Factors Determining the Attenuation of Chlorinated Aliphatic

Jun 17, 2009 - (18) Kuhn, T.; Hamonts, K.; Dijk, J. A.; Kalka, H.; Stichler, W.;. Springael, D.; Dejonghe, W.; Meckenstock, R. U. Assessment of the in...
6 downloads 0 Views 2MB Size
Environ. Sci. Technol. 2009, 43, 5270–5275

Factors Determining the Attenuation of Chlorinated Aliphatic Hydrocarbons in Eutrophic River Sediment Impacted by Discharging Polluted Groundwater K E L L Y H A M O N T S , †,§ T H O M A S K U H N , ‡,⊥ MIRANDA MAESEN,† JAN BRONDERS,† RICHARD LOOKMAN,† HARALD KALKA,# LUDO DIELS,† RAINER U. MECKENSTOCK,‡ D I R K S P R I N G A E L , § A N D W I N N I E D E J O N G H E * ,† Flemish Institute for Technological Research (VITO), Separation and Conversion Technology, Boeretang 200, 2400 Mol, Belgium, Helmholtz Zentrum Mu ¨ nchen, Institute for Groundwater Ecology, Ingolsta¨dter Landstrasse 1, 85764 Neuherberg, Germany, Umwelt-und Ingenieurtechnik GmbH (UIT), Zum Windkanal 21, 800140 Dresden, Germany, and Catholic University of Leuven, Division Soil and Water Management, Kasteelpark Arenberg 20, 3001 Heverlee, Belgium

Received December 18, 2008. Revised manuscript received April 18, 2009. Accepted May 13, 2009.

This study explored the potential of eutrophic river sediment to attenuate the infiltration of chlorinated aliphatic hydrocarbon (CAH)- polluted groundwater discharging into the Zenne River near Brussels, Belgium. Active CAH biodegradation by reductive dechlorination in the sediment was suggested by a high dechlorination activity in microcosms containing sediment samples and the detection of dechlorination products in sediment pore water. A unique hydrogeochemical evaluation, including a δ2H and δ18O stable isotope approach, allowed to determine the contribution of different abiotic and biotic CAH attenuation processes and to delineate their spatial distribution in the riverbed. Reductive dechlorination of the CAHs seemed to be the most widespread attenuation process, followed by dilution by unpolluted groundwater discharge and by surface water mixing. Although CAHs were never detected in the surface water, 26-28% of the investigated locations in the riverbed did not show CAH attenuation. We conclude that the riverbed sediments can attenuate infiltrating CAHs to a certain extent, but will probably not completely prevent CAHs to discharge from the contaminated groundwater into the Zenne River.

Introduction Chlorinated aliphatic hydrocarbons (CAHs) such as tetrachloroethene (PCE), trichloroethene (TCE), and 1,1,1trichloroethane (1,1,1-TCA) are common groundwater pollutants due to their widespread use as solvents and degreasing agents (1). By discharging into surface waters, CAH plumes can represent a continuous source of diffuse contamination * Corresponding author phone: +3214335176; fax: +3214580523; e-mail: [email protected]. † Flemish Institute for Technological Research (VITO). ‡ Helmholtz Zentrum Mu ¨ nchen. # Umwelt-und Ingenieurtechnik GmbH (UIT). § Catholic University of Leuven. ⊥ Present address: CSIRO Land and Water, Private Bag 2, Glen Osmond SA 5064, Australia. 5270

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 14, 2009

and impose environmental risks (2-6). However, before reaching the surface water, the CAHs have to pass hyporheic zone sediments which form an interface between groundwater and surface water. Hyporheic sediments are storage zones for organic carbon (7) and are often characterized by sharp physical and chemical gradients, enabling a broad spectrum of metabolic processes (8). Due to the high diversity and productivity of organisms (7-9), the nutrient and pollutant flow through those river sediments can be impacted via biotic processes (7, 8), in addition to abiotic processes such as sorption and dilution. In anaerobic eutrophic river sediments, the high organic matter content, in combination with strong reducing conditions and the presence of infiltrating CAHs, could favor CAH dechlorination activity since hydrogen or other compounds, produced from the fermentation of organic substrates, serve as electron donor for halorespiring organisms (10). However, although CAHhalorespiring activity has been detected in anaerobic river sediments (5), information on the relative contribution of biotic and abiotic degradation processes to the attenuation of CAHs in infiltrating groundwater is scarce. In this study, the intrinsic CAH attenuation capacity of anaerobic eutrophic river sediment was investigated in a part of the Belgian Zenne River which receives CAH-polluted groundwater from a nearby aquifer. In contrast to previous studies (2, 5, 6, 11), batch biodegradation tests were combined with hydrogeochemical groundwater, pore water, and surface water analyses, including stable water isotope analyses, to examine which processes determine CAH attenuation in the riverbed. In addition, the spatial distribution of these processes in the riverbed was delineated and their relative contribution to the observed CAH attenuation determined.

