WEATHERING EFFECTS ON BIODEGRADATION AND TOXICITY OF HYDROCARBONS IN GROUNDWATER

Transcription

1 WEATHERING EFFECTS ON BIODEGRADATION AND TOXICITY OF HYDROCARBONS IN GROUNDWATER A Master s Thesis Presented to the Faculty of California Polytechnic State University San Luis Obispo In partial fulfillment of the requirements for the degree of Master of Science in Civil and Environmental Engineering By Marie Gabrielle Dreyer June 2004

2 COPYRIGHT OF MASTER S THESIS I hereby grant permission for the reproduction of this thesis in its entirety or any of its parts, without further authorization, provided acknowledgement is made to the author(s) and advisor(s). Marie G. Dreyer Date ii

3 MASTER S THESIS APPROVAL Title: Author: Weathering Effects on Biodegradation and Toxicity of Hydrocarbons in Groundwater Marie Gabrielle Dreyer Date Submitted: June 2004 THESIS COMMITTEE MEMBERS: Dr. Yarrow Nelson Date Dr. Nirupam Pal Date Dr. Chris Kitts Date iii

4 ABSTRACT WEATHERING EFFECTS ON BIODEGRADATION AND TOXICITY OF HYDROCARBONS IN GROUNDWATER MARIE G. DREYER This study examined the effect of weathering on hydrocarbon biodegradation and toxicity at a former oil field near Guadalupe, California. Soil and groundwater at this site contains residual diesel-range hydrocarbons from refinery products used to dilute the viscous crude oil at this site to facilitate pumping. Natural attenuation is being considered at this site as a means of remediating residual hydrocarbons (diluent) in soil and groundwater following more active remediation techniques. To provide the lines of evidence required for use of natural attenuation at this site, this research was undertaken to determine if the hydrocarbons continue to be biodegradable after extensive weathering in the field. To investigate changes in diluent characteristics with aging, a total of thirty four groundwater samples were collected from across the site with a range of total petroleum hydrocarbon (TPH) concentrations. It was assumed that samples with low TPH concentrations were more weathered than samples with high TPH concentrations. The biodegradability of TPH in each sample was determined by measuring respiration rates (through CO 2 production) and twenty day TPH degradation rates measured in the laboratory. Changes in TPH composition with weathering were evaluated using gas chromatography (GC) with simulated distillation (SIMDIS) to determine equivalent carbon chain length. Total organic carbon (TOC) analysis was used to determine concentrations of other organic material in the groundwater samples. Finally, Microtox toxicity was measured to determine the effect of weathering of the hydrocarbon mixtures on toxicity. Respirometry experiments for groundwater samples from twenty wells indicated the presence of microbial activity as measured over time intervals of six or twenty days. An initial phase of constant microbial activity was seen within the first 36 hours of measurements, followed by a second phase of declined activity by most samples. This decline in degradation, when samples presumably contain lower (than initially measured) TPH concentrations, would be expected for first order kinetics where rate increases with increasing concentration. In fact, twenty day biodegradation rates were directly proportional to initial TPH concentrations (R 2 = ) indicative of first-order kinetics. The first-order rate constant was day -1 based on all groundwater samples analyzed. Two samples with low initial TPH concentrations exhibited first-order rate constants significantly lower than expected from the first order rate constant, which is possibly due to low nutrient concentrations. TPH degradation rates and CO 2 respiration rates decreased with increasing downgradient distance from contaminant source. This is expected because TPH concentrations also iv

5 decreased with increasing distance from the source, so lower rates would be expected based on first order kinetics. However, by plotting first order rate constant versus distance from source zone for four plumes it was seen that first-order rate constants decreased from 5 to 46 % along plume transects. This suggests the first order rate constants decrease significantly with weathering. This indicates a reduction of hydrocarbon biodegradability in groundwater down gradient from the source, however biodegradation rates are capable of degradation at a reduced rate. Microtox toxicity decreased significantly with decreasing TPH concentrations. However, some samples with similar TPH concentration exhibited widely different toxicity suggesting hydrocarbon composition affects toxicity. All samples with initially high toxicity exhibited dramatic decreases in toxicity during twenty days of biodegradation in the laboratory. Samples with low initial toxicity exhibited a slight decrease or, in some instances, an increase in toxicity. These findings indicate a difficulty in biodegrading samples below a certain threshold of toxicity. However, the initial toxicity concentrations for these samples were low enough that many were considered (by the analyzer) as nontoxic. After the period of most active biodegradation (twenty days) slight amounts of both TPH and toxicity remained. These results may be helpful for determining appropriate remediation endpoints. The rapid decreases in toxicity and TPH concentrations for samples with initially high values indicates the use of natural attenuation may still be feasible regardless of residual TPH or toxicity concentrations. v

6 ACKNOWLEDGEMENTS I want to give a special thanks to the following special individuals: Dr. Yarrow Nelson for his constant guidance and patiently sitting though my senioritis. Dr. Nirupam Pal and Dr. Chris Kitts for their support. Bob Pease (BFJ Services), for his continued support in all these efforts at the Guadalupe site. My fellow graduate students, good luck to you all. Mom, Dad and Kenny for their constant I think I can attitudes. Al, Meryll, Jimmy, Bunkim and Dave for their insightful comments and keeping things fun. Eric, for being you. And to Paul Lundegard, Gonzalo Garcia and the Unocal team, for the opportunity to work on this amazing site and for funding this phenomenal study. vi

7 TABLE OF CONTENTS LIST OF TABLES... x LIST OF FIGURES... xi INTRODUCTION... 1 PROJECT SCOPE... 3 BACKGROUND Principles behind Monitored Natural Attenuation (MNA) Current Regulations for MNA Lines of Evidence Remediation Objectives Monitoring Unsuccessful Remediation Advantages of MNA Limitations of MNA Guadalupe Restoration Project (GRP) Site, Guadalupe, California Characterization of Dissolved Phase Diluent Contamination Biodegradation Analyses Previous Work on TPH Biodegradation at GRP CO 2 Production as a Measurement of Biodegradation Total Organic Carbon Toxicity as a Measurement of Biodegradation D Gas Chromatography Analysis at Woods Hole Oceanographic Institute.. 17 vii

8 MATERIALS AND METHODS Groundwater Samples Respirometry Experiments Respirometry Methods Respirometry Sample Preparation for Subset Respirometry Sample Preparation for Subset Total Organic Carbon (TOC) Measurement TOC Methods TOC Sample Preparation Toxicity Experiments Toxicity Methods Microtox Sample Preparation Total Petroleum Hydrocarbon (TPH) Analysis Twenty Day Biodegradation Rate Analysis RESULTS AND DISCUSSION Respirometry Results Respirometry Results for Subset 1A Respirometry Results for Subset 1B Respirometry Results for Subset CO 2 Respiration Rate vs. Initial TPH Concentration TOC Results Day Biodegradation Rates Simulated Distillation (SIMDIS) Analysis viii

9 5.4 Effect of Distance from Source on Biodegradation Day Biodegradation Rate vs. Distance from Source Zone CO 2 Respiration Rate vs. Distance from Source Toxicity Results Toxicity vs. Distance from Source Comparison to Similar Studies CONCLUSIONS RECOMMENDATIONS REFERENCES APPENDIX A (Tables of Raw Data) APPENDIX B (BC Laboratory Analyses) APPENDIX C (Microtox Toxicity Procedures) ix

