Hydrological Sciences -Journal- des Sciences Hydrologiques,3S,4, 8/1993 309

Natural degradation rates of BTEX compounds and naphthalene in a sulphate reducing groundwater environment*

JOSEPH THIERRIN Centre d'Hydrogéologie, Université de Neuchâtel, Rue E.-Argand II, CH-2007 Neuchâtel, Switzerland

GREGORY B. DAVIS, CHRIS BARBER, BRADLEY M. PATTERSON, FRIDERIK PRIBAC, TERRY R. POWER & MICHAEL LAMBERT CSIRO, Division of Water Resources, Private Bag, PO Wembley, Western Australia 6014

Abstract Field and laboratory evidence show natural degradation of toluene, ethylbenzene, m-, p- and o-xylenes, 1,3,5-trimethylbenzeneand naphthalene in sulphate reducing groundwater conditions of the Bassendean Sands in the Perth basin, Western Australia. Natural degradation rates were obtained from a groundwater tracer test with deuterated organic compounds injected into a dissolved hydrocarbon plume, down-gradient of a leaking underground storage tank at an urban service station. These were compared with similar data obtained from modelling of the whole contaminant plume itself and also with data obtained from large-scale laboratory column experiments with ground­water spiked with BTEX compounds. Toluene degradation rate was 200 to 500 times higher in the anaerobic laboratory columns than in the field. Degradation rates in the tracer test compared well with model-derived field estimates.

Taux de dégradation naturelle des composés BTEX et de la naphtalène dans un environnement sulfo-réducteur dans l'eau souterraine Résumé Des essais en laboratoire et sur le terrain ont démontré la dégradation des toluène, benzène d'éthyle, m- p- et o-xylènes, benzène de 1,3,5-triméthyle et naphtalène dans l'eau souterraine en condition sulfo-réductrice, dans les sables du Bassendean, bassin de Perth, Australie occidentale. Les taux naturels de biodégradation ont été obtenus à partir d'essais de traçage avec des composés organiques deutérés, injectés dans le panache d'hydrocarbures dissous, en aval de fuites d'un réservoir souterrain de station service en milieu urbain. Ces résultats ont été comparés avec des données similaires obtenues par modélisation de tout le panache de contamination ainsi qu'avec des données obtenues à partir d'expériences sur des colonnes à grande échelle en laboratoire, avec de l'eau souterraine dosée avec les composés BTEX. Le benzène n'a été que peu dégradé, autant sur le terrain qu'en laboratoire, en

*Paper presented at theIn-Situ Bioremediation Symposium '92, Niagara-on-the-Lake, Ontario, Canada, 20-24 September 1992. Open for discussion until 1 February 1994

310 Joseph Thierrin et al.

condition anaérobie. Le taux de dégradation du toluène était de 200 à 500 fois supérieur dans les colonnes anaérobies en laboratoire, que sur le terrain. Les taux de dégradation observés lors de l'essai de traçage se sont avérés comparables aux taux estimés par modélisation pour le terrain.

INTRODUCTION

Assessing the natural degradation of BTEX compounds (benzene, toluene, ethylbenzene and xylenes) in anoxic groundwater is of great interest for under­standing natural processes and developing bioremediation techniques. Degrada­tion of some of these compounds has been described for denitrifying conditions (Zeyer et al., 1986; Major et al, 1988; Hutchins étf a/., 1991a; 1991b) for iron reducing conditions (Lovley & Lonergan, 1990) and for methanogenic condi­tions (GrbiC-Galid & Vogel, 1987). Few data have yet been reported for sulphate reducing environments, although Edwards et al. (1991; 1992) have described toluene degradation in batch experiments under sulphate reducing conditions. Fewer studies still have reported estimates of degradation at a range of scales, which are reported here.

Data described here were observed in groundwater of the Bassendean Sands which form the main Quaternary aquifer of the Swan Coastal Plain in Perth, Western Australia (Fig. 1). In these sands, inorganic fertiliser amend­ments and septage are mainly responsible for relatively high (20-100 mg l"1) sulphate contents (Gerritse et al., 1990). On the other hand, degradation of organic matter in the saturated zone rapidly consumes molecular oxygen as well as nitrate (when surface amendments are not too high) and sulphate reducing conditions often occur.

INVESTIGATIONS

Natural degradation likely to be due to biological activity has been demonstrated with the following three methods: (a) large-scale laboratory column experiments with groundwater spiked with

BTEX compounds; (b) a groundwater tracer test with deuterated organic compounds within a

contaminated zone; and (c) mathematical modelling and mass balance calculations on the basis of an

extended assessment of a contaminant plume, within which the tracer test was conducted.

