- Email: [email protected]
Application of in situ biosparging to remediate a petroleum-hydrocarbon spill site: Field and microbial evaluation
Available online at www.sciencedirect.com
Chemosphere 70 (2008) 1492–1499 www.elsevier.com/locate/chemosphere
Application of in situ biosparging to remediate a petroleum-hydrocarbon spill site: Field and microbial evaluation C.M. Kao a, C.Y. Chen a, S.C. Chen b, H.Y. Chien a, Y.L. Chen a
b,*
Institute of Environmental Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan Department of Biotechnology, National Kaohsiung Normal University, Kaohsiung, Taiwan
b
Received 30 April 2007; received in revised form 12 August 2007; accepted 14 August 2007 Available online 22 October 2007
Abstract In this study, a full-scale biosparging investigation was conducted at a petroleum-hydrocarbon spill site. Field results reveal that natural attenuation was the main cause of the decrease in major contaminants [benzene, toluene, ethylbenzene, and xylenes (BTEX)] concentrations in groundwater before the operation of biosparging system. Evidence of the occurrence of natural attenuation within the BTEX plume includes: (1) decrease of DO, nitrate, sulfate, and redox potential, (2) production of dissolved ferrous iron, sulfide, methane, and CO2, (3) decreased BTEX concentrations along the transport path, (4) increased microbial populations, and (5) limited spreading of the BTEX plume. Field results also reveal that the operation of biosparging caused the shifting of anaerobic conditions inside the plume to aerobic conditions. This variation can be confirmed by the following field observations inside the plume due to the biosparging process: (1) increase in DO, redox potential, nitrate, and sulfate, (2) decrease dissolved ferrous iron, sulfide, and methane, (3) increased total cultivable heterotrophs, and (4) decreased total cultivable anaerobes as well as methanogens. Results of polymerase chain reaction, denaturing gradient gel electrophoresis, and nucleotide sequence analysis reveal that three BTEX biodegraders (Candidauts magnetobacterium, Flavobacteriales bacterium, and Bacteroidetes bacterium) might exist at this site. Results show that more than 70% of BTEX has been removed through the biosparging system within a 10-month remedial period at an averaged groundwater temperature of 18 C. This indicates that biosparging is a promising technology to remediate BTEX contaminated groundwater. 2007 Elsevier Ltd. All rights reserved. Keywords: Biosparging; Natural attenuation; BTEX; PCR; DGGE
1. Introduction Accidental releases of petroleum products from pipelines and fuel-oil storage tanks are among the most common causes of groundwater contamination. Petroleum hydrocarbons contain benzene, toluene, ethylbenzene, and xylene isomers (BTEX), the major components of fuel oils (especially gasoline), which are hazardous substances regulated by many nations. At many BTEX spill sites, the residual amount of BTEX exists in a pure liquid phase (commonly referred as non-aqueous-phase liquids) within pore spaces or fractures. The slow dissolution of residual BTEX results in a contaminated plume of groundwater. *
Corresponding author. Tel.: +886 7 717 2930; fax: +886 7 525 4449. E-mail address: [email protected] (Y.L. Chen).
0045-6535/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.08.029
Given that it is often not possible to locate and remove the residual BTEX, remediation must focus on preventing further migration of the dissolved contamination. This plume control must be maintained for a long period of time. Therefore, more economic approaches are desirable for groundwater remediation to provide for long-term control of contaminated groundwater. Bioremediation is an attractive remediation option because of its economic benefit (Sutherland et al., 2004; Chen et al., 2005; Chen et al., 2006). Recently, intrinsic bioremediation has been considered as one of the potential methods for the cleanup of petroleum-hydrocarbon contaminated sites. If the intrinsic bioremediation rate is limited by in situ environmental factors (e.g., oxygen, nutrients, and electron acceptors), enhanced in situ bioremediation can be applied to stimulate contaminants biodegradation
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
(Schirmer et al., 2003; Schmidt et al., 2004). BTEX are biodegradable under both aerobic and anaerobic conditions. Nevertheless, rates of BTEX biodegradation under aerobic conditions are higher than those under anaerobic conditions (Deeb et al., 2003; Moreels et al., 2004). Organic compounds are removed more completely under aerobic biodegradation. Moreover, microbiological investigations of aquifer sediments have revealed the presence of microbial communities capable of degrading a broad range of naturally occurring and xenobiotic compounds under a broad range of environmental conditions (Kao et al., 2003, 2005). Based on the above discussion, in situ aerobic bioremediation is a feasible technology to clean up BTEX contaminated sites if oxygen can be provided to the subsurface economically. Biosparging is an effective mechanism for removal of volatile organic compounds (VOCs) including BTEX (Adams and Reddy, 2003; Wu et al., 2005; Brar et al., 2006). Biosparging functions by injecting air at a low rate into the aquifer below the zone of contamination. At a relatively close well spacing, the injected air promotes oxygenation of the aquifer as necessary to promote aerobic biodegradation. In this study, the purpose of the biosparging system is to stimulate aerobic biodegradation of BTEX. The objectives of the field-scale study were to (1) evaluate the effectiveness of in situ biosparging as a method for controlling of BTEX at the fuel-oil spill site, and (2) determine the dominant native microorganisms at different locations of the contaminated aquifer through microbial identification. The selected site in this study is an abandoned petrochemical manufacturing facility where petroleum products were produced. Leakage of a petrochemical pipeline and has resulted in site groundwater contamination with BTEX. Approximately 125 soil gas samples, 67 soil samples, 27 one-time GeoprobeTM groundwater samples, and 23 groundwater samples from monitor wells were collected and analyzed to determine the local hydrogeology and delineate the BTEX plume during a two-year site investigation period (data not shown). Site investigation results show that the components of site soils are mainly sands and silty sands. The water table is generally found at depths ranging from 2 to 2.5 m below ground surface (bgs). The groundwater flows from northeast to southwest at a velocity of 6.5 cm d1 and a hydraulic conductivity of 0.001 cm s1. Fig. 1 presents the site map showing the contaminant source area, groundwater flow direction, biosparging wells, and soil and groundwater sampling locations used in this study. Results from previous site investigation studies reveal that the contaminants resulted in an approximately 780 m long and 270 m wide plume from source area to downgradient. 2. Materials and methods 2.1. The performance of biosparging system and sampling condition The biosparging system consisted of six biosparging wells (injection points) (well screen at 5.7–6 m bgs), air
1493
N
Fig. 1. Site map showing the contaminant source area, groundwater flow direction, biosparging wells, soil vapor sampling wells, and the soil and groundwater sampling locations.
