A ground-water tracer test with deuterated compounds for monitoring in situ biodegradation and retardation of aromatic hydrocarbons

A ground-water tracer test with deuterated compounds for monitoring in situ biodegradation and retardation of aromatic hydrocarbons

Thierrin, Joseph

Introduction

Development of ground-water cleanup technologies such as in situ bioremediation, requires methods for understanding the behavior of contaminants in the underground environment. Usually laboratory batch or column experiments are undertaken in order to determine biodegradation rates and sorption isotherms of contaminants (Hutchins et al., 1991). However, field data on natural degradation and transport rates of organic contaminants are often needed for reliable plume modeling and pollution management. Such data can be obtained with in situ column systems (e.g., Acton et al., 1989), by tracer test experiments in uncontaminated aquifers (e.g., Patrick and Barker, 1985; Roberts et al., 1986; Lemon et al., 1989), or by assessment of the relative extent of the chemicals in a polluted zone (e.g., Schwarzenbach et al., 1983).

The main objective of this study was to quantify intrinsic in situ biodegradation rates of selected aromatic hydrocarbons using deuterium labeled benzene, toluene, p-xylene, and naphthalene as pollution tracers within a contaminant plume. It also allowed the determination of local aquifer parameters and retardation coefficients for the organic tracers. This natural gradient ground-water tracer test was part of a larger CSIRO (Commonwealth Scientific and Industrial Research Organization) research program on local ground-water contamination management and on characterization of organic contaminants behavior in ground water where anoxic conditions prevail. As the test was performed within a contaminated zone where local microorganisms were already adapted to the contaminated environment, it was assumed that biodegradation of the introduced deuterium-labeled compounds would begin without any lag period.

Site and Contaminant Plume Description

Location and Hydrogeology

The test was performed near a petrol service station in the Perth metropolitan area, Western Australia, where a four-years-long leakage of gasoline from an underground storage tank has led to ground-water contamination by soluble aromatic hydrocarbons, principally BTEX compounds (Figure 1). (Figure 1 omitted) The hydrocarbons contaminate ground water in a marginal zone of the Bassendean sands (Pleistocene), an important aquifer formation, which serves as a source of drinking water in the Perth region. On the test site, the aquifer material consists of medium to fine dune sands, mainly composed of quartz and containing between 0.08 and 0.6% organic carbon. The sand aquifer is 7 to 12 m thick with a 6 m thick saturated zone and overlies a thick (> 4 m) aquitard clay layer of the Guildford formation. The local ground water flows towards the southeast with a mean gradient of 0.0039 m/m. Hydraulic conductivities range from 2.1 to 3.4 x 10 sup -4 m/s with a mean of 3.1 x 10 sup -4 m/s (Barber et al., 1991).

Ground-Water and Contaminant Plume Characteristics

The main ground-water characteristics inside and outside of the contaminant plume are listed in Table 1. (Table 1 omitted) Regionally and also on the test site, ground water has a very low dissolved oxygen content(

The contaminant plume extends beyond 420 m, downgradient of the point source, with a thickness of 0.5 to 3 m and a width of 20 to 40 m (Figures 1 and 2). (Figure 2 omitted) Major dissolved pollutants are benzene, toluene, ethylbenzene, the xylene isomers, trimethylbenzenes, and naphthalene (Table 1). Figure 2 represents vertical sections of benzene, toluene, and sulfate concentrations along the center of the contaminant plume as of April 1991. It shows that toluene is below detection levels in ground water beyond 250 m, and a zone of depleted sulfate concentrations appears throughout the contaminant plume. Compared to benzene, the concentrations of all other observed contaminants strongly decrease as distance from the source increases (Figure 3). (Figure 3 omitted) Retardation and dilution are responsible for only a small part of this relative decrease. Thus, this figure can be seen as a relative degradation measure of the hydrocarbons (Davis et al., 1994). These data suggest that reductive reactions (low Eh) inside the contaminant plume lead to a production of C02 (reflected in (Characters omitted)) and H sub 2 S, as well as a consumption of (Characters omitted) O sub 2 , Fe sup ++ and (Characters omitted). Sulfate is thought to be used as the main electron acceptor during the. biodegradation of toluene, naphthalene, and the xylene isomers (Davis et al., 1994). However, at the contaminant source (service station), Fe sup ++ is produced in an iron-reductive zone, with concentrations up to 23.6 mg/l (Table 1). It then disappears further downgradient in the contaminant plume as the presence of H sub 2 S begins to be detected.