Experimental Section Site Description. In an industrial area in Vilvoorde, Belgium, a 1.4 km-wide groundwater plume contaminated with PCE, TCE, and 1,1,1-TCA, originating from several sources, flows to the Zenne River in a northwesterly direction (Supporting Information (SI) Figure S1A). Those pollutants are reductively dechlorinated in the plume and the products of this process, i.e., vinyl chloride (VC), cis-dichloroethene (cis-DCE), 1,1dichloroethane (1,1-DCA), and chloroethane (CA) discharge into the Zenne. This CAH plume reaches the right riverbank while unpolluted groundwater approaches the opposite riverbank (Figure 1 and SI Figure S1A). The Zenne River is dammed in this area at both sides by steel walls, resulting in a vertical discharge of the groundwater into the riverbed (SI Figure S1C). The Zenne receives domestic sewage at various locations, which creates highly eutrophic conditions in the surface water and the riverbed. Bed materials range from fine to coarse grained sand and gravel to silt and slick. Except during extreme precipitation events, when the water level of the Zenne River can exceed the groundwater level, the Zenne drains groundwater from both sides of the river up to 7245 m3/day over the studied 2 km indicated in SI Figure S1A (12). While groundwater in the CAH-polluted aquifer is primarily recharged by precipitation, groundwater recharge in the unpolluted aquifer is supplemented by seepage from the nearby canal “Brussels-Scheldt” (SI Figure S1A). Hydraulic conductivity in the CAH-polluted aquifer is 1-3 m/day, the regional hydraulic gradient ∼0.0015 m/m and the calculated groundwater velocity is in the order of 2-60 m/year (12). Additional hydrogeological characteristics of the test site are provided in S1 of the Supporting Information. Sample Collection. The collection and preservation of groundwater, surface water, pore water, and sediment 10.1021/es8035994 CCC: $40.75

 2009 American Chemical Society

Published on Web 06/17/2009

and stored at 20 °C. Headspace samples of 400 or 250 µL were taken at regular intervals for analyzing CAHs or methane, ethene, and ethane, respectively, as described above. After two headspace analyses, vials were refilled with 650 µL nitrogen gas. The organic carbon content of homogenized sediment sections was calculated as the fraction of dry matter (dm) that was removed at 550 °C, after drying the sediment overnight at 105 °C.

Results and Discussion

FIGURE 1. Site map showing the CAH plume (hatched area) discharging into the Zenne River, pore water sampling locations (asterisks) in the selected test area, and location of monitoring wells (dots) and screen points (diamonds). Arrows indicate flow directions of the CAH plume and Zenne River. samples are reported in detail in section S2 of the Supporting Information. Briefly, sediment was sampled using a 4 cm diameter piston sampler. We took 260 mL pore water samples by vacuum pumping using a 3 cm diameter stainless steel lance equipped with a polyurethane sampling tubing, a tip containing a porous polyethylene filter, and probes for temperature and conductivity measurements. In May 2005, July 2005, and May 2006, pore water samples were collected in a selected test area at two or three different depths in the riverbed (i.e., 20 cm and where possible 60 and 80-120 cm) along three parallel lines of 45 m length, located 2 m from the right riverbank, 2 m from the left riverbank, and in the middle of the river (Figure 1). In total, 175 riverbed pore water samples were taken. Analytical Methods. CAH concentrations in water samples were analyzed using a Thermo Finnigan Trace GC-MS equipped with a DB5-MS column (J&W Scientific), with an analytical accuracy of 10%. Methane, ethene, and ethane concentrations were measured using a Varian GC-FID (CP3800) with a Rt-U plot column (J&W Scientific). Headspace analyses of the microcosms were performed on a Varian GCFID (CP-3800) equipped with a Rt-U plot column for the detection of methane, ethene, and ethane, or a split-splitless injector followed by a Rt-X column (Restek) and a DB-1 column (J&W Scientific) for analysis of CAHs. Detailed information on the GC-analysis procedures is reported in SI S3. Stable hydrogen and oxygen isotope compositions of water samples were determined by isotope ratio monitoring mass spectrometry as described by Faye et al. (13). All isotope values are cited in the δ notation in reference to the ViennaStandard Mean Ocean Water (VSMOW) standard, with an analytical accuracy of 0.15‰ for δ18O and 1.0‰ for δ2H (4 σ level). Concentrations of sulfate, nitrate, and nitrite were analyzed by ion chromatography using a Dionex DX-120 ion chromatograph equipped with a Dionex AS14A column. Analytical data reported in this article are averages from duplicate water samples. Batch Biodegradation Test. Homogenized sediment samples (37.5 g) from 0-20, 20-60, and 60-100 cm depth, and 70 mL groundwater from monitoring well SB-1 (located at 28.2 m from the test area and sampled at 10 m depth, Figure 1) were added to sterilized 160 mL glass vials under a 100% nitrogen atmosphere. Vials were sealed using aluminum crimp caps containing Teflon-lined butyl-rubber septa. Initial CAH concentrations in the microcosms were in situ groundwater concentrations of 800 µg/L VC, 230 µg/L cis-DCE, and 30 µg/L 1,1-DCA. Abiotic controls were set up similarly and inhibited by adding 800 µL formaldehyde (0.42%). All microcosm set-ups were prepared in triplicate