compressors, flow indicators, inline regulators, and pressure gauges. The air flow was approximately 0.06– 0.17 m3 min1 (2–6 cfm) for each biosparging well. Although a total of 23 monitor wells have been installed at this site, three monitor wells (MW1, MW2, and MW3) were selected as the representative wells to assess the potential of BTEX biodegradation by native microorganisms via biosparging processes. MW1 was located at the source area, MW2 was located at the downgradient area along the groundwater flow path, and MW3 was located in the background area (Fig. 1). Twelve soil vapor sampling wells (screened in the unsaturated zone) located within the plume were selected to collect the soil gas samples to evaluate the amount of BTEX loss due to volatilization (Fig. 1). The collected soil vapor was analyzed periodically by an organic vapor analyzer with a photoionization detector (Model TVA1000B, Thermo Environmental Instruments Inc., USA). All selected wells were sampled bimonthly during the 10-month investigation period. The first sampling event was conducted before the start of the biosparging process (day 0). For each sampling event, four 40-mL VOC vials were filled with the groundwater collected from each monitor well via the flow cell unit and micro-purge technique following the standard sampling procedure (NIEA, 2003). Groundwater sample from one of the vials
1494
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
was analyzed for BTEX concentrations. The second vial was analyzed for verification. The third and fourth vials were used for dilution purposes for samples with high BTEX concentrations. Groundwater samples from the three monitor wells were collected and analyzed for organic compounds and geochemical indicators, including BTEX, methane, CO2, inorganic nutrients, anions, pH, redox potential (Eh), groundwater temperature, and DO (Kao and Wang, 2000; APHA, 2001). Organic compound analyses were performed in accordance with US EPA Method 502.2, using a Varian 3800 GC. Methane was analyzed on a Shimadzu GC-9A GC using headspace techniques (Kao and Wang, 2000). Ion chromatography (Dionex) was used for inorganic nutrients and anions ðNO 3 , NO2 , 2 3 SO4 , and PO4 Þ analyses (APHA, 2001). Total iron and ferrous iron were analyzed by Hach DR/400 Spectrophotometer using US EPA Method 8008 and Method 8146, respectively. DO, Eh, pH, CO2, and temperature were measured in the field. Two MP120 pH Eh1 meters (MettlerToledo) were used for pH and Eh measurements. A WTW DO meter (Oxi 330) was used for DO and temperature measurements, and a Hach digital titrator cartridge was used for CO2 measurements. 2.2. Manipulation of aquifer sediments Aquifer sediments were collected from soil borings SB1, SB2, and SB3, which were located adjacent to MW1, MW2, and MW3, respectively (Fig. 1). Soil borings SB1, SB2, and SB3 were located adjacent to monitor wells MW1, MW2, and MW3, respectively, both spatially and horizontally. All SBs were collected below the groundwater table in the saturated zone. Sediments were gathered at the same time while monitor wells were sampled. Aquifer sediments from SB1, SB2, and SB3 were collected on day 0 and 300. Collected aquifer sediments were used for soil organic content analysis and microbial enumeration and identification study. Soil organic content was analyzed quarterly and was determined by burning the samples at 550 C and calculating the organic content as the preburn weight minus the postburn weight (NIEA, 2002). Microbial enumeration was performed to determine the number of total cultivable heterotrophs, total cultivable heterotrophic anaerobes (total anaerobes), and methanogens. Total plate counts were conducted using plate count agar (Difco) to assess the approximate size of the total cultivable heterotrophic bacterial in soil samples using the spread plate method (APHA, 2001). Prepared plates were incubated at 30 C for 48 h, then counted for colony forming unit (CFU). The analytical methods for total cultivable heterotrophic anaerobes and methanogens are described by Kao and Wang (2000) and enumerated using five-tube MPN assay. The total anaerobe tubes contained media and were score positive based on optical density. The methanogen tubes contained 20% H2 and 80% CO2 in the headspace, and were score positive based on the production of methane.
2.3. Denaturing gradient gel electrophoresis (DGGE) Total bacterial DNAs from 1 g of collected soil samples were extracted with a Soil Genomic DNA Purification kit (GeneMark Co., Taiwan) for detecting the community dynamics in the process of BTEX degradation. Bacterial 200-bp fragments of 16S rDNA V3 region for subsequent denaturing gradient gel electrophoresis (DGGE) analysis were amplified with the primer sets (341f, forward: 5 0 CCTACGGGAGGCA GCAG-3 0 containing a GC clamp of 40-nucleotide GC-rich sequence; 534r, reversed: 5 0 ATTACCGCGGCTGCTGG-3 0 ) (Chen, 2005). The mixtures of polymerase chain reaction (PCR) contained 10 ng of DNA extract, 4 pmol of each primer, and 5 U of Taq polymerase (Takara, Shiga, Japan) in final concentrations of 2.5 mM of MgCl2 and 0.12 mM of deoxyribonucleoside triphosphates in PCR buffer. The PCR amplification was conducted for 35 cycles: denaturation at 94 C for 1 min, an initial annealing temperature of 65.8 C decreased by 1 C per cycle until it reached 55.8 C, followed by 25 additional cycles at 55.8 C, and extension at 72 C for 2 min. The equal concentration of each amplified PCR products (2.5 lg) was furthermore performed with DGGE using a Bio-Rad DCode system (Bio-Rad, Hercules, CA, USA), as described by the manufacturer. The 10% polyacrylamide gel with a 30–60% denaturant gradient was used and electrophoresis was performed at 60 C and 70 V for 14 h. The gels were then stained with SybrGreenIand photographed. 2.4. Banding analysis and phylogenic analysis The relative intensity of amplified bands in gels was analyzed with Phoretix 1D software (Nonlinear Dynamics, Newcastle upon Tyne, NE1, UK). The PCR-amplified products were electro-eluted from gel and then sequenced by MdBio, Inc. in Taiwan. Those sequences were evaluated by using the basic local alignment search tool to determine the closest relatives in the GenBank databases (http:// www.ncbi.nlm.nih.gov). Alignment of nucleotide sequences of PCR-amplified products generated a matrix of similarity coefficients with Neighbor-Joining method (Saitou and Nei, 1987). The dendrogram based on these similarity coefficients was plotted with UPGMA (unweighted pair-group method with arithmetic mean) method for clustering. Clustal X software and Jukes and Cantor distances model were applied for the Phylogeny Tree analysis (Watanabe et al., 2000). 3. Results and discussion 3.1. Change of BTEX concentration after the operation of biosparging Groundwater samples were collected from monitor wells MW1, MW2, and MW3, which were located at the source area, downgradient area, and background area of the
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
ben-MW1 tol-MW1 ethylben-MW1 xyl-MW1 ben-MW2 tol-MW2 ethylben-MW2 xyl-MW2
1000
-1
Conc. (μg l )
10000
100 10 1 0.1 0
60
120
180
240
300
Time (d)
Fig. 2. Variations in BTEX concentrations (log scale) versus the sampling time after the onset of the biosparging process at MW1 and MW2 (ben: benzene; tol: toluene; ethylben: ethylbenzene; xyl: xylene isomers).
plume, respectively. Fig. 2 presents variations in BTEX concentrations (log scale) versus the sampling time at MW1 and MW2 after the onset of the biosparging process. Results show that the BTEX concentrations were significant decreased after the operation of biosparging mainly due to the enhanced aerobic biodegradation. Part of the BTEX loss might be also due to the vaporization. Because the measured soil gas results from twelve soil gas sampling wells were insignificant (<46 ppm) (data not shown) before and after the biosparging process, the amount of BTEX loss due to volatilization would be insignificant. Table 1 shows the results of groundwater analyses in MW1, MW2, and MW3, and microbial enumeration in SB1,
1495
SB2, and SB3 on day 0 and 300 after the operation of biosparging. The occurrence of aerobic biodegradation of BTEX due to the air injection can be confirmed by the increased population of total cultivable heterotrophs and decreased population of total anaerobes as well as methanogens in sediments SB1 and SB2. Before the operation of biosparging, natural attenuation mechanisms were the major causes of the decrease in groundwater contaminant concentrations through mixed physical, chemical, and biological processes. Field results show that the detected Eh and DO near the source area were low, which reflects the reduced conditions in the most contaminated zone. Moreover, high CO2 concentrations were observed in the plume, which indicates that significant microbial activity and natural bioremediation occurred in this area. The averaged phosphate concentrations ranged from 0.8 to 3.2 mg l1 during the investigation period. Thus, the observed natural occurring phosphate in the subsurface would not be the limiting factor for the growth of bacteria. The lower nitrate and sulfate concentrations within the plume reveal that both nitrate and sulfate were used as the electron acceptors after the depletion of oxygen in the contaminated zone. The production of sulfide in MW1 also confirmed the occurrence of sulfate reduction process. High ferrous concentrations were detected in MW1 indicating that ferric irons might have also been used as the electron acceptors around the source area.
