Environ. Sci. Techno/. 1995, 29, 2339-2345
Experiences of a Large-Scale Application of 1,tDichloroethane Degrading Microorganisms for Groundwater Treatment G E R H A R D STUCKI* A N D M A R K U S THUER Ciba-Geia, Environmental Process Technoloa, INS-2074.14, CH-4133 Schweizerhnlle,Switzerland FIGURE 1. Process flow diagram of the groundwater treatment.
1,2-Dichloroethane (DCA) mineralizing microorganisms, enriched and isolated under laboratory conditions, were successfully inoculated into a full-scale groundwater purification plant to treat 5-20 m3 of water/h at 8-12 "C. The groundwater contained DCA as the single contaminant in the range of 200015000 pg/L. It had to be treated to below 10 pg/L. The treatment plant, consisting of two sand filtersfollowed by two granular activated carbon adsorbers, was biologically modified bythe inoculation of DCA degraders and the supply of H202 and nutrients. A total of 5000 kg of activated carbon was consumed during the first month of operation to reach the required DCA levels. This amount decreased as the biological process developed. After 2years and with the implementation of a rotating biological contactor as an additional process step,the exchange of activated carbon became redundant. The microbial DCA removal was followed by its disappearance, the production of chloride ions, the concomitant reduction of pH, and the requirement for H202. So far, 1930 kg of DCA has been removed during the 5-year remediation period.
Introduction The detection of newhacteriaand the amount ofknowledge collectedoverthelast tens ofyearsofusingmicroorganisms to degrade chlorinated synthetic chemicals is impressing. Initially,the expectations of their degradationpotential was high. So far, however, most publications reported their application were related to investigations carried out on laboratory or pilot scales (1-4). Reports describing the full-scale application of microorganisms for remediation purposes or groundwater cleanup were usually limited to sites contaminated with naturally produced persistent organic chemicals, such as fuel (5,6). The slow progress in using bacteria to degrade chlorinated compounds could he based to numerous reasons: innovative technologies were not competitive with conventional chemical andlor physical methods; the application of the microorganisms * Correspondingauthor e-mailaddress:
[email protected]. ciha.cam; phone: +41 61 468 3044: fax: +41 61 468 3006.
0013-936xF)510929-2339$09.00~
0 1995 American Chemical Society
under field conditions simply failed; and additionally, people working in the applied fields were often not scientifically motivated to share their experiences gained from the new technologies. In this paper, we present the results of a pump-andtreat groundwater remediationwith a successfullarge-scale application of microorganisms able to degrade 12-dicbloroethane (DCA). Both aerobic and anaerobic microbial degradation and mineralization of DCA by CulNIeS have been reported by several research groups (7-11). Laboratory-scaleexperiments designed to evaluate the application potential of aerobic strains for groundwater remediations were assessed before the bacteria were inoculated into the reactor [ I Z , 1 3 ) . It was shown that microorganisms could convert DCA to the 100pg1L range if they were fixed onto suitable surfaces, such as sintered glass or expanded clay. Therefore, it was concluded that the biological degradation process had to he combined with an adsorption process if a target limit of 10 pg/L DCA was to he reached. The results obtained in the large-scale plant exceeded allexpectations since thetargetlimitwas met evenwithout the use of activated carhon. We report here about 5 years of experiences gathered at a plant that consisted of conventional treatment steps and was modified so that special microorganisms could develop their full DCA degrading potential. As a result of the biological activity, the exchange of exhausted granular activated carhon became redundant, and thus the process became economically much more interesting.
Materials and Methods Description of Groundwater Treatment Plant. The groundwater treatment plantwas initiallydesigned to treat 20 m3/h groundwater by a conventional dual media filtration process followed by carhonadsorption. The plant was erected on a platform, which was built on piles over swampy ground, and was covered with a fixed tent. The water was pumped first from the pond and later from the well gallery by individual pumps to two parallel sand filters (Preussag, Hamburg, Germany), each with a volume of 5 m3, They were followed by two filters Med with granular activated carbon run in series, and each had an empty bed volume of 10 m3 (Figure 1). The sand filters were backwashed every 2 weeks, and the sludge was collected and thickened in an old storage tank. The supernatant was pumped hack to the enqance of the plant while the sediment was disposed of by an authorized disposal
VOL. 29. NO. 9.1995 I ENVIRONMENTAL SCIENCE &TECHNOLOGY m 2339
