Cyclodextrin-Enhanced In Situ Flushing of Multiple-Component

Flushing of Multiple-Component. Immiscible Organic Liquid. Contamination at the Field Scale: Analysis of Dissolution Behavior. JOHN E. MCCRAY †,§. ...
0 downloads 0 Views 79KB Size
Environ. Sci. Technol. 1999, 33, 89-95

Cyclodextrin-Enhanced In Situ Flushing of Multiple-Component Immiscible Organic Liquid Contamination at the Field Scale: Analysis of Dissolution Behavior J O H N E . M C C R A Y †,§ A N D M A R K L . B R U S S E A U * ,†,‡ Departments of Hydrology and Water Resources and of Soil, Water, and Environmental Science, The University of Arizona, Tucson, Arizona 85721-0038

There is great interest in the potential use of solubilityenhancing agents for subsurface remediation of non-aqueousphase organic liquid (NAPL) contamination. Cyclodextrin was demonstrated to be effective for NAPL removal during a recent pilot-scale field study. The study provides an opportunity to investigate the mechanisms controlling mass transfer between a multicomponent NAPL and an enhancedflushing agent solution at the field scale. A relationship is developed to describe enhanced dissolution of a multiplecomponent NAPL and is used to analyze the field data. While NAPL dissolution behavior was generally complex during the cyclodextrin flush, the initial peak and final effluent concentrations for most of the target contaminants were within a factor of 2 of the equilibrium values predicted using the ideal enhanced-dissolution theory. This suggests that the dissolution of the multicomponent NAPL during the cyclodextrin flush may be approximately treated, at least for practical purposes, as an ideal, equilibrium process. It appears that the dissolution theory successfully predicted the observed behavior for this system. Thus, it may be useful for assisting in the planning, design, and evaluation of other enhanced-flushing applications involving multicomponent NAPL.

Theory An expression for equilibrium partitioning of a component between water and a multiple-component NAPL may be derived from basic thermodynamic principles and is given by (23):

C Wi ) C oWi(XNi/XoNi)(γNi/γoNi)(γoWi/γWi)

Introduction The contamination of groundwater by hazardous organic chemicals and the associated risks to humans and the environment have become issues of great importance. Nonaqueous-phase liquids (NAPLs) are a common cause of groundwater pollution and occur in the subsurface at numerous contaminated sites. According to the National Research Council, the presence of NAPL is the single most important factor limiting site cleanup (1). Due to the welldocumented limitations of pump-and-treat remediation, alternative methods for removal of subsurface NAPL contamination are a focus of current research (1, 2). Enhanced in situ flushing is an innovative remediation technique that * Corresponding author phone: (520)621-1646; fax: (520)621-1647; [email protected]. † Department of Hydrology and Water Resources. § Present address: Department of Geology and Geological Engineering, Berthoud Hall, Colorado School of Mines, Golden, CO 80401-1887. ‡ Department of Soil, Water, and Environmental Science.10.1021/es980117b CCC: $18.00 Published on Web 11/14/1998

is currently attracting a great deal of attention. In this technique, dissolution of NAPL into the aqueous phase is enhanced by the addition of a solubility-enhancement agent to the flushing fluid. Cosolvents (e.g., alcohols) and surfactants are examples of reagents that have been proposed for this purpose. Effective design and evaluation of enhanced in situ flushing systems requires an understanding of the mechanisms controlling NAPL dissolution. Most laboratory research for enhanced-solubility flushing has involved single-component NAPLs (e.g., refs 3-10). Such research has contributed greatly to our understanding of enhanced NAPL dissolution. However, many hazardous waste sites contain multiplecomponent NAPLs (e.g., refs 11-20). Examples of these types of wastes include chlorinated solvents, coal tars, pesticides, and various petroleum-based fuels. Relatively few studies have examined the enhanced dissolution of multicomponent NAPLs, particularly at the field scale (11-16). However, fieldscale studies are essential for evaluating the performance potential of new remediation technologies (21), as well as for advancing our understanding of relevant processes. Recently, cyclodextrin, an enhanced-flushing agent that is a sugar-based molecule, was shown to be effective at remediating an aquifer contaminated with a multicomponent NAPL during a pilot-scale field test (12, 15). During this study, cyclodextrin induced very large increases in the aqueous concentrations of all the target contaminants, ranging from nearly 100 to over several thousand times the concentrations achieved in a water flush conducted immediately prior to the cyclodextrin flush (12, 22). The purpose of the present work is examine the NAPL dissolution behavior observed during the pilot-scale field test and to determine if it can be treated as an ideal, equilibrium process. This will be accomplished by comparing the measured aqueous contaminant concentrations to those predicted using an enhanced-dissolution theory combined with laboratory data. The factors influencing mass transfer of contaminants between the multicomponent NAPL and the cyclodextrinflushing solution will be briefly discussed.