LARGE SCALE LABORATORY COLUMN EXPERIMENTS

Large scale column experiments were performed at the CSIRO Division of

Degradation rates ofBTEX compounds in sulphate reducing groundwater 311

Water Resources in the Perth Laboratories. They were used to try to simulate in a physical model the different groundwater conditions on the Perth coastal plain in order to assess the natural degradation potential of BTEX and PCE compounds in aerobic and anaerobic groundwater (Barber et al., 1991).

These experiments and column settings are described in Patterson et al. (1993). The 150 mm diameter and 2.0 m long columns were fitted with eight sampling ports. The first port sampled influent groundwater to the column. The other ports were positioned 130, 330, 530, 730, 1030 and 1630 mm from Port 1. Typical characteristics of groundwater used for this experiment are given in Table 1.

Only toluene showed marked concentration decreases and only then after an acclimation period (lag) of 40 days. The decreases occurred during 150 h of residence time of spiked groundwater in the sand column (Fig. 2). Sulphate was also partly removed during the first 100 h residence time in the sand column. Subsequently, sulphate remained constant until discharged from the

Joseph Jhierrin et al.

Benzene vaîues for July 1ï

^ § > loood Ut [ | 1 000 - 10 MO (ig/l

[; ; ; 110- 100qLg/l

fgj Mulîiport mi

m Slotted Bo

/- Water tabl

^ 1 Tracer tes

T NORTH

Fïg. 1(6) Distribution of benzene in groundwater down gradient of a petrol service station, Swan Coastal Plain, Western Australia.

Fig. 1(c) Distribution of toluene in groundwater showing reduced extent of toluene relative to benzene (Fig. 1(b)).

Degradation rates ofBTEX compounds in sulphate reducing groundwater 313

Sulphate ion values for Juîy

| | > 20 mg, [ |10-20

0-10mg/!

(£> Multiport adre

s Slotted Bo E

/" Water tab!'

r j Tracer test £

BBS Open drair

T NORTH

Mg. 2(irf) Distribution of sulphate in groundwater illustrating reduced concentrations within the BTEX plume (Fig. 1(b)).

end of the column (325 h residence time). For this relatively short residence time, any significant degradation should lead to a 20% concentration decrease of toluene at port 8 (two times uncertainty on measurements), which corres­ponds to a degradation rate of 1.2 x 10"7 s"1, or a half life of 42 days (after

Table 1 Mean values for inorganic chemical content and physical data of groundwaters where simultaneous degradation of BTEX compounds and sulphate reduction were observed

Parameter

NH4+ [mg r1]

F e + + [ m g r ' ]

N03--N [mgl4]

SO," -S [mgl"1]

HCO3- [mgl1]

EC [#tS cm1]

P H

Eh [mv]

DO [mg I4]

Tracer test site

Mean

0.12

0.37

0.08

21

37

644

5.9

- 7 2

0.28

Standard deviation

0.10

0.52

0.02

11

23

164

0.2

43

0.30

Contaminant plume

Mean

0.34

1.4

0.09

20.3

47

597

5.7

- 2 4

0.5

Standard deviation

0.44

2.15

0.4

9.7

31

113

0.3

83

0.4

Columns

Mean

-0.65

0.2

14

0

498

5.0

-0.3

Number of samples 25 110

314 Joseph Thierrin et al.

acclimation). No significant degradation of benzene, ethylbenzene, p-xyleneor o-xylene was detected. Therefore, half-lives of degradation for these species are listed in Table 4 as " > 42 days".

Figure 3 shows that the simultaneous degradation of toluene and sulphate occurred after an approximate lag period of 40 days. After 13 and 31 days since the experiment began, no significant degradation (> 200 fig Ï1 decrease) of toluene took place between Ports 1 and 2, from an initial concentration of 1000 fig l"1 at Port 1. After this lag period, measurements at 57, 92, 131, 184 and 230 days after the experiment began showed that the amount of toluene and sulphate removed between Ports 1 and 2 is proportional, with twice as much sulphate-S removed as toluene. The degradation rate for toluene, listed in Table 4 was calculated by numerical simulation with a 1-D solute transport code.

