Biodegradation Rates of Weathered Hydrocarbons in Controlled Laboratory Microcosms and Soil Columns Simulating Natural Attenuation Field Conditions

Transcription

1 Biodegradation Rates of Weathered Hydrocarbons in Controlled Laboratory Microcosms and Soil Columns Simulating Natural Attenuation Field Conditions 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 C. Robin Cunningham October 2004

2 AUTHORIZATION FOR REPRODUCTION 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). C. Robin Cunningham Date ii

3 APPROVAL PAGE TITLE: BIODEGRADATION RATES OF WEATHERED HYDROCARBONS IN CONTROLLED LABORATORY MICROCOSMS AND SOIL COLUMNS SIMULATING NATURAL ATTENUATION FIELD CONDITIONS AUTHOR: C. ROBIN CUNNINGHAM DATE SUBMITTED: OCTOBER 2004 Dr. Yarrow Nelson Advisor Signature Dr. Chris Kitts Committee Member Signature Dr. Nirupam Pal Committee Member Signature iii

4 ABSTRACT Biodegradation Rates Of Weathered Hydrocarbons In Controlled Laboratory Microcosms And Soil Columns Simulating Natural Attenuation Field Conditions C. Robin Cunningham Controlled laboratory microcosms and soil columns were used to observe biodegradation of hydrocarbon-contaminated groundwater under conditions mimicking natural attenuation field conditions at a former oil field near Guadalupe, CA. Diesel range oil (DRO) was used as a diluent to facilitate pumping the viscous crude oil during oil production at this site from Leaking tanks and pipes used for the containment and transportation of diluent contaminated the soil and groundwater directly beneath the Guadalupe site. Following active remediation measures such as excavation, the feasibility of using natural attenuation by native microbial species to remediate the levels of diluent is currently being investigated. This laboratory study was undertaken using soil columns with soil and groundwater from the site and carboys with groundwater and no soil to ascertain the degradation kinetics and to evaluate the ability of the local microbial species to bioremediate hydrocarbons under conditions more closely matched to site conditions than previous studies. The experiments were also designed to determine if soil-associated bacteria accelerate biodegradation. Laboratory microcosms were set up using vertical soil columns in triplicate with soil collected from the site. Microcosms without soil were set up in 12-liter carboys in duplicate. Diluent-contaminated groundwater from Well 204-A of the Guadalupe site was recirculated through the soil columns and was constantly stirred and aerated in the carboys during simultaneous 150-day experiment periods. Biodegradation rates were iv

5 determined by measuring the total petroleum hydrocarbon (TPH) concentration over time. Also monitored during the 150 day period was toxicity by Microtox Vibrio fischeri bacteria, terminal restriction fragment (TRF) bacterial assays, dissolved oxygen, ph, nutrient levels (SO 2-4, NO - 3, NO - 2, NH 3 & PO - 4 ) and total organic carbon (TOC). Observed decreases in TPH concentration were similar for all setups with the exception of an azide-inhibited control soil column, which showed an increase in TPH concentration. The largest TPH decrease observed was for the soil columns. For the 150-day experiment beginning with an initial TPH concentration of 2300 ± 100 µg/l, the TPH concentration in the azide control increased to 4600 µg/l, the two carboy setups dropped to 930 ± 84.9 µg/l, and the remaining two soil columns dropped to 760 ± 0.0 µg/l indicating soil biofilms enhanced long-term biodegradation. Although the observed difference was not large, these results show a statistically significant effect of the presence of soil, suggesting the soil increased the bioremediation both for the first 30 days and the last 120 days under these conditions. Since TPH did not decrease in the azide-inhibited control, adsorption onto soil particles is unlikely and observed TPH loss in the soil columns can be attributed to biodegradation by bacterial cells. Hydrocarbon biodegradation followed first-order kinetics for the first 30 days for both soil columns and carboys, with first-order rate constants of ± day -1 and ± day -1, respectively. After 30 days the biodegradation rates decreased and appeared to follow zero-order kinetics. Toxicity decreased rapidly during the first 15 days to the point where samples were non-toxic or no EC 50 values could be calculated. v

6 ACKNOWLEDGEMENTS This project was funded by the Unocal Corporation through the Environmental Biotechnology Institute at California Polytechnic State University in San Luis Obispo. The author would like to acknowledge the assistance of Chris Kitts, Alice Hamrick, and all of the biology students. The author wishes to thank the Guadalupe Unocal employees, especially Paul Lundegard for valuable discussions and Bob L. Pease of BFJ Services for collecting and coordinating all the samples in the field. The author thanks Campus Market for having $3.95 turkey sandwiches and Topango s for the $1.09 bean and rice burrito, which provided the necessary sustenance for so many long days and late nights working in the labs. I would like to thank my girlfriend Kristi, who supported and put up with me during my graduate studies. Lastly, I would like to acknowledge the everlasting, continuous support, mentorship, and friendship of Dr. Yarrow Nelson; without his support I would have not been interested or even started pursuing a Master s Thesis. Thanks for the hikes, bike rides, all of the advice and the wonderfully positive attitude toward life in general, especially during those critical times fighting cancer and really never showing that it got to you. vi

7 TABLE OF CONTENTS LIST OF TABLES...viii LIST OF FIGURES... ix CHAPTER 1 INTRODUCTION... 1 CHAPTER 2 BACKGROUND Site Information Remediation Field Experiments Bioremediation at Guadalupe Natural Attenuation Background Natural Attenuation Mechanisms and Process Definitions Limitations of the Natural Attenuation Process Natural Attenuation of Hydrocarbons Natural Attenuation Studies at Guadalupe CHAPTER 3 METHODS & MATERIALS Contaminated Groundwater Selection Soil Selection Soil Column Experimental Setup Soil Column Experiment Operation Non-Soil Carboy Experiment Set-up and Operation Groundwater Sampling and Experiment Start Up Sampling Procedures CHAPTER 4 RESULTS & DISCUSSION TPH Results for Soil Columns and Microcosms TPH Composition changes during biodegradation Toxicity Results Nutrient Concentrations TRF Results TOC Results CHAPTER 5 CONCLUSIONS CHAPTER 6 REFERENCES Appendix A Zymax-TPH Measurements: The Advantage of Using GC/MS Appendix B Time-course TPH Analysis Chromatographs and Simulated Distillation Results vii