 1998 American Chemical Society

(1)

The attachment of the superscript “o” on a term denotes that the term represents a single-component NAPL and water system. When a term does not include the “o” superscript, it refers to a multicomponent NAPL and water system. The subscript “i” represents a particular component in the NAPL mixture, the superscript “N” denotes the NAPL phase, the superscript “W” denotes the aqueous (water) phase, C is concentration [mass/L3], XNi is NAPL phase mole fraction of component i, and γi is the activity coefficient of component i in the associated phase (N or W). Equation 1 is a modified form of eq 7 in Burris and MacIntyre (23), where aqueousphase mole fractions have been converted to mass-pervolume concentrations. The activity coefficient terms account for nonideal partitioning effects that result from interactions between organic constituents in the NAPL and aqueous phases (24). The γNi may be larger than unity (nonideal condition) when the organic chemical is significantly dissimilar in shape, size, or polarity from the average NAPL mixture, especially when VOL. 33, NO. 1, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

89

the NAPL phase mole fraction of the constituent is small (23-25). However, it is often assumed that components in the NAPL phase behave ideally (i.e., γNi ) 1). For a singlecomponent NAPL, XoNi and γoNi are equal to 1 when the partitioning of water into the NAPL phase is negligible (or has minimal impact) (23-25). Thus, for ideal conditions, it is not necessary to include the terms XoNi, γoNi, and γNi in eq 1. The value for γoWi is generally greater than 104 for hydrophobic organic compounds (24). The γoWi term may differ from the γWi term when interactions among organic cosolutes in the aqueous phase are significant. For an ideal system, it is assumed that these interactions are negligible and that γoWi is equal to γWi. This is generally valid for the dilute aqueous solutions associated with typical NAPL-water systems (23, 25). Thus, for ideal conditions, the aqueousphase activity coefficient terms also disappear from eq 1. The result of implementing the above assumptions is the well-known form of eq 1 called Raoult’s law (C Wi ) C oWiXNi). Raoult’s law has been derived from thermodynamic principles in a different manner but using the same assumptions, by Banerjee (25) and Schwarzenbach et al. (24). The Raoult’s law-based approach has been used to successfully predict aqueous-phase concentrations of compounds (or partition coefficients) for gasoline (26), diesel fuel (19), and coal-tar (27) systems. Enhanced solubilization of a single-component NAPL into the aqueous phase is obtained by the effect of the solubilization agent on the aqueous activity of the contaminant. The agent used for this research, cyclodextrin, is a cyclic oligosaccharide, or sugar, formed from the degradation of starch by bacteria (28). Cyclodextrins are similar in effect to both cosolvents and surfactants but have significant differences that preclude classification in either category. The apparent solubilities of single organic compounds in aqueous cyclodextrin solutions have been observed to increase linearly with the concentration of cyclodextrin (29, 30). The influence of cyclodextrin on the apparent solubility of an organic compound is described by (29):

C Ai ) C Wi(1 + KcwCH) ) C WiEi

(2)

where C Ai is the equilibrium apparent aqueous solubility of the organic solute in the cyclodextrin-enhanced aqueous phase, Kcw is the partition coefficient of the solute between cyclodextrin and water, CH is the aqueous concentration of cyclodextrin, and Ei is the solubility-enhancement factor. The superscript “A” is used to denote a general aqueous phase containing water, dissolved cyclodextrin, and dissolved organic chemical. As discussed in ref 29, eq 2 is based on complexation theory. An equation analogous to eq 1 can be developed to describe the influence of an enhanced-solubility agent on the equilibrium aqueous concentration (C Ai) of a multicomponent NAPL constituent by incorporating eq 1 into eq 2 (substituting for C Wi):

