Monitoring trichloroethene remediation at an iron permeable reactive barrier using stable carbon isotopic analysis

Download Monitoring trichloroethene remediation at an iron permeable reactive barrier using stable carbon isotopic analysis

Post on 05-Sep-2016

212 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

<ul><li><p>cEnviroMetal Technologies, Inc., 745 Bridge St. W., Suite 7, Waterloo, ON, Canada, N2V 2G6d</p><p>more enriched d13C values compared to the upgradient mean. In addition, d13C values for thedegradation products of TCE, cis-dichloroethene and vinyl chloride, show fractionation patterns</p><p>Journal of Contaminant Hydrology 78 (2005) 313325</p><p>www.elsevier.com/locate/jconhydURS Corporation, 8181 East Tufts Ave., Denver, CO, 80237, United StateseAFCEE/ ERD-FEW, 300 Vesle Drive, Building 367, F.E. Warren AFB, WY, 82005, United States</p><p>Received 27 May 2004; received in revised form 31 May 2005; accepted 31 May 2005</p><p>Abstract</p><p>Stable carbon isotopic analysis, in combination with compositional analysis, was used to evaluate</p><p>the performance of an iron permeable reactive barrier (PRB) for the remediation of ground water</p><p>contaminated with trichloroethene (TCE) at Spill Site 7 (SS7), F.E. Warren Air Force Base,</p><p>Wyoming. Compositional data indicated that although the PRB appeared to be reducing TCE to</p><p>concentrations below treatment goals within and immediately downgradient of the PRB,</p><p>concentrations remained higher than expected at wells further downgradient (i.e. N9 m) of thePRB. At two wells downgradient of the PRB, TCE concentrations were comparable to upgradient</p><p>values, and d13C values of TCE at these wells were not significantly different than upgradient values.Since the process of sorption/desorption does not significantly fractionate carbon isotope values, this</p><p>suggests that the TCE observed at these wells is desorbing from local aquifer materials and was</p><p>present before the PRB was installed. In contrast, three other downgradient wells show significantlyMonitoring trichloroethene remediation at an iron</p><p>permeable reactive barrier using stable carbon</p><p>isotopic analysis</p><p>Nancy VanStone a,b, Andrzej Przepiora c, John Vogan c,</p><p>Georges Lacrampe-Couloume a, Brian Powers d, Ernesto Perez e,</p><p>Scott Mabury b, Barbara Sherwood Lollar a,*</p><p>aStable Isotope Laboratory, University of Toronto, 22 Russell St., Toronto, ON, Canada, M5S 3B1bDepartment of Chemistry, University of Toronto, 80 St. George St., Toronto, ON, Canada, M5S 3H60169-7722/$ -</p><p>doi:10.1016/j.</p><p>* Correspon</p><p>E-mail addding author. Tel.: +1 416 978 0770; fax: +1 416 978 3938.see front matter D 2005 Elsevier B.V. All rights reserved.</p><p>jconhyd.2005.05.013</p><p>ress: bslollar@chem.utoronto.ca (B. Sherwood Lollar).</p></li><li><p>expected for the products of the reductive dechlorination of TCE. Since concentrations of both TCE</p><p>and degradation products drop to below detection limit in wells within the PRB and directly below it,</p><p>these downgradient chlorinated hydrocarbon concentrations are attributed to desorption from local</p><p>aquifer material. The carbon isotope values indicate that this dissolved contaminant is subject to local</p><p>degradation, likely due to in situ microbial activity.</p><p>D 2005 Elsevier B.V. All rights reserved.</p><p>Keywords: Carbon isotopes; Groundwater remediation; Permeable reactive barrier; Reductive dechlorination;</p><p>Trichloroethene; Zero-valent iron</p><p>1. Introduction</p><p>Permeable reactive barriers (PRBs) constructed of elemental iron have emerged as an</p><p>effective passive remediation method for ground water contaminated with a range of</p><p>contaminants, mainly chlorinated hydrocarbons (CHCs) (e.g. Agrawal and Tratnyek,</p><p>1996; Alowitz and Scherer, 2002; Gillham and OHannesin, 1994; Hozalski et al., 2001;</p><p>Nam and Tratnyek, 2000). A PRB is an in situ engineered zone of reactive material placed</p><p>across the path of contaminated ground water. The major advance of PRBs over other</p><p>ground water remediation approaches is the lack of above ground structures, low operation</p><p>and maintenance cost and the enhanced remediation efficiency, particularly compared with</p><p>pump-and-treat systems. There are currently more than 90 PRBs installed at CHC</p><p>contaminated sites (OHannesin, 2003). Many studies have investigated the pathways,</p><p>mechanisms, kinetics and longevity of CHC degradation using elemental iron (see Scherer</p><p>et al., 2000; Tratnyek, 2002).