Benzene Toluene Sulfate

100 150 200 250 300 350 Residence time [hours] of water in the column

Data points are mean values of measurements at 57, 92,131,184 and 230 days after experiment began, for ports 1 to 8 of the column corresponding to increasing residence time of the spiked water in the sand column. Mean initial concentrations (Co) were 1710 ± 460 //g I 1 , 970 ± 250//g I 1 and 8.85 ± 1.5 mg I 1 for benzene, toluene and sulphate-S.

Fig. 2 Relative evolution of benzene, toluene and sulphate in groundwater in large scale column ofBassendean Sand.

GROUNDWATER TRACER TEST

The natural gradient groundwater tracer test experiment is described in detail in Thierrin et al. (1992). A volume of 400 1 of groundwater from the contami­nant plume described below (see also Fig. 1) was pumped out of a single bore­hole and spiked with fully deuterated benzene (2 g), toluene (2 g), p-xylene (2 g) and naphthalene (0.5 g) as well as bromide (120 g KBr) as a non-reactive tracer. Average groundwater chemistry at the tracer test site is given in

Degradation rates ofBTEX compounds in sulphate reducing groundwater 315

=• 1200

0-i • 1 1 1 1 i — i 1 1 1 (-0.0

0 50 100 150 200 250

Time [days]

Lower values of Co-C for toluene at days 180 and 230 are related to lower initial concentrations (Co) of toluene.

Fig. 3 Concentration decrease of toluene and sulphate in ground­water between Ports 1 (Co) and 2 (C) in a column ofBassendean Sand.

Table 1. This water was then injected in the same borehole inside the contami­nant plume. Care was taken that during the whole pump, mix and injection process, groundwater and solutions were kept under nitrogen atmosphere. Tracer displacement was principally monitored 1 m down hydraulic-gradient from the injection point with one multiport borehole and at 17 m with a line of 7 stainless steel multiport boreholes placed 0.5 m apart across the flow direction and with vertical port spacing of 0.25 m (bores L-R, Fig. 4). The deuterated compounds were chosen because of their relatively low cost, the stability of the C-D bonds and the ease to extract and analyse in the same run as for the non-deuterated species. They were of highest purity grade (99.6, 99.6, 99.0 and 99.0% for benzene-d6, toluene-d8, p-xylene-dl0 and naphtha-lene-d8 respectively) and were purchased from MSD Isotopes, Montreal, Canada.

Mass balance of the organic tracers was calculated from breakthrough (Fig. 5) at 17 m down-gradient of the injection point. Calculated recovery percentages were respectively 68, 69, 48, 56 and 15% for bromide, benzene-d6, toluene-d8, p-xylene-dl0 and naphthalene-d8. Except for benzene which had a similar recovery percentage to bromide, a significant loss of the deuterated organic tracers compared to bromide was observed.

First order degradation rates were calculated for each port depth from breakthrough in bores O, P and Q (Thierrin et al., 1992). Data listed in Table 4 correspond to average values of half-lives of anoxic degradation.

vu

in TZ O)

F O o o

•1.5 £

3

316 Joseph Thierrin et al.

Observation row at 5 m. 0.5 m distant bores with 12 0.25 m (C, D, E) and 0.5 m (B) port spacing

injection bore^ Close survey multiport bore A

Observed flow direction 152 degrees

Observation row at 17 m. 0.5 m distant bores with 12 0.25 m port spacing

EH4 o

Observation row at 10 m. 0.7 m distant bores with 11 0.5 m port spacing

Expected flow line accorded to the main direction of the contamination Angle 146.75 degrees

Expected flow direction for tracer lest, according to a previous tracer test: 134.0 degrees

o MP2

o Multiport bore with 0.5 m vertical port spacing o Multiport bore with 0.25 m vertical port spacing

t- Multiport bore number for the tracer test MP2, EH4, MP3: Bores for contaminant plume observation

5 m NORTH

Section (projection)

Ground level 20.7 m

20 m

Groundwater table

15 m

Bassendean sand

Impervious clay

10 m

21.29

20 m

• 15m

*• 10m

Fig. 4 Layout of the natural gradient groundwater tracer test, 80 m down-gradient of the leaking underground storage tank at an urban service station (location on Fig. 1).