10 LIST OF TABLES Table 1. All Groundwater Samples from the Guadalupe Restoration Project Site Used in Experimentation Table 2. Initial TPH concentrations for all groundwater samples, as reported by Zymax Labs (detection limit = < 100 ug/l). For comparison, deionized water registered as non-detect for all carbon chain lengths using the same Zymax analytical methods Table 3. Initial TOC Results for all 34 Groundwater Samples Table Day TPH Biodegradation Rates for Subset 1 and 3. All TPH concentrations were reported by Zymax Labs Table 5. Twenty Day TOC Results for Subset 1 and Table 6. Plume Analysis for four plumes. A table summarizing these variables for all groundwater samples can be seen in Appendix A. Plume and distance values were based on data from P. Lundegard (Unocal Corp.) Table 7. Initial Microtox Results for all 34 Groundwater Samples Table 8. An Example of the Data Spreadsheet Provided by Microtox Analyzer (using sample 204-A) Table 9. Twenty Day Microtox Results for Subset 1 and x

11 LIST OF FIGURES Figure 1. United States Environmental Protection Agency (EPA) depiction of attaining cleanup goals using natural attenuation and engineered methods Figure 2. Photograph of GRP site. The site is an intricate weave of pipelines, sand dunes, sensitive costal plants and animals, and contaminated groundwater Figure 3. Location of Groundwater Wells Sampled in this Study. (Map courtesy of K. Schroeder) Figure 4. Respirometer components. Left: Sample pump, sample dryer column, and infrared sensor. Right: Expansion interface, condensing air dryer, waterbath, and temperature-controlled water recirculator Figure 5. Respirometry sample set-up for groundwater samples Figure 6. Shimadzu TOC-5000A analyzer. During inactivity, the sample port sits in a mixture of ph=2 deionized water. Insert: Photograph of the analyzer main menu Figure 7. Microtox 500 Analyzer by Strategic Diagnostics, Inc. setup. Insert: Close-up of test vials used for sample dilution Figure day biodegradation period for Subset 1. Samples were stirred and kept refrigerated at a constant temperature of 19 C Figure 9. Total CO 2 Production for Subset 1A over 6 days Figure 10. Total CO 2 Production for Subset 1B over 6 days Figure 11. Cumulative CO 2 Production over 20 days for Subset Figure 12. CO 2 Respiration Rate vs. Initial TPH Concentrations for Subset 1 and Figure 13. CO2 Respiration Rate vs. Initial TPH Concentrations for Subset 1 and 3 on Samples below 7,000 ug/l TPH Figure 14. Lineweaver-Burk Plot for Subset 1 and 3. R 2 value indicates a poor correlation Figure 15. Initial TOC Concentrations vs. Initial TPH Concentrations for all 34 Groundwater Samples xi

12 Figure 16. CO 2 Respiration Rate vs. Initial TOC Concentration for all 34 Groundwater Wells Figure 17. TPH Concentrations for Subset 1 and 3. Measurements were taken initially and at 20 days of biodegradation. All analyses were performed by Zymax Labs Figure Day TPH Biodegradation vs. Initial TPH Concentrations for Subset 1 and 3. Trendline and R-squared = indicates a direct proportionality in data Figure Day TPH Biodegradation vs. Initial TPH Concentrations for Low TPH Concentration samples Figure 20. Correlation between CO 2 Respiration Rate and 20 Day Biodegradation Rate. Used to identify the validity of respirometry as indicator for TPH biodegradation Figure 21. Carbon Chain Distribution for Sample G3-2 (Initial TPH = 12,500 ug/l) Figure 22. Carbon Chain Distribution for Sample F4-1 (Initial TPH = 1,350 ug/l) Figure 23. Location of Four Plumes Analyzed. Plumes analyzed are circled in red Figure Day Biodegradation Rate vs. Distance from Source Zone for Selected Plumes Figure 25. Specific TPH Utilization Rate vs. Distance from Source Zone for Selected Plumes. Rate was calculated by dividing degradation rate by TPH concentration Figure 26. CO 2 Respiration Rate vs. Distance from Source Figure 27. Initial Microtox EC 50 vs. Initial TPH Concentration for all 34 Groundwater Samples Figure 28. A Sample Plot for % Effect vs. Concentration (using sample 204-A) Figure 29. Toxicity as measured by Initial % Effect at Full Concentration vs. Initial TPH Concentration for all groundwater samples Figure 30. % Effect at Full Concentration for Subset 1 and 3. Measurements were taken initially and at 20 days of biodegradation Figure 31. Toxicity (Initial % Effect) vs. Distance from Source xii

13 CHAPTER 1 INTRODUCTION Natural attenuation is a method of remediation reliant on natural biological, chemical and physical processes to biodegrade or otherwise reduces concentrations of contaminants at hazardous waste sites. It is a remediation method rapidly increasing in popularity because of its capability to remediate a contaminated site with little to no disturbance. Use of natural attenuation is currently being investigated at the Guadalupe Restoration Project (GRP) site. Hydrocarbon mixtures similar to diesel fuel were used at this former oil field as a diluent for facilitating pumping of the viscous crude oil extracted at Guadalupe. Unfortunately, large quantities of this diluent were accidentally released and led to extensive soil and groundwater contamination at the site. The diluent is a moderately viscous compound containing volatile organic compounds (VOCs) and polycyclic aromatic hydrocarbons (PAHs). From the 1950s to about 1990 diluent was used to ease the transport of viscous heavy crude oil in pipelines throughout the site. Leaks occurred at various times and volumes. However, some were sufficient enough in quantity to percolate through the sand dunes into the groundwater table. Plumes sizes vary from 1 to 100,000 m 3 (hundreds of gallons to a few million gallons) of diluent. The GRP site is home to a number of endangered or sensitive species and flora. As a highly sensitive site requiring restoration, natural attenuation seems like a good remediation method for the GRP site. For natural attenuation to be successful, biodegradation must be sustained, even at low contaminant concentrations (weathered material). To help determine the feasibility of 1

14 natural attenuation at GRP, an evaluation of physical and chemical factors affecting sustainability of natural attenuation of dissolved phase diluent is being made by examining biodegradation and toxicity. Together with Unocal, the Environmental Biotechnology Institute (EBI) and Dr. Yarrow Nelson, this project is expected to further the research into sustainable natural attenuation at the GRP site. 2

15 CHAPTER 2 PROJECT SCOPE Fundamental to the question of ongoing natural attenuation at the Guadalupe site is sustainability. Previous tests at the site have shown initial rapid biodegradation (following first-order kinetics), followed by an asymptotic curve of reduced activity (simulating a zero-order kinetic reaction) (Cunningham, 2004; Waudby, 2003). Given the consistency of this trend, it is important to explore possible reasons behind the pattern. Therefore, a key area of research is to examine the trends of biodegradation and toxicity for GRP groundwater samples initially and after weathering in the field. Critical questions to be answered include: What causes the observed initial first-order kinetic hydrocarbon biodegradation to diminish after 20 days? Are nutrients limiting biodegradation? Are easily degraded components of the diluent preferentially biodegraded leaving a more recalcitrant compound? Is toxicity reduced through biodegradation? This research addresses the effects of weathering on biodegradability and toxicity using a combination of field and laboratory tests. Samples of groundwater at varying stages of biodegradation were collected from a series of thirty four monitoring wells. Groundwater samples collected at varying distances from the source are expected to have varying degrees of weathering and stages of degradation. The field samples were analyzed for total petroleum hydrocarbon (TPH) (including simulated distillation (SIMDIS) integration), respiration rate, TPH degradation rate (over twenty days) and Microtox toxicity to infer changes in diluent characteristics with aging during natural attenuation. 3