company after it had been analyzed for DCA. The granular activated carbon adsorbers were also back-washed when the pressure drop over one unit exceeded 1.5 bar, which occurred about twice a year. The following procedure was applied when the carbon adsorption capacity of the first carbon adsorber was exhausted or when the effluent DCA level of the final adsorber exceeded the required cleanup level of 10 pg/L DCA. The top layer of the first adsorber was transferred to the top section of the second one to save some of the DCA degrading bacteria, because most of the DCA degrading activity was felt to be in this part of the adsorber. M e r the activated carbon of the first adsorber had been replaced, the water flow through the carbon filters was changed, and the adsorber with the regenerated carbon became the second one of the groundwater treatment process. The treatment plant was inoculated with DCA degrading microorganisms from the beginning of the water treatment. An H202 and a nutrient feed station was installed prior to the treatment of the well water from the gallery (Figure 1). Two years after startup, a rotating biological contactor (RBC) (Norddeutsche Seekabelwerke, Nordenham, Germany)with a water volume of 8 m3 and a covered headspace was installed as an additional process unit in front of the sand filters to increase the biological degradation potential of the plant. At the same time, the processing of the backwash water was optimized by the addition of a sludge sedimentation tank with avolume of 1.5m3. The electrical power requirement for the whole plant is in the range of 5 i 1 kW. Microbial Inoculation and Maintenance of Biological Activity. The treatment plant was initially inoculated with cells of Xanthobacter autotrophicus GJ 10 (strain 43050 of the American Type Culture Collection, Rockville, MD) ( 7 ) and Pseudomonas sp. DE 1 (8). These microorganisms were grown aerobically at 30 "C in a laboratory fermenter according to Janssen et al. (7)and harvested by centrifugation. Three grams (as wet weight) of a mixture of these microorganisms and 0.6 L of a bacterial solution, collected from a biologically active carbon column used in earlier studies served as inoculum (12). The bacterial paste and the solution were mixed and suspended with water to a total volume of 2 L. Each sand filter received 0.5 L of the bacterial suspension. One litre was added to the influent of the first carbon adsorber. The total DCA degradation potential of the inoculum was estimated at 2 g of DCA/day, which was about 4% of the daily DCA load fed to the plant during the startup phase by treating low contaminated pond water. To maintain the microbial activity,the groundwater was charged with 0.6 mg/L (NH4)2HP04at the entrance to the sand filters. This concentration was increased to 2-5 mg/L after 2 years of operation due to higher DCA feed concentrations. A solution of H202(30%,wlw, Degussa, Frankfurt, Germany) was added at the same dosing site in such amounts that oxygen concentrations of between 2 and 6 mg/L were reached at the plant effluent. Measurements during the first 2 years indicated that Ha02 was decomposed quantitativelyto H20 and 0 2 in the sand filters. Analytical Methods. Oxygen (Oximeter90, WTWGmbH, Weilheim, Germany), pH (Xerolyt electrode, Mettler-Toledo AG, Urdorf, Switzerland; pHmeter 632, Metrohm, Herisau, Switzerland),and temperature of influent and effluent were manually determined at regular intervals. Samples (1 L) for DCA analysis were collected regularly in headspacefree bottles from the feed and the outlet of each of the 2340
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 29. NO. 9.1995
@ed I
-
residential area
€4@
€2.
f i plant river
300 m
FIGURE 2. Situation of the remediation site.
process units. Theywere analyzed by a certified laboratory (CLL, Chemisches Laboratorium Liibeck, Germany) using the headspace technique. For routine analysis, the limit of quantitation of DCA was at 10 pg/L, which was equal to the discharge limit for DCA. For special cases, the quantification limit was lowered to 5 pg/L. Chloride ions were quantified by ion chromatography by the same laboratory.
Background Local and HydrogeologicalDescriptionof Site. The source of the groundwater pollution is located at a former pharmaceutical production plant where DCA served as the single solvent for the extraction of pancreatin from dried and grained calf's stomachs. Its production lasted from the early 1950s until 1987. The extraction procedure was the main activity at the site, and therefore DCA was the major synthetic chemical used. The pollution was detected when the plant was decommissioned. Groundwater flows in an easterly direction, passing an area of private homes. The chemical contamination was detected in the top aquifer only, ranging from 3 to 15 m below the surface. It was expected that most of the contaminant, with a density of 1.25kg/L (20 "C),would run along the bottom of the aquifer toward a pond and a river 350 m east of the production site (Figure 2). There, the aquifer flows into these surfaces waters. The river itself serves as a drinking water recharge area for the nearby city and ultimatelydischarges into the Eastern sea 18km further north. Groundwater analysis of different bore holes confirmed DCA as the only contaminant. The iron and manganese concentrations were slightly below 0.5 mglL. Many private nonpotable wells in the residential area east of the former plant contained DCA in the range of 10-200 pg/L. These wells were closed, and the water was analyzed once a year. While the pond contained 80-100pg/L DCA, the chemical was not detected in the river. A groundwater well close to the production site (well B201) showed DCA levels in the range of 60 000 pg/L (Table 1). Nonaqueous phase liquid was not observed. As part of the remediation project, additional bore holes (B228, B233, B234, etc.) were drilled in a southeast direction between the plant and the pond. These bore holes showed concentrations of DCAup to twice the level of B201 (Table 1). The highest levels were observed at bore holes B201, B234, and E2, a well of the well gallery (Figure2). Along these wells, the hydrogeologists assumed the presence of a high permeability channel in the subsurface.
TABLE 1
Development of DCA levels in Groundwater Observation Wellsa well
1988
1989
1990
1991
1992
7000 1900 1000 8201 63000 3000 640 2200 420 8220 900 1500 19 11