More detailed description of the site and the contaminant plume can be found in Barber et al. (1991) and Davis et al.(1994).

Materials and Methods

Deuterated Compounds

Fully deuterated benzene (benzene-d6), toluene (toluene-d8), p-xylene (p-xylene-d10), and naphthalene (naphthalene-d8) were chosen because of their affordable cost, the stability of the carbon-deuterium (C-D) bonds and the ease of identification and analysis by GC-MS. These compounds were representative of the major pollutants of the contaminant plume (Table 1). It was assumed that deuterated and nondeuterated species have a very similar behavior concerning solute transport, sorption-desorption, and biodegradation. The highest purity grade compounds (99.6, 99.6, 99.0, and 99.0% for benzene-d6, toluene-d8, p-xylene-d sub 10 , and naphthalene-d8, respectively) were purchased from MSD Isotopes, Montreal, Canada.

To the best of our knowledge there is no previous published work describing the use of deuterated organics as ground-water tracers.

Implementation of the Test

Injection of the tracers took place on July 15, 1991 in a single 40 mm borehole slotted between 1.1 and 2.3 m below the ground-water table (Figure 4). (Figure 4 omitted) This depth coincided with the upper half of the BTEX contaminated plume. The injection borehole was emplaced two months prior to tracer injection to allow stabilization of chemical conditions in the vicinity of the installations.

For preparation of the injection solution, 2 g each of benzene-d6, toluene-d8, and p-xylene-d10, as well as 0.5 g of naphthalene-d8 were mixed together and split into six equal parts which were each combined with 20 g KBr in six 4-liter bottles of oxygen-free sterile water. Each 4-1iter bottle was then kept for two hours in an ultrasonic bath for dissolution of the deuterated organic compounds.

On the test site, two 200-liter steel drums were filled with contaminated water pumped from the borehole which served subsequently for solution injection. Three of the six 4-1iter bottles containing the deuterated organics were siphoned under N sub 2 pressure into each drum and mixed with an electric pump (10 liters/min for one hour) previously placed in each drum. Then the 400 liters of tracer spiked ground water were injected into the same borehole. Constant mixing within the borehole was assured by means of a small electric pump. During pumping, mixing, and injection, a strict nitrogen environment was maintained so that neither water nor solutions came into contact with air.

In order to achieve minimal gravity flow of the injected solution, the concentration of bromide (0.3 g/l KBr) was empirically determined according to the results of previous tracer tests in a similar hydrogeological situation (Thierrin et al., 1992a). Injection solution concentrations were 207 mg/l for bromide and 5.2, 4.62, 3.87, and 1.2 mg/l for deuterated benzene, toluene, p-xylene, and naphthalene, respectively. These concentrations represent the mean values of the injection solution (eight samples taken during injection from the feeding line), and correspond to injected masses (M sub inj ) of 82.8 g, 2.1 g, 1.85 g, 1.55 g, and 0.5 g, respectively.

Monitoring of the Tracer Plume

Migration of the tracer was monitored at 19 locations with stainless steel multiport boreholes located at 1 m distance downgradient of the injection borehole and also along rows at 5, 10, and 17 m distance (Figure 4). Stainless steel tubes were used in order to avoid sorption of the organics onto the walls of the multilevel samplers. Complete breakthrough of the tracers could be observed at the 17-m line, but not at the 5 and 10-m distances because of monitoring difficulties and measurement failures. At 17-m distance, a row of 7 stainless steel multiport samplers placed 0.5 m apart and perpendicular to ground-water flow direction, allowed collection of information in the Y-Z-time field. Most of the multiport bores had 12 ports (2 mm internal diameter and 0.2 m port height) placed at 0.25-m vertical intervals.

Sampling and Analysis

Ground-water samples for organic analysis were taken without air contact into 50-ml glass syringes which fitted into threeways taps connected to each port tube. After discarding the first 200 ml of ground water flushed through the sampling port, a 10-ml sample was transferred into a 20-ml glass tube, then acidified with 30 mu-l 1 N HCl and spiked with 100 mu-l of an aqueous internal standard solution (1.0 mg/l l,2ibromoethane). The organic compounds were extracted in the field with 2 ml of diethylether by vigorous shaking for 2 minutes.