16 Water chemistry of these samples was also fully characterized to test for nutrient concentrations. Microbial characterization is being evaluated in a companion study by the Department of Microbiology at California Polytechnic State University, San Luis Obispo (Cal Poly, SLO). This project is part of a larger natural attenuation study of weathering effects on biodegradation currently being conducted by Yarrow Nelson and Chris Kitts. A companion laboratory study simulated field conditions using laboratory soil columns to examine temporal changes in one sample of hydrocarbon-rich groundwater (Cunningham, 2004). Another companion study is using an advanced two-dimensional gas chromatography method to provide detailed chemical analyses of the weathered hydrocarbons. 4

17 CHAPTER 3 BACKGROUND Natural attenuation is a powerful remediation alternative for many sites with contaminated soil and groundwater. Often, in the face of engineered processes, natural attenuation can be overlooked. As a process relying on naturally occurring activities, it is typically criticized as being too slow (Figure 1). Figure 1. United States Environmental Protection Agency (EPA) depiction of attaining cleanup goals using natural attenuation and engineered methods. Though the history of natural attenuation is relatively short, various sites are already benefiting from the effectiveness of this technology. Agencies such as the U.S. EPA 5

18 have teamed together to create standard methods for treating contaminated sites through natural attenuation processes (AFCEE, 2002). 3.1 Principles behind Monitored Natural Attenuation (MNA) Natural attenuation has recently become a popular method of remediation. By 1995, using natural attenuation to treat contaminated soils became a favored option at U.S. Superfund sites. Approximately 29,000 (or 28 %) of all sites currently use natural attenuation. It proves even more valuable for contaminated groundwater sites, being the most favored above all other technologies. Approximately 17,000 sites, or about 47 % of all contaminated groundwater sites, use this technology for treatment in the U.S. (USEPA, October 2003; Tulis, 2002). Natural attenuation works by using natural biological, chemical, and physical processes to remediate and treat contaminants in soils and groundwater. It is considered a passive or non-invasive remediation method, meaning it works without significant intervention or associated harm to the environment. The successful remediation of many sites has proven this technology capable of diminishing inorganic and organic contaminants, the most notable being petroleum-based compounds (AFCEE, 2002). Because of its name and non-invasive approach, people often mistake natural attenuation as the No Further Action (NFA) approach. A NFA site is one deemed protective of human health and the environment. On the contrary, a natural attenuation site is by definition not protective of those resources. It must still be actively characterized, assessed for risks, and monitored, but without relying solely on engineered remediation 6

19 processes. Confusion between these two processes has led to the renaming of natural attenuation to monitored natural attenuation (MNA) (USEPA, October 2003; CPEO, 2003; Tulis, 2002). MNA is a culmination of various treatment processes. These processes (in one way or another) fit the description of providing ample treatment of contaminated soils and groundwater while still remaining relatively non-invasive. Some of these processes are listed below (AFCEE, 2002). Dilution and Dispersion - the lowering of contaminant concentrations as the contaminants migrate away from the source. Absorption or Adsorption - the reduction of environmental contaminants due to contaminant incorporation and adhesion to soil particles. Volatilization - the reduction of environmental contaminants through vaporization or evaporation into the atmosphere. Chemical Transformation - the decomposition of contaminants through a series of naturally occurring chemical reactions. Biodegradation or Bioremediation - the decomposition of environmental contaminants by soil microorganisms. Extensive studies into the successful use of natural attenuation have been done by the United States Geological Survey (USGS) agency. At many sites, USGS was able to quantitatively demonstrate the ability of microorganisms to actively consume toxic compounds and convert them into CO 2. In one year a USGS site in South Carolina, 7

20 suffering from leaky military fuel storage facilities, was able to reduce contaminant concentrations seventy-five percent. Another study at a site in Bemidji, Minnesota confirmed that crude oil was rapidly degraded by indigenous microbial populations. Further success of natural attenuation at the site was also witnessed when contaminated groundwater was hindered from further spreading as biodegradation rates came into equilibrium with rates of contaminant leaching. A chlorinated solvents contaminated site in New Jersey showed the adaptability of microorganisms. At this site, microbes utilized readily available chlorinated compounds as oxidants when other oxidants were not present. Another site in New Jersey contaminated with gasoline showed the importance of microbial populations in the unsaturated zone toward biodegradation. Together these sites laid the technical foundation for USGS scientists to continue their consideration for the use of natural attenuation at contaminated sites (USGS, 1997) Current Regulations for MNA Site contamination has become more problematic as responsible parties watch their remediation costs increase. Though every site is specific to its area and history, soil and groundwater contamination is a national problem. As such, the EPA, Air Force, Army, Navy and Coast Guard as well as industrial players have all put forward their ideas of a successful MNA program. Though each program has distinctive differences, they are bound by the following requirements: providing lines of evidence, listing remediation objectives, a continual monitoring program, and a backup plan if remediation by natural attenuation proves unsuccessful (AFCEE, 2002). 8

21 Lines of Evidence Sites deemed favorable for the use of natural attenuation must first meet at least one of the following lines of evidence (AFCEE, 2002). 1. Historical groundwater and/or soil chemistry data may be used to demonstrate a clear and meaningful trend of decreasing contaminant mass and/or concentration has been observed over time at appropriate monitoring or sampling ports. 2. Hydrogeologic and geochemical modeling data can be used to demonstrate indirectly the type(s) of natural attenuation processes active at the site and the rate these processes will reduce contaminant concentrations to required levels. 3. Data from field or microcosm studies (conducted in or with actual contaminated site media) may be used to demonstrate the occurrence of a particular natural attenuation process at the site and its ability to degrade the contaminants of concern Remediation Objectives Once the site is deemed appropriate for MNA, a plan must be created with clearly defined objectives. Objectives include identifying remediation levels or performance requirements, determining points of compliance, and establishing an acceptable timeframe for remediation (Tulis, 2002). Additionally, the MNA plan must comply with any state groundwater or soil use classification and remediation standards (USEPA, March 2003). 9

22 Monitoring Given the generally slow progression of MNA, long-term monitoring is essential. This type of monitoring ensures the processes are performing up to par and are meeting expected remediation goals. To facilitate the process, frequent (usually quarterly) sampling takes place to determine current site conditions, plume migrations, byproduct creation, and any increased risks to human health or the environment. Analysis generally includes geochemical, hydro-geological, and microbiological changes of the site and contaminants. Ultimately, monitoring reveals whether MNA is working or if a more active means of remediation must be investigated (AFCEE, 2002) Unsuccessful Remediation Should MNA fail, it can become necessary to throw in the towel. More than likely, should MNA fail, an applicable engineered process will be employed to complete the job (USEPA, March 2003) Advantages of MNA MNA has several key advantages over engineered processes. The use of naturally occurring microorganisms and soil/groundwater interactions generates less remediation waste, reduces human exposure to contaminants, and limits environmental disturbance. Little to no equipment is needed to successfully operate a MNA site, resulting in less equipment failure and downtime. By limiting reliance on equipment, MNA significantly reduces remediation costs, particularly when compared to more active remediation technologies. As a simple, yet effective, technology MNA can be the sole restoration 10