8 LIST OF TABLES Table 3.1 Historical Groundwater Characterization Data for Well 204-A Table 4.1 TPH Concentrations of Initial Final Samples Table 4.2 TPH Concentration over time for soil columns and non-soil carboys Table 4.3 Simulated Distillation Values of Initial and Final TPH Table 4.4 Initial Microtox Data for Day Table 4.5 Microtox % Effect Data Table 4.6 Inorganic Nutrient Analysis of Well 204-A Groundwater, Summary of Analytical Results Table B.1 10 Day Carbon Data Table B.2 20 Day Carbon Data..71 Table B.3 30 Day Carbon Data..72 Table B.4 50 Day Carbon Data..72 Table B.5 90 Day Carbon Data..73 Table B Day Carbon Data viii

9 LIST OF FIGURES Figure 2.1 Guadalupe Site Location in California (Yahoo! Maps 2004)... 4 Figure 2.2 Former Guadalupe Oil Field looking southeastward from the Pacific... 7 Figure 2.3 A closer view of the Guadalupe dunes ecosystem... 7 Figure 3.1 Guadalupe Site Map Figure 3.2 Inset Location Detail for Well 204-A Figure 3.3 Guadalupe Site Map Legend Figure 3.4 Groundwater from Guadalupe Well 204-A Figure 3.5 Schematic of Soil Column Groundwater Flow Figure 3.6 Experimental Column Setups Figure 3.7 Soil Column Reservoir System Figure 3.8 Column overflow drawtube being crafted by Robin Cunningham Figure 3.9 Experimental Carboy Setups Figure 3.10 SDI M500 Microtox Analyzer Figure 3.11 Shimadzu TOC-5000A Figure Day Time-course TPH Concentrations Figure 4.2 Ln TPH Concentration over 150-Day Experiment Period Figure 4.3 Individual Carbon Range 30-Day Degradation Constants for Soil Columns.. 42 Figure 4.4 Individual Carbon Range 30-Day Degradation Constants for Carboys Figure 4.5 Summary of Average Final TPH Results for Soil Columns and Carboys w/o Soil Figure 4.6 Averaged Simulated Distillation Curve Results for Soil Columns #1 & # Figure 4.7 Averaged Simulated Distillation Curve Results for Carboys #1 & # Figure 4.8 Simulated Distillation Curve Results for Soil Column #3 (Azide Inhibited).. 47 Figure 4.9 Percentage of Initial Carbon Range Removed after 150 Days Figure 4.10 Initial Microtox Data for Day Figure 4.11 Soil Column Nutrient Levels versus Time Figure 4.12 Average of Carboys Nutrient Levels versus Time Figure 4.13 Sulfate Concentrations for 150-Day Experiment Figure 4.14 TOC versus Time for Days Figure B.1 10 Day TPH Chromatograph of Soil Column #1 with TPH ~ 1400 µg/l Figure B.2 10 Day TPH Chromatograph of Soil Column #2 with TPH ~ 1400 µg/l Figure B.3 10 Day TPH Chromatograph of Carboy #1 with TPH ~ 1300 µg/l Figure B.4 10 Day TPH Chromatograph of Carboy #2 with TPH ~ 1400 µg/l Figure B.5 20 Day TPH Chromatograph of Soil Column #1 with TPH ~ 970 µg/l Figure B.6 20 Day TPH Chromatograph of Soil Column #2 with TPH ~ 1000 µg/l Figure B.7 20 Day TPH Chromatograph of Carboy #1 with TPH ~ 1300 µg/l Figure B.8 20 Day TPH Chromatograph of Carboy #2 with TPH ~ 1400 µg/l Figure B.9 30 Day TPH Chromatograph of Soil Column #1 with TPH ~ 880 µg/l Figure B Day TPH Chromatograph of Soil Column #2 with TPH ~ 850 µg/l Figure B Day TPH Chromatograph of Carboy #1 with TPH ~ 1300 µg/l Figure B Day TPH Chromatograph of Carboy #2 with TPH ~ 1100 µg/l Figure B Day TPH Chromatograph of Soil Column #1 with TPH ~ 940 µg/l ix

10 Figure B Day TPH Chromatograph of Soil Column #2 with TPH ~ 640 µg/l Figure B Day TPH Chromatograph of Carboy #1 with TPH ~ 790 µg/l Figure B Day TPH Chromatograph of Carboy #2 with TPH ~ 810 µg/l Figure B Day TPH Chromatograph of Soil Column #1 with TPH ~ 830 µg/l Figure B Day TPH Chromatograph of Soil Column #2 with TPH ~ 640 µg/l Figure B Day TPH Chromatograph of Carboy #1 with TPH ~ 960 µg/l Figure B Day TPH Chromatograph of Carboy #2 with TPH ~ 1100 µg/l Figure B Day TPH Chromatograph of Soil Column #1 with TPH ~ 760 µg/l Figure B Day TPH Chromatograph of Soil Column #2 with TPH ~ 760 µg/l Figure B Day TPH Chromatograph of Carboy #1 with TPH ~ 990 µg/l Figure B Day TPH Chromatograph of Carboy #2 with TPH ~ 870 µg/l Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 2800 µg/l 92 Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 2700 µg/l 92 Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 2900 µg/l 93 Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 2900 µg/l 93 Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 3600 µg/l 94 Figure B Day Chromatograph of Soil Column #3(Control) w/ TPH ~ 4600 µg/l x