Degradation rates ofBTEX compounds in sulphate reducing groundwater 317

so

UO|]BJJU8OUO0 UOI1EJJU03U0O

UOIJBJJU30UO0 uojienueouoo s k,

318 Joseph Thierrin et al.

ASSESSMENT OF CONTAMINANT PLUME AND MATHEMATICAL MODELLING

Barber etal. (1991) describe the groundwater plume contaminated with BTEX compounds down-gradient of a leaking underground storage tank at a petrol station (Fig. 1). At this site, the aquifer consists of fine to medium aeolian sand. Its hydrogeological characteristics are summarized in Table 2. Assess­ment of hydraulic conductivity from boreholes throughout the contaminant plume (Fig. 1) shows a relatively homogeneous sand with hydraulic conduc­tivity K = 1.0 to 3.2 x 1G4 m s"1 and the presence of some less conductive lenses (K = 0.1 to 1.0 x 10~4 m s"1), particularly in the zone of water table fluctuation where precipitation of iron hydroxides takes place.

Table 2 Characteristics of the water-table aquifer at the tracer test site

Thickness Approx 6 m

Hydraulic conductivity 2.7 to 3.0 x 10 4 m s '

Groundwater velocity 140 + 20 m year"1

Mean specific yield 0.28 + 0.02

Longitudinal dispersivity 0.026 + 0.001 m

Transverse dispersivity 0.0034 + 0.0006 m

Organic carbon content 0.08 to 0.6%

Assessment of the contamination was conducted with 34 slotted boreholes and 11 stainless steel multiport boreholes each with 12 ports at 0.5 m vertical spacing (Fig. 1). Sampling of the multiports was carried out in April, May, July, September and December 1991 (Tables 1 and 3) giving an indication of the 3-D extent and time evolution of the contamination. Table 3 shows the average of highest contaminant concentrations from 5 samplings between April and Decem­ber 1991 at multiport boreholes along the centre of the contaminant plume. These values are part of the data used for computation of the degradation rates.

Table 3 Average of highest contaminant concentrations from 5 samplings between April and December 1991

Compound

Benzene (jig Vs)

Toluene (jxg 1"')

Ethylbenzene (p,g l"1)

m- & p-Xylene (jig V1)

o-Xylene (jig l"1)

1,3,5-Trimethylbenzene

Naphthalene (jig l"1) 0*gl

Distance"

20 m

31200

62 400

16 500

18 000

7 600

') 520

350

80 m

19 500

19 000

1 800

4 900

2 400

204

140

170 m

10 300

1 980

890

2 800

950

110

40

260 m

8 800

1 200

490

1 350

450

100

40

420 m

6 500

0

250

450

0

0

0

a: Distance from the source along the centre of the contaminant plume.

Degradation rates ofBTEX compounds in sulphate reducing groundwater 319

Mathematical modelling was performed with the analytical transport model SOLUTE (Sellmeijer, 1990) which allows advection, dispersion, diffusion, retardation and first-order degradation computation. For this model, groundwater flow velocity (150 m year"1), effective porosity (0.28) and aquifer thickness (6 m) were kept constant. These data are the mean values determined by hydrogeological assessment of the area and two natural gradient ground­water tracer tests. Retardation coefficients due to sorption for benzene (1.02), toluene (1.05), ethylbenzene (1.4), m- o- & p-xylenes (1.12), 1,3,5-trimethyl-benzene (1.4) and naphthalene (1.32) were calculated from the tracer test with deuterated organic compounds (Thierrin et al., 1992).

Calibration of the model was carried out for benzene (assumed conserva­tive) by adapting longitudinal, transverse and vertical dispersivities as well as source width and rate. For model calculations, it was assumed that input from the leaking underground storage tank was constant in time. Input rate was calcu­lated from mean concentrations (11 400 jtg F1) and plume thickness (1.35 m) in two multiport bores near the service station (Fig. 1) and an assumed source width of 12 + 2 m. Approximate input load estimated with these data and mean aquifer characteristics (Table 2) was 7.7 ± 1 kg year"1. After calibration of the transport model on benzene data, optimal values were obtained for total load (8.3 kg year"1), source width (10 m), first-order degradation rate for benzene (lower than 5 x 10"9 s"1), longitudinal dispersivity (0.08 m), transverse disper-sivity (0.015 m) and vertical dispersivity (0.0015 m). Transverse and longitu­dinal dispersivities for the plume were 3 and 4.3 times higher than the data calculated for the tracer test. The difference in dispersivity estimates may be due to scale-dependence (Robbins & Domenico, 1985) and/or groundwater flow direction changes with time (Thierrin et al., 1992). Model calculations suggest that the leakage from die source has continued for a period of 4 years to develop the current plume with a constant groundwater velocity of 150 m year"1. They also show that groundwater contamination by benzene at concentrations higher than 500 /ig l"1 could reach a distance of 3.7 km.