</p><p>Assessment of iron PRB performance is based on monitoring CHC concentrations,</p><p>along with pH and Eh, and major inorganic constituents, in ground water well transects</p><p>across the PRB. Often, iron PRBs are installed within existing contaminant plumes and</p><p>therefore elevated concentrations of contaminants are observed downgradient of PRBs for</p><p>some time after the system has been installed, depending on the extent of initial</p><p>contamination, ground water flow rates, desorption rates and type of the aquifer material.</p><p>The persistence of CHCs downgradient of iron PRBs is expected because significant</p><p>volumes of ground water may be needed for the remnant contaminants to be desorbed and</p><p>flushed from downgradient aquifer materials (Heneman et al., 2001; Powell et al., 1998). It</p><p>is difficult, however, to evaluate PRB performance in situations where elevated</p><p>contaminant concentrations downgradient of the PRB remain for longer than expected</p><p>periods (Powell et al., 1998). This could arise for three reasons: (1) incomplete degradation</p><p>within the PRB, (2) hydraulic bypass underneath or around the PRB, and (3) slower than</p><p>expected desorption of remnant contaminants trapped within aquifer materials. Distin-</p><p>guishing between these processes is important for the assessment of PRB performance and</p><p>consequently developing new techniques that help resolve these processes is of</p><p>considerable interest.</p><p>Fractionation of stable carbon isotopes has been observed for chlorinated ethenes</p><p>N. VanStone et al. / Journal of Contaminant Hydrology 78 (2005) 313325314during processes such as biodegradation (Bloom et al., 2000; Hunkeler et al., 1999;</p><p>Sherwood Lollar et al., 2001; Slater et al., 2001), oxidation by permanganate (Hunkeler et</p></li><li><p>whereas carbon isotopic fractionation associated with degradative processes is on the orderof tens of permil (x).Carbon compound specific isotopic analysis (CSIA) is used to measure carbon isotope</p><p>fractionation and has been shown to be a powerful tool for identifying in situ</p><p>biodegradation in the field for contaminants such as perchloroethene (PCE) and</p><p>trichloroethene (TCE) (Hunkeler et al., 1999; Sherwood Lollar et al., 2001; Song et al.,</p><p>2002; Vieth et al., 2003). Several studies have shown large and reproducible carbon</p><p>isotopic fractionation during the degradation of chlorinated ethenes on electrolytic and cast</p><p>iron (Bill et al., 2001; Dayan et al., 1999; Slater et al., 2002; VanStone et al., 2004), but</p><p>these observations have not been applied in the field to date. The objective of this study</p><p>was to investigate the use of carbon stable isotopic analysis to help evaluate PRB</p><p>performance at a field installation for the remediation of TCE. Routine performance</p><p>monitoring of the PRB consisted of quantification of TCE and its products upgradient,</p><p>within and downgradient of the iron PRB. In addition, carbon CSIAwas carried out in the</p><p>vicinity of the PRB to investigate the source of elevated chlorinated ethene concentrations</p><p>observed downgradient of the installation.</p><p>2. Site background</p><p>A PRB constructed of iron filings was installed in 1999 at the F.E. Warren Air Force</p><p>Base, Wyoming, Spill Site 7 (SS7) to remediate ground water contaminated with TCE.