Delineation of the CAH Influx Zones in the Zenne Riverbed. A test area was selected in the river where the highest concentrations of VC, cis-DCE and 1,1-DCA were detected in the inflowing groundwater (SI S4). This test area stretches from reference point post 26 to 45 m upstream (Figure 1). CAH concentrations from pore water samples collected in May 2005, July 2005, and May 2006 all indicated that the CAH plume discharges into the riverbed over the entire length and width of the test area. While VC, 1,1-DCA and CA were detected in the pore water throughout the studied area, the occurrence of cis-DCE was limited to an area between 20 and 45 m upstream of post 26 (Figure 2, SI Figure S3). Concentrations of CAHs were spatially variable but rather constant in time. Around post 26, concentrations ranged from nondetectable up to 81 µg/L VC and 100 µg/L 1,1-DCA (in May 2005), while the highest concentrations were measured between 25 and 45 m upstream of post 26, up to 2600 µg/L VC (in July 2005), 495 µg/L cis-DCE (in May 2006), 169 µg/L 1,1-DCA (in May 2005), and 61 µg/L CA (in May 2006). The variable CAH concentration pattern in the riverbed can be explained by the discharge of a heterogeneous CAH plume over the investigated area (data not shown). The highest CAH concentrations were generally detected in the riverbed near the right riverbank, where the CAH plume reaches the river, while the lowest CAH concentrations were measured in pore water sampled near the left bank of the Zenne, where unpolluted groundwater discharges. Such a discharge of CAH-polluted groundwater across the width of the river has been previously reported (3-6). Conant et al. (5) suggested that the widening of the plume was attributed to layers of low hydraulic conductivity in the riverbed and anisotropy of geological deposits. Despite the discharging CAH plume, CAHs were never detected in the Zenne surface water, probably due to massive dilution. Evaluation of the Microbial CAH Degradation Potential of the River Sediment. Since a spatially heterogeneous CAH plume discharges over the entire length and width of the riverbed, batch degradation tests were performed with sediments from four different locations in the test area and from three different depths in the riverbed. Complete reductive dechlorination of 2 µM cis-DCE and 8-13 µM VC to nontoxic ethene and ethane occurred within 13-46 days in sediment obtained from the top 20 cm of the riverbed, whereas 28-46 days were needed in sediments sampled from 20-60 or 60-100 cm depth (SI Figure S4). Dechlorination of cis-DCE and VC occurred in sediment obtained from locations with low (up to 81 µg/L VC at post 26) and high (up to 1400 µg/L VC 20 m upstream of post 26) in situ pore water concentrations. In addition, reductive dechlorination of 0.19-0.27 µM 1,1-DCA within 13-46 days was observed for 9 of the 12 tested positions in the riverbed (data not shown). CA that was produced from 1,1-DCA, was subsequently converted into ethane (data not shown). No removal of the CAHs was observed in the abiotic controls (data not shown). The high organic carbon content in the upper layer (on average 1.73% dm at 0-20 cm depth versus 0.76% and 0.50% dm at, respectively, 20-60 and 60-100 cm depth) correlated with the fastest biodegradation of CAHs. Since no electron donors were amended to the microcosms, hydrogen or other compounds produced from the fermentation of natural VOL. 43, NO. 14, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5271

FIGURE 2. Interpolated VC concentrations measured in pore water samples that were collected at three different depths in the test area in July 2005 (A) and May 2006 (B) at the riverbed locations indicated by white dots. Detected VC concentrations were interpolated using log transformation kriging analysis. Numbers indicate the occurring CAH attenuation process at that location. 1, Dilution by surface water infiltration. 2, Dilution by unpolluted groundwater discharge. 3, Reductive dechlorination. 4, No apparent CAH attenuation. 3a-c are categories within reductive dechlorination that are explained in the text.