23 alternative or utilized in conjunction with more active technologies (USEPA, March 2003; AFCEE, 2002; Tulis, 2002) Limitations of MNA Although MNA has an extensive list of advantages, it also has limitations. Of primary concern is time of completion. Typically, MNA requires a longer time frame to achieve established remediation goals (Figure 1). Site evaluation and providing lines of evidence are often complex and costly with MNA. Often, site characteristics change over time and may require the implementation of a more active remediation method. Required longterm monitoring of MNA can also extend the time of completion. Prolonged times of planning, evaluation and other publicly-viewed inactivity may cause misinterpretation from the public. This uncertainty delays future land uses and property transfers, and can create liability issues (USEPA, March 2003; AFCEE, 2002; Tulis, 2002). 3.2 Guadalupe Restoration Project (GRP) Site, Guadalupe, California About 30 miles south of San Luis Obispo lies a site greatly contaminated by petroleum hydrocarbons used as diluent. The GRP site is a former oil field which was active from 1950 to about The composition of the crude oil at the site is such that an oil thinner (diluent) was needed to facilitate pumping. During that time, diluent leaked and percolated through the dune sand and along the groundwater table. Plumes are located throughout the site and range in size from hundreds to a few million gallons. During the period of production no remedial efforts were made, and diluent continued to flow into the soil and groundwater at the site. Now, almost 50 years later, agencies are finally 11

24 holding Unocal responsible for their actions. As part of the lawsuit issued on them, they must restore the site to safe conditions both for humans and the natural ecosystem (Catts et al., 2003). Figure 2. Photograph of GRP site. The site is an intricate weave of pipelines, sand dunes, sensitive costal plants and animals, and contaminated groundwater. Since the lawsuit, Unocal has tried various remediation technologies including biosparging, soil vapor extraction, pump and treat, landfarming, phytoremediation, excavation and steam injection. Each has shown varying degrees of success. In particular, excavation prevented the infiltration of diluent into the nearby Pacific Ocean and the Santa Maria River. As a polishing step, Unocal is exploring the use of natural attenuation. Though it has been some years, the Unocal Corporation is still trying to provide lines of evidence for its use at the Guadalupe site. 12

25 3.2.1 Characterization of Dissolved Phase Diluent Contamination The diluent used at Guadalupe was a complex mixture of hydrocarbons with equivalent chain lengths from C10 to C30. The majority (90 %) of the hydrocarbons are within equivalent chain lengths of C14 to C30. These ranges are consistent with diesel and kerosene hydrocarbons. It has a specific gravity of 0.9 at 60 F and ranges in TPH concentrations from 910,000 to 990,000 mg/kg. Problems with biodegradation are encountered in part due to the forty one identifiable polycyclic aromatic hydrocarbons (PAH) (Lundegard and Garcia, 2001). Testing by Zymax Envirotechnology, San Luis Obispo, California (Zymax Labs) identified naphthalene as being the PAH in greatest abundance (accounting for 90 % of the sum total of PAHs). The presence of PAH residues in petroleum spill sites are problematic to soil remediation. Their persistence is largely due to the difficulty of breaking ring structures (Wang and Bartha, 1990). The presence of PAHs can increase the toxicity of a hydrocarbon mixture (Kropp and Fedorak, 1998). However, given the right conditions for microbial degradation, ph control, nutrient balance, aeration and mixing, bioremediation can be a very cost-effective remediation procedure for PAHs (Dragun, 1998). The variety of PAHs and the variance in chain lengths in petroleum products have caused researchers to question the feasibility of biodegradation at oil spill sites. Typically, short chain hydrocarbons (C10 to C18) are more bioavailable, which directly correlates to their ease of biodegradation (Nocentini et al., 2000; Siddiqui and Adams, 2001; Wang and Bartha, 1990). This finding is significant considering equivalent carbon chain lengths at 13

26 the GRP site reached C30. Most sites experience rapid first-order biodegradation near the source zone (Lundegard and Johnson, 2003; Yerushalmi et al., 2003). However, at farther distances biodegradation slows considerably (Lundegard and Johnson, 2003). At farther distances, petroleum deposits are considered weathered by both microorganisms and physical processes. Other groundwater contaminated sites witnessed a residual fraction of contaminants which remained undegraded at these weathering locations (Huesemann, 1997; Nocentini et al., 2000). 3.3 Biodegradation Analyses Many methods are available for measuring hydrocarbon biodegradation rates in the laboratory including direct TPH measurements, respirometry, TOC and toxicity. Each of these methods is described in the following subsections Previous Work on TPH Biodegradation at GRP Previous experiments at Cal Poly, SLO using GRP diluent measured the biodegradation of diluent contaminated groundwater. Waudby tested biodegradation and the effect of nutrient addition and dissolved oxygen supply using respirometry and TPH analysis (Waudby, 2003 and 2004). Scott (2003 and 2004) tested similar parameters to Waudby with the addition of Microtox toxicity. However, her tests were performed on leachate from land treatment units, rather than groundwater. Larson measured BOD and COD on eight different groundwater samples ranging in TPH concentrations from 4200 to 29,000 ug/l. He used the ratio of BOD to COD as an 14

27 indication of biodegradability. Larson found that there was no decrease in BOD/COD ratio with decreasing TPH concentration. However, he did report a slight reduction in biodegradability with distance from source zone. But, BOD and COD are very indirect measures of biodegradability and TPH concentration (Larson, 2003). All of these experiments helped lay the groundwork for this study. By applying Scott and Waudby s respirometry methods and Scott s Microtox toxicity analysis to Larson s methodology of testing a wide array of samples, this study is better able to assess the affects of biodegradation on TPH contaminated groundwater CO 2 Production as a Measurement of Biodegradation One method of measuring microbial biodegradation is through respiration. An aerobic organism respires by continuously consuming O 2 and producing CO 2. Measurement of CO 2 is considered a reliable evaluation method and has been used extensively in laboratory settings (Hollender et al., 2003; Miles and Doucette, 2001; Namkoong et al., 2001; Siddiqui and Adams, 2001; Whyte et al., 2001). Diesel fuel biodegradation can be strongly correlated to CO 2 production rates. Using dodecane as the TPH source, the theoretical stoiciometric estimate is 3.11 mg of CO 2 is produced from the biodegradation of 1 mg of TPH (Waudby, 2003). Previous research (in the laboratory at Cal Poly) has shown dissolved phase diluent capable of biodegrading in groundwater when kept continuously stirred at a temperature of 19 F. Rapid 15

