fraction) had decreased approximately 99% to 100% in each of the three wells. While the cis-1,2-dce and vinyl chloride concentrations did increase over this period, the cis-1,2-dce concentrations subsequently decreased by more than 93% from their peak. Figure 2 presents the molar concentrations of TCE and cis-1,2-dce in groundwater samples collected from bedrock wells MS-7 and MS-28. TCE has not been detected in either well since June 2010. The detection of methane in all three wells demonstrates that methanogenic conditions were achieved in bedrock and particularly in the overburden formations, which was also verified by the negative oxidation-reduction potential (ORP) values and minimal dissolved oxygen levels measured in groundwater. These conditions are conducive for robust ERD rates to occur. The beta version of ERD ENHANCED migrated with groundwater and created groundwater conditions suitable for ERD at least 200 feet (60 m) downgradient of the injection wells (i.e., at well MW- 9), where TCE concentrations decreased from a baseline of 640 μg/l to consistently <300 μg/l after 2008. The concentration of cis-1,2-dce in groundwater at well MW-9 significantly increased during the ERD program from non-detect baseline concentrations to 180 μg/l, and vinyl chloride was detected there for the first time in 2008.

Moles/Liter 0.0005 0.0004 0.0003 1st Injection 2nd Injection 0.0002 0.0001 0 Jun-07 Dec-07 Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 TCE (MS-7) DCE (MS-7) TCE (MS-28) DCE (MS-28) Figure 2 cvoc concentrations (moles/liter) in bedrock groundwater The longevity of the additive in bedrock was demonstrated by the elevated total organic carbon (TOC) concentrations that existed in the formation for at least 4 years following the initial injections, which is consistent with the longevity (up to 9 years) observed at other sites. The longevity of the additive in the overburden was approximately 2 years, likely due to the recharge of oxygenated surface water into the highly permeable overburden, which limited the longevity but not the effectiveness of the anaerobic dechlorination. Conclusions The beta version of ERD ENHANCED successfully created the desired biogeochemical conditions to destroy residual cvoc source mass and its daughter byproducts in bedrock, as well as in the overburden. Based on a nearly 4-year period with no measurable rebound in TCE concentrations following the initial injection, chemically reducing conditions were maintained in the groundwater system, which resulted in the nearly complete destruction of source mass as well as the downgradient plume. Case Study No. 2 Watertown, Wisconsin Background and Geologic Conditions - In 1989, an investigation began at a Watertown, Wisconsin farm site that received 900 drums (~550,000 pounds [~250,000 kg]) of spent solvent during a 6-month period between 1970 and 1971. The spent solvent, which consisted primarily of TCE, was dumped onto the ground surface at three locations along the flanks of a drumlin, resulting in contamination of the site s drinking water well and local water-supply aquifer, the St. Peter Sandstone. Approximately 80 to 100 feet (25 to 30 m) of glacial till overlies the St. Peter Sandstone. A discontinuous layer of sand and gravel up to 23-feet (7 m) thick overlies the sandstone primarily along the flanks of the drumlin. Given the limited timeframe between release and remedial action, TCE migration in groundwater did not extend more than 200 feet (61 m) from each of the source areas, and potential receptors, including a wetland and local drinking water wells, were more than a half mile from the source area, or were located upgradient of the site.

A remediation system was installed, which consisted of hydraulic control in the sandstone bedrock utilizing two groundwater extraction wells, groundwater extraction from the glacial till utilizing fifteen extraction wells, and soil vapor extraction (SVE) from twelve, nested SVE wells with a sparge well installed at each SVE-well cluster. In 2002, after about 10 years of remedial action, a Remedial Strategy Review was conducted on behalf of the responsible party to evaluate the appropriateness of the existing environmental efforts at the site. The Conceptual Site Model was reviewed with respect to information gained during remediation, and the remedial system was evaluated with respect to the progress made towards meeting remedial objectives. It was concluded that, although there was progress in groundwater cleanup in the shallow sandstone, TCE concentrations in many monitoring wells had not decreased significantly. The review revealed that TCE mass recoveries consisted of approximately 2,500 pounds (1,135 kg) from the two bedrock groundwater extraction wells, approximately 1,000 pounds (455 kg) from the 15 glacial till extraction wells, and approximately 40,000 pounds (18,150 kg) from the SVE system. Thus, less than 10% of the estimated TCE release had been recovered after 10 years of active remediation. It was further concluded that the TCE recovery rates had become asymptotic for the groundwater extraction systems, and that TCE soil vapor recovery was rapidly declining by approximately 50 percent (%) each year. Based on these findings, deactivation of each of the three remedial systems was recommended. However, due to the potential for long-term off-site migration, ERD was recommended through the introduction of a proprietary ERD additive into the sandstone aquifer. This green technology would allow for the destruction of contaminant mass in situ without the continuous use of electrical equipment and anticipated replacement of aging remedial components. Groundwater Quality Groundwater analytical results suggested that DNAPL migrated through 80 to 100 feet (25 to 30 m) of clayey glacial till and into the sandstone aquifer, with TCE concentrations up to 990,000 μg/l or 90 percent of the solubility of TCE in the till and up to 75,000 μg/l in the sandstone. Due to the limited migration timeframe and low groundwaterflow velocities (estimated to be ~35 feet/year [11 m/year]) at this relatively large site, the potential for off-site migration was deemed minimal. Green Remediation Program Implementation The recommendation to use ERD to limit plume migration beyond the TCE source area was presented to the Wisconsin Department of Natural Resources (WDNR) in late 2002. After consideration of the rationale, WDNR approved the recommended approach. The existing remedial system was deactivated in October 2002, and the monitoring well network was expanded with the addition of six monitoring wells. Semiannual groundwater monitoring events were conducted to evaluate post-active remediation groundwater quality and migration potential. Six injection wells screened in the bedrock were installed and remedial additive injections into the sandstone aquifer were performed in late 2004, using approximately 30,000 pounds (9,145 kg) of the beta version of ERD ENHANCED. Table 2 summarizes the TCE, cis-1,2-dce, and vinyl chloride (VC) concentrations detected in groundwater samples from the three closest (at distances of one to two years travel time) monitoring wells downgradient of the injection wells, before and after additive injections in September and October 2004.