</p><p>The TCE originates from a defunct on-site liquid oxygen manufacturing facility. A grease</p><p>trap at the plant was the source of organic solvents (primarily TCE) to a surface drainage</p><p>ditch, resulting in migration of the solvents through the vadose zone to local ground water</p><p>(Heneman et al., 2001). Construction of the PRB was completed in October 1999. A</p><p>schematic of the SS7 PRB is shown in Fig. 1, with the direction of ground water flow</p><p>indicated. The PRB is 173 m long, has a flow through thickness of 1.2 m and a saturated</p><p>depth of 4.6 m below the historic low ground water level, with a low-permeability clay cap</p><p>installed on the PRB to minimize ground water flow over the top of the treatment PRB.al., 2003; Poulson and Naraoka, 2002), and reductive dechlorination on zero-valent iron</p><p>(Bill et al., 2001; Dayan et al., 1999; Slater et al., 2002; VanStone et al., 2004). Carbon</p><p>isotopic fractionation results from differences in the rates of reaction for 13C- and 12C-</p><p>bearing molecules. The differences in reaction rates for the different isotopes are due to</p><p>mass-dependent differences in activation energies for the respective reactions (Fry, 1971).</p><p>In general, for most chlorinated ethenes and aromatic compounds, the lighter isotope (12C)</p><p>reacts faster than the heavier isotope (13C), leading to fractionation, and enrichment of the</p><p>heavy isotope in the remaining reactant as the reaction proceeds. Likewise, this leads to</p><p>enrichment in the light isotope with respect to the parent compound in the products of the</p><p>reaction. Laboratory experiments have shown that carbon isotopic fractionation for</p><p>chlorinated ethenes is not significant (i.e. b0.5x) for non-degradative processes such asdissolution, vaporization or adsorption under equilibrium conditions (Dempster et al.,</p><p>1997; Harrington et al., 1999; Poulson and Drever, 1999; Slater et al., 1999, 2000),</p><p>N. VanStone et al. / Journal of Contaminant Hydrology 78 (2005) 313325 315The PRB was installed at the site as an interim remediation measure and it does not extend</p><p>to the bottom of the contaminated shallow aquifer. The PRB is located within 20 to 60 m</p></li><li><p>N. VanStone et al. / Journal of Contaminant Hydrology 78 (2005) 313325316of a perennial creek. The SS7 PRB contains approximately 1.6106 kg of iron filings. Toaccommodate the variations in ground water velocity and CHC concentration expected</p><p>along the line of PRB installation, the amount of iron in the PRB was varied by using a</p><p>different proportion of iron/sand in each of the three sections: Section 1 with 100% iron,</p><p>Section 2 with 25% iron/75% sand mix, and Section 3 with 38% iron/62% sand mix.</p><p>After the PRB was installed in October 1999, the system was left undisturbed for 6</p><p>months to allow the ground water conditions to equilibrate before implementing a</p><p>performance monitoring program (Heneman et al., 2001). The monitoring wells discussed</p><p>in this study are shown in Fig. 1 and represent only a portion of the wells at the site. These</p><p>monitoring wells are made of PVC and have 1.5 m screens. There are a series of</p><p>monitoring wells at the centre of each segment of the treatment PRB (Transects 1, 2 and</p><p>3), which are shown in expanded view in Fig. 1. In Transect 1, wells 101A and 101B are</p><p>located within a meter of the upgradient side of the PRB at depths of 8.5 and 4.5 m,</p><p>respectively, with well 101A located about 1 m below the PRB and 101B located within</p><p>Fig. 1. Schematic Map of SS7 PRB. Gray-filled circles indicate the position of monitoring wells that contained</p><p>concentrations of contaminants above the detection limits (d.l.) for isotopic analysis, and black-filled circles</p><p>indicate wells that did not contain adequate concentrations for CSIA. The treatment PRB is approximately 173 m</p><p>long. Base map is to scale (1 cm=10 m). Hatched lines show expanded view of transects 1, 2 and 3. Section 1 is</p><p>1.2 m thick, Section 2 is 0.3 m thick and Section 3 is 0.5 m thick. Expanded views show sampling wells located at</p><p>the centre of each main section (Transects 1 to 3). See text for detailed description of each monitoring well. Wells</p><p>indicated (*) are at a depth at least 1 m below the PRB.