Table 1 Concentrations of BTEX and indicating parameters in monitor wells on day 0 and 300 after the operation of biosparging Monitor Well Location
MW1 Sourcea
MW1 Source
MW2 Downb
MW2 Down
MW3 BKc
MW3 BK
Days after biosparging Benzene (lg l1) Toluene (lg l1) Ethylbenzene (lg l1) Xylenes (lg l1) DO (mg l1) Nitrate (mg l1) Total iron (mg l1) Ferrous iron (mg l1) Sulfate (mg l1) Sulfide (lg l1) Carbon dioxide (mg l1) Methane (mg l1) pH Redox potential (mV) Ammonia nitrogen (mg l1) Phosphate (mg l1) Temperature (C) Soil organic content (%) Total cultivable heterotrophs (cell g1) Total anaerobes (cell g1) Methanogens (cell g1)
0 190 6430 125 244 0 0.1 17.0 15.5 9 18 221 12.2 7.2 254 2.1 0.92 24.3 0.51 (SB1) 7.6 · 105 (SB1)e 1.3 · 105 (SB1) 3.4 · 102 (SB1)
300 41 124 38 66 2.1 3.3 0.9 0.4 27 10 265 0.12 6.8 18 1.3 0.14 25.2 0.47 (SB1) 8.2 · 107 (SB1) 7.4 · 102 (SB1) 2.4 · 101 (SB1)
0 25 657 17 33 0.3 0.5 3.2 2.8 19 12 158 2.5 7.1 81 1.2 0.8 24.6 0.42 (SB2) 8.8 · 106 (SB2) 8 · 104 (SB2) 2.1 · 102 (SB2)
300 3 14 0.8 0.5 1.1 4.6 0.5 0.3 23 8 184 0.02 6.7 54 0.28 1.23 25.1 0.33 (SB2) 7.1 · 107 (SB2) 8.2 · 102 (SB2) 1.3 · 101 (SB2)
0 BDLd 0.4 BDL BDL 2.4 6.2 1.9 0.1 31 2 124 0.02 7.4 189 4 1.16 24.1 0.28 (SB3) 1.2 · 105 (SB3) 2.3 · 102 (SB3) –f (SB3)
300 BDL 0.2 BDL BDL 2.9 4.7 1.2 0.2 28 4 98 0.01 7.3 238 3.8 1.7 25 0.35 (SB3) 4.2 · 105 (SB3) 1.7 · 102 (SB3) – (SB3)
a b c d e f
Source: sample collected at the source area. Down: sample collected at the downgradient area. BK: sample collected at the background area. BDL: below detection limit. SB: microbial enumeration was performed using soil sediments. ‘‘–’’: not detected.
1496
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
Moreover, relative higher concentration of methane was also detected in MW1. This indicates that mixed anaerobic biodegradation processes occurred within the most contaminated zone. The decrease in BTEX concentrations from MW1 to MW2 suggests the occurrence of intrinsic attenuation of BTEX. Results show that significant amount of total cultivable heterotrophs and total anaerobes (>105 cells per g of soil) were detected in SB1 and
SB2 soil samples collected from the contaminated areas. Compared to the background soil sample (SB3), SB1 and SB2 contained more bacterial population. The increased bacterial population might be due to the supplement of carbon sources (petroleum hydrocarbons) to the subsurface microorganisms. The observed soil organic content was low (<0.51%) (Table 1), and thus, the natural organic carbon would not be the major cause of the significant increase in microbial population in the soil samples after biosparging process. This indicates that higher BTEX concentration caused the increased bacterial population. Field results also indicate that the anaerobic biodegradation patterns were the dominant intrinsic biodegradation processes. Because bioremediation rates under aerobic conditions are generally higher those under anaerobic conditions, air injection into the subsurface would enhance the BTEX biodegradation rates. Results from the field investigation also reveal that the biosparging process caused significant changes in environmental conditions. Increased DO in the groundwater activated the aerobic microorganisms and enhanced the BTEX removal rates. Effects of biosparging on the variations in indicating parameters and enhancement of aerobic biodegradation include: (1) increase in DO, CO2, redox potential, nitrate, and sulfate within the plume; (2) decrease in pH, dissolved ferrous iron, sulfide, and methane within the plume; (3) increased total cultivable heterotrophs within the plume; and (4) decreased total anaerobes and methanogens within the plume.
Table 3 Comparison of the nucleotide sequences of 16S rDNA of 22 specific microorganisms in SB1 with the database from Gene Bank Strain
Fig. 3. DGGE profiles of the PCR-amplified 16S rDNA for soils collected from SB1, SB2, and SB3 on day 0 and 300.
Table 2 Variations in the intensities of the selected 22 strains versus time in SB1 Intensity Strain
D0
D 300
Strain
D0
D 300
1 2 3 4 5 6 7 8 9 10 11
0 0 0 0 23 712 0 11 302 7249 41 349 11 230 22 587
21 008 12 134 16 584 32 445 15 447 27 836 40 227 50 776 10 442 9847 21 305
12 13 14 15 16 17 18 19 20 21 22
17 445 52 441 13 519 14 213 0 7354 16 432 0 10 258 0 11 243
33 258 14 579 17 151 16 473 17 374 17 499 5127 30 197 9903 10 248 9972
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Microorganisms
Similarity (%)
Ensifer sp. (AF229863) Geitlerinema sp. (AF132780) Uncultured bacterium (AY917343) Uncultured bacterium (AB195911) Uncultured bacterium (AJ133615) Uncultured green non-sulfur bacterium (AJ441228) Sulfurihydrogenibium sp. (DQ906006) Boyliae praeputiale (AY101388) Pseudomonas saccharophila (AF368755) Variovorax paradoxus (AF451851) Uncultured bacterium (AJ289998) Rubrivivax gelatinosus (AF487435) Methylobacterium sp. (AY436812) Sulphate reducing bacterium (AJ300515) Green non-sulfur bacterium (AF027032) b-Proteobacterium (U46748) Methylobacterium sp. (M95655) Rhodothermus marinus (AF217499) Bacterium Ellin6089 (AY234741) Candidauts Magnetobacterium bavaricum (X71838) Uncultured bacterium (AJ441230) Aquificales str. (AF255597)
90 87 100 96 100 96 98 96 92 96 99 93 93 97 95 94 98 95 98 99 98 96
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
1497
Fig. 4. The Phylogeny tree for illustrating relationships among 22 microbial strains collected from SB1. The bootstrap value, as determined from 1000 bootstrap samples, is presented at each node (in percent). Only bootstrap values of P50% are presented.
1498
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
Fig. 5. The UPGMA dendrogram for illustrating relationships among microorganisms collected from SB1, SB2, and SB3 at different time points.