28 first order biodegradation occurred during the first twenty days, leading into a period of slower degradation (Scott, 2003 and 2004; Waudby, 2003 and 2004) Total Organic Carbon Total Organic Carbon (TOC) is a measurement of the total amount of organic carbon present in a sample. This type of carbon measurement differs from those made using the respirometer. CO 2 production (as measured by the respirometer) is based on microbial activity. TOC measures the amount of organic carbon in a sample regardless of its biodegradability (Micro-Oxymax, 1993; Shimadzu). These values can be used as a measurement against respiration to show the presence (if any) of other potential CO 2 producing organics Toxicity as a Measurement of Biodegradation Toxicity is the measurement of how lethal a contaminant is when a known microorganism is exposed to the contaminant. Often this can be considered the backbone to a natural attenuation sustainability study. A contaminant too toxic to microorganisms in the environment may significantly increase the time needed to attenuate a contaminated site fully and will impede the return of a healthy ecosystem. A priori understanding of biodegradation promotes the expectation that by-products will be less toxic than parent compounds. However, there is also the possibility of toxicity increasing during biodegradation (Belkin and Steiber, 1994). In a study by Belkin and Steiber (1994), the formation of at least one toxic fungal metabolite was found as a result of biodegradation of PAHs. Toxicity measurements in the current study are made using a Microtox Analyzer (Belkin et al., 1994; Wang and Bartha, 1989; Wang et al., 1990; Yerushalmi et al., 2003). 16

29 D Gas Chromatography Analysis at Woods Hole Oceanographic Institute Petroleum products are a complex mixture of hydrocarbons, often containing hundreds of components. Traditional gas chromatography has only at best been able to present a hump-o-graph showing the presence of unresolved complex mixtures (UCM). Slight improvements were made when gas chromatography was coupled with mass spectrometry. Chris Reddy and Bob Nelson of the Woods Hold Oceanographic Institute (WHOI) have developed a method chemical analysis using dual gas chromatogram columns. The pairing of gas chromatography (GCxGC) has proved the most promising of all techniques in identifying unresolved complex mixtures of hydrocarbons. GCxGC is capable of separating components in complex mixtures an order of magnitude greater than ordinary means (GC alone or GC/MS). This is accomplished by directly injecting analytes from the first gas chromatogram column into the second gas chromatogram by way of a modulator. In this way, there is no net mass loss between columns and the compounds can be separated on the basis of both volatility and degree of polarity (Frysinger et al., 1999; Frysinger and Gaines, 2000; Frysinger and Gaines, 2001). 17

30 CHAPTER 4 MATERIALS AND METHODS 4.1 Groundwater Samples Thirty-four groundwater samples were collected for analysis from various wells throughout the GRP site. Figure 3 is a map locating the 34 groundwater wells sampled for this study. These sites were chosen by Dr. Paul Lundegard (Unocal Corp.) on the basis of historical total petroleum hydrocarbons (TPH) values and locations relative to source zones. Bob Pease (BFJ Services) headed the collection efforts throughout the week of November 11 th through 19 th, 2003 for the first twenty eight samples. An additional six wells were sampled on March 29 th, 2004 (described in detail below). Upon collection, approximately 1 L of each sample was delivered directly to Zymax Labs for TPH and simulated distillation (SIMDIS) with carbon chain identification (CCID) analysis. Duplicate TPH analyses were done for all samples that were used in respirometry and 20-day biodegradation experiments. Nutrient analyses were performed initially on all samples by BC Laboratory (Bakersfield, CA.). Sulfate, nitrite as N, nitrate as N, and orthophosphate were analyzed using the EPA Standard Method and ammonia-nitrogen using the EPA Standard Method. Characterization results can be found in Appendix A. Another 3 L s of each sample were delivered to Dr. Yarrow Nelson s laboratory for respirometry, 20-day biodegradation measurements, total organic carbon and toxicity analysis. Samples were also sent to Dr. Chris Kitts laboratory for evaluation of the microbial community using terminal restriction fragment (TRF) analyses. 18

31 For fifteen of the initial set of wells all parameters were analyzed. The second batch of thirteen wells was analyzed for all parameters except respirometry and 20-day biodegradation. The third set of six samples was added to the study to provide samples with lower TPH concentrations. These samples were extracted from GRP site wells known to have low TPH levels (< 2 mg/l). Analyses of these samples were identical to Subset 1, with the exception of respirometry. Subset 3 was the last batch of samples analyzed and was not restricted by time. Therefore, Subset 3 samples were allowed to run for a full 20-day period on the respirometer. All samples used are listed in Table 1. 19

32 Figure 3. Location of Groundwater Wells Sampled in this Study. (Map courtesy of K. Schroeder) 20

33 Table 1. All Groundwater Samples from the Guadalupe Restoration Project Site Used in Experimentation. Subset Delivery Date Wells 1A November 12, 2003 G3-2 H1-3 G3-1 G4-3 A8-6 F4-1 A8-8 A8-7 1B November 18, A 206-C 207-B H2-3C 209-C 209-D A8-5 2 November 18, A 207-C 207-D 208R-C H13-4 H5-6 I12-2 I5-5 J8-11 K11-1 H11-1 L11-1 M4-4 3 March 29, 2004 F14-4 I3-1 TB8-2 D12-1 J2-1 L1A-1 21

34 4.2 Respirometry Experiments Respirometry was used to measure CO 2 production for sample Subsets 1 and 3. Subset 1 was run for six days, while Subset 3 was run for twenty days. The variation in time was due to time restrictions. A review of the experiments can be seen below Respirometry Methods Microorganism respiration was measured using a Micro-Oxymax open cell respirometer (Columbus Instruments, Columbus, Ohio). Open cell systems are able to supply fresh air to experiments which undergo biological and chemical processes. Components of the system include a sample pump, sample dryer column packed with magnesium percholorate (Mg(ClO 4 ) 2 ), infrared sensor, expansion interface (expandable to 10 sampling ports), condensing air dryer, and an air drying column (for ambient air) containing CaSO 4 (see Figure 4). A Columbus Micro Systems compatible IBM computer connects to the respirometer and utilizes version 6.0 hardware components and version 6.06d and 6.09b upgraded software. CO 2 production is measured using a single beam, non-dispersive infrared CO 2 /CH 4 sensor with a range of 0.0 to 1 %. 22

35 Figure 4. Respirometer components. Left: Sample pump, sample dryer column, and infrared sensor. Right: Expansion interface, condensing air dryer, waterbath, and temperature-controlled water recirculator. Six experiments were run on the respirometer. The first experiment ran without complications. However, problems were encountered while running the second experiment. After consultation with the manufacturer, the sample port was shipped to Columbus Instruments, Inc. for repair. No further complications were encountered after the sample port was returned Respirometry Sample Preparation for Subset 1 At the time of experimentation nine test ports (out of ten) were functioning on the respirometer. Thus, two batches of eight (Subset 1A) then seven (Subset 1B) samples were run. A deionized water blank was run with each batch to provide a control. To accommodate for these batched tests, Unocal spread out the delivery of both batches by one week. 23

36 CO 2 production was monitored for a period of six days for each experiment for Subset 1. Each sample port was connected to a two-liter glass media bottle with 2 liters of diluentcontaminated groundwater sample. To keep the samples well oxygenated and promote microbial degradation, the bottles were continuously stirred by magnetic stirrers. To match field conditions, the samples were kept in a water bath regulated at a constant temperature of 19 C. No nutrients were added to any of the samples during measurement of respiration. Figure 5 shows the sample set-up for the respirometer runs. Figure 5. Respirometry sample set-up for groundwater samples. 24