11 CHAPTER 1 INTRODUCTION The Guadalupe Oil Field is located on the central coast of California near the city of Santa Maria. Historically, crude oil pumped from this site was extremely viscous and unable to be pumped. It was thinned using a refined mid-range crude known as diluent from a nearby oil refinery to facilitate pumping. The diluent, a mixture of hydrocarbons in the range of C -C (Lundegard et al., 2001), used at the site was transported through pipes to the oil field and stored in tanks. During the transfer of diluent, many of these pipes, tanks and fittings leaked or spilled diluent into the sand dunes at 90 cataloged leakage sites with between 8.5 and 20 million gallons of leaked diluent. The soil and groundwater at the site is thus heavily contaminated with diluent. The site is comprised of more than 2,700 acres of sand dunes adjacent to the Santa Maria River channel and the Pacific Ocean. Due to the extensive cleanup efforts at the site, many traditional and novel remediation methods have been demonstrated. Horizontal biosparging, steam extraction, excavation, pump and treat, phytoremediation, and land treatment units have been evaluated at the site. These methods have achieved different levels of success for removing the diluent contaminant. Endangered species such as the red-legged frog and snowy plover thrive at the site, and thus it is important that the natural dune ecosystems remain undisturbed as much as possible during remediation activities. Natural attenuation, which is the use of in-situ 1

12 native microorganisms to aid or assist the degradation or extraction of contaminants, is an attractive technology for the Guadalupe site because excavation is not required, extensive and expensive equipment is not used and the low cost monitoring are a plus for all parties. Natural attenuation is the perfect culmination of all the remediation technologies used. It acts as a final polishing step after the more active remediation methods have been completed. The trace amounts and low concentrations of residual contaminants are well suited for final treatment by natural attenuation processes. It has been difficult to evaluate the efficacy of natural attenuation in the field due to the non-uniformity of the site. Differences of soil contamination levels and composition, groundwater flow rates and direction, variability of TPH measurements and sampling locations all contribute to the complexity of evaluating natural attenuation in the field. Due to the uncertainty of variables in the field, bench scale laboratory experiments were conducted to control these variables while investigating natural attenuation. The experiments described in this thesis were designed to determine if native microbes could facilitate biodegradation naturally and to determine rates of diluent biodegradation. In order to determine if soil bacteria enhance biodegradation, TPH degradation rates for soil columns were compared to non-soil carboys on a bench scale level. This comparison showed evidence of TPH biodegradation enhancement using soil as compared to no soil. One of the key issues at the Guadalupe site is the determination of timeframes for hydrocarbon degradation or simply degradation rates. Can laboratory experiments accurately reflect the actual changes in concentration, composition and toxicity of 2

13 contaminates at the site? Our strategy is to do the best we can. The research presented in this paper investigates the observed natural rate of hydrocarbon degradation by laboratory experiments under conditions very closely matched to the field conditions. Diluent-contaminated groundwater was recirculated through soil columns and stirred in glass carboys and the biodegradation rates of the diluent contaminant were measured. TPH was measured at the beginning and at the end of each experiment as well as monitored at regular intervals. A control was run to check that there was no loss of contaminant to the experimental apparatus. The influence of the bacterial soil biofilms on biodegradation rate was also examined by comparing biodegradation rates with and without soil. The major difference between the two systems was the addition of Guadalupe soil to the columns. Our goal was to compare the rates of TPH degradation and observe the differences for the two independent systems. 3

14 CHAPTER 2 BACKGROUND 2.1 Site Information The Guadalupe Oil Field is located near the town of Guadalupe, on the central coast of California as shown in Figure 2.1. This area is approximately halfway between San Francisco and Los Angeles. Guadalupe Site N Figure 2.1 Guadalupe Site Location in California (Yahoo! Maps 2004) The site is composed of more than 2,700 acres (1,100 hectares) of migrating and fixed sand dunes (Figure 2.2). To the north of the site is the Mobil Coastal Reserve, to the 4

15 south is the Santa Maria River Channel and to the west is the Pacific Ocean. Access to the site by the public has been restricted for over 50 years due to oil production and now due to remediation activities. This restriction has preserved one of the most intact coastal sand dune ecosystems left in California as shown in Figure 2.3. Ultimately site management will be handed over to the Fish and Game Department under a conservation easement and the site will become a dedicated nature preserve. Eventually all roads, piping and remaining remediation equipment will be removed from the site and it will be allowed to return to its natural state. The depth to groundwater fluctuates as the sand dunes change in elevation, with depths to groundwater ranging from feet (0 40 m) below the dune surface. Approximately 20 feet (6 m) below the water table surface is a confining clay layer. This creates a shallow aquitard under the dunes. The ground water surfaces seasonally, creating small marshes and wetland areas in the valleys of the sand dunes. The groundwater at the site generally flows from east to west at approximately 1 foot per day (30 cm/day). In order to pump the crude oil at the site, it was necessary to pump diluent, a diesel range oil (DRO), into the well to thin the crude. The diluent hydrocarbon mixture was produced at a nearby oil refinery (and other sources) and was pumped from the refinery through pipes in the dunes and stored in two central diluent tanks at the field. Many of these pipes, tanks and fittings were buried by the migrating dunes, corroded by the salty ocean air and leaked. Today these pipes and tanks are attributed with 90-cataloged 5

16 leakage sites and having 8.5 million gallons (32,176,000 L) of diluent leaked during 50 years of petroleum production at the site ( In January 1988, the oil field reached peak capacity with 215 production wells pulling heavy crude from approximately 3,000 feet (915 m) below the ground. In 1990 Unocal employees and local surfers noticed a petroleum smell emanating from the beach and in February 1990 Unocal reported oil on the beach and suspended field operations. At this point Unocal permanently discontinued the use of diluent. In a 1998 court decision, the State of California required Unocal to pay a $43.8 million settlement to include funding for restoration, replacement and rehabilitation efforts involving natural resources related to the dune sand aquifer petroleum contamination. Unfortunately, the future costs for remediation were not included in the settlement. Since the cleanup and abatement order of 1998, an estimated 145 miles (233,355 m) of pipeline and millions of gallons of diluent have been removed from the site. Today the site is becoming one of the most monitored and characterized spill sites in the nation with over 800 active groundwater monitoring wells and 3,000 soil borings logged. 6

17 Figure 2.2 Former Guadalupe Oil Field looking southeastward from the Pacific Figure 2.3 A closer view of the Guadalupe dunes ecosystem 7