3.2. Change of microbial community structures after the operation of biosparging To determine if microbial community patterns in environment were changed due to the biosparging process, the PCR-DGGE was performed to investigate the dominant microorganisms on BTEX biodegradation. On day 300, the numbers of DGGE bands in SB1 or SB2 sediments were greater than that in the background sediments SB3 (Fig. 3). Results of the microbial enumeration of soil sediments reveal that the total number of cultivable heterotrophs in SB1 and SB2 was substantially increased on day 300 (Table 1). As shown in Table 2, the amounts and diversity of organisms were dramatically changed on day 300 after the operation of biosparging. In the instance of SB1 sediment sample, a total of 22 bands on DGGE profiling were clear in an appearance of their intensities after a 300-day operation (Fig. 3). Results also show that the intensities of eight bands (5, 10, 11, 12, 14, 15, 20, and 22) had slight changes (<2-fold). Moreover, eight novel bands (1, 2, 3, 4, 6, 16, 19, and 21) were observed. Intensities of three bands (7, 8, and 17) were significantly increased (>2-fold), and intensities of another three bands (9, 13, and 18) were significantly decreased (>2-fold). Results also show that bands 5, 9, 11, and 13 were among the dominant bands on day 0 under the anaerobic conditions (appearance of anaerobic condition in highly contaminated areas). However, bands 1, 4, 6, 7, 8, 11, 12, and 19 became dominant on day 300 under aerobic conditions after the onset of biosparging process. Band 11 was the only one, which was able to remain high intensity during the shifting of redox conditions from anaerobic to aerobic environment. Although the band intensity provides us information about the dominance of the bacteria in the soil environment, it might not be necessarily a valid estimate of changes in population density. To determine the meaning of representatives for bacterial species, the bands of DGGE profiles in SB1 were eluted and then amplified and sequenced for their nucleotide sequences of 16S rDNA variable V3 regions. As shown in
Table 3, the identities of the nucleotide sequences of 22 dominant bands are shown to be in a range of 87–100% of specific microorganisms as compared to database of GeneBank (Table 3). Using the similarity coefficients in 16S cDNA gene sequences, an UPGAM dendrogram allocated 22 specific microorganisms in this population into two major separate phylogenetic clusters (A and B) (Fig. 4). In Cluster A, the representatives of bands 2, 3, 5, 6, 7, 8, 20, 21, and 22 were closely related to Candidauts magnetobacterium. In Cluster B, strains 4 and 11 were closely related to Flavobacteriales bacterium and Bacteroidetes bacterium. C. magnetobacterium, F. bacterium, and B. bacterium have been reported to able to biodegrade petroleum hydrocarbons under aerobic conditions (Greene et al., 2000; Prince, 2000; Watanabe et al., 2000; Duarte et al., 2001; Fiorenza and Rifai, 2003). This indicates that BTEX degrading bacteria in SB1 could be positively selected after a 300-d operation of biosparging process. The dominant microorganisms involving in BTEX degradation could be exploited and isolated for their application on the bioremediation of BTEX-contaminated sites. Fig. 5 presents the UPGMA dendrogram for illustrating relationships among microorganisms collected from SB1, SB2, and SB3 at different time points (0, 60, 120, 180, 240, or 300 d after the onset of the biosparging). Results show that the microbial species in those three areas could be grouped into two major phylogenetic clusters. SB1 and SB2, which were located at source and downgradient areas, contained similar genetic information in DNA and are somewhat related. Moreover, the microbial species in SB3 were not closely related to SB1 and SB2. This indicates that the released BTEX have caused the variations in the dominant microbial species within the BTEX plume. Results also show that both sampling location and sampling time would cause the shifting of microbial species. 4. Conclusions Results from this field-scale study indicate that natural attenuation was the major cause of the decrease in BTEX
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
concentrations in groundwater before the operation of biosparging system. Evidence of the occurrence of natural attenuation includes the following: (1) decrease of DO, nitrate, sulfate, and Eh within the plume, (2) production of dissolved ferrous iron, sulfide, methane, and CO2 within the plume, (3) decreased BTEX concentrations along the transport path, (4) increased microbial populations within the plume, and (5) limited spreading of the BTEX plume. Field results also reveal that the operation of biosparging caused the shifting of anaerobic conditions inside the plume to aerobic conditions. This variation can be confirmed by the following field observations inside the BTEX plume due to the biosparging process: (1) increase in DO, Eh, nitrate, and sulfate, (2) decrease in dissolved ferrous iron, sulfide, and methane, (3) increased total cultivable heterotrophs, and (4) decreased total anaerobes as well as methanogens. Results show that the aerobic biodegradation was the dominant degradation processes of BTEX after the operation of biosparging. According to the results from GeneBank, three microorganisms, C. magnetobacterium, F. bacterium, and B. bacterium, which can biodegrade BTEX under aerobic conditions might exist at this site. Results also reveal that DGGE and nucleotide sequence techniques provide a guide for microbial ecology, which can be used as an indication of the trend of the biodegradation process. Moreover, the significant decrease (more than 70%) in BTEX concentrations within the plume also indicates that biosparging might be a very promising technology to remediate BTEX contaminated groundwater. Further field investigation is a necessity to confirm the removal efficiency and removal mechanisms of BTEX in groundwater via the biosparging process. Results from this study provide us insight into the characteristics of aerobic biodegradation of BTEX. The knowledge and comprehension obtained in this study will be helpful in designing a biosparging system for the bioremediation of BTEX-contaminated site. Acknowledgements This study was funded by National Science Council in Taiwan. Additional thanks to Mr. C.Y. Hsieh and L.W. Wang of National Sun Yat-Sen University for their assistance throughout this project. References Adams, J.A., Reddy, K.R., 2003. Extent of benzene biodegradation in saturated soil column during air sparging. Ground Water Monit. Remediat. 23, 85–94. APHA, 2001. Standard methods for the examination of water and wastewater. American Public Health Association, 21th ed. APHAAWWA-WEF, Washington, DC, USA. Brar, S.K., Verma, M., Surampalli, R.Y., Misra, K., Tyagi, R.D., Meunier, N., Blais, J.F., 2006. Bioremediation of hazardous wastes – a
1499
review. Pract. Period. Hazard. Toxic. Radioact. Waste Manage. 10, 59–72. Chen, K.F. 2005. Evaluation of the biodegradability of MTBE in groundwater. Ph.D. Dissertation, Institute of Environmental Engineering, National Sun Yat-Sen University, Taiwan. Chen, K.F., Kao, C.M., Wang, J.Y., Chien, C.C., 2005. Natural attenuation of MTBE at two petroleum-hydrocarbon spill sites. J. Hazard. Mater. 125, 10–16. Chen, K.F., Kao, C.M., Chen, T.Y., Weng, C.H., Tsai, C.T., 2006. Intrinsic bioremediation of MTBE-contaminated groundwater at a petroleum-hydrocarbon spill site. Environ. Geol. 50, 439–445. Deeb, R.A., Chu, K.H., Shih, T., Linder, S., Suffet, I.H., Kavanaugh, M.C., Alvarez-Cohen, L., 2003. MTBE and other oxygenates: environmental sources, analysis, occurrence and treatment. Environ. Eng. Sci. 20, 433–447. Duarte, G.F., Rosado, A.S., Seldin, L., de Araujo, W., van Elsas, J.D., 2001. Analysis of bacterial community structure in sulfurous-oilcontaining soils and detection of species carrying dibenzothiophene desulfurization gene. Appl. Environ. Microb. 67, 1052–1062. Fiorenza, S., Rifai, H.S., 2003. Review of MTBE biodegradation and bioremediation. Biorem. J. 7, 1–35. Greene, E.A., Kay, J.G., Jaber, K., Stehmeier, L.G., Voodouw, G., 2000. Composition of soil microbial communities enriched on a mixture of aromatic hydrocarbons. Appl. Environ. Microb. 66, 5282–5289. Kao, C.M., Wang, C.C., 2000. Control of BTEX migration by intrinsic bioremediation at a gasoline spill site. Water Res. 34, 3413–3423. Kao, C.M., Chen, Y.L., Chen, S.C., Yeh, T.Y., Wu, W.S., 2003. Enhanced PCE dechlorination by biobarrier systems under different redox conditions. Water Res. 37, 4885–4894. Kao, C.M., Huang, W.Y., Chang, L.J., Chien, H.Y., Hou, F., 2005. Application of monitored natural attenuation to remediate a petroleum-hydrocarbon spill site. Water Sci. Technol. 53 (2), 321–328. Moreels, D., Bastiaens, L., Ollevier, F., Merckx, R., Diels, L., Springael, D., 2004. Evaluation of the intrinsic methyl tert-butyl ether (MTBE) biodegradation potential of hydrocarbon contaminated subsurface soils in batch microcosm systems. FEMS Microb. Ecol. 49, 121–128. NIEA, National Institute of Environmental Analysis, 2002. Methods of Soil Analysis, Taiwan Environmental Protection Administration, Taiwan. NIEA, National Institute of Environmental Analysis, 2003. Methods of Soil and Groundwater Sampling, Taiwan Environmental Protection Administration, Taiwan. Prince, R.C., 2000. Biodegradation of methyl tertiary-butyl ether (MTBE) and other fuel oxygenates. Crit. Rev. Microb. 26, 163–178. Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. Schirmer, M., Butler, B.J., Church, C.D., Barker, J.F., Nadarajah, N., 2003. Laboratory evidence of MTBE biodegradation in Borden aquifer material. J. Contam. Hydrol. 60, 229–249. Schmidt, T.C., Schirmer, M., Weib, H., Haderlein, S., 2004. Microbial degradation of methyl tert-butyl ether and tert-butyl alcohol in the subsurface. J. Contam. Hydrol. 70, 173–203. Sutherland, J., Adams, C., Kekobad, J., 2004. Treatment of MTBE by air stripping, carbon adsorption, and advanced oxidation: technical and economic comparison for five groundwaters. Water Res. 38, 193–205. Watanabe, K., Kodama, Y., Syutsubo, K., Harayma, S., 2000. Molecular characterization of bacterial populations in petroleum-contaminated groundwater discharged from underground crude oil storage cavities. Appl. Environ. Microb. 66, 4803–4809. Wu, Y.W., Huang, G.H., Chakma, A., Zeng, G.M., 2005. Separation of petroleum hydrocarbons from soil and groundwater through enhanced bioremediation. Energy Sources 27, 221–232.