37 4.2.3 Respirometry Sample Preparation for Subset 3 Ideally, continuous respirometry would be run over the entire biodegradation period of twenty days, but due to time constraints the samples in Subset 1 were only run for six days. Continuous respirometry for Subset 3 was run for a full twenty days so that the TPH loss in twenty days could be correlated with CO 2 production. For Subset 3, the respirometer was restarted after every 6 day period for three consecutive runs. These runs were followed by the remaining period of two days. All other set up conditions for Subset 3 were identical to Subset Total Organic Carbon (TOC) Measurement TOC measurements were made to identify the presence of organic carbon. This method varies from the respirometer by measuring all organic carbon, rather than focusing on microbial production alone. Sample preparation was the same for all Subsets TOC Methods TOC measurements were made using a Shimadzu TOC-5000A analyzer, as seen in Figure 6. Inorganic carbon (HCO - 3 and CO 2-3 ) is removed from samples by adding ph=2 deionized water (using HCl) to the sample to acidify all inorganics. All initial CO 2 is then purged from the sample using nitrogen gas. Organic carbon is then measured by thermally oxidizing all organic carbon to CO 2 and then measuring the CO 2 production by non-dispersive infra-red analysis. All samples were run against a calibration curve of ph=2 diluted samples of 1, 5, 10, and 20 mg/l of glucose standard. Each sample is 25

38 placed under a sampling port and allowed to run for two minutes performing a rotation of sparging, measuring and washing. The final step of washing readies the sampling port for the next measurement. Midway through experimentation the inline sparger stopped functioning. Manual sparging was then used by sparging each sample with an aquarium pump for one minute. Figure 6. Shimadzu TOC-5000A analyzer. During inactivity, the sample port sits in a mixture of ph=2 deionized water. Insert: Photograph of the analyzer main menu. 26

39 4.3.2 TOC Sample Preparation Each test sample was a combination of 5 ml of ph=2 deionized water and 5 ml of groundwater sample. Diluting the samples gave TOC readings which were half of the original concentration. Measurements were multiplied by two before recording in the results section. TOC was measured for all samples in duplicate. 4.4 Toxicity Experiments Toxicity is the measurement of how lethal a contaminant is to a known organism. For these experiments, all toxicity measurements were made using the Microtox method which measures toxicity to the bioluminescent bacteria Vibrio Fisheri (bioluminescent bacteria) Toxicity Methods Toxicity measurements were made using the Microtox 500 Analyzer by Strategic Diagnostics, Inc. (Newark, DE). The use of this analyzer has been approved as a regulatory test for the estimation of toxicity for oil well drilling sump fluids by the Alberta Energy and Utilities Board (AZUR, 2004). All reagents, test solutions, and the MicrotoxOmni software were also provided by Strategic Diagnostics, Inc. This analyzer determines toxicity for a series of sample dilutions and interpolates to find the effective concentration where 50% of the bioluminescent test bacteria are killed. This value is termed an EC 50. It is important to remember a high EC 50 value represents a low toxicity and vice versa. 27

40 To ensure proper operation, a 100 mg/l phenol solution standard was tested occasionally. According to manufacturers, the standard should result in an EC 50 of between 13 to 26 mg/l Microtox Sample Preparation 2.5 ml of each sample was transferred to a vial for analysis. Dilutions were made in accordance to the Basic Test and performed manually in test vials set within the analyzer (Figure 7). The Basic Test was performed in accordance to the methods outlined by Strategic Diagnostics, Inc. and can be seen in Appendix C. Tests were performed on all subsets (including initial and 20 day biodegradation for Subset 1 and 3) and run in duplicate. Figure 7. Microtox 500 Analyzer by Strategic Diagnostics, Inc. setup. Insert: Close-up of test vials used for sample dilution. 28

41 4.5 Total Petroleum Hydrocarbon (TPH) Analysis TPH concentrations were determined using gas chromatography by Zymax Labs in accordance with EPA standards. Initial samples collected from monitoring wells at the GRP site by BFJ Services scientists were taken directly to Zymax Labs for analyses. Twenty day biodegraded samples were collected by EBI scientists and shipped directly to Zymax Labs for analyses. Samples were kept at a temperature of 4 C until analyses were performed. TPH analyses were performed by gas chromatography with mass spectrophotometry detection (GC/MS). Samples are first extracted into methylene chloride (MCl) using EPA Method After extraction, analyses are performed using an amended State of California EPA Method TPH is then measured against diluent standards over an analytical range of C10-C Twenty Day Biodegradation Rate Analysis To quantify biodegradation rates, TPH measurements were made after twenty days of incubation. These measurements were made for Subsets 1 and 3. Subset 1 was constantly stirred and incubated for fourteen days (after six days in the respirometer) in an incubator at 19ºC (Figure 8). Subset 3 was constantly stirred and kept at a constant temperature of 19ºC in the waterbath recirculator attached to the respirometer apparatus. No nutrients or inoculum were added to any of the samples for the biodegradation experiments. All samples were analyzed in duplicate by 29

42 Zymax Labs using EPA standardized gas chromatography tests. TOC and Microtox toxicity were measured in duplicate at the end of twenty days. Figure day biodegradation period for Subset 1. Samples were stirred and kept refrigerated at a constant temperature of 19 C. 30

43 CHAPTER 5 RESULTS AND DISCUSSION The results are described in five sections. Section 5.1 describes the results from the three major respirometry experiments. Section 5.2 presents the results from the TOC experiments. Twenty day biodegradation rates were determined by measuring the change in TPH concentrations over 20 days. In Section 5.3, these measurements were compared to initial TPH concentrations. Within the 34 wells sampled, 12 were identified as belonging to four different plumes. In Section 5.4 comparisons were made between these plumes 20 day biodegradation rates, respiration rates and initial toxicity to distance from source. Section 5.5 describes the results from toxicity experiments. The groundwater samples used in these experiments are listed in Table 2 along with their measured initial TPH concentrations. Additional analytical information on the wells is included in Appendix A. 31

44 Table 2. Initial TPH concentrations for all groundwater samples, as reported by Zymax Labs (detection limit = < 100 ug/l). For comparison, deionized water registered as nondetect for all carbon chain lengths using the same Zymax analytical methods. Initial TPH Concentration (ug/l) Sample Rep. 1 Rep. 2 Ave. Std. Dev. Subset 1A A (8) A A F G G G H Subset 1B 204-A (7) 206-C B H2-3C C D A Subset A 2800 (13) 207-C D R-C ND H H I I J K H L M Subset 3 F (6) I3-1 N/A N/A 49 N/A TB D12-1 N/A N/A 49 N/A J L1A-1 N/A N/A 49 N/A 32

45 5.1 Respirometry Results Respirometry experiments were conducted to measure the microbial activity for 21 of the groundwater samples. Subset 1 was comprised of 15 wells and Subset 3 contained 6 wells. Respirometry was not measured for Subset 2 samples because of budget and time constraints. O 2 respiration rates were omitted due to O 2 sensor component malfunction Respirometry Results for Subset 1A Total CO 2 production rates over a period of 6 days are shown in Figure 9 (TPH = 1,350 to 12,500 ug/l see Table 2). The 2 L sample of deionized water tested as the control did not produce any measurable CO 2 over the six day period. This result confirms the respirometer was in proper working order and verifies that any observed CO 2 evolution of the samples was from the groundwater constituents. CO 2 production was observed for all 8 groundwater samples, indicating microbial activity over the six day period (Figure 9). An initial phase of constant microbial activity was seen within the first 36 hours of measurements. This phase was followed by a second phase of declined activity by most samples. Groundwater from well G3-2 was able to maintain a high CO 2 production rate during the entire six days. This sustained activity may be attributed to the high TPH concentrations measured in sample G3-2 (12,500 ug/l). Conversely, sample F4-1 with the lowest TPH concentration (1,350 ug/l) exhibited a leveling off of CO 2 production sooner than other samples. These trends in Figure 9 point to increased activity with increased TPH concentrations because samples containing higher levels of TPH produced CO 2 more readily than those with low TPH concentrations. 33