Normal and deuterated organic compounds were analyzed in 1 mu-l of the diethylether extract using a Hewlett-Packard 5890 gas chromatograph with a 5970 mass-selective detector (detection limit of 10 mu-g/l). The BP-1, 25 m x 0.32 mm id, 1.0-mu-m film-thickness capillary column was directly interfaced with the electron-impact ion source (70 eV). Temperatures of the gas chromatograph were: 220degC at the injection port and transfer line, 40degC in the oven for the first 1.5 minutes and then increased to 200degC at 10degC/minute. The data were acquired by selective ion monitoring (SIM) with dwell times of 50 ms.

Br sup – was measured by HPLC (conductivity detector) with a detection limit of 0.1 mg/l.

Computation of Transport and Degradation Rates

Longitudinal and transverse dispersivities as well as ground-water velocities were calculated for each 25-cm depth interval from bromide data by transport simulation of a tracer pulse injection in a 2-D uniform flow field with constant dispersivities (Sauty and Kinzelbach, 1988). The solution for a pulse introduced at time t = 0 and location x, y = 0, 0 is given by:

C(x, y, t) = M sub 0 /4-pi-nh(alpha sub L alpha sub T ) sup 1/2 1/ut exp{(x – ut) sup 2 /4alpha sub L ut – y sup 2 /4alpha sub T ut}exp{-lambda-t} (1)

with C: concentration; x, y: coordinates of measurement point with x-axis aligned parallel to the flow direction; n: effective porosity; h: thickness of aquifer; alpha sub L and alpha sub T : longitudinal and horizontal transverse dispersivities; lambda: decay rate; u: convective velocity of the nonreactive tracer; and M sub 0 : mass of tracer injected (M sub inj ) per unit section at time t = O.

Since complete breakthrough of the tracers could be monitored only at the 17 m distance, a first-order degradation model was applied to best represent mass losses of the organic compounds over time, where

lambda = -1/t x ln(M sub t /M sub 0 ) (2)

with M sup t == M x R: total mass of tracer (in solution and sorbed) flowing through an aquifer section, according to the linear sorption model of equation (4); with R: retardation factor; and M: mass of tracer in the aqueous phase flowing through this aquifer section. M can be computed for the entire plume section as:

(Equation 3 omitted)

with i: horizontal transverse direction (y); j: vertical direction (z); k: time; x: horizontal distance from injection point; Delta-y: transverse influence distance of each port = distance between boreholes (0.5 m); and Delta-z: vertical influence distance of each port = port spacing (0.25 m).

R = Nsol + Nsorb/Nsol == u sub water /u sub tracer (4)

where Nsol and Nsorb are the number of molecules in solution and respectively sorbed onto solid particles and U sub water a U sub bromide .

The recovery fraction f of each compound was determined as:

f = M sub t /M sub inj . (5)

Equations (1) and (2) were applied for each depth interval Delta-z for degradation rates calculations of the deuterated organic tracers, assuming negligible vertical dispersivity. First-order degradation rates can be expressed as h [days sup -1 ] or in a more expressive way as half-life of first-order degradation t sub 1/2 = -ln(O.5)/lambda[days].

Results

Transport of the Tracers and Aquifer Characteristics

Aquifer characteristics (Table 2), retardation factors (Table 3), and mass balance for the organic tracers were calculated from breakthrough data at 17 m downgradient of the injection point [equations (1), (3), and (4)]. (Tables 2 and 3 omitted) Figure 5 represents time-depth concentration contours of the five tracers in the control bore P coinciding with the center of the tracer plume, 17-m downgradient of the injection point. Contour lines were chosen to be 0.5%, 5%, and 50% of the injection solution concentrations. The tracers reached bore P at different times depending on their retardation and the depth variations of the hydraulic conductivities, which vary as a function of depth by a factor of 1.6, from 2.1 10 sup -4 to 3.4 10 sup -4 m/s over the whole thickness of the tracer plume. Bromide and benzene-d6 show very similar breakthrough. Toluene-d6, p-xylene-d10, and naphthalene-d8 undergo progressively more retardation and different degrees of mass losses. Similar retardation of the tracers and vertical variations of hydraulic conductivities were observed in multiports O and Q, situated 0.5 m away and at each side of bore P (data not shown). The slight vertical variations of hydraulic conductivities suggest that a larger apparent dispersivity would be calculated if the tracer breakthrough was monitored in fully slotted rather than in multilevel boreholes.