18 2.2 Remediation Field Experiments Various types of biological treatment (bioremediation) exist for hydrocarboncontaminated soils. Examples are pump and treat, land farming, phytoremediation, natural attenuation, enhanced natural attenuation and bioaugmentation. These contaminant removal techniques can work efficiently when matched with the appropriately contaminated soils. Bioremediation techniques stimulate the microorganism population and use the contaminants found in soils as a food and energy source by creating a favorable environment for the microorganisms. Generally, this means providing some combination of oxygen and nutrients and ph. If necessary, microorganisms adapted for degradation of the specific contaminants are applied to enhance the process. This process is better known as bioaugmentation. This technique, involves the use of microbial cultures specially bred or cultured for degradation of a variety of contaminants and sometimes for survival under unusually severe environmental conditions (Mishra et al., 2004). Sometimes, microorganisms from the remediation site are collected, separately cultured, and returned to the site as a means of rapidly increasing the microbial population at the site (Cunningham et al., 2004). Usually, an attempt is made to isolate and accelerate the growth of the population of natural microorganisms preferentially feeding on the contaminants at the site. In some situations different microorganisms may be added at different stages of the remediation process because the contaminants in abundance change as the degradation process proceeds. This provides a diverse microbial population more adaptable to changing conditions. 8

19 Another bioremediation technique used is land treatment, which is a bioremediation technology in which contaminated soils, sediments, biosolids, or sludges are mixed with water and are tilled to allow interaction with the soil and climate at the site. The soils are periodically tilled to aerate the waste. Organic bulking agents (alfalfa, straw, or sawdust) can be applied to the soil mixture to accelerate biodegradation rates (Rivera-Espinoza et al., 2004). The bulking agents help aerate soil and can provide necessary nutrients. The applied water, contaminated soil and microbes interact dynamically as a system to degrade, transform, and immobilize waste constituents. Soil conditions are often controlled to optimize the rate of contaminant degradation. Some conditions normally controlled include moisture content (usually by irrigation or spraying), aeration by tilling the soil with a predetermined frequency and ph adjustment (buffered near neutral ph by adding crushed limestone or agricultural lime) (Shen et al., 1994). Nutrients such as nitrogen, phosphorus and potassium (NPK) are also added as needed to amend nutrient deficient soils. Enhancements in bioremediation of hydrocarbons can be made with addition of simple commercial (NPK) fertilizers (Hess et al., 1996). These additions stimulate the microbial activity and provide much needed sustenance to depleted bacteria. Several field-wide techniques used for the treatment of contaminated soils at Guadalupe use bioremediation methods Bioremediation at Guadalupe At the Guadalupe site many different pilot-scale remediation experiments have been conducted, including biosparging steam extraction, phytoremediation, and full-scale 9

20 excavation of contaminated soil into land treatment units. The projects and experiments have helped to reduce the overall hydrocarbon contamination levels at the site. But these projects come at a high cost and can do damage to the natural ecosystems present. One simple solution to this problem is monitored natural attenuation. Natural attenuation can be used as a polishing step after these treatments in addition to serving as a remediation alternative. 2.3 Natural Attenuation Background Hydrocarbons are some of the most frequently occurring environmental contaminants. Hydrocarbon-degrading microorganisms are present in most ecosystems where contaminants may serve as organic carbon sources. Most environments are enriched with oil-degrading microbial communities soon after contamination (Margesin et al., 2000). Several environmental factors influence the intensity of hydrocarbon biodegradation in soil including: indigenous microbial populations involved, availability of nutrients, oxygen, ph, temperature, water content, bioavailability of contaminants, and soil properties. Often in contaminated areas, the conditions for microbial degradation may be limited. Bioremediation techniques accelerate the naturally occurring biodegradation by optimizing conditions for biodegradation through aeration, addition of nutrients and control of ph and temperature (Atlas et al., 1992). To assess the results of biological decontamination, it is necessary to measure the remaining hydrocarbon content in soil, and to observe the microbial processes (Joergensen et al., 1995). Microorganisms (bacteria) degrade contaminants and little to no residual treatment is required. However, some compounds may be decomposed into more toxic by-products during the bioremediation process but this has not been shown for hydrocarbons. Short chain 10

21 hydrocarbons are more preferentially degraded as compared to long chain hydrocarbons (Scott 2003). The biological aspects of the bioremediation processes are typically implemented at lower costs compared to more complex non-biological treatment processes. Although not all organic compounds are amenable to bioremediation, techniques have been successfully used to remediate soils contaminated by petroleum hydrocarbons (Atlas 1995). Treatability or feasibility studies are used to determine whether bioremediation would be effective in a given situation. The extent of study can vary depending on the nature of the contaminants and the characteristics of the site. For sites contaminated with common petroleum hydrocarbons (e.g., gasoline and/or other readily degradable compounds), it is usually sufficient to examine representative samples for the presence and level of an indigenous population of microbes, nutrient levels and soil characteristics (ph, porosity, and moisture). Once the indigenous microbial populations and processes are identified, they can be used as an indicator of degradation (Zytner et al., 2001) Natural Attenuation Mechanisms and Process Definitions Natural attenuation is the process of carefully monitoring a site for contaminant concentrations and determining if the natural degradation processes are proceeding at a consistent and timely rate. Monitored natural attenuation (MNA) will usually only be considered as a remedial action if the following conditions are found to hold: 1) There is a clear indication that the site currently poses no unacceptable risk to human health or the environment; 2) There is no active point source contamination; 3) Contaminant plume 11

22 contours are static or retreating; 4) Geochemical and/or hydrological data suggest a strong likelihood that attenuation processes are operative at the site; 5) Remediation goals are completed in an acceptable time frame (Restoration 1999). Static or retreating plumes are often a sign of ongoing biogeochemical attenuation in the subsurface. However, full-scale MNA implementation will ultimately require a clear understanding of the specific processes leading to those conditions. Not only must the specific processes be identified, they must also be quantified to the extent where their long-term reliability can be assured. Natural attenuation pathways are typically contaminant-specific and site-specific. Moreover, their identification and quantification will involve a practical investment of time and effort. It is, therefore, reasonable to focus MNA implementation efforts on only those sites where natural attenuation is likely to be feasible, and where it will account for a significant reduction in contaminant availability over time Limitations of the Natural Attenuation Process One possible limitation of MNA is the timeframe required for complete remediation. Since natural attenuation has no national standards by which to be gauged, there is no clear timeframe for degradation. However, the EPA has recently developed some preliminary guidelines for degradation timeframes. These guidelines give general information on the degradation timelines for classes of contaminants. For the guidelines to be applicable, each potential natural attenuation site must be inspected carefully to determine the site characteristics, degradation parameters and chemical composition of contaminants. 12