Degradation rates for toluene, ethylbenzene, m- and p-xylenes, o-xylene, 1,3,5-trimethylbenzene and naphthalene (Table 4) were estimated using the model with relevant input rates (14.1,4.65, 13.7, 6.65, 1.0 and 0.41 kg year"1, respectively), estimated from mean values in the two multiport bores closest to the source.

DISCUSSION

Table 4 summarizes the results obtained by the three investigation methods on natural degradation of dissolved aromatic compounds in anoxic groundwater where sulphate reduction occurs.

Good agreement between tracer test and plume modelling was found for toluene as well as for p-xylene (tracer test) and m- and p-xylene (model). In

320 Joseph Thierrin et al.

Table 4 Half-lives (in days) for first order degradation rates of gasoline compounds in anoxic groundwater in Bassendean Sands where sulphate reduction takes place

Parameter

Benzene

Toluene

Ethylbenzene

p-xylene

m- & p-xylene

o-xylene

1,3,5-trimethylbenzene

Naphthalene

Method

Tracer test*

no (> 800)

100 + 40

-(d)

225 ± 75

---33 + 6

Plumeb

no (> 800)

120 + 25

230 ± 30

-170 ± 10

125 ± 10

180

160 ± 20

Columns0

no (> 42)

0.3 + 0.1

no (> 42)

no (> 42)

-n o ( > 42)

--

(a) values of deuterated compounds from the tracer test; (b) preliminary data obtained after modelling the whole contaminant plume; (c) data for degradation within the laboratory columns (because of the maximum residence time of 325 h for water in the column and +10% maximum error on organic measurements, the lowest first order degradation rate which is possible to determine with this experiment is 1.9 X 10"7 s"', corresponding to a half-life of 42 days; and (d) no: no degradation was observed; - the compound was not investigated.

contrast to observations made by Hutchins et al. (1991b) and Edwards et al. (1992), o-xylene showed quicker degradation in the plume than the other xylene isomers. Naphthalene seemed to degrade more rapidly during the tracer test than in the contaminant plume. However deuterated naphthalene input for the tracer test was probably too low and values measured at 17 m distance were too close to the detection limit of the method, so this estimate is based on limited data. Stronger degradation of naphthalene could also have happened at the tracer test site which was slightly off the centre of the contaminant plume, where sulphate is present in relatively high amounts. No significant degradation of benzene was observed under anaerobic conditions. Ethylbenzene and 1,3,5-trimethylbenzene showed slow but significant degradation in the contaminant plume.

Toluene degradation in the column experiment happened after a lag period of approximately 30 to 40 days. Its rate was between 200 and 500 times higher than in the contaminant plume. This is possibly the result of a concen­tration effect. Initial concentrations in the contaminant plume at the service station were between 30 000 and 150 000 jig l"1 and in the column only 1000 ng T1. Another possibility is that the bacterial population acclimated to toluene degradation had better growth conditions in the column than in the field (due to trace oxygen content of the water and absence of inhibitory effects from reaction products such as H2S). It is emphasised that similar groundwater and contaminant conditions prevailed for the tracer test and the contaminant plume with high contaminant concentrations (Table 3). In contrast, the column experi­ment used uncontaminated water spiked with BTEX compounds at low initial concentrations (1000 ng I'1) and therefore reflects different conditions.

Degradation rates listed in Table 4 are useful to compare relative

Degradation rates ofBTEX compounds in sulphate reducing groundwater 321

degradability of different hydrocarbon compounds in the same environment and are a help for pollution management. From available data, it is not possible to determine how the degradation occurs, and this is to be investigated in further studies at the test site.

CONCLUSIONS

The studies reported show that toluene, ethylbenzene, m- and p-xylenes, o-xylene, 1,3,5-trimethylbenzene and naphthalene degradation occurs at different rates, simultaneously with sulphate reduction in groundwater. The use of deuterated compounds for in-plume investigations provided good field estimates for natural degradation of aromatic hydrocarbons. This work adds to the list of contributions (Acton et al., 1989; Edwards et al., 1992; Hutchins et al., 1991a; 1991b; Lemon et al., 1989) showing that benzene is persistent under anoxic groundwater conditions. The results suggest that pilot field-scale estimates of degradation rates were close to those obtained at field scale, but too great a reliance on small-scale laboratory tests should be questioned. They also indicate that even small leakages of fuel, such as that at the test site can have a significant impact on groundwater.

Acknowledgments The authors acknowledge the Water Authority of Western Australia, the Australian Institute of Petroleum and the Swiss National Science Foundation for funding this research. Thanks are given to Alison Wells who contributed to this research.