FIGURE 3. Isotopic compositions (δ18O versus δ2H) of Zenne surface water, surface water of the canal Brussels-Scheldt, and all polluted and unpolluted groundwater samples collected during this study (A). Plot of δ18O versus δ2H for groundwater and pore water collected in July 2005 (B) or May 2006 (C). δ18O or δ2H values of pore water samples plotting outside the polluted groundwater cluster, significantly differ from the highest δ18O or δ2H values measured for this groundwater in July 2005 or May 2006, by more than 2 times the respective analytical accuracy. organic substrates present in the river sediment likely served as electron donor for CAH dechlorination (10). Hydrogeochemical Evaluation of Factors Determining CAH attenuation in the River Sediment. In addition to biodegradation, dilution by infiltrating surface water or by discharging unpolluted groundwater are possible CAH attenuation processes in the Zenne riverbed. To discriminate between the different processes, the origin of the riverbed pore water was identified. Therefore, stable hydrogen and oxygen isotope compositions of surface water, pore water, and groundwater from both the polluted and unpolluted aquifer were used as natural tracers (Figure 3). CAH-polluted groundwater exhibited significantly lower δ18O (-7.51 to -7.03‰) and δ2H (-51.6 to -47.8‰) values than unpolluted groundwater (δ18O of -6.39 to -5.87‰ and δ2H of -44.5 to -41.2‰). The latter values plot close to values obtained for surface water of the nearby canal (Figure 3A), indicating that seepage from this canal is primarily responsible for the isotope signature of the unpolluted groundwater. In Zenne surface water samples, δ18O values ranged between -7.32 and -6.58‰ and δ2H values between -48.8 and -44.9‰, after excluding three outliers, two of which could be related to extreme rainfall events (Figure 3A). The isotope signature of 5272

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 14, 2009

the Zenne surface water seems to be determined by the isotopic composition of inflowing wastewater, discharging groundwater plumes and precipitation. Of the pore water sampled in July 2005 (n ) 53) and May 2006 (n ) 64), respectively, 72% and 67% of the isotope values plot within the range of the isotope values of the polluted groundwater, taking the analytical accuracy of the isotope analyses into account (Figure 3B and C). This indicates that the pore water mainly consists of CAH-polluted groundwater. Pore water samples plotting outside the polluted groundwater cluster (Figure 3B and C) contain a large portion of unpolluted groundwater or Zenne surface water. The δ18O or δ2H values of these last pore water samples significantly differ (by more than two times the respective analytical accuracy) from the highest δ18O or δ2H values measured for the CAH-polluted groundwater in July 2005 or May 2006. This means that the pore water δ18O and δ2H values are more positive than -6.73‰ for δ18O or -45.8‰ for δ2H in July 2005 and -6.95‰ for δ18O or -46.5‰ for δ2H in May 2006, respectively. By inspecting δ18O and δ2H pore water data in vertical profiles in the riverbed in conjunction with pore water concentrations of CAHs, their transformation products, and methane, four major CAH attenuation processes could be

a Pore water samples significantly differ from the highest δ18O or δ2H values measured for the CAH-polluted groundwater in July 2005 or May 2006, by more than two times the respective analytical accuracy. b This criterion was not used to identify the CAH attenuation process. c A decrease in the CAH concentration of more than 20% was considered significant since it corresponds to a difference of more than two times the analytical accuracy for the applied GC-MS method.

no yes -

-

no no no -

yes

no no yes no yes -

-b yes

no

no no no no yes, at all sampled depths yes, not necessarily at all sampled depths

3a. reductive dechlorination 1. surface water infiltration pore water criteria

δ2H, δ18O significantly differ from the CAH-polluted groundwatera [CH4] > 1000 µg/L [CAH]20 cm depth < [CAH]60, 120 cm depth (>20% reductionc) increase in concentration of CAH transformation products from 120 to 20 cm depth decrease in [CAH] of >20%c below 120 cm depth during cross-riverbed flow from right to left riverbank

4. no CAH attenuation in top 1.2 m of the riverbed 3c. reductive dechlorination below top 1.2 m of the riverbed without observed increase in ethene or ethane 3b. reductive dechlorination within top 1.2 m of the riverbed without observed increase in ethene or ethane 2. unpolluted groundwater discharge