For mass balance calculations, linear instantaneous and reversible sorption equilibrium was assumed for the deuterated organic tracers. Retardation factors (R) calculated from equation (4) are listed in Table 3. These values appear slightly lower than those determined by Barker et al. (1989) for benzene (1.1), toluene (1.2), and p-xylene (1.4) in the Borden aquifer. Calculated recovery percentages [equation (5)] amount to 68, 69, 48, 56, and 15% for bromide, benzene-d6, toluene-d8, p-xylene-d10, and naphthalene-d8, respectively. In equation (3), an effective porosity of 0.28 +/0.02 was used. This was determined by laboratory column measurements on 32 repacked 1-liter samples from the contaminated site (Thierrin et al., 1992a). Calculated effective porosity for 100% recovery of bromide would be 0.40. This value appears to be high compared to published data for similar environments. Several causes can lead to this difference. We estimated that the observation network and schedule were not tight enough for observing the totality of the tracer breakthrough. For instance, incomplete mass recovery of the tracers may be attributed to the relative poor transverse definition of this narrow tracer plume (approximately 1.75 m width), intercepted in four multiport boreholes “only.” On the other hand, aquifer material at the site may not be optimally packed because the pore space could have been widened by dissolution of carbonated and iron-bearing minerals, since the ground water is undersaturated with respect to CaCO sub 3 and Fe.

Degradation of the Deuterated Compounds

Except for benzene, the deuterated organic tracers lost significant mass compared to bromide. This mass loss can be caused by sorption, dispersion, and degradation. Taking sorption and dispersion into account with equation (1), it is possible to compute the amount of tracer lost through degradation. At this site, biodegradation is thought to play the main role.

Table 4 lists the half-lives of the first-order degradation for each compound at each depth interval. (Table 4 omitted) They were calculated according to equations (1) and (2) with the following M sub 0 values measured in the injection solution and controlled in the close survey multiport bore A, 1 m downgradient of the injection borehole: 82.8 g for bromide and 2.1 g, 1.85 g, 1.55 g, and 0.5 g for deuterated benzene, toluene, p-xylene, and naphthalene, respectively. These data show that in the slightly oxic zone which is also the upper edge of the contaminant plume rapid degradation of toluene, p-xylene, and naphthalene takes place. Benzene also degrades, but much more slowly (half-life of 75 to 160 days). In the anoxic zone, between 15.55 and 16.8 m AHD (Australian Height Data, i.e., altitude over sea level), benzene does not show any indication of degradation. At these depths, its estimated mass recovery was slightly greater than that of bromide. From these calculations, we conclude that toluene, p-xylene, and naphthalene undergo natural degradation with half-lives of 70 to 140, 150 to 300, and 20 to 40 days, respectively, under the field anoxic conditions.

Discussion

To the best of our knowledge, there has been no previous work which reports a field technique capable of determining BTX and naphthalene degradation within a contaminated ground-water plume under natural flow conditions. We document here the transport characteristics and mass losses for deuterium labeled benzene, toluene, p-xylene, and naphthalene as a function of time. Except for naphthalene, the tracer test data are very consistent with figures on natural degradation of the pollutants computed by mathematical modeling of the whole contaminant plume, using benzene as a conservative tracer (Table 5). (Table 5 omitted) Compared to model calculations for the entire contaminant plume, naphthalene mass loss is five to six times more rapid during the tracer test than in the contaminant plume (Thierrin et al., 1992b; Davis et al., 1994). This could be caused by nonlinear sorption mechanisms or by preferred biodegradation of naphthalene at the test site. Nevertheless, results of this study suggest that naphthalene degrades faster and more completely than benzene in anaerobic conditions.