C Ai ) C oWiXNiγNi(γoAi/γAi)Ei

(3)

and assuming that XoNi and γoNi are equal to 1, as stated above. The “W” superscripts on the activity coefficient terms have been changed to “A” to denote a general aqueous phase. In this formulation, the term γNi(γoAi/γAi) represents all potential NAPL phase and aqueous-phase nonidealities for component i due to mixture-dependent effects, including those related to the presence of the reagent. For ideal dissolution, the value for Ei in a multicomponent NAPL system is the same as that measured for singlecomponent NAPL systems (Eoi). However, some nonideal effects, such as competition between cosolutes for the cyclodextrin cavity, may cause the enhancement factor for 90

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 1, 1999

a multicomponent NAPL system to differ from that for a single-component NAPL system. The assumption that cosolutes have a negligible effect on the aqueous-phase activity coefficients of a component may not be valid for enhanced dissolution because of the significant increase in the concentrations of organic solutes in the aqueous phase. However, for the cyclodextrin system, the increased concentration is due to the partitioning of solute molecules into the cyclodextrin cavity. This essentially shields them from direct contact with water molecules, which suggests that an assumption of “ideality” may remain valid. For enhanced-solubility agents in general, the γNi term may vary somewhat from that for a water-only system due to partitioning of the agent to the NAPL phase. However, cyclodextrin does not partition significantly to the NAPL phase (12). For ideal dissolution, γNi and (γoAi/γAi) are equal to unity, Ei is equal to Eoi, and eq 3 reduces to a form analogous to Raoult’s law for ideal enhanced dissolution:

C Ai ) XNiC oWiEoi

(4)

Site Description The enhanced-solubilization complexing-sugar flush (CSF) field experiment was conducted at the OU1 site at Hill AFB in Layton, UT. The experiment was part of a multipleuniversity project intended to test innovative remediation technologies (see ref 12). The unit of concern is a shallow aquifer (averaging about 9 m in depth) that consists of fineto-coarse sand interbedded with gravel and some silt. A 60m-thick clay layer forms a bottom boundary for the aquifer. The porosity at the test area is approximately 20% (31). The experiment was conducted in an enclosed cell (3 m × 5 m in area) to minimize potential migration of normally sparingly soluble contaminants that could experience enhanced solubilization and transport in the presence of cyclodextrin and to facilitate mass balance and performance assessment (see ref 12 for details). The site is severely contaminated from many sources including chemical disposal pits, fire training areas, landfills, and waste oil pits (12). The treatment cell is adjacent to (or within) one of the disposal pits and was emplaced in what is considered to be a NAPL source area. The initial NAPL saturation was estimated from partitioning tracer studies to be 12.6% (31). Table S1, found in the Supporting Information, lists many of the contaminants detected in aquifer samples collected from OU1 in the vicinity of the treatment area. Of the many different compounds in the NAPL, 12 “target contaminants” were chosen prior to the experiment for the purpose of evaluating remediation effectiveness. These targets (with acronyms listed in parentheses) are trichloroethene (TCE), 1,1,1-trichloroethane (TCA), naphthalene (NAP), o-xylene (o-XYL), m,p-xylene (p-XYL) measured as p-xylene, toluene (TOL), benzene (BENZ), ethylbenzene (EB), 1,2-dichlorobenzene (DCB), 1,2,4-trimethylbenzene (TMB), decane (DEC), and undecane (UND). These compounds were selected to provide a representative subset of the NAPL constituents of concern present within the study area. On the basis of estimates of NAPL phase mole fractions for each target contaminant (to be discussed in the results section), we estimate that these compounds comprise slightly less than 10% of the total NAPL within the cell. The remainder appears to be comprised primarily of higher-molar-mass jetfuel components, other chlorinated hydrocarbons, and possibly relatively insoluble, pitchlike components. Additional information regarding the site may be found in other works (12-14).