Table 2 - cvoc concentrations (μg/l) in bedrock groundwater wells in Wisconsin Well No. Compound Baseline Post-ERD ENHANCED Injection Reduction from Peak 8/04 3/05 10/05 2/06 11/06 3/07 10/07 11/08 MW-40D TCE 5,500 2,700 3,900 3,850 4,900 3,500 124 80 98.5% DCE 1,300 2,400 6,900 8,620 5,900 4,700 2,800 63.9 99.3% VC <5 <5 <10 <10 26 90 48.3 6.3 93.0% MW-41D MW-42D TCE 26 85 160 84 5 2 <0.38 4.7 97.1% DCE <0.8 <0.8 1 29 11 8 46.7 60.1 VC <1 <1 <1 <1 <1 <0.3 <0.3 <0.3 TCE 43 35 13 6 11 4 3.2 0.73 98.3% DCE <0.8 <0.8 <0.8 5 9 4 1.6 0.35 VC <1 <1 <1 <1 <1 <1 <0.3 <0.3 The TCE concentration at well MW-40D decreased initially from 5,500 μg/l to 2,700 µg/l and then increased likely due to a surfactant/co-solvent effect owing to the additive that enhances source mass dissolution, which is a common observation for ERD ENHANCED source area applications. Between November 2006 and November 2008, TCE concentrations decreased from 4,900 μg/l to 80 μg/l; a decrease of more than 98%. The concentration of cis-1,2-dce, a byproduct of TCE dechlorination, but less readily dechlorinated than TCE, increased from 1,300 μg/l to 8,620 μg/l by February 2006, before decreasing more than 99 percent to 63.9 μg/l in less than 3 years, as dechlorination continued. Similarly, the concentrations of vinyl chloride, a degradation byproduct of cis-1,2-dce, initially increased from <5 μg/l to 90 μg/l prior to decreasing by 93% to 6.3 μg/l. Total cvocs, on a molar basis, decreased by almost 99% from the peak of 1.2x10-3 moles/liter (mol/l) in February, to 1.5x10-6 mol/l in November 2008. The increase and then decrease in cis-1,2-dce and vinyl chloride concentrations was anticipated at the site due to the generation and subsequent degradation of TCE daughter products. The concentrations of ethene, a byproduct of vinyl chloride dechlorination, increased in well MW-40D from below detection to 55 μg/l in March 2007, which demonstrates that dechlorination was going to completion as a result of ERD. The TCE concentrations in well MW-41D initially increased from 26 μg/l to 160 μg/l and then decreased to below analytical reporting limits before rebounding slightly. The cis-1,2-dce concentrations generally increased throughout the post-injection period, consistent with the favorable conditions for dechlorination. At well MW-42D, which had low concentrations inconsistent with the presence of residual DNAPL, TCE and cis-1,2-dce were observed to decrease nearly continuously throughout the post-injection monitoring period, given the lack of source mass, and eventually approaching the analytical reporting limit. Vinyl chloride was not detected in samples from this well. Conclusions The implementation of ERD using the proprietary additive was highly effective at the site and allowed for the continued, effective remediation of TCE impacts utilizing this green technology, and termination of the costly, former electrical power-based system. The effectiveness of the remediation at this site improved dramatically while the remedial cost decreased significantly due to elimination of the operation and maintenance of an active