23 2.3.3 Natural Attenuation of Hydrocarbons Hydrocarbons are subject to natural attenuation processes because they are readily biodegradable, at least under aerobic conditions (Mills et al., 2003). The rate of microbial degradation of contaminants is influenced by the specific contaminants present, temperature, oxygen supply, nutrient supply, ph; the availability of the contaminant to the microorganism (clay soils can adsorb contaminants making them unavailable to the microorganisms), the concentration of the contaminants (high concentrations may be toxic to the microorganism), the presence of substances toxic to the microorganism (e.g., mercury; or inhibitors to the metabolism of the contaminant). Oxygen plays a vital role in bioremediation processes, particularly for hydrocarbons. Although microorganisms biodegrade hydrocarbons both under aerobic and anaerobic conditions, organic constituents have a more rapid rate of biodegradation under aerobic conditions as compared to anaerobic conditions (Atlas 1981). Therefore, aerobic processes are considered to be more important for the bioremediation of petroleum products. Anaerobic conditions may be used to degrade highly contaminated soils (Hess et al., 1996), which can be followed by aerobic treatment to complete the biodegradation of the partially degraded compounds plus the other contaminants. The general process of aerobic biodegradation is summarized in the following equation: Bacteria + Organics + O 2 + Nutrients CO 2 + H 2 O + Biomass + Byproducts + Energy In reference to current studies, the organics for this experiment are hydrocarbons. 13

24 Aerobic biodegradation of hydrocarbons involves the incorporation of oxygen by microbial cells via oxygenase enzymes to break the hydrocarbon bonds (Suthersan 1997). Therefore, consistent aeration is beneficial in stimulating the complete biodegradation of petroleum products. Stoichiometric analysis can be used to estimate the theoretical amount of oxygen required to aerobically degrade a given quantity of hydrocarbon. Generally, the degradation of 1 mg of a medium length hydrocarbon requires 3 to 4 mg of oxygen (Dragun 1998). A general formula used to describe the stoichiometry of hydrocarbon degradation is: C x H y + [x+(y/4)] O 2 x CO 2 + (y/2) H 2 O For example: If dodecane (C 12 H 26 ) is used to represent the total petroleum hydrocarbons (TPH) present in the samples, the aerobic biodegradation can be written as: C 12 H O 2 12 CO H 2 O The molecular weight of dodecane is 170 mg/mmol and oxygen is 32 mg/mmol. For aerobic biodegradation of 1 mmol of dodecane, the mass of dodecane consumed is thus 170 mg/mmol x 1 mmol = 170 mg dodecane. The mass of oxygen consumed is 32 mg/mmol x 18.5 mmol = 592 mg O 2, and the mass of carbon dioxide produced is 44 mg/mmol x 12 mmol = 528 mg CO 2 Therefore 592 mg O 2 / 170 mg/mmol dodecane = 3.48 mg O 2 /mg dodecane are theoretically consumed. When the same calculations are performed for the biodegradation of for octadecane (C 18 H 38 ) as the representative TPH compound, the ratio of oxygen to hydrocarbon consumed is 3.46 mg O 2 to 1 mg TPH. Thus, approximately 3.5 mg of oxygen is required aerobically to biodegrade 1 mg of TPH (dodecane) to CO 2 and H 2 O. These two chemical equations only account for respiration, and do not take 14

25 into consideration the amount of reactants consumed in microbial cell synthesis. Only part of the energy obtained by the oxidation reaction can be used for the decomposition of the contaminants. The remaining energy is used for the production of new cells and for cell maintenance. Therefore cell synthesis would increase expected O 2 consumption and CO 2 production Natural Attenuation Studies at Guadalupe Natural attenuation field experiments at the Guadalupe site currently consist of the quarterly analysis and monitoring of the 800+ monitoring wells and the soil borings as well as ongoing laboratory studies at California Polytechnic State University. Each area of the site is specifically monitored based on the contaminant concentration and location. The individual areas with high concentrations of contaminants are usually monitored more closely. In addition to Unocal monitoring the site, local water quality boards and the USGS are an integral part of the evaluation of the natural attenuation process at Guadalupe. Past natural attenuation work has been performed by Paul Lundegard and Paul Johnson and is detailed in the Source Zone NA Investigation at the Former Oil Field report (Lundegard et al., 2003). This paper defines the source zone as petroleum-impacted soils that are either in direct contact with groundwater, or are in indirect contact through dissolution and soil moisture infiltration. Natural attenuation laboratory work was also performed by Marie Dreyer (Dreyer 2004) simultaneously along with the work presented in this thesis. The strategy behind her 15

26 project was to determine if there was a correlation between TPH concentrations and TPH weathering. She carefully analyzed TPH concentrations for 21 wells at the former Guadalupe oil field over a period of 20 days in non-soil 2 L media bottles. The main outcome of her research showed TPH degradation rates of day -1 with an average rate of day -1. Strong correlations were made between biodegradability of TPH as a function of weathering. One assumption of the work was that low TPH concentrations were more biodegraded or weathered. This assumption was not proven and showed the need for a more closely monitored laboratory TPH weathering soil column experiments. Jason Waudby (Waudby 2003) performed experiments on biosparged groundwater at Guadalupe in which he concluded nutrient amendment and/or bioaugmentation were not limiting factors in the biodegradability of TPH. Thus adding nutrients or bacterium to treat the groundwater at Guadalupe were not necessary, at least for the groundwater he worked with. For the time-course experiment detailed in this thesis no nutrient amendments or bioaugmentation were performed. Evan Larson showed sustainability of natural attenuation based on the ratio of biochemical oxygen demand/ chemical oxygen demand (BOD/COD). 16