Materials and Methods The enhanced in situ flushing technique used in the field experiment is based on the use of cyclodextrin as the

enhanced-solubilization agent. For the field experiment, the cyclodextrin was dissolved in water and pumped into the contaminated target zone using a horizontal flow field. A line of four injection wells and a line of three extraction wells, both normal to the direction of flow, were used to generate a steady-state flow field. The injection and extraction wells (5.1-cm diameter) were fully screened over the saturated thickness. Technical-grade cyclodextrin was used for the experiment and was delivered in dehydrated form (Cerestar USA Inc., Lot 8028). This technical-grade product was comprised of approximately 90% pure hydroxypropyl-βcyclodextrin (HPCD) and 10% production byproducts (mainly hydrated ash). The cyclodextrin was then mixed with potable water to achieve an HPCD concentration of 10%. Approximately 8 pore volumes of the 10% cyclodextrin solution (approximately 65 500 L total) was pumped through the cell at a rate of 4.54 L/min ((5%) for 10 days. The flow was controlled such that the rate was equal for each well. The total flow rate was monitored and adjusted to maintain the water table in the enclosed cell at 5.4-5.7 m below ground surface (bgs). After the eighth day, flow was ceased for 1 day to investigate the potential for rate-limited dissolution. An additional 2 days of cyclodextrin flushing was conducted after the flow interruption period. At the end of the experiment, about 5 pore volumes of cyclodextrin-free water was flushed through the cell to remove the cyclodextrin. Aqueous samples were collected at each extraction well (E51, E52, E53) during the CSF to monitor for target contaminant and cyclodextrin concentrations. Samples were collected hourly for the first 36 h of the experiment and every 3-4 h thereafter. The samples were analyzed by a gas chromatograph (GC-FID) (Shimadzu, GC-17A) with a headspace autosampler (Tekmar, 7000). With this method, QA/ QC standards were met for 10 of the 12 compounds. Benzene and TCA presented analytical difficulties for the aqueous CSF samples due to coeluting compounds. Therefore, the aqueous-phase CSF concentrations for these two compounds are considered unreliable. In addition, benzene could not be detected consistently in water-flush samples, and DEC and UND were below detection limits in all water-flush samples. Soil-core concentrations of the targets were measured for core samples collected before and after the CSF. The core samples were placed in 40-mL vials containing 5 mL of acid and 5 mL of methylene chloride and delivered to the Environmental Laboratory at Michigan Technological University for analysis. The samples were analyzed by a gas chromatograph (GC) (Hewlett-Packard 5890 with capillary column)-mass spectrometer (Hewlett-Packard 5970), equipped with an automatic sampler (Hewlett-Packard ALS 7673). QA/QC standards for the soil-core analysis were met for all 12 target compounds. Subsamples were taken throughout the remediated zone at vertical increments of about 0.25-1 m, depending on the amount of soil recovered with the coring equipment. Additional details on the experimental setup, core collection, and sampling and analytical protocols for soil-core and aqueous-phase samples may be found in McCray and Brusseau (12) and McCray (32). Calculation of NAPL Phase Mole Fractions for Target Contaminants. An estimate of the NAPL phase mole fractions for the target contaminants must be obtained prior to attempting meaningful quantitative analysis of multicomponent NAPL dissolution (see eqs 3, 4). The determination of mole fractions can be very difficult for complex NAPL mixtures at contaminated sites. However, the measured soilcore contaminant concentrations, along with other measured or estimated parameters, enable us to obtain an estimate of this important parameter for use in subsequent analysis. The mole fractions (XNi) were calculated for each component as

TABLE 1. Comparison of Percent Change for Mole Fraction and Concentration during Complexing-Sugar Flush (CSF) % changee

X Ni target

pre-CSFa,b

post-CSFa

in X Ni

in Cmc

TCE BENZ TOL DCB TCA NAP EB o-XYL m,p-XYL TMB UND DEC

0.0015 0.0001 0.0046 0.0186 0.0006 0.0013 0.0006 0.0028 0.0010 0.0013 0.0330 0.0089

0.0002 1). Some compounds may have experienced selective complexation into the cyclodextrin cavities, resulting in values of Ei that are not equal to Eio (contrary to the assumption used in eq 4). The value for Ei could not be measured for the in situ organic mixture because the exact composition was not known. Thus, Eio values from single-component batch experiments were used. Significant differences between Ei and Eio could result in the MIR values deviating from unity. Finally, other processes, such as biodegradation and nonuniform flow, may have resulted in reduced measured concentrations, further complicating the analysis. However, given the chemical complexities of the field study, as well as the analytical uncertainties inherent in such a study, it is very difficult to explicitly quantify the effect of all potential nonideal dissolution processes. The potential for rate-limited dissolution during this experiment was investigated by interrupting flow for 1 day after 8 days of flushing. The static conditions would presumably allow more time for dissolution to occur. Thus, an increase in the effluent concentration following flow interruption would indicate that NAPL dissolution may be ratelimited rather than instantaneous. Figure 1 illustrates that there was not a significant increase in the effluent concentrations of the target contaminants after restarting flow (on day 9). Generally, these results indicate that the NAPL dissolution in the presence of the cyclodextrin solution was near equilibrium during the CSF, which is consistent with the results of the analysis presented above. However, under typical field conditions, this technique may not detect changes in local mass-transfer processes because the average mass-transfer behavior of the flushed NAPL zone will dominate the evolution of the effluent concentrations. In addition, the duration of the no-flow period may not have been sufficient to allow time for contaminants experiencing rate-limited dissolution to achieve equilibrium. Therefore, these results should not be interpreted to mean that dissolution of NAPL was instantaneous everywhere within the cell.