Chemosphere 70 (2008) 1492–1499 www.elsevier.com/locate/chemosphere
Application of in situ biosparging to remediate a petroleum-hydrocarbon spill site: Field and microbial evaluation C.M. Kao a, C.Y. Chen a, S.C. Chen b, H.Y. Chien a, Y.L. Chen a
b,*
Institute of Environmental Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan Department of Biotechnology, National Kaohsiung Normal University, Kaohsiung, Taiwan
b
Received 30 April 2007; received in revised form 12 August 2007; accepted 14 August 2007 Available online 22 October 2007
Abstract In this study, a full-scale biosparging investigation was conducted at a petroleum-hydrocarbon spill site. Field results reveal that natural attenuation was the main cause of the decrease in major contaminants [benzene, toluene, ethylbenzene, and xylenes (BTEX)] concentrations in groundwater before the operation of biosparging system. Evidence of the occurrence of natural attenuation within the BTEX plume includes: (1) decrease of DO, nitrate, sulfate, and redox potential, (2) production of dissolved ferrous iron, sulfide, methane, and CO2, (3) decreased BTEX concentrations along the transport path, (4) increased microbial populations, and (5) limited spreading of the BTEX plume. Field results also reveal that the operation of biosparging caused the shifting of anaerobic conditions inside the plume to aerobic conditions. This variation can be confirmed by the following field observations inside the plume due to the biosparging process: (1) increase in DO, redox potential, nitrate, and sulfate, (2) decrease dissolved ferrous iron, sulfide, and methane, (3) increased total cultivable heterotrophs, and (4) decreased total cultivable anaerobes as well as methanogens. Results of polymerase chain reaction, denaturing gradient gel electrophoresis, and nucleotide sequence analysis reveal that three BTEX biodegraders (Candidauts magnetobacterium, Flavobacteriales bacterium, and Bacteroidetes bacterium) might exist at this site. Results show that more than 70% of BTEX has been removed through the biosparging system within a 10-month remedial period at an averaged groundwater temperature of 18 C. This indicates that biosparging is a promising technology to remediate BTEX contaminated groundwater. 2007 Elsevier Ltd. All rights reserved. Keywords: Biosparging; Natural attenuation; BTEX; PCR; DGGE
1. Introduction Accidental releases of petroleum products from pipelines and fuel-oil storage tanks are among the most common causes of groundwater contamination. Petroleum hydrocarbons contain benzene, toluene, ethylbenzene, and xylene isomers (BTEX), the major components of fuel oils (especially gasoline), which are hazardous substances regulated by many nations. At many BTEX spill sites, the residual amount of BTEX exists in a pure liquid phase (commonly referred as non-aqueous-phase liquids) within pore spaces or fractures. The slow dissolution of residual BTEX results in a contaminated plume of groundwater. *
Corresponding author. Tel.: +886 7 717 2930; fax: +886 7 525 4449. E-mail address: [email protected] (Y.L. Chen).
0045-6535/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2007.08.029
Given that it is often not possible to locate and remove the residual BTEX, remediation must focus on preventing further migration of the dissolved contamination. This plume control must be maintained for a long period of time. Therefore, more economic approaches are desirable for groundwater remediation to provide for long-term control of contaminated groundwater. Bioremediation is an attractive remediation option because of its economic benefit (Sutherland et al., 2004; Chen et al., 2005; Chen et al., 2006). Recently, intrinsic bioremediation has been considered as one of the potential methods for the cleanup of petroleum-hydrocarbon contaminated sites. If the intrinsic bioremediation rate is limited by in situ environmental factors (e.g., oxygen, nutrients, and electron acceptors), enhanced in situ bioremediation can be applied to stimulate contaminants biodegradation
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
(Schirmer et al., 2003; Schmidt et al., 2004). BTEX are biodegradable under both aerobic and anaerobic conditions. Nevertheless, rates of BTEX biodegradation under aerobic conditions are higher than those under anaerobic conditions (Deeb et al., 2003; Moreels et al., 2004). Organic compounds are removed more completely under aerobic biodegradation. Moreover, microbiological investigations of aquifer sediments have revealed the presence of microbial communities capable of degrading a broad range of naturally occurring and xenobiotic compounds under a broad range of environmental conditions (Kao et al., 2003, 2005). Based on the above discussion, in situ aerobic bioremediation is a feasible technology to clean up BTEX contaminated sites if oxygen can be provided to the subsurface economically. Biosparging is an effective mechanism for removal of volatile organic compounds (VOCs) including BTEX (Adams and Reddy, 2003; Wu et al., 2005; Brar et al., 2006). Biosparging functions by injecting air at a low rate into the aquifer below the zone of contamination. At a relatively close well spacing, the injected air promotes oxygenation of the aquifer as necessary to promote aerobic biodegradation. In this study, the purpose of the biosparging system is to stimulate aerobic biodegradation of BTEX. The objectives of the field-scale study were to (1) evaluate the effectiveness of in situ biosparging as a method for controlling of BTEX at the fuel-oil spill site, and (2) determine the dominant native microorganisms at different locations of the contaminated aquifer through microbial identification. The selected site in this study is an abandoned petrochemical manufacturing facility where petroleum products were produced. Leakage of a petrochemical pipeline and has resulted in site groundwater contamination with BTEX. Approximately 125 soil gas samples, 67 soil samples, 27 one-time GeoprobeTM groundwater samples, and 23 groundwater samples from monitor wells were collected and analyzed to determine the local hydrogeology and delineate the BTEX plume during a two-year site investigation period (data not shown). Site investigation results show that the components of site soils are mainly sands and silty sands. The water table is generally found at depths ranging from 2 to 2.5 m below ground surface (bgs). The groundwater flows from northeast to southwest at a velocity of 6.5 cm d1 and a hydraulic conductivity of 0.001 cm s1. Fig. 1 presents the site map showing the contaminant source area, groundwater flow direction, biosparging wells, and soil and groundwater sampling locations used in this study. Results from previous site investigation studies reveal that the contaminants resulted in an approximately 780 m long and 270 m wide plume from source area to downgradient. 2. Materials and methods 2.1. The performance of biosparging system and sampling condition The biosparging system consisted of six biosparging wells (injection points) (well screen at 5.7–6 m bgs), air
1493
N
Fig. 1. Site map showing the contaminant source area, groundwater flow direction, biosparging wells, soil vapor sampling wells, and the soil and groundwater sampling locations.