treatment system. Due to a change in the corporate status of the responsible party, groundwater monitoring at the site was discontinued in 2009, which prevented long-term evaluation of the ERD program. Case Study No. 3 Southern New Hampshire Background and Hydrogeology A municipal WWTP was closed in 1986; however, illegal dumping at the site resulted in cvoc impacts to the overburden and bedrock groundwater systems. Site stratigraphy consists of about 40 to 75 feet (12 to 23 m) of overburden (fine sand and silt, overlying clayey silt, coarse sandy glacial till, and silty glacial till). Metamorphic bedrock consisting of moderately fractured and slightly weathered granofels and schist underlies the till. Leakage from the upper overburden groundwater system through the tills recharges the shallow bedrock groundwater system. Groundwater contamination, consisting principally of dissolved-phase cvocs, persists in the upper and lower overburden and in the underlying shallow bedrock. The major contaminant transport mechanisms within each hydrogeologic unit are advection and hydrodynamic dispersion, with groundwater flow towards a river. Groundwater Quality Overburden and bedrock wells were installed in 1988. Historical cvoc concentrations in bedrock well GZ-3B were as high as 11,000 µg/l TCE, 49,200 µg/l cis-1,2-dce, and 660 µg/l 1,1-dichlorethane (1,1-DCA). Overburden concentrations were higher, with TCE concentrations >1,000,000 µg/l in some wells and cis-1,2-dce concentrations up to 79,390 µg/l. Groundwater concentrations in the bedrock had decreased by as much as an order of magnitude between 1988 and the initiation of ERD injections in 2001, but the baseline total cvoc concentration in bedrock was still 4,740 µg/l, which was several orders of magnitude greater than New Hampshire s regulatory standards. Green Remediation Program Implementation ERD was initiated at the site using a beta version of ERD ENHANCED for injections into the overburden in 2000. Additional additive was injected into the bedrock groundwater system later in 2001. As shown on Table 3, cvoc concentrations in groundwater decreased dramatically following additive injection into the bedrock groundwater system. In particular, the concentration of cis- 1,2-DCE decreased over an order of magnitude (from 3,949 µg/l to 85 µg/l) within the first year following the injection. Concentrations continued to decrease and, by December 2003, VC was the only compound detected at concentrations exceeding regulatory standards (11 µg/l versus a standard of 2 µg/l). No cvocs were detected at concentrations exceeding regulatory standards after 2003, including vinyl chloride. The total cvoc concentration in December 2006 was 11 µg/l, a decrease of greater than 99.7% during the more than 5 years since the injections. Total cvoc concentrations decreased further to 5.7 µg/l by November 2011. Importantly, there have been no rebounds in cvoc concentrations since the 2001 injection.

Table 3 - cvoc concentrations (μg/l) in bedrock groundwater well in New Hampshire cvoc Baseline Post-Injection Sampling Reductions from Peak Aug-01 Sep-01 Dec-01 Jun-02 Oct-02 Dec-03 Dec-06 Nov-08 Nov-11 TCE 250 170 <20 <1 <2 <2 <1 <1 <1 >99.6% 1,2-DCE 3,949 4,111 1,700 85 14 6 1.9 <2 <1 >99.9% 1,1-DCE 25 <10 <20 <1 <1 <5 <1 <1 <1 >96.0% VC 440 110 31 0.098 <2 11 1.6 <1 <1 >99.7% 1,1-DCA 69 31 <20 <11 <2 <2 6.5 5.9 3.7 >94.6% Total cvocs 4,733 4,422 1,751 183 30 23 11 5.9 5.7 >99.8% Conclusions The beta version of ERD ENHANCED successfully dechlorinated cvocs in fractured bedrock, and destroyed cvoc source mass in the overburden groundwater system. Summary/Lessons Learned. Each of the three case studies demonstrated that positive results from the ERD injections were observed in bedrock groundwater proximate to injection wells, with reduction in TCE concentrations approaching 100% observed in both bedrock and overburden groundwater systems. The beta version of ERD ENHANCED at each of the three sites was effective in establishing the desired biogeochemical conditions required to reductively dechlorinate cvocs, even in fractured bedrock. An understanding of the fractured bedrock groundwater system and proper delivery of bioremediation additive is critical for a successful ERD remedy. But as demonstrated in these examples, residual DNAPL in fractured bedrock can be remediated to meet regulatory standards. ERD is a green technology with a small carbon footprint that uses no energy-consuming equipment after the initial injection, with some additive remaining effective for 5 years or more. The longevity of this remedial additive in the formation, relative to many other additives, can result in more cost-effective remediation by reducing or eliminating the need to perform additional injections to achieve the desired reduction in cvoc concentrations. This paper was prepared for the proceedings of The Eighth International Conference on Remediation of Chlorinated and Recalcitrant Compounds Conference; Monterey, California; May 2012.