CAH attenuation process

TABLE 1. Criteria Used for Discrimination between the Major CAH Attenuation Processes Occurring at the Zenne Riverbed Locations Depicted in Figure 2

identified according to the criteria summarized in Table 1. The spatial distribution of these processes is depicted in Figure 2. Table 2 shows typical pore water data for each type of attenuation process. Surface water infiltration (process 1 in Table 1 and Figure 2) was identified as the major CAH attenuation process at 22 and 24% of the investigated riverbed locations in July 2005 and May 2006, respectively. For the type example (Table 2), δ18O and δ2H pore water values at 20 cm depth were indicative of surface water and no ethene, ethane, or CA could be detected. In contrast, at 80 cm depth, the pore water isotope composition was indicative of upwelling polluted groundwater and high VC, cis-DCE and 1,1-DCA concentrations were present (Table 2). Therefore, the sharp decrease in the CAH concentration and their transformation products from 17 µM at 80 cm depth to 0.6 µM at 20 cm depth was attributed to dilution by surface water. Surface water infiltration occurs locally in the Zenne riverbed due to the riverbed topography (14) and occasionally during or after extreme precipitation events in which the water level of the Zenne rises above the groundwater level and causes a short flow reversal. Surface water infiltration zones were further characterized by high methane (8064 ( 7207 µg/L) pore water concentrations (n ) 13), compared to 563 ( 340 µg/L methane measured at the riverbed locations exhibiting other CAH attenuation processes (n ) 112) (see Table 2 for type examples). Input of dissolved organic matter from the surface water could have stimulated methanogenic archaea. Dilution by unpolluted groundwater discharge (process 2 in Figure 2) was suggested as the major attenuation process at 26 and 24% of the sampled locations in July 2005 and May 2006, respectively. At those locations, δ18O and δ2H pore water values of all investigated depths plotted outside the polluted groundwater cluster (Figure 3B, 3C), without displaying evidence of surface water infiltration at all depths based on methane concentrations (Table 1). In addition, no increase in ethene, ethane, and CA concentrations was detected between pore water sampled at 120 and 20 cm depth. As expected, dilution by unpolluted groundwater mainly occurred at the left side of the river where the unpolluted groundwater approaches and discharges into the river, but this process was also observed in the middle of the river and near the right riverbank close to post 26 (Figure 2). For the type example (Table 2), 320 µg/L VC and 20 µg/L 1,1-DCA were detected at 60 cm depth in the riverbed near the left riverbank, whereas 2100-2600 µg/L VC and 135-140 µg/L 1,1-DCA were measured in discharging polluted groundwater in the middle of the Zenne and near the right riverbank in this river transect perpendicular to the riverbanks (data not shown). Dilution of the upwelling pollutants by unpolluted groundwater discharge caused the lower concentrations of VC and 1,1-DCA near the left riverbank. Reductive dechlorination of the CAHs (process 3 in Figure 2) was suggested as the major attenuation process at 30 and 40% of the investigated riverbed locations in July 2005 and May 2006, respectively. This process was divided into three categories, depending on the detected concentrations of ethene or ethane and the riverbed depth where reductive dechlorination occurred (Table 1). At 4% of the sampled locations in July 2005 and at 16% in May 2006, clear and direct evidence of reductive dechlorination during upwelling of the CAH-polluted groundwater was obtained (process 3a in Figure 2). δ18O and δ2H pore water values were indicative for polluted groundwater (Figure 3B and C) and the CAH concentrations significantly decreased toward the sediment surface with a concomitant increase in the concentration of the corresponding transformation products (Table 1), as shown for the type example in Table 2. At other locations, depending on the sampling time, δ18O and δ2H pore water values also plotted within the polluted VOL. 43, NO. 14, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

5273

TABLE 2. Type Examples of Locations and Typical Pore Water Characteristics Corresponding to Each of the CAH Attenuation Processes Defined in the Zenne River Sediment type example

1. surface water infiltration

2. unpolluted groundwater discharge

3a. reductive dechlorination

sampling campaign

July 2005

July 2005

May 2006

distance upstream of post 26

35 m, near right riverbank

depth in sediment VC (µg/L) cis-DCE (µg/L) 1,1-DCA (µg/L) CA (µg/L) ethane (µg/L) ethene (µg/L) mass balance (µM)b methane (µg/L) δ2H (‰) δ18O (‰)

type example

30 m, near left riverbank

25 m, midriver

20 cm

80 cm

20 cm

60 cm

20 cm

60 cm

100 cm

25 12 5