compressors, flow indicators, inline regulators, and pressure gauges. The air flow was approximately 0.06– 0.17 m3 min1 (2–6 cfm) for each biosparging well. Although a total of 23 monitor wells have been installed at this site, three monitor wells (MW1, MW2, and MW3) were selected as the representative wells to assess the potential of BTEX biodegradation by native microorganisms via biosparging processes. MW1 was located at the source area, MW2 was located at the downgradient area along the groundwater flow path, and MW3 was located in the background area (Fig. 1). Twelve soil vapor sampling wells (screened in the unsaturated zone) located within the plume were selected to collect the soil gas samples to evaluate the amount of BTEX loss due to volatilization (Fig. 1). The collected soil vapor was analyzed periodically by an organic vapor analyzer with a photoionization detector (Model TVA1000B, Thermo Environmental Instruments Inc., USA). All selected wells were sampled bimonthly during the 10-month investigation period. The first sampling event was conducted before the start of the biosparging process (day 0). For each sampling event, four 40-mL VOC vials were filled with the groundwater collected from each monitor well via the flow cell unit and micro-purge technique following the standard sampling procedure (NIEA, 2003). Groundwater sample from one of the vials
1494
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
was analyzed for BTEX concentrations. The second vial was analyzed for verification. The third and fourth vials were used for dilution purposes for samples with high BTEX concentrations. Groundwater samples from the three monitor wells were collected and analyzed for organic compounds and geochemical indicators, including BTEX, methane, CO2, inorganic nutrients, anions, pH, redox potential (Eh), groundwater temperature, and DO (Kao and Wang, 2000; APHA, 2001). Organic compound analyses were performed in accordance with US EPA Method 502.2, using a Varian 3800 GC. Methane was analyzed on a Shimadzu GC-9A GC using headspace techniques (Kao and Wang, 2000). Ion chromatography (Dionex) was used for inorganic nutrients and anions ðNO 3 , NO2 , 2 3 SO4 , and PO4 Þ analyses (APHA, 2001). Total iron and ferrous iron were analyzed by Hach DR/400 Spectrophotometer using US EPA Method 8008 and Method 8146, respectively. DO, Eh, pH, CO2, and temperature were measured in the field. Two MP120 pH Eh1 meters (MettlerToledo) were used for pH and Eh measurements. A WTW DO meter (Oxi 330) was used for DO and temperature measurements, and a Hach digital titrator cartridge was used for CO2 measurements. 2.2. Manipulation of aquifer sediments Aquifer sediments were collected from soil borings SB1, SB2, and SB3, which were located adjacent to MW1, MW2, and MW3, respectively (Fig. 1). Soil borings SB1, SB2, and SB3 were located adjacent to monitor wells MW1, MW2, and MW3, respectively, both spatially and horizontally. All SBs were collected below the groundwater table in the saturated zone. Sediments were gathered at the same time while monitor wells were sampled. Aquifer sediments from SB1, SB2, and SB3 were collected on day 0 and 300. Collected aquifer sediments were used for soil organic content analysis and microbial enumeration and identification study. Soil organic content was analyzed quarterly and was determined by burning the samples at 550 C and calculating the organic content as the preburn weight minus the postburn weight (NIEA, 2002). Microbial enumeration was performed to determine the number of total cultivable heterotrophs, total cultivable heterotrophic anaerobes (total anaerobes), and methanogens. Total plate counts were conducted using plate count agar (Difco) to assess the approximate size of the total cultivable heterotrophic bacterial in soil samples using the spread plate method (APHA, 2001). Prepared plates were incubated at 30 C for 48 h, then counted for colony forming unit (CFU). The analytical methods for total cultivable heterotrophic anaerobes and methanogens are described by Kao and Wang (2000) and enumerated using five-tube MPN assay. The total anaerobe tubes contained media and were score positive based on optical density. The methanogen tubes contained 20% H2 and 80% CO2 in the headspace, and were score positive based on the production of methane.
2.3. Denaturing gradient gel electrophoresis (DGGE) Total bacterial DNAs from 1 g of collected soil samples were extracted with a Soil Genomic DNA Purification kit (GeneMark Co., Taiwan) for detecting the community dynamics in the process of BTEX degradation. Bacterial 200-bp fragments of 16S rDNA V3 region for subsequent denaturing gradient gel electrophoresis (DGGE) analysis were amplified with the primer sets (341f, forward: 5 0 CCTACGGGAGGCA GCAG-3 0 containing a GC clamp of 40-nucleotide GC-rich sequence; 534r, reversed: 5 0 ATTACCGCGGCTGCTGG-3 0 ) (Chen, 2005). The mixtures of polymerase chain reaction (PCR) contained 10 ng of DNA extract, 4 pmol of each primer, and 5 U of Taq polymerase (Takara, Shiga, Japan) in final concentrations of 2.5 mM of MgCl2 and 0.12 mM of deoxyribonucleoside triphosphates in PCR buffer. The PCR amplification was conducted for 35 cycles: denaturation at 94 C for 1 min, an initial annealing temperature of 65.8 C decreased by 1 C per cycle until it reached 55.8 C, followed by 25 additional cycles at 55.8 C, and extension at 72 C for 2 min. The equal concentration of each amplified PCR products (2.5 lg) was furthermore performed with DGGE using a Bio-Rad DCode system (Bio-Rad, Hercules, CA, USA), as described by the manufacturer. The 10% polyacrylamide gel with a 30–60% denaturant gradient was used and electrophoresis was performed at 60 C and 70 V for 14 h. The gels were then stained with SybrGreenIand photographed. 2.4. Banding analysis and phylogenic analysis The relative intensity of amplified bands in gels was analyzed with Phoretix 1D software (Nonlinear Dynamics, Newcastle upon Tyne, NE1, UK). The PCR-amplified products were electro-eluted from gel and then sequenced by MdBio, Inc. in Taiwan. Those sequences were evaluated by using the basic local alignment search tool to determine the closest relatives in the GenBank databases (http:// www.ncbi.nlm.nih.gov). Alignment of nucleotide sequences of PCR-amplified products generated a matrix of similarity coefficients with Neighbor-Joining method (Saitou and Nei, 1987). The dendrogram based on these similarity coefficients was plotted with UPGMA (unweighted pair-group method with arithmetic mean) method for clustering. Clustal X software and Jukes and Cantor distances model were applied for the Phylogeny Tree analysis (Watanabe et al., 2000). 3. Results and discussion 3.1. Change of BTEX concentration after the operation of biosparging Groundwater samples were collected from monitor wells MW1, MW2, and MW3, which were located at the source area, downgradient area, and background area of the
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
ben-MW1 tol-MW1 ethylben-MW1 xyl-MW1 ben-MW2 tol-MW2 ethylben-MW2 xyl-MW2
1000
-1
Conc. (μg l )
10000
100 10 1 0.1 0
60
120
180
240
300
Time (d)
Fig. 2. Variations in BTEX concentrations (log scale) versus the sampling time after the onset of the biosparging process at MW1 and MW2 (ben: benzene; tol: toluene; ethylben: ethylbenzene; xyl: xylene isomers).
plume, respectively. Fig. 2 presents variations in BTEX concentrations (log scale) versus the sampling time at MW1 and MW2 after the onset of the biosparging process. Results show that the BTEX concentrations were significant decreased after the operation of biosparging mainly due to the enhanced aerobic biodegradation. Part of the BTEX loss might be also due to the vaporization. Because the measured soil gas results from twelve soil gas sampling wells were insignificant (<46 ppm) (data not shown) before and after the biosparging process, the amount of BTEX loss due to volatilization would be insignificant. Table 1 shows the results of groundwater analyses in MW1, MW2, and MW3, and microbial enumeration in SB1,
1495
SB2, and SB3 on day 0 and 300 after the operation of biosparging. The occurrence of aerobic biodegradation of BTEX due to the air injection can be confirmed by the increased population of total cultivable heterotrophs and decreased population of total anaerobes as well as methanogens in sediments SB1 and SB2. Before the operation of biosparging, natural attenuation mechanisms were the major causes of the decrease in groundwater contaminant concentrations through mixed physical, chemical, and biological processes. Field results show that the detected Eh and DO near the source area were low, which reflects the reduced conditions in the most contaminated zone. Moreover, high CO2 concentrations were observed in the plume, which indicates that significant microbial activity and natural bioremediation occurred in this area. The averaged phosphate concentrations ranged from 0.8 to 3.2 mg l1 during the investigation period. Thus, the observed natural occurring phosphate in the subsurface would not be the limiting factor for the growth of bacteria. The lower nitrate and sulfate concentrations within the plume reveal that both nitrate and sulfate were used as the electron acceptors after the depletion of oxygen in the contaminated zone. The production of sulfide in MW1 also confirmed the occurrence of sulfate reduction process. High ferrous concentrations were detected in MW1 indicating that ferric irons might have also been used as the electron acceptors around the source area.