46 Cumulative CO 2 Production (ug/ml) A8-7 A8-8 G3-2 F4-1 H1-3 A8-6 G4-3 G3-1 DI H Time Elapsed (hours) Figure 9. Total CO 2 Production for Subset 1A over 6 days. 34

47 5.1.2 Respirometry Results for Subset 1B Subset 1B grouped 7 groundwater samples with TPH concentrations ranging from 455 to 22,500 ug/l (see Table 2). Total CO 2 production rates over a period of 6 days are shown in Figure 10. The deionized control sample again did not produce any measurable CO 2 over the six day period. Similar trends from Subset 1A were observed in Subset 1B. Constant-rate microbial activity was observed for the first 36 hours of measurements followed by a period of decreased activity (Figure 10). The trends of high CO 2 respiration with high TPH concentrations from Subset 1A were again seen in this subset, with the exception of samples 209-D and H2-3C. Sample 209-D exhibited a high level of CO 2 inconsistent with its relatively low initial TPH concentration. Conversely, sample H2-3C exhibited a low level of CO 2 given the samples high initial TPH concentration. 35

48 Cumulative CO 2 Respiration (ug/ml) D 204-A 207-B DI H C 206-C H2-3C A Time Elapsed (hours) Figure 10. Total CO 2 Production for Subset 1B over 6 days. 36

49 5.1.3 Respirometry Results for Subset 3 Subset 3 contained 6 groundwater samples with TPH concentrations ranging from 49 (below non-detect levels) to 2,700 ug/l (see Table 2). These samples were collected to provide samples with low TPH concentrations. The experiment was run in three consecutive six day intervals followed by a two day interval. Total CO 2 production rates over a period of 20 days are shown in Figure 11. As observed in earlier respirometry experiments, the deionized water control did not produce measurable CO 2 over the six day period. Even though Subset 3 samples had low TPH concentrations they followed the same trends seen in the previous subsets: decreased TPH concentrations cause a decrease in microbial activity, causing samples to respire less. The same trend was seen in Subset 3, with the exception of sample L1A-1. Sample L1A-1 produced a higher than expected amount of CO 2 relative to its initially low level of TPH concentration. Unlike the previous two subsets, respiration was measured for Subset 3 for a total of 20 days. Figure 11 shows a successive decrease of CO 2 production as more time elapses. TPH concentrations were not taken after each interval. However, assuming these samples follow trends seen in previous experiments, it can be assumed TPH concentrations decreased successively over each interval. 37

50 Cumulative CO 2 Production (ug/ml) J2-1 TB8-2 F14-4 DI H2O D12-1 L1A-1 I Time Elapsed (hours) Figure 11. Cumulative CO 2 Production over 20 days for Subset 3. 38

51 5.1.4 CO 2 Respiration Rate vs. Initial TPH Concentration Respiration rates were correlated with initial TPH concentration for all samples measured with respirometry. Average initial CO 2 respiration rates for Subsets 1A and 1B were calculated by dividing the final accumulated CO 2 respiration value by total time (141 hours). Subset 3 ran for 20 days, but for this calculation, only the initial 6 day value was used to be consistent. Although CO 2 production rates varied over the 6 day period, average respiration rates were used for this correlation. The correlation of increased CO 2 production with increasing TPH concentrations is seen in Figure 12. CO 2 respiration rates increased linearly with initial TPH concentrations, up to 7,000 ug/l (R 2 = 0.873) (Figure 12). Above 7,000 ug/l TPH concentration, the respiration rates did not increase with increasing TPH, rather a leveling off was seen typical of Michaelis-Menten growth kinetics (Figure 13) (Shuler and Kargi, 1992). A Lineweaver-Burk plot is used to indicate the compatibility of data to Michaelis-Menten kinetics by plotting the double reciprocal of reaction rate and substrate concentration. From this plot, a higher linear correlation would indicate a strong fit for Michaelis-Menten growth kinetics. Figure 14 indicates the data for Subset 1 and 3 do not closely follow Michaelis-Menten kinetics (R 2 = 0.692). The scatter in the measured respiration rates is likely due to the inconsistent TOC levels of the samples, as in sample G3-2. Organic matter, in the form of something other than TPH, is likely to have affected CO 2 production rates (see next Section). 39

52 G3-2 CO 2 Respiration Rate (ug/ml-hr) DI H2O 204-A F4-1 L1-A-1 A B J2-1 F14-4 I3-1D12-1 TB8-2 A8-6 A C G D H1-3 G3-1 H2-3C Initial TPH (ug/l) 209-C A8-5 Figure 12. CO 2 Respiration Rate vs. Initial TPH Concentrations for Subset 1 and 3. 40

53 CO 2 Respiration Rate (ug/ml-hr) y = 7E-05x R 2 = Initial TPH (ug/l) Figure 13. CO2 Respiration Rate vs. Initial TPH Concentrations for Subset 1 and 3 on Samples below 7,000 ug/l TPH. 41

54 / Respiration Rate y = x R 2 = / Initial TPH Figure 14. Lineweaver-Burk Plot for Subset 1 and 3. R 2 value indicates a poor correlation. 42

55 5.2 TOC Results TOC analysis was used to measure the total organic carbon present in all 34 groundwater samples. This method varies from the respirometer by measuring all organic carbon, rather than focusing on microbial production alone. Initial TOC concentrations of the groundwater samples are tabulated in Table 3. TOC concentrations ranged from 13 to 42 mg organic carbon / L of sample. Figure 15 identifies that a correlation between initial TOC versus initial TPH concentrations does not exist (R 2 = 0.156). This indicates groundwater samples vary in natural organic matter content. It should also be noted that samples with very low TPH concentrations exhibited a wide range of TOC concentrations from 10 to 35 mg/l. TOC levels in the samples may have influenced respiration rates measured for some of the samples. Sample 209-D contained high TOC levels (Table 3) which may have contributed to a higher CO 2 production rate (Figure 10). However, there was no correlation observed between TOC levels and CO 2 production rates (Figure 16). 43

56 Table 3. Initial TOC Results for all 34 Groundwater Samples. Sample TPH (ug/l) Initial TOC Concentrations (mg/l) Rep. 1 Rep. 2 Ave. Std. Dev. Subset 1A A (8) A A F G G G H Subset 1B 204-A (7) 206-C B H2-3C C D A Subset A (13) 207-C D R-C ND H H I I J K H L M Subset 3 F (6) I TB D J L1A

57 45 40 K11-1 I12-2 G C Initial TOC (mg/l) I3-1 F14-4 D12-1 J2-1 L1A B A8-7 TB8-2 A8-6 H5-6 A8-8 H13-4 G C H2-3C L D 207-D H1-3 M4-4 G4-3 y = x R 2 = A A H C 207-2A F4-1 I R-C J Initial TPH (ug/l) Figure 15. Initial TOC Concentrations vs. Initial TPH Concentrations for all 34 Groundwater Samples. 45