27 CHAPTER 3 METHODS & MATERIALS In these experiments diluent-contaminated groundwater from Guadalupe was passed through soil columns and stirred in non-soil carboys to estimate natural attenuation rates in the field and to see if bacterial biofilms on soil particles increase hydrocarbon biodegradation. Changes in the microbial community consuming the hydrocarbons were also examined using terminal restriction fragment (TRF) analysis. To examine the effects of the bacterial biofilms on the hydrocarbons it was necessary to eliminate as many variables as possible and to run controls to compare against. These goals were accomplished by recirculating groundwater under controlled conditions and monitoring TPH concentration in the groundwater over time. All laboratory conditions were matched as closely as possible to Guadalupe field conditions. 3.1 Contaminated Groundwater Selection The concentration and source of water to be used was carefully selected. The water used for the experiment was contaminated groundwater from the Guadalupe Site. A pure mixture of hydrocarbons was not used because the purity of the mixture would not mimic the weathered nature of the hydrocarbons in groundwater at the site. Free-product samples were also not used due to the potential lack of native hydrocarbon-degrading bacterial populations. The use of the actual groundwater from the site provided the greatest comparison to field conditions. For these reasons site groundwater was collected directly from Well 204-A (Figure 3.1). An enlarged detail map of the well site is found in Figure 3.2 and an explanation of site detail in Figure 3.3. The water from this well 17

28 Well 204-A See Figure 3.2 See Figure 3.3 Figure 3.1 Guadalupe Site Map 18

29 historically has had TPH concentrations of 5-7 mg/l (Table 3.1) and was expected to contain native bacterial populations capable of hydrocarbon degradation. Water was drawn from this well for use in experiments on November 18, 2003 and had a goldenbrown color (Figure 3.4). Several well casing volumes were pulled prior to sampling to ensure the samples were representative of the groundwater. The depth to groundwater was 85 feet and the temperature of the groundwater was 19.4 C. It had a ph of 7.06, a conductivity of 0.51 mω (milliohms) and a dissolved oxygen concentration of 0.46 mg/l. The samples had dissolved methane (CH 4 ) concentration of 0.18 mg/l, a carbon dioxide level of mg/l and dissolved iron concentration of 8.4 mg/l. Two 1-Liter samples were obtained and taken to Zymax Labs for an array of analyses. The TPH concentrations measured were unusually low for the two initial samples at 2400 µg/l and 2200 µg/l. These samples were mixed before use in laboratory experiments and the average of the sample (2300 µg/l) was used as the initial TPH concentration. This was lower than the historical TPH concentrations but was sufficient for these experiments. This lower TPH concentration is most likely due to a historical decrease in the concentration of TPH for Well 204-A due to changing groundwater levels. 19

30 Well 204-A Figure 3.2 Inset Location Detail for Well 204-A Figure 3.3 Guadalupe Site Map Legend Figure 3.4 Groundwater from Guadalupe Well 204-A 20

31 Summary of Groundwater TPH Analytical Results for Well 204-A Guadalupe Restoration Project LFR All results in milligrams per Liter (mg/l) Area Monitoring Date Sample TPH TPH TPH Carbon Range Well Sampled Name Method PQL Min Max EPA 8015 Diluent 204-A 08/03/ A DIES Tanks 204-A 03/11/94 GW204A GCMS DIES A 02/25/95 GW204-A GCMS DILU A 10/25/95 GW204-A GCMS DILU A 01/17/96 GW204-A GCMS DILU A 04/12/96 GW204-A GCMS DILU A 06/24/96 GW204-A GCMS DILU A 11/03/96 GW204-A GCMS DILU A 01/22/97 GW204-A GCMS DILU A 07/14/97 GW204-A GCMS DILU A 10/10/97 GW204-A GCMS DILU A 02/19/98 GW204-A GC/MS DILU A 11/21/98 GW204-A GC/MS DILU A 01/16/01 GW204-A GS/MS DILU A 09/16/03 GW204-A GC/MS DILU A 11/18/03 GW204-A GCMS DILU A 11/18/03 GW204-A GCMS DILU Notes: TPH = Total Petroleum Hydrocarbon PQL = Practical Quantitation Limit Samples Analyzed by Zymax Envirotechnology: ND = Not Detected Above Laboratory Detection Limits Table 3.1 Historical Groundwater Characterization Data for Well 204-A 3.2 Soil Selection The source of soil and TPH content of soil to be used was also carefully selected. The soil was chosen so that native hydrocarbon-degrading bacteria might be present. A soil concentration of 0 mg/kg TPH was desired to guaranty that no TPH would be leached into the recirculation water returned to the reservoirs. The soil selected was sand from Ken Hoffman s prior phytoremediation laboratory study (Hoffman 2003). This soil was used in a microcosm, which showed nearly complete biodegradation of TPH hydrocarbons. The TPH of the soil was near zero but active bacteria were likely to be 21

32 present because of the recently observed TPH biodegradation in this soil. Therefore the soil used for the current study was collected from the remnants of the experimental phytoremediation setup. The willow roots from the phytoremediation experiment were removed and the sand was sieved with a #40 (U.S.) standard soil sieve. This soil was used in hopes of having the same native soil and bacteria as that in the field. 3.3 Soil Column Experimental Setup Each individual column setup (see Figure 3.5 and Figure 3.6) consists of three parallel 4- Liter glass columns (10 cm outer diameter and 60 cm long), and a 4-Liter glass reservoir. Groundwater was recirculated using a Masterflex peristaltic pump drive (1-100 rpm) (Model ) with Viton tubing. (Figure 3.7) The influent peristaltic pump tubing used Masterflex Viton microbore two-stop tubing (blue/blue stops) which had an inner diameter (I.D.) of approximately inches (1.65 mm) with flowrates of ~ 3.00 ml/minute. The effluent peristaltic pump tubing used Masterflex Viton microbore twostop tubing (purple/white stops) which was approximately inch (2.79 mm) I.D. for the effluent water with flowrates of ~ 6.00 ml/minute. The effluent return flowrate was higher than the inlet flowrate, which maintained a constant liquid head at the top of each soil column. The inlet of the reservoir s glass drawtubes were placed near the bottom of the reservoir so flows could be maintained during low level reservoir conditions. The reservoir drawtubes (Figure 3.7) were made from 50 cm of 6 mm I.D. Pyrex glass tubing. The column overflow drawtubes (Figure 3.6) were constructed of 25 cm length of 9 mm I.D. Pyrex glass tubing. Both drawtubes were bent to be in U shaped configurations. Small pieces of Masterflex L/S Tygon F-4040-A Fuel Grade Tubing 22