Summary The results of this paper illustrate that dissolution of a multicomponent NAPL during a cyclodextrin flush at the 94

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 33, NO. 1, 1999

field scale was approximately an ideal, equilibrium process and that enhanced-dissolution theory could be combined with laboratory data to interpret and approximately predict the field-scale dissolution behavior of the multicomponent NAPL. While dissolution was apparently an ideal, equilibrium process during this experiment, it may not be for other cases, especially if the treatment zone is significantly larger than the NAPL-contaminated zone. An in-depth understanding of the dissolution behavior of multicomponent NAPL can contribute greatly to remediation design and risk-based decision making. This may be particularly true when specific contaminants in a multicomponent NAPL are the remediation targets, because the dissolution behavior of certain contaminants may deviate significantly from the expected behavior of the majority of the contaminants. The dissolution phenomena discussed here may contribute to understanding the dissolution behavior of organic contaminants in solutions of other enhanced-solubility agents, such as surfactants or cosolvents.

Acknowledgments This material is based upon work funded by the U.S. EPA under Cooperative Agreement CR-822024 to The University of Arizona with funds provided through the Strategic Environmental Research and Development Program (SERDP). This document has not been subject to agency review; mention of trade names or commercial products does not constitute endorsement or recommendation for use. The authors would especially like to thank the following: George Reed and Matthew Romberger of Cerestar USA Inc., who were responsible for donation of the cyclodextrin used in the field experiment, and Carl Enfield and Lynn Wood of the U.S. EPA for their considerable support for the project. Special thanks are extended to Bill Blanford, Ken Bryan, Brent Cain, and Gwynn Johnson for their assistance in this research.

Supporting Information Available Tables S1-S3 and Figures S1 and S2 (8 pages). Ordering information can be found on any current masthead page.

Literature Cited (1) National Research Council (NRC). Alternatives for Groundwater Cleanup; National Academy Press: Washington, DC, 1994. (2) Palmer, C. D.; Fish, W. USEPA, EPA/540/S-92/001, 1992. (3) Imhoff, P. T.; Gleyzer, S. N.; McBride, J. F.; Vancho, L. A.; Okuda, I.; Miller, C. T. Environ. Sci. Technol. 1995, 29 (8), 1966-1976. (4) Shiau, B.-J.; Sabatini, D. A.; Harwell, J. H. Ground Water 1994, 32 (4), 561-569. (5) Peters, C. A.; Luthy, R. G. Environ. Sci. Technol. 1994, 28 (7), 1331-1340. (6) Pennell, K. D.; Abriola, L. M.; Weber, W. J. Environ. Sci. Technol. 1993, 27, 2332-2340. (7) Mason, A. R.; Kueper, B. H. Environ. Sci. Technol. 1996, 30 (11), 3205-3215. (8) Sabatini, D.; Knox, R.; Harwell, J. USEPA\600\S-96\002, 1996. (9) Roy, S. B.; Dzombak, D. A.; Ali, M. A. Water Environ. Res. 1995, 67 (4), 4-15. (10) Brandes, D.; Farley, K. J. Water Environ. Res. 1993, 65 (7), 869878. (11) Abdul, A. S.; Gibson, T. L.; Ang, C. C.; Smith, J. C.; Sobczynski, R. E. Ground Water 1992, 30 (2), 219-231. (12) McCray, J. E.; Brusseau, M. L. Environ. Sci. Technol. 1998, 32 (9), 1285-1293. (13) Knox, R. C.; Sabatini, D. A.; Shiau, B.; Harwell, J. H. Proceedings of the 213th American Chemical Society National Meeting; American Chemical Society: Washington, DC, 1997. (14) Rao, P. S. C.; Annable, M. D.; Sillan, R. K.; Dai, D.; Hatfield, K.; Graham, W. D.; Wood, A. L.; Enfield, C. G. Water Resour. Res. 1997, 33 (12), 2673-2686. (15) Brusseau, M. L.; McCray, J. E.; Johnson, G. R.; Wang, X. J.; Enfield, C.; Wood, A. L. In Field Testing of Innovative Subsurface Remediation and Characterization Technologies; Brusseau, M.