58 CO 2 Respiration Rate (ug/ml-hr) Initial TOC Concentration (mg/l) Figure 16. CO 2 Respiration Rate vs. Initial TOC Concentration for all 34 Groundwater Wells. 46

59 Day Biodegradation Rates The most direct measure of active biodegradation is measurement of TPH changes over time. For 21 wells (Subsets 1 and 3) TPH analyses were made in duplicate by Zymax Labs initially and again after 20 days of biodegradation in the lab at 19 C. Biodegradation rates were calculated by dividing the change in TPH concentration (initial minus 20 day) over the 20 day time interval (Table 4). Figure 17 outlines the degradation of TPH for those two time intervals for Subset 1 and 3. Significant reductions in TPH concentration were observed for all samples with measurable initial TPH concentrations. Table Day TPH Biodegradation Rates for Subset 1 and 3. All TPH concentrations were reported by Zymax Labs. Sample Subset 1 and 3 Initial TPH (ug/l) Final TPH (20 day) (ug/l) 20 Day Biodeg. Rate (ug/l-day) Subset 1A A (8) A A F G G G H Subset 1B 204-A (7) 206-C B H2-3C C D A Subset 3 F (6) I3-1 NA NA NA TB D12-1 NA NA NA J L1A-1 NA NA NA 47

60 After twenty days of biodegradation TOC concentrations were still mg/l (Table 5). These measurements were only taken for those samples which underwent respirometry (Subset 1 and 3). As expected, TOC levels decreased after the 20 day degradation period. The 20 day biodegradation rates are plotted as a function of initial TPH concentrations in Figure 18. The R 2 value of 0.92 for the 20 day biodegradation rate versus initial TPH concentration in Figure 18 shows a direct proportionality between 20 day biodegradation rate and initial TPH concentrations. This indicates that TPH biodegradation in groundwater at the site follows first-order biodegradation kinetics. From the slope of the trendline in Figure 18, a first order rate constant of day -1 (R 2 = 0.92) was observed. Figure 19 focuses in on groundwater samples with low concentrations, as they pose the greatest concern with biodegradability after weathering. It is observed that samples 204- A, A8-8 and A8-6 fall below the trendline. Hydrocarbons in groundwater in the vicinity of these monitoring wells appear to biodegrade, but with a lower first order rate constant. This difference may be attributed to numerous factors. Initial and final equivalent carbon chain length analyses did not show a discrepancy between these and other samples tested (Appendix A). However, when nutrients were compared, these samples lacked a greater quantity of nutrients than other samples (Appendix B). For these samples, Nitrate, Nitrite and ortho-phosphate levels were below the practical quantification limits. Nutrients are necessary for microbial cell growth and enzyme production necessary for biodegradation. Without nutrients, biodegradation can still occur at a limited rate due to the natural recycling of elements (USEPA, 1995). The lack of nutrients in samples 204-A, A8-8 and 48

61 A8-6 may have led to the lower rates of biodegradation observed. Overall, samples with low TPH concentrations (first order rate constant = day -1 ) followed similar first order kinetics as high TPH samples which indicates that biodegradability of hydrocarbons is not hindered by chemical changes between high and low TPH concentrations. Sample F14-4 also showed slightly lower TPH biodegradation than expected from the first order plot. From Appendix B, this sample also contained low levels of Nitrate, Nitrite and ortho-phosphate. To quantify the validity of using respirometry as an indicator of TPH biodegradation rate, a graph of CO 2 production versus 20 day TPH biodegradation was plotted. Figure 20 indicates a poor correlation between the two measurement techniques (R 2 = ). Direct TPH measurement over 20-days measures TPH at given time intervals to examine biodegradation, whereas respirometry relies on the production of CO 2 for measurement. The indirect method of measuring biodegradation through respirometry may have lead to the variability seen in some samples during those experiments. 49

62 Table 5. Twenty Day TOC Results for Subset 1 and 3. After 20 Day Biodegradation Final TOC Concentration (mg/l) Sample TPH Rep. 1 Rep. 2 Ave. Std. Dev. Subset 1A A (8) A A F G G G H Subset 1B 204-A (7) 206-C B H2-3C C D A Subset 3 F (6) I TB D J L1A

63 25000 Inital Day TPH Concentration (ug/l) I3-1 D12-1 L1A-1 F14-4 J B F4-1 A A TB8-2 A8-8 A C G D H1-3 G3-1 H2-3C G C A8-5 Well Figure 17. TPH Concentrations for Subset 1 and 3. Measurements were taken initially and at 20 days of biodegradation. All analyses were performed by Zymax Labs. 51

64 600 A Day Biodegredation Rate (ug/l-day) G4-3 G3-1 H1-3 H2-3C 209-C G3-2 y = x R 2 = C TB D A8-6 A8-7 F4-1 A8-8 J A 207-B F Initial TPH (ug/l) Figure Day TPH Biodegradation vs. Initial TPH Concentrations for Subset 1 and 3. Trendline and R- squared = indicates a direct proportionality in data. 52

65 C 120 y = x R 2 = Day Biodegredation Rate (ug/l-day) F4-1 A8-7 TB8-2 A8-8 A A 20 0 F14-4 J B Initial TPH (ug/l) Figure Day TPH Biodegradation vs. Initial TPH Concentrations for Low TPH Concentration samples. 53

66 CO 2 Respiration Rate (ug/ml-hr) y = x R 2 = Day Biodegradation Rate (ug/l-day) Figure 20. Correlation between CO 2 Respiration Rate and 20 Day Biodegradation Rate. Used to identify the validity of respirometry as indicator for TPH biodegradation. 54

67 5.3.1 Simulated Distillation (SIMDIS) Analysis SIMDIS shows the equivalent carbon chain length distribution for the hydrocarbons in each groundwater sample (Appendix A). Samples G3-2 and F4-1 had the highest and lowest concentrations of TPH respectively (Table 5), thus their carbon chain distributions are shown as examples in Figure 21 and 22. These charts show similar distributions with short chain hydrocarbons peaking in the range of C After twenty days, significant biodegradation was observed for all equivalent carbon ranges. This suggests there is not a recalcitrant component identifiable through SIMDIS analysis. These observations were similar for all groundwater samples studied (Appendix A). 55

68 Initial 20-Day TPH C10-12 C12-14 C14-16 C16-18 C18-20 C20-24 C24-28 C28-32 C32-36 C36-40 Carbon Chain Lengths Figure 21. Carbon Chain Distribution for Sample G3-2 (Initial TPH = 12,500 ug/l) Initial 20-Day 300 TPH C10-12 C12-14 C14-16 C16-18 C18-20 C20-24 C24-28 C28-32 C32-36 C36-40 Carbon Chain Lengths Figure 22. Carbon Chain Distribution for Sample F4-1 (Initial TPH = 1,350 ug/l). 56

69 5.4 Effect of Distance from Source on Biodegradation The groundwater sample set used in this study contained four recognizable plume configurations (Figure 23). Each plume studied was given a number designation (Plume 1-4). Three parameters were analyzed based on plume and distance from source: 20 day biodegradation rate, CO 2 respiration rate and initial toxicity (in Section 5.5). The data used in these analyses are summarized in Table 6. Figure 23. Location of Four Plumes Analyzed. Plumes analyzed are circled in red. 57