During this test, the degradative potential of the indigenous microorganisms has not been investigated. Nevertheless, the reduction of (Characters omitted) and the production of H sub 2 S and CO sub 2 accompanying the degradation of the contaminants strongly suggest that anaerobic biodegradation in sulfate reducing conditions is the main cause of mass removal of the tracers. Ongoing research also suggests that iron plays an important role in this system as an electron acceptor (Fe sup +++ ) for hydrocarbons biodegradation and as a neutralization agent (Fe sup ++ ) for H sub 2 S which could be toxic for microorganisms involved in this process. Baedecker et al. (1993), Edwards et al. (1992), and Beller et al. (1992) report biodegradation of hydrocarbons and toluene under anoxic and sulfate-reducing conditions in experiments involving soil microcosms and enrichment cultures. They also describe the importance of iron in this process. Edwards et al. (1992) observed preferential biodegradation of the xylene isomers and relative persistence of benzene and ethylbenzene.

Of particular concern during the ground-water tracer test was the release of the organic tracers (deuterated benzene, toluene, p-xylene, and naphthalene) in the environment. Since the test had significance only when performed within an already contaminated aquifer volume, additional hazard caused by the test was strongly diminished. Furthermore, the test required small amounts of deuterium labeled compounds compared to the pollutant load already present. In this case, a total of 6.0 g of deuterated compounds were injected into the contaminant plume. This appears insignificant compared to the total of more than 100 kg of BTEX dissolved in the whole contaminant plume. Locally, the injection slug had a cylindrical volume of approximately 1.2-m height and 1.2-m diameter, forming a section of 1.4 m sup 2 in the plan perpendicular to the groundwater flow. During the total duration of the tracer test (71 days), 6 g of deuterated tracers, and an estimated 185 g of contaminant BTEX flowed through this same 1.4 m sup 2 aquifer section.

In order to avoid loss of isotopically labeled tracers in the environment, forced gradient tracer tests could be conducted with subsequent treatment of the recovered water. This configuration would need less monitoring effort since flow lines would be easier to predict but underground residence times of the tracers would be significantly reduced. In-plume tracing could also be performed with sup 13 C-labeled compounds with potential application to a wider range of contaminant tracers and with the possibility to detect degradation products. As for D-labeled compounds, sup 13 C-labeled organics can be measured and scanned with GC-MS techniques in the same run as the unlabeled compounds.

Summary and Conclusions

Natural degradation rates of gasoline compounds inside a contaminated plume of anoxic ground water were successfully determined by comparing mass losses of deuterated tracers with a conservative tracer (bromide) during a ground-water tracer test. The results confirm degradation rates determined through mathematical modeling of the whole contaminant plume. The main advantage of this method is the ability to monitor natural degradation and retardation of pollutants and to provide important field data for contamination management. In this way, the deuterated compounds are used like internal standards. This method has a great potential for assessment of in situ intrinsic or enhanced bioremediation.

Results suggest that in a plume of BTEX contaminated ground water with sulfate reduction as well as CO sub 2 and H sub 2 S production, toluene, naphthalene, and p-xylene significantly degrade, probably due to bacterial activity. Under these conditions, benzene did not show any significant degradation.

A large variety of organic contaminants can be synthesized as deuterium or sup 13 C labeled compounds. They could be used in analogous ways as presented here for assessment of sorption and degradation within a contaminated underground area.

Acknowledgments

The authors acknowledge the Water Authority of Western Australia, the Swiss National Science Foundation, and the Australian Institute of Petroleum for funding this research.

Thanks are given to Terry R. Power, Bradley M. Patterson, Michael Lambert, and Alison Wells who contributed to this research.

Paul V. Roberts kindly provided helpful comments on earlier drafts of this paper.

References

Acton, D. W., J. F. Barker, and C. I. Mayfield. 1989. Enhanced in situ biodegradation of aromatic and chlorinated aliphatic compounds in a leachate-impacted aquifer. Proc. NWWA 3rd National Outdoor Action Conference and Exposition, Orlando, FL, May 22-25, 1989. pp. 124-136.

Baedecker, M. J., I. M. Cozzarelli, R. P. Eganhouse, D. I. Siegel, and P. C. Benett. 1993. Crude oil in shallow sand and gravel aquifer. 3. Biogeochemical reactions and mass balance modeling in anoxic groundwater. Applied Geochemistry. v. 8, pp. 569-586.