</p></li><li><p>projects (OHannesin and Gillham, 1998).Performance monitoring of the SS7 PRB revealed a decrease in CHC concentration</p><p>across the PRB from upgradient to non-detectable values within the PRB and immediately</p><p>downgradient (Heneman et al., 2001). In contrast, in wells further downgradient of the</p><p>PRB (e.g. wells 186, 700B, 707B and 708 located 9 to 12 m from the PRB) concentrations</p><p>of both TCE and cDCE were comparable to upgradient concentrations and were suggested</p><p>to be due to desorption of TCE from aquifer materials downgradient of the PRB (Heneman</p><p>et al., 2001). Corresponding increases in pH and decreases in Eh, carbonate alkalinity,</p><p>Ca2+, and Mg2+ were noted within the PRB and immediately downgradient in comparison</p><p>to upgradient wells (Heneman et al., 2001). These trends are expected due to reducing</p><p>conditions created by iron corrosion and are consistent with those noted at other PRB sites</p><p>(OHannesin and Gillham, 1998). The values of iron-sensitive parameters (i.e. Eh and pH)</p><p>and constituents (Ca2+, alkalinity) measured in wells 700B, 707B, and 173B were similar</p><p>to the upgradient values. It was unclear whether the persistent high concentrations of TCE</p><p>and cDCE in the downgradient wells were due to incomplete degradation of TCE within</p><p>the PRB, or to continued desorption from the aquifer material (Heneman et al., 2001).</p><p>Based on hydraulic heads measured at each monitoring well, ground water flow paths in</p><p>the vicinity of the PRB were not significantly altered from pre-PRB ground water flow</p><p>records (Heneman et al., 2001), ruling out hydraulic bypass of the PRB as the source of</p><p>elevated downgradient concentrations of CHCs.</p><p>3. Methods</p><p>Samples for concentrations of chlorinated hydrocarbons and carbon isotopic analysis</p><p>were taken in June 2002 (Table 1). Four 40 mL samples from each well were preserved</p><p>with several drops of undiluted HCl to inhibit microbial activity, and stored without</p><p>headspace. Additional water samples were taken from each well for analysis of TCE,</p><p>cDCE and VC concentrations by EPA method 502.1. The detection limit for this analysis</p><p>is 0.05 ug/L. All samples for carbon CSIA were packed on ice and shipped to the Stable</p><p>Isotope Laboratory at the University of Toronto, Canada, and analyzed within 1 month of</p><p>sampling. Detection limits for isotopic analysis varies from compound to compound but isthe ground water plume transected by the PRB. Well 102 is located in the middle of the</p><p>PRB. Wells 103A and 103B are located approximately 1 m downgradient of the PRB at</p><p>depths of 8.5 and 4.5 m, respectively, with 103A located about 1 m below the PRB and</p><p>103B located within the ground water plume intercepted by the PRB. Well 700B is located</p><p>9 m downgradient from the PRB. The wells for the other 2 transects (Transect 2 and 3) are</p><p>similarly located with respect to the PRB.</p><p>Periodic sampling of the groundwater monitoring wells has been conducted since April</p><p>2000 for concentrations of CHCs (TCE, cDCE, trans-dichloroethene and VC) using</p><p>standard EPA method 502.1, and several inorganic indicators of water quality (Heneman et</p><p>al., 2001). There were consistent decreases in Eh, sulfate, calcium and magnesium</p><p>concentrations and increases in pH consistent with those documented at other PRB</p><p>N. VanStone et al. / Journal of Contaminant Hydrology 78 (2005) 313325 317between 10 and 30 ug/L for TCE, cDCE and VC. One day prior to analysis, 2 of each 40</p><p>mL water samples were transferred to 160 mL glass bottles with 40 g of NaCl to facilitate</p></li><li><p>Table 1</p><p>Concentrations and stable carbon isotopic compositions of VC, cDCE and TCE from monitoring wells at</p><p>F.E.Warren AFB spill site 7 iron-filings PRB</p><p>Well TCE cDCE VC</p><p>(Ag/L) d13C (x) (Ag/L) d13C (x) (Ag/L) d13C (...</p></li></ul>