REFERENCES

Acton, D. W.j Barker, J. F. &Mayfield, C. I. (1989) Enhanced in situ biodégradation of aromatic and chlorinated aliphatic compounds in a leachate-impactedaquifer. In: Proc. NWWA 3rd National Outdoor Action Conf. and Exposition, Orlando, Florida, USA, 22-25 May 1989, 535-549.

Barber, C , Davis, G. B., Thierrin, J., Bates, L., Patterson, B. M., Pribac, F., Gibbs, R., Power, T., Briegel, D., Lambert, M. & Hosking, J. (1991) Final report forproiect on assessment of the impact of pollutants on groundwater Beneath urban areas. Report 91 /22 CSIRO, Div. of Water Resources, Perth, Western Australia.

Edwards, E. A., Wills, L. E., Grbic-Galic, D. & Reinhard, M. (1991) Anaerobic degradation of toluene and xylene: evidence for sulfate as the terminal electron acceptor. In: In Situ Biorecla-mation. Applications and Investigations for Hydrocarbon and Contaminated Site Remediation. ed. R. E. Hinchee& R. F. Olfenbuttel, Butterworth-Heinemann, San Diego, California, USA, 463-471.

Edwards, E. A., Wills, L. E., Reinhard,M. & Grbic-Galic, D. (1992) Anaerobicdegradationof toluene and xylene by aquifer microorganisms under sulfate-reducing conditions. Appl. Environ. Microbiol. 58(3), 794-800.

Gerritse, R. G., Barber, C. & Adeney, J. A. (1990) The impact of residential urban areas on ground­water quality: Swan Coastal Plain, Western Australia. CSIRO Water Resources Series 3, 26.

Grbic-Galic, D. & Vogel.T. M. (1987) Transformation of toluene and benzeneby mixed methanogenic cultures. Appl. Environ. Microbiol. 53(2), 254-260.

Hutchins, S. R., Downs, W. C , Wilson, J. T., Smith, G. B., Kovacs, D. A., Fine, D. D., Douglas, R. H. &Hendrix, D. J. (1991a) Effectsof nitrate addition on biorestoration of fuel-contaminated aquifer: Field demonstration. Ground Water 29(4), 571-580.

Hutchins, S. R., Sewell, G. W., Kovacs, D. A. & Smith, G. A. (1991b) Biodégradation of aromatic hydrocarbons by aquifer microorganisms under denitrifying conditions.environ. Sci. Technol.

322 Joseph Thierrin et al.

25(1), 68-76. Lemon, L. A., Barbaro, J. R. & Barker, J. F. (1989) Biotransformation of BTX under anaerobic

denitrifying conditions: Elevation of field observations. In: Focus Conference on Eastern Regional Groundwater Issues. 17-19 Oct. 1989, Kitchener, National Water Well Association, Ontario, Canada. 213-227.

Lovley, D. R. & Lonergan, D. J. (1990) Anaerobic oxidation of toluene, phenol and p-cresol by the dissimilatory iron-reducing organism, GS-l5.Appl. Environ. Microbiol. 56, 1858-1864.

Major, D. W., Mayfield, C. I. & Barker, J. F. (1988) Biotransformation of benzene by denitrification in aquifer sand. Ground Water 26(1), 8-14.

Patterson, B. M., Pribac, F., Barber, C , Davis, G. B. & Gibbs, R. (1993) Biodégradation of aqueous soluble organic compounds using large scale columns./. Contamin. Hydrol. (in press).

Robbins, G. A. & Domenico, P. A. (1985) Determining dispersion coefficients and sources of model-dependent scaling. J. Hydrol. 75, 195-211.

Sellmeijer, J. B. (1990) SOLUTE, analytical transport model. October 1990,ReportNoSE-702161(in Dutch) Delft Geotechnics, Netherlands.

Thierrin, J., Davis, G. B., Barber, C. & Power, T. R. (1992) Use of deuterated organic compounds as groundwater tracers for determination of natural degradation rates within a contaminated zone. In: 6th International Conference on Water Tracing, ed W. Hôtzel & B. Reichert, Karlsruhe, Germany; Balkema, Rotterdam.

Zeyer, J., Kuhn, E. P. & Schwarzenbach, R. P. (1986) Rapid microbial mineralisation of toluene and l,3-dimethylbenzeneintheabsenceofmolecularoxygen.,4/>/>/. Environ. Microbiol. 52(4), 944-