Barber, C., G. B. Davis, J. Thierrin, L. Bates, B. M. Patterson, F. Pribac, R. Gibbs, T. Power, D. Briegel, M. Lambert, and J. Hosking. 1991. Final report for project on “Assessment of the impact of pollutants on groundwater beneath urban areas.” Report No. 91/22, CSIRO, Div. Water Resources, Perth, Australia. 57 pp., 32 fig., 17 tables.

Barker, J. F., G. C. Patrick, E. A. Sudicky, C. I. Mayfield, and R. W. Gillham. 1989. The fate and persistence of aromatic hydrocarbons dissolved in groundwater: Results from controlled field experiments. Pacific Section American Assoc. of Petroleum Geologists, Annual Convention of the National AAPG, Palm Springs, May 10-12, 1989. 14 pp.

Beller, H. R., D. Grbic-Galic, and M. Reinhard. 1992. Microbial degradation of toluene under sulfate-reducing conditions and the influence of iron on the process. Applied and Environmental Microbiology. v. 58, no. 3, pp. 786-793.

Davis, G. B., J. Thierrin, and C. Barber. 1994. The variability and natural degradation of a BTEX plume in anaerobic sulphate-rich groundwater. J. Contaminant Hydrology. (paper submitted).

Edwards, E. A., L. E. Wills, M. Reinhard, and D. Grbic-Galic. 1992. Anaerobic degradation of toluene and xylene by aquifer microorganisms under sulfate-reducing conditions. Applied and Environmental Microbiology. v. 58, no. 3, pp. 794-800.

Hutchins, S. R., G. W. Sewell, D. A. Kovacs, and G. A. Smith. 1991. Biodegradation of aromatic hydrocarbons by aquifer microorganisms under denitrifying conditions. Environ. Sci. Technol. v. 25, no. 1, pp. 68-76.

Lemon, L. A., J. R. Barbaro, and J. F. Barker. 1989. Biotransformation of BTX under anaerobic denitrifying conditions: Evaluation of field observations. Focus Conference on Eastern Regional Groundwater Issues, Oct. 17-19, 1989. Kitchener, Ontario, Canada, National Water Well Assoc. pp. 213-227.

Patrick, G. C., and J. F. Barker. 1985. A natural-gradient tracer study of dissolved benzene, toluene and xylenes in ground water. Proc. of the Second Canadian/American Conference on Hydrogeology: “Hazardous Wastes in Ground Water, A Soluble Dilemma,” Banff, Alberta, Canada, June 25-29, 1985, National Water Well Assoc., Dublin, OH, USA, pp. 141-147.

Roberts, P. V., M. N. Goltz, and D. M. Mackay. 1986. A natural gradient experiment on solute transport in a sand aquifer. 3.–Retardation estimates and mass balances for organic solutes. Water Resources Research. v. 22, no. 13, pp. 2047-2058.

Sauty, J. P. and W. Kinzelbach. 1988. Computer Aided Tracer Test Interpretation Code “CATTI” Version 2.0. Bureau de Recherches Geologiques et Minitres, B.P. 6009, 45060 Orleans, France.

Schwarzenbach, R. P., W. Giger, E. Hoehn, and J. K. Schneider. 1983. Behavior of organic compounds during infiltration of river water to groundwater. Environ. Sci. Technol. v. 17, no. 8, pp. 472-479.

Thierrin, J., G. B. Davis, C. Barber, T. R. Power, B. M. Patterson, and M. Lambert. 1992a. Groundwater tracer tests within a BTEX contaminated zone in order to determine aquifer parameters and natural degradation rates of organic compounds. CSIRO, Div. Water Res. Perth. Report No. 92/5. 14 pp., 7 figures, 14 tables.

Thierrin, J., G. B. Davis, C. Barber, B. M. Patterson, F. Pribac, T.R. Power, and M. Lambert. 1992b. Natural degradation rates of BTEX compounds in a sulfate reducing groundwater environment. Hydrological Sciences Journal. v. 38, no. 4, pp. 309-322.

Copyright Ground Water Publishing Company May 1995

Provided by ProQuest Information and Learning Company. All rights Reserved.