Table 1 Concentrations of BTEX and indicating parameters in monitor wells on day 0 and 300 after the operation of biosparging Monitor Well Location
MW1 Sourcea
MW1 Source
MW2 Downb
MW2 Down
MW3 BKc
MW3 BK
Days after biosparging Benzene (lg l1) Toluene (lg l1) Ethylbenzene (lg l1) Xylenes (lg l1) DO (mg l1) Nitrate (mg l1) Total iron (mg l1) Ferrous iron (mg l1) Sulfate (mg l1) Sulfide (lg l1) Carbon dioxide (mg l1) Methane (mg l1) pH Redox potential (mV) Ammonia nitrogen (mg l1) Phosphate (mg l1) Temperature (C) Soil organic content (%) Total cultivable heterotrophs (cell g1) Total anaerobes (cell g1) Methanogens (cell g1)
0 190 6430 125 244 0 0.1 17.0 15.5 9 18 221 12.2 7.2 254 2.1 0.92 24.3 0.51 (SB1) 7.6 · 105 (SB1)e 1.3 · 105 (SB1) 3.4 · 102 (SB1)
300 41 124 38 66 2.1 3.3 0.9 0.4 27 10 265 0.12 6.8 18 1.3 0.14 25.2 0.47 (SB1) 8.2 · 107 (SB1) 7.4 · 102 (SB1) 2.4 · 101 (SB1)
0 25 657 17 33 0.3 0.5 3.2 2.8 19 12 158 2.5 7.1 81 1.2 0.8 24.6 0.42 (SB2) 8.8 · 106 (SB2) 8 · 104 (SB2) 2.1 · 102 (SB2)
300 3 14 0.8 0.5 1.1 4.6 0.5 0.3 23 8 184 0.02 6.7 54 0.28 1.23 25.1 0.33 (SB2) 7.1 · 107 (SB2) 8.2 · 102 (SB2) 1.3 · 101 (SB2)
0 BDLd 0.4 BDL BDL 2.4 6.2 1.9 0.1 31 2 124 0.02 7.4 189 4 1.16 24.1 0.28 (SB3) 1.2 · 105 (SB3) 2.3 · 102 (SB3) –f (SB3)
300 BDL 0.2 BDL BDL 2.9 4.7 1.2 0.2 28 4 98 0.01 7.3 238 3.8 1.7 25 0.35 (SB3) 4.2 · 105 (SB3) 1.7 · 102 (SB3) – (SB3)
a b c d e f
Source: sample collected at the source area. Down: sample collected at the downgradient area. BK: sample collected at the background area. BDL: below detection limit. SB: microbial enumeration was performed using soil sediments. ‘‘–’’: not detected.
1496
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
Moreover, relative higher concentration of methane was also detected in MW1. This indicates that mixed anaerobic biodegradation processes occurred within the most contaminated zone. The decrease in BTEX concentrations from MW1 to MW2 suggests the occurrence of intrinsic attenuation of BTEX. Results show that significant amount of total cultivable heterotrophs and total anaerobes (>105 cells per g of soil) were detected in SB1 and
SB2 soil samples collected from the contaminated areas. Compared to the background soil sample (SB3), SB1 and SB2 contained more bacterial population. The increased bacterial population might be due to the supplement of carbon sources (petroleum hydrocarbons) to the subsurface microorganisms. The observed soil organic content was low (<0.51%) (Table 1), and thus, the natural organic carbon would not be the major cause of the significant increase in microbial population in the soil samples after biosparging process. This indicates that higher BTEX concentration caused the increased bacterial population. Field results also indicate that the anaerobic biodegradation patterns were the dominant intrinsic biodegradation processes. Because bioremediation rates under aerobic conditions are generally higher those under anaerobic conditions, air injection into the subsurface would enhance the BTEX biodegradation rates. Results from the field investigation also reveal that the biosparging process caused significant changes in environmental conditions. Increased DO in the groundwater activated the aerobic microorganisms and enhanced the BTEX removal rates. Effects of biosparging on the variations in indicating parameters and enhancement of aerobic biodegradation include: (1) increase in DO, CO2, redox potential, nitrate, and sulfate within the plume; (2) decrease in pH, dissolved ferrous iron, sulfide, and methane within the plume; (3) increased total cultivable heterotrophs within the plume; and (4) decreased total anaerobes and methanogens within the plume.
Table 3 Comparison of the nucleotide sequences of 16S rDNA of 22 specific microorganisms in SB1 with the database from Gene Bank Strain
Fig. 3. DGGE profiles of the PCR-amplified 16S rDNA for soils collected from SB1, SB2, and SB3 on day 0 and 300.
Table 2 Variations in the intensities of the selected 22 strains versus time in SB1 Intensity Strain
D0
D 300
Strain
D0
D 300
1 2 3 4 5 6 7 8 9 10 11
0 0 0 0 23 712 0 11 302 7249 41 349 11 230 22 587
21 008 12 134 16 584 32 445 15 447 27 836 40 227 50 776 10 442 9847 21 305
12 13 14 15 16 17 18 19 20 21 22
17 445 52 441 13 519 14 213 0 7354 16 432 0 10 258 0 11 243
33 258 14 579 17 151 16 473 17 374 17 499 5127 30 197 9903 10 248 9972
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Microorganisms
Similarity (%)
Ensifer sp. (AF229863) Geitlerinema sp. (AF132780) Uncultured bacterium (AY917343) Uncultured bacterium (AB195911) Uncultured bacterium (AJ133615) Uncultured green non-sulfur bacterium (AJ441228) Sulfurihydrogenibium sp. (DQ906006) Boyliae praeputiale (AY101388) Pseudomonas saccharophila (AF368755) Variovorax paradoxus (AF451851) Uncultured bacterium (AJ289998) Rubrivivax gelatinosus (AF487435) Methylobacterium sp. (AY436812) Sulphate reducing bacterium (AJ300515) Green non-sulfur bacterium (AF027032) b-Proteobacterium (U46748) Methylobacterium sp. (M95655) Rhodothermus marinus (AF217499) Bacterium Ellin6089 (AY234741) Candidauts Magnetobacterium bavaricum (X71838) Uncultured bacterium (AJ441230) Aquificales str. (AF255597)
90 87 100 96 100 96 98 96 92 96 99 93 93 97 95 94 98 95 98 99 98 96
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
1497
Fig. 4. The Phylogeny tree for illustrating relationships among 22 microbial strains collected from SB1. The bootstrap value, as determined from 1000 bootstrap samples, is presented at each node (in percent). Only bootstrap values of P50% are presented.
1498
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
Fig. 5. The UPGMA dendrogram for illustrating relationships among microorganisms collected from SB1, SB2, and SB3 at different time points.