33 (yellow-tinted) were used as connectors for the Pyrex glassware (Figure 3.7). This allowed for some flexible joints in the tubing system. Column Overflow Tubing (Glass) 9 mm ID Glass Tubing # 20 Stopper Glass Column Vol. ~ 4 Liter Guadalupe Sand Column Stand (Wood) Glass Fiber Filter Peristaltic Pump Effluent Tubing Influent Tubing Reservoir Figure 3.5 Schematic of Soil Column Groundwater Flow 23

34 Setup #1 Setup #2 Setup #3 Figure 3.6 Experimental Column Setups 24

35 Thermometer Return Port Reservoir Drawtubes Recirculated Groundwater Flow Temperature Control Tubing 3 L Reservoir Tygon Tubing Pump Drive Head Viton Tubing Figure 3.7 Soil Column Reservoir System 25

36 It was critical that the flow rate and velocity of the recirculating water through the soil columns be similar to field measurements (approximately 1 ft/day or 30.5 cm/day). For this reason flow rates to mimic site conditions were estimated and used as shown below: Model flow estimation: Q = n * v * A Q = Flow (Liters/day) n = Porosity = Assumed to be 0.42 for Guadalupe site sand (Personal correspondence with Don Eley, Registered Geologist for LFR Associates) v = Actual Groundwater Velocity at site (cm/day) => Site data = meters/year = cm/day (1ft/day) A = Column Cross-Sectional Area (cm 2 ) = π*(d) 2 /4 d = Inner diameter of column (cm) = 10 cm Q = Flow = 0.42*30.48 cm/day*π*(10 cm) 2 /4 = cm 3 /day = 1.05 L/day Therefore, water was recirculated at a rate of 1.0 L/day, to mimic site hydraulic conditions. This was measured and checked by taking the heights of water in the columns at different times and determining the change in distance per change in time. The Darcy velocity of the water flowing in the columns filled without sand was cm/day (0.42 foot/day) by adjusting the pump-controller flowrate setting on the dial to number 1.5. Glass would provide the best tubing conduit for the recirculated groundwater, because plastic has the potential to absorb hydrocarbons when in contact for extended periods of 26

37 time. Thus, the outcome of the experiment (TPH concentrations) could be dramatically altered if plastic tubing was used. Knowing this, the setup was constructed almost entirely of Pyrex glass tubing. This required an extensive amount of preparation work to construct the experimental setups. Most of this prep time was spent learning to blow glass and perfecting the bends in the glass tubing. If the bends are not made properly the glass is susceptible to cracking, the cross-sectional area is reduced, flowrates could decrease and velocities could increase. Luckily the learning curve was fast and only about 40 hours worth of time was spent behind the torch (Figure 3.8). Each section of glass and each bend were carefully measured to fit perfectly into the soil column stand (Figure 3.6). Figure 3.8 Column overflow drawtube being crafted by Robin Cunningham The columns were filled with Guadalupe site sand (from the previous phytoremediation experiments) to within 10 cm (4 inches) of the top of the columns. A single Masterflex 27

38 peristaltic pump with two eight-channel heads was used to provide the three columns in each soil-column setup with groundwater from the reservoir. The groundwater was pumped from the bottom of the column to the top (opposite of gravity flow). This was designed to have the soil in the columns in contact with groundwater at all times. The outlet of the columns was at the top and overflow drawtubes were installed to pump the effluent back into the reservoir. The effluent pump tubing was larger than the influent pump tubing so the water level on the top of the soil column was always maintained because the effluent return pump had a higher flowrate than the influent. The reservoir container was a four-l glass sample jar with a wide mouth ( L) (Figure 3.7). The range of water levels in the reservoir was due to scheduled periodic sampling. Reservoir water levels were marked on the outside of the reservoirs after each sample was taken. Any loss of water due to evaporation was compensated by the addition of de-ionized water to the reservoir to maintain the reservoir liquid at a proper level. The quantity of water in the reservoir was kept to a minimum to insure the majority of groundwater was kept in contact with soil in the columns. Each reservoir was aerated using an airstone and continuously stirred with a Scinics Co. Multistirrer (Model MC 303). The temperatures of the columns and the reservoirs were kept constant at 19 C. This water/ground temperature has been documented at the Guadalupe site during pumping of Well 204-A. The columns and reservoirs were wrapped with 0.5-inch I.D. vinyl tubing (Figure 3.7) connected to a Fisher Scientific Isotemp temperature-controlled recirculating waterbath (Model 1006S). The total length of temperature-control tubing was approximately 500 linear feet. The temperature-control tubing was external to the 28

39 columns and reservoirs and was independent of the groundwater recirculation tubing. De-ionized (DI) water was added as needed to the waterbath reservoir in order to keep pumping levels sufficient and flowing. The waterbath unit was equipped with a digital temperature display in degrees Celsius. Additional checks for temperature were monitored by thermometer in the last reservoir (Figure 3.7-Soil column #1 reservoir). The reservoir for soil column setup #1 was the reservoir closest to the return port on the temperature-controlled waterbath. The temperature of this reservoir being the last in line would potentially be the most varied from the waterbath reservoir and therefore needed to be monitored most closely. The temperature was constant at 19 C ± 1 for the duration of the experiment. The designs for both soil and non-soil setups were partially-open to the atmosphere. Volatile emissions from the soil columns and carboys were not a concern because a companion study in our laboratory showed no volatilization of diluent hydrocarbons under similar conditions (Elliot 2002). Thus, due to the non-volatile nature of diluent, portions of the experimental setups were left unsealed and degradation was measured from change in TPH concentration of the recirculated water. As an extra precautionary measure carbon-traps were installed in the experimental setups as described above. 3.4 Soil Column Experiment Operation Biodegradation rates were measured with 3 sets of (3) 4-L columns (Figure 3.6 Experimental Column Setups). The first two sets of columns contained Guadalupe Soil with active bacteria. The third set of columns served as a control to determine if TPH 29

40 loss could be due to abiotic means or adsorption onto soil particles. This set of columns was filled with autoclaved soil originally from the Guadalupe site (Figure 3.1). The soilfilled control columns were autoclaved 5 times for two hours at a temperature of 121 C. The extra precaution in the autoclaving procedure was performed to insure no live bacteria existed in the soils. The groundwater used for the control set of columns (Soil Column Setup #3) was inhibited with 10,000 ppm sodium azide, which was added to the groundwater at the start of the experiment. Sodium azide is a known bacterial respiration inhibitor (Margesin et al., 2000), which essentially suffocates the bacteria and prevents them from consuming carbon. On November 18, 2003 a preliminary design check was run to evaluate the integrity of the soil column setups and attempt to determine initial pump flowrates. The setups had to be tested for water-tightness and pump flowrate conditions (i.e. what setting on the pump controller should be used). Dissolved oxygen (D.O.) and ph were measured for each column in each setup throughout the experiment s 150-day duration. The main purpose of monitoring and recording the dissolved oxygen levels was to determine if sufficient oxygen was present which is necessary for aerobic conditions. Sufficient oxygen levels were maintained and the flowrate of air entering the reservoirs through the air pump was not changed. The ph was only monitored and not adjusted during the experiment. 30

41 During the 150-day experiment water levels in the reservoirs routinely dropped due to sampling. To combat the ever lowering of the reservoir groundwater levels the third column in each soil column setup was taken off-line and allowed to drain into the reservoir. To aid the draining the pumping flow was reversed until the column appeared dry (usually several days). This was again performed for the second column as more water samples were removed (around Day 90). Each sampling required 1-2 Liters of groundwater. It would have been less difficult to use a much larger reservoir volume, but this groundwater would not have been in contact with soil and soil biofilms. 3.5 Non-Soil Carboy Experiment Set-up and Operation Biodegradation rates were also measured with two 12-L carboys without soil (Figure 3.9). Each carboy filled to capacity could allow a larger volume of groundwater (12-20 Liters) compared to each soil column setup with ~ 10 Liters of groundwater. The two carboys were housed inside of a large refrigerated Fisher Scientific Isotemp temperaturecontrolled incubator for the duration of the 150-day experiment and are being maintained post graduation to determine more long-term degradation rates. The temperature of the incubator housing the carboys was kept at 19 C to mimic temperatures of the site and to match the temperature of the soil columns. The carboys were continuously stirred and aerated using an air pump connected to an airstone placed inside the carboys. A carbon trap was installed on Carboy #2 as a volatility measure to check for TPH loss from evaporation (Figure 3.9). The carboys were sampled simultaneously along with the soil columns on Days 10, 20, 30, 50, 90 and 150. The non-soil carboys were used as a 31

42 complementary comparison to the soil columns during the experimental time-course analysis. Air Tube Carb Trap Air Pump Carboy #1 Carboy #2 Stir Plate Figure 3.9 Experimental Carboy Setups 3.6 Groundwater Sampling and Experiment Start Up The water used in this experiment was water taken from a groundwater monitoring well at Guadalupe (204-A) as described above. This water had been stored in (5)-20 Liter containers refrigerated in the laboratory for a few days before use. Before use in any experimental apparatus the 5 containers were dumped into a clean stainless steel tank and 32

43 mixed completely. This was to ensure the homogeneity of the samples used for the experiments. The initial samples taken in the field during the pumping of Well 204-A and the groundwater from the 5 mixed containers were assumed to be homogeneous in nature. Approximately 10 Liters of contaminated groundwater was added to each of the (3) soil column setups. Water levels were allowed to adjust in the columns and reservoirs due to evaporation and/or soil pore activity for the first few days. After acclimation, each reservoir was filled with less than 1 L of groundwater to maximize capacity. 3.7 Sampling Procedures Groundwater was recirculated in the columns and carboys for a total of 150 days. Routine sampling was performed on Days 10, 20, 30, 50, 90 & 150. Initial analyses and all results for Day 0 were performed on samples collected at the site during initial pumping and the data is used in the results section. The sampling procedures for determining TPH concentration required approximately 1 Liter of water placed in a 1 Liter amber sample bottle and then samples were sent to Zymax Labs (San Luis Obispo, CA) for analyses. TPH analyses of all groundwater samples used in the 150-day experiment were performed by Zymax Envirotechnology, Inc. (San Luis Obispo, CA). The samples were analyzed for TPH gas chromatography with mass spectrophotometry (GC/MS) (State of California method similar to EPA Method 8015), after extraction in methylene chloride (EPA Method 3510). The TPH was quantified against diluent standards, over an 33

44 analytical carbon-chain range of C 8 C 40 with a practical quantitation limit (PQL) of 50 µg/l. A detailed description of their method is included in Appendix A. For each sample date, Microtox toxicity tests (ASTM Standard D-5660), were performed using a SDI M500 Analyzer (Figure 3.10) and total organic carbon (TOC) levels were measured using a Shimadzu (Model TOC-5000A) (Figure 3.11). The toxicity and TOC samples required approximately 50 ml of water in a small sample container. The initial Microtox and TOC analyses for Day 0 was performed by Marie Dreyer (Dreyer 2004). The initial and 90-day microtox test used the 45 % basic concentration dilution test, whereas the remander of the test runs used the 81.9 % basic concentration dilution tests. The 81.9 % basic test was recommended by the Azur Microtox software technical support as being better suited for groundwater toxicity tests due to their low or diluted toxicity. The 45 % basic test is better suited for wastewater and sludge toxicity testing and uses a maximum 45 % of the original concentration for the sample dilutions. Figure 3.10 SDI M500 Microtox Analyzer Figure 3.11 Shimadzu TOC-5000A 34