(16) (17) (18) (19) (20) (21)

(22)

(23) (24)

(25) (26)

L., Sabatini, D., Gierke, J. S., Annable, M., Eds.; American Chemical Society: Washington, DC, 1998. Jawitz, J. W.; Annable, M. D.; Rao, P. S. C.; Rhue, D. Environ. Sci. Technol. 1998, 32 (4), 523-530. Mercer, J. W.; Cohen, R. M. J. Contam. Hydrol. 1990, 6, 107163. Mackay, D.; Shiu, W. Y.; Maijanen, A.; Feenstra, S. J. Contam. Hydrol. 1991, 8, 23-42. Lee, L. S.; Hagwell, M.; Delfino, J.; Rao, P. S. Environ. Sci. Technol. 1992, 26 (11), 2104-2109. Bordon, R. C.; Kao, C. M. Water Environ. Res. 1992, 64 (1), 2836. National Research Council (NRC). Innovations in Ground Water and Soil Cleanup; National Academy Press: Washington, DC, 1997. McCray, J. E.; Bryan, K.; Cain, R. B.; Blanford, W.; Johnson, G.; Brusseau, M. L. In Field Testing of Innovative Subsurface Remediation and Characterization Technologies; Brusseau, M. L., Sabatini, D., Gierke, J. S., Annable, M., Eds.; American Chemical Society: Washington, DC, 1998. Burris, D. R.; MacIntyre, W. G. Environ. Toxicol. Chem. 1985, 4, 371-377. Schwarzenbach, R. P.; Gschwend, P. M.; Imboden, D. M. Environmental Organic Chemistry; Wiley-Interscience: New York, 1993. Banerjee, S. Environ. Sci. Technol. 1984, 18 (8), 587-591. Cline, P. V.; Delfino, J. J.; Rao, P. S. C. Environ. Sci. Technol. 1991, 25 (5), 914-920.

(27) Lee, L. S.; Rao, P. S. C.; Okuda, I. Environ. Sci. Technol. 1992, 26 (11), 2110-2115. (28) Bender, M. L.; Komiyama, M. Cyclodextrin Chemistry; SpringerVerlag: New York, 1978. (29) Wang, X.; Brusseau, M. L. Environ. Sci. Technol. 1993, 27 (12), 2821-2825. (30) Bizzigotti, G. O.; Reynolds, D. A.; Kueper, B. H. Environ. Sci. Technol. 1997, 31 (2), 472-478. (31) Brusseau, M. L.; Nelson, N. T.; Cain, R. B. In Field Testing of Innovative Subsurface Remediation and Characterization Technologies; Brusseau, M. L., Sabatini, D., Gierke, J. S., Annable, M., Eds.; American Chemical Society: Washington, DC, 1998. (32) McCray, J. E. Ph.D. Dissertation, The University of Arizona, 1998. (33) Karickhoff, S. W.; Brown, D. S.; Scott, T. A. Water Res. 1979, 13, 241-248. (34) Wiedemeier, T.; Wilson, J. T.; Kampbell, D. H.; Miller, R. N.; Hansen, J. E. Technical report prepared for the Air Force Center for Environmental Excellence, San Antonio, TX, 1995. (35) Jury, W. A.; Gardner, W. R.; Gardner, W. H. Soil Physics, 5th ed., John Wiley and Sons: New York, 1991. (36) Priddle, M. W.; MacQuarrie, K. T. B. J. Contam. Hydrol. 1994, 15, 27-56. (37) Mongtomery, J.; Welkom, L. Ground Water Chemicals Desk Reference; Lewis Pub.: Chelsea, MI, 1989.

Received for review February 4, 1998. Revised manuscript received June 11, 1998. Accepted October 14, 1998. ES980117B

VOL. 33, NO. 1, 1999 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

95