3.2. Change of microbial community structures after the operation of biosparging To determine if microbial community patterns in environment were changed due to the biosparging process, the PCR-DGGE was performed to investigate the dominant microorganisms on BTEX biodegradation. On day 300, the numbers of DGGE bands in SB1 or SB2 sediments were greater than that in the background sediments SB3 (Fig. 3). Results of the microbial enumeration of soil sediments reveal that the total number of cultivable heterotrophs in SB1 and SB2 was substantially increased on day 300 (Table 1). As shown in Table 2, the amounts and diversity of organisms were dramatically changed on day 300 after the operation of biosparging. In the instance of SB1 sediment sample, a total of 22 bands on DGGE profiling were clear in an appearance of their intensities after a 300-day operation (Fig. 3). Results also show that the intensities of eight bands (5, 10, 11, 12, 14, 15, 20, and 22) had slight changes (<2-fold). Moreover, eight novel bands (1, 2, 3, 4, 6, 16, 19, and 21) were observed. Intensities of three bands (7, 8, and 17) were significantly increased (>2-fold), and intensities of another three bands (9, 13, and 18) were significantly decreased (>2-fold). Results also show that bands 5, 9, 11, and 13 were among the dominant bands on day 0 under the anaerobic conditions (appearance of anaerobic condition in highly contaminated areas). However, bands 1, 4, 6, 7, 8, 11, 12, and 19 became dominant on day 300 under aerobic conditions after the onset of biosparging process. Band 11 was the only one, which was able to remain high intensity during the shifting of redox conditions from anaerobic to aerobic environment. Although the band intensity provides us information about the dominance of the bacteria in the soil environment, it might not be necessarily a valid estimate of changes in population density. To determine the meaning of representatives for bacterial species, the bands of DGGE profiles in SB1 were eluted and then amplified and sequenced for their nucleotide sequences of 16S rDNA variable V3 regions. As shown in
Table 3, the identities of the nucleotide sequences of 22 dominant bands are shown to be in a range of 87–100% of specific microorganisms as compared to database of GeneBank (Table 3). Using the similarity coefficients in 16S cDNA gene sequences, an UPGAM dendrogram allocated 22 specific microorganisms in this population into two major separate phylogenetic clusters (A and B) (Fig. 4). In Cluster A, the representatives of bands 2, 3, 5, 6, 7, 8, 20, 21, and 22 were closely related to Candidauts magnetobacterium. In Cluster B, strains 4 and 11 were closely related to Flavobacteriales bacterium and Bacteroidetes bacterium. C. magnetobacterium, F. bacterium, and B. bacterium have been reported to able to biodegrade petroleum hydrocarbons under aerobic conditions (Greene et al., 2000; Prince, 2000; Watanabe et al., 2000; Duarte et al., 2001; Fiorenza and Rifai, 2003). This indicates that BTEX degrading bacteria in SB1 could be positively selected after a 300-d operation of biosparging process. The dominant microorganisms involving in BTEX degradation could be exploited and isolated for their application on the bioremediation of BTEX-contaminated sites. Fig. 5 presents the UPGMA dendrogram for illustrating relationships among microorganisms collected from SB1, SB2, and SB3 at different time points (0, 60, 120, 180, 240, or 300 d after the onset of the biosparging). Results show that the microbial species in those three areas could be grouped into two major phylogenetic clusters. SB1 and SB2, which were located at source and downgradient areas, contained similar genetic information in DNA and are somewhat related. Moreover, the microbial species in SB3 were not closely related to SB1 and SB2. This indicates that the released BTEX have caused the variations in the dominant microbial species within the BTEX plume. Results also show that both sampling location and sampling time would cause the shifting of microbial species. 4. Conclusions Results from this field-scale study indicate that natural attenuation was the major cause of the decrease in BTEX
C.M. Kao et al. / Chemosphere 70 (2008) 1492–1499
concentrations in groundwater before the operation of biosparging system. Evidence of the occurrence of natural attenuation includes the following: (1) decrease of DO, nitrate, sulfate, and Eh within the plume, (2) production of dissolved ferrous iron, sulfide, methane, and CO2 within the plume, (3) decreased BTEX concentrations along the transport path, (4) increased microbial populations within the plume, and (5) limited spreading of the BTEX plume. Field results also reveal that the operation of biosparging caused the shifting of anaerobic conditions inside the plume to aerobic conditions. This variation can be confirmed by the following field observations inside the BTEX plume due to the biosparging process: (1) increase in DO, Eh, nitrate, and sulfate, (2) decrease in dissolved ferrous iron, sulfide, and methane, (3) increased total cultivable heterotrophs, and (4) decreased total anaerobes as well as methanogens. Results show that the aerobic biodegradation was the dominant degradation processes of BTEX after the operation of biosparging. According to the results from GeneBank, three microorganisms, C. magnetobacterium, F. bacterium, and B. bacterium, which can biodegrade BTEX under aerobic conditions might exist at this site. Results also reveal that DGGE and nucleotide sequence techniques provide a guide for microbial ecology, which can be used as an indication of the trend of the biodegradation process. Moreover, the significant decrease (more than 70%) in BTEX concentrations within the plume also indicates that biosparging might be a very promising technology to remediate BTEX contaminated groundwater. Further field investigation is a necessity to confirm the removal efficiency and removal mechanisms of BTEX in groundwater via the biosparging process. Results from this study provide us insight into the characteristics of aerobic biodegradation of BTEX. The knowledge and comprehension obtained in this study will be helpful in designing a biosparging system for the bioremediation of BTEX-contaminated site. Acknowledgements This study was funded by National Science Council in Taiwan. Additional thanks to Mr. C.Y. Hsieh and L.W. Wang of National Sun Yat-Sen University for their assistance throughout this project. References Adams, J.A., Reddy, K.R., 2003. Extent of benzene biodegradation in saturated soil column during air sparging. Ground Water Monit. Remediat. 23, 85–94. APHA, 2001. Standard methods for the examination of water and wastewater. American Public Health Association, 21th ed. APHAAWWA-WEF, Washington, DC, USA. Brar, S.K., Verma, M., Surampalli, R.Y., Misra, K., Tyagi, R.D., Meunier, N., Blais, J.F., 2006. Bioremediation of hazardous wastes – a
1499
review. Pract. Period. Hazard. Toxic. Radioact. Waste Manage. 10, 59–72. Chen, K.F. 2005. Evaluation of the biodegradability of MTBE in groundwater. Ph.D. Dissertation, Institute of Environmental Engineering, National Sun Yat-Sen University, Taiwan. Chen, K.F., Kao, C.M., Wang, J.Y., Chien, C.C., 2005. Natural attenuation of MTBE at two petroleum-hydrocarbon spill sites. J. Hazard. Mater. 125, 10–16. Chen, K.F., Kao, C.M., Chen, T.Y., Weng, C.H., Tsai, C.T., 2006. Intrinsic bioremediation of MTBE-contaminated groundwater at a petroleum-hydrocarbon spill site. Environ. Geol. 50, 439–445. Deeb, R.A., Chu, K.H., Shih, T., Linder, S., Suffet, I.H., Kavanaugh, M.C., Alvarez-Cohen, L., 2003. MTBE and other oxygenates: environmental sources, analysis, occurrence and treatment. Environ. Eng. Sci. 20, 433–447. Duarte, G.F., Rosado, A.S., Seldin, L., de Araujo, W., van Elsas, J.D., 2001. Analysis of bacterial community structure in sulfurous-oilcontaining soils and detection of species carrying dibenzothiophene desulfurization gene. Appl. Environ. Microb. 67, 1052–1062. Fiorenza, S., Rifai, H.S., 2003. Review of MTBE biodegradation and bioremediation. Biorem. J. 7, 1–35. Greene, E.A., Kay, J.G., Jaber, K., Stehmeier, L.G., Voodouw, G., 2000. Composition of soil microbial communities enriched on a mixture of aromatic hydrocarbons. Appl. Environ. Microb. 66, 5282–5289. Kao, C.M., Wang, C.C., 2000. Control of BTEX migration by intrinsic bioremediation at a gasoline spill site. Water Res. 34, 3413–3423. Kao, C.M., Chen, Y.L., Chen, S.C., Yeh, T.Y., Wu, W.S., 2003. Enhanced PCE dechlorination by biobarrier systems under different redox conditions. Water Res. 37, 4885–4894. Kao, C.M., Huang, W.Y., Chang, L.J., Chien, H.Y., Hou, F., 2005. Application of monitored natural attenuation to remediate a petroleum-hydrocarbon spill site. Water Sci. Technol. 53 (2), 321–328. Moreels, D., Bastiaens, L., Ollevier, F., Merckx, R., Diels, L., Springael, D., 2004. Evaluation of the intrinsic methyl tert-butyl ether (MTBE) biodegradation potential of hydrocarbon contaminated subsurface soils in batch microcosm systems. FEMS Microb. Ecol. 49, 121–128. NIEA, National Institute of Environmental Analysis, 2002. Methods of Soil Analysis, Taiwan Environmental Protection Administration, Taiwan. NIEA, National Institute of Environmental Analysis, 2003. Methods of Soil and Groundwater Sampling, Taiwan Environmental Protection Administration, Taiwan. Prince, R.C., 2000. Biodegradation of methyl tertiary-butyl ether (MTBE) and other fuel oxygenates. Crit. Rev. Microb. 26, 163–178. Saitou, N., Nei, M., 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406–425. Schirmer, M., Butler, B.J., Church, C.D., Barker, J.F., Nadarajah, N., 2003. Laboratory evidence of MTBE biodegradation in Borden aquifer material. J. Contam. Hydrol. 60, 229–249. Schmidt, T.C., Schirmer, M., Weib, H., Haderlein, S., 2004. Microbial degradation of methyl tert-butyl ether and tert-butyl alcohol in the subsurface. J. Contam. Hydrol. 70, 173–203. Sutherland, J., Adams, C., Kekobad, J., 2004. Treatment of MTBE by air stripping, carbon adsorption, and advanced oxidation: technical and economic comparison for five groundwaters. Water Res. 38, 193–205. Watanabe, K., Kodama, Y., Syutsubo, K., Harayma, S., 2000. Molecular characterization of bacterial populations in petroleum-contaminated groundwater discharged from underground crude oil storage cavities. Appl. Environ. Microb. 66, 4803–4809. Wu, Y.W., Huang, G.H., Chakma, A., Zeng, G.M., 2005. Separation of petroleum hydrocarbons from soil and groundwater through enhanced bioremediation. Energy Sources 27, 221–232.