Bioremediation of Recalcitrant Compounds

Citation preview

Bofioremediation Relcalcitrant Compounds

Bofioremediation Relcalcitrant Compounds EDITED BY

JEFFREY W. TALLEY

Boca Raton London New York Singapore

A CRC title, part of the Taylor & Francis imprint, a member of the Taylor & Francis Group, the academic division of T&F Informa plc.

Published in 2005 by CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2005 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group No claim to original U.S. Government works Printed in the United States of America on acid-free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number-10: 1-56670-656-4 (Hardcover) International Standard Book Number-13: 978-1-56670-656-8 (Hardcover) Library of Congress Card Number 2005043923 This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data Bioremediation of recalcitrant compounds / edited by Jeffrey Talley. p. cm. Includes bibliographical references and index. ISBN 1-56670-656-4 1. Organic compounds--Biodegradation. 2. Organochlorine compounds--Biodegradation. 3. Bioremediation. I. Talley, Jeffrey. TD196.O73B558 2005 628.5--dc22

2005043923

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com Taylor & Francis Group is the Academic Division of T&F Informa plc.

and the CRC Press Web site at http://www.crcpress.com

Preface This book summarizes many of the results of a 7-year research effort conducted by the Federal Integrated Biotreatment Research Consortium (FIBRC). The purpose of the work presented in this book was to develop bioremediation technologies for soil, sediment, and groundwater contaminated with chlorinated solvents, polychlorinated biphenyls (PCBs), and polycyclic aromatic hydrocarbons (PAHs). The Strategic Environmental Research and Development Program (SERDP) sponsored this project under its cleanup thrust area and assigned it project number CU-720. The U.S. Army Engineer Research and Development Center (ERDC) directed the FIBRC research program, which was entitled “Biotreatment: Flask to Field Initiative.” Active membership of the FIBRC whose work is represented in this book consisted of the following organizations: U.S. ERDC, Environmental Laboratory, Waterways Experiment Station, Vicksburg, MS U.S. Army Natick Research, Development and Engineering Center, Natick, MA U.S. Army Corps of Engineers (USACE), Baltimore District, Baltimore, MD U.S. Naval Research Laboratory, Washington, D.C. U.S. Naval Command, Control and Ocean Surveillance Center Research, Development, Test and Evaluation Division (NRaD), San Diego, CA U.S. Environmental Protection Agency (USEPA), Environmental Research Laboratory, Athens, GA USEPA Robert S. Kerr Laboratory, Ada, OK Great Lakes and Mid-Atlantic Hazardous Substance Research Center (GLMAC), Ann Arbor, MI In addition, the following organizations participated in the FIBRC in an advisory capacity: ERDC, Cold Regions Research Engineering Laboratory (CRREL), Hanover, NH

ERDC, Construction Engineering Research Laboratory (CERL), Champaign, IL U.S. Army Environmental Center, Aberdeen Proving Ground, MD U.S. Department of Energy (DOE), Argonne National Laboratory, Argonne, IL I was the FIBRC director for most of the work, although Dr. Mark Zappi, P.E. (ERDC/Mississippi State University) and Dr. Kurt Preston (ERDC/Army Research Office) served as the initial directors. Dr. Rakesh Bajpai (University of Missouri–Columbia) served as an interim director for 1 year. Thank you, Mark, Kurt, and Rakesh, for providing great leadership. My executive assistant was Ms. Deborah Felt, Applied Research Associates (ARA), whose daily contributions were invaluable … thanks, Debbie. To my lead authors, Dr. Jim Tiedje (Michigan State University), Dr. Hap Prichard (Naval Research Laboratory), and Dr. Guy Sewell (EPA Robert S. Kerr Research Laboratory/East Central University), your long hours of work are greatly appreciated. Equally important is the list of other coauthors and contributors. My thanks to all of you. Thank you, SERDP, for funding this work and giving us great folks to work with, such as Ms. Cathy Vogel (SERDP’s program manager for the cleanup thrust area), Dr. Femi Ayorinde (program manager), and Mr. Bradley Smith and Dr. John Harrison (SERDP directors). Special thanks to ERDC and Daniel E. Averett for their support of this project. Thank you, Technical Advisory Committee (all great scholars and too many to mention here) and Mr. Richard Conway (SERDP project shepherd), for keeping us focused to produce this work. Thank you, Dr. Xiangru Zhang (University of Notre Dame) and Mr. Tim Ruggaber (University of Notre Dame) for your valuable assistance in proofreading this manuscript! Finally, thank you to all the staff at CRC Press and Taylor & Francis. Without your prodding, this book would never have been completed.

Jeffrey W. Talley Notre Dame, Indiana

The Editor Dr. Jeffrey W. Talley, P.E., is an Assistant Professor of Bioengineering and Environmental Engineering in the Department of Civil Engineering and Geological Sciences, University of Notre Dame, Notre Dame, Indiana. He specializes in the treatment of contaminated surface water, groundwater, soil, and sediment. His research interests include the examination of physiochemical and microbial processes for application to waste reduction and treatment. Special interests are phase partitioning and the treatment and fate of hydrophobic organic compounds (Dioxins, PCBs, PAHs, DDT), other tightly bound pollutants (TNT, RDX, HMX), and select inorganic contaminants (Hg, Pb, Cr, and As) in the environment. He is especially interested in the integration of engineering, microbial ecology, and toxicology for purposes of enhancing detection, characterization, and remedial strategies. Professor Talley teaches Introduction to Environmental Engineering and Science, Hazardous Waste Management and Design, and Physiochemical Processes and Treatment of Pollutants. Professor Talley is noted for his innovative applications of thermal programmed desorption mass spectrometry (TPD/MS) for the assessment of pollutants and his contributions to the development of treatment technologies with focus on field remediation. His work involving the bioavailability of PAHs in sediments was part of the team project honored as 1999 SERDP Research Project of the Year (Cleanup) and best research presented (poster) at the 2000 Gordon Research Conference on Environmental Science. His recent collaborative work involving the detection and analyses of toxic heavy metals and organic acids in herbal dietary supplements was part of a team project honored as best research presented at the 2004 International Symposium on Recent Advances in Pharmacology. Prior to his appointment at Notre Dame, Dr. Talley spent 20 years in design, consulting, and military positions involving more than 100 different environmental sites throughout the United States and abroad. In 2003, Professor Talley was in the Middle East with the U.S. Army Corps of Engineers conducting civil and environmental engineering projects throughout Kuwait and Iraq. Of special significance was his environmental work with Task Force Restore Iraqi Oil (TF RIO), where he assisted in the assessment and remediation recommendations for multiple oil-waste impacted sites in Iraq.

Contributors Dr. Michael Annable Department of Environmental Engineering and Science University of Florida–Gainesville Gainesville, Florida Dr. Herbert Fredrickson U.S. Army Engineer Research and Development Center Environmental Laboratory at Waterways Experiment Station Vicksburg, Mississippi John S. Furey U.S. Army Engineer Research and Development Center Environmental Laboratory at Waterways Experiment Station Vicksburg, Mississippi Dr. Lance D. Hansen U.S. Army Engineer Research and Development Center Cold Regions Research and Engineering Laboratory Hanover, New Hampshire Dr. John Hind Maryland Biotechnology Institute Baltimore, Maryland Desirée P. Howell RMT, Inc. Jackson, Mississippi

Dr. William Jones Maryland Biotechnology Institute Baltimore, Maryland Dr. Joanne Jones-Meehan Naval Research Laboratory Washington, D.C. Susan C. Mravik USEPA-RSK Lab Ada, Oklahoma Cathy Nestler Applied Research Associates, Inc., Southern Division Vicksburg, Mississippi Dr. Kurt D. Pennell School of Civil and Environmental Engineering Georgia Institute of Technology Atlanta, Georgia Dr. Hap Prichard Naval Research Laboratory Washington, D.C. Dr. Guy Sewell East Central University Department of Environmental Health Sciences Ada, Oklahoma

Dr. Randy Sillan LFR Levine Fricke Emeryville, California William Straube Geo-Centers, Inc. Washington, D.C. Dr. Jeffrey W. Talley, P.E. Department of Civil Engineering and Geological Sciences University of Notre Dame Notre Dame, Indiana Dr. James M. Tiedje Center for Microbial Ecology Michigan State University East Lansing, Michigan

Dr. Tamara V. Tsoi Center for Microbial Ecology Michigan State University East Lansing, Michigan Dr. Altaf Wani U.S. Army Engineer Research and Development Center Vicksburg, Mississippi Kevin Warner LFR Levine Fricke Emeryville, California Dr. A. Lynn Wood USEPA-RSK Lab Ada, Oklahoma

Introduction This book provides an authoritative state-of-the-art biotreatment review for three key contaminant groups: chlorinated solvents, polychlorinated biphenyls (PCBs), and polycyclic aromatic hydrocarbons (PAHs). Issues such as availability, toxicity, and treatability are discussed along with a summary of the latest bioremediation technologies. Special innovative research and development projects are presented for each contaminant group. These projects are the results of a 7-year concerted effort by the Strategic Environmental Research Development Program’s (SERDP) Federal Integrated Biotreatment Research Consortium funded by the Department of Defense (DOD), Department of Energy (DOE), and Environmental Protection Agency (EPA). The consortium’s objective was to develop field-ready biotechnologies. The technologies developed through this program yielded successful field or large-scale lab demonstrations for each contaminant group. Cosolvent extraction of chlorinated solvents was validated at a field site, and a guidance document explaining the technology was developed. Bioaugmentation-enhanced PAH degradation was compared to traditional land-farming methods and provided new insight on how to optimize biotreatment. Genetically engineered microorganisms (GEMs) that enhanced PCB degradation were developed and field tested, and a GEM guidance document was written. These projects discuss both the science and engineering challenges that were encountered as each project advanced from the flask to the field. They serve as useful guides for the implementation of any new bioremediation technology.

Contents Chapter 1 Introduction to recalcitrant compounds .............................................................1 Jeffrey W. Talley Chapter 2 Toxicological exposure of bound recalcitrant compounds ............................ 11 Herbert Fredrickson, John S. Furey, and Jeffrey W. Talley Chapter 3 Roadblocks to the implementation of biotreatment strategies .....................33 Jeffrey W. Talley Chapter 4 The federal integrated biotreatment research consortium (flask to field).........................................................................................................51 Jeffrey W. Talley Chapter 5 Chlorinated solvent contaminated soils and groundwater: field application of the solvent extraction residual biotreatment technology ......................................................................................59 Guy Sewell, Susan C. Mravik, A. Lynn Wood, Michael Annable, Randy Sillan, and Kevin Warner Chapter 6 Enhancing PCB bioremediation .......................................................................147 James M. Tiedje, Tamara V. Tsoi, Kurt D. Pennell, Lance D. Hansen, Altaf Wani, and Desirée P. Howell Chapter 7 Polycyclic aromatic hydrocarbons (PAHs): improved land treatment with bioaugmentation......................................................................215 Hap Prichard, Joanne Jones-Meehan, Cathy Nestler, Lance D. Hansen, William Straube, William Jones, John Hind, and Jeffrey W. Talley

Chapter 8 Future needs for research and development .................................................301 Jeffrey W. Talley Index .....................................................................................................................305

chapter one

Introduction to recalcitrant compounds Jeffrey W. Talley Contents 1.1 Introduction ....................................................................................................1 1.2 Relevance.........................................................................................................2 1.3 Biodegradation and bioavailability.............................................................3 1.4 The sequestration of recalcitrant compounds ...........................................4 References.................................................................................................................7

1.1 Introduction Bioremediation is defined by the U.S. Environmental Protection Agency (EPA) as a managed or spontaneous process in which microbiological processes are used to degrade or transform contaminants to less toxic or nontoxic forms, thereby remedying or eliminating environmental contamination (EPA, 1994). These microbiological processes may reduce hydrocarbon concentrations in various types of soils and sediments to levels that no longer pose an unacceptable risk to the environment or human health (Linz and Nakles, 1997). However, hydrocarbons that remain in treated soils and sediments still might not meet stringent regulatory levels, even if they represent site-specific, environmentally acceptable endpoints (NRC, 1997). This unresolved issue of the availability of residual hydrocarbon contaminants is the focus of this work. There is a great need to understand contaminant soil–sediment interactions and their effect on bioavailability and toxicity (NCR, 1997). This is especially true for recalcitrant compounds. The adherence and slow release of recalcitrant compounds from soils and sediments is an obstacle to

1

2

Bioremediation of Recalcitrant Compounds

remediation (NCR, 1994; Moore et al., 1989) and is challenging our concepts about cleanup standards and risks (Alexander, 1995). This is particularly the case for biological treatment of recalcitrant compounds, in which one of the most important site-specific factors is the availability of the compounds held within solids and how this affects treatment rates and acceptable toxicological endpoints. Biostabilization is a newly developed concept that could significantly benefit the remediation process for soils and sediments contaminated with recalcitrant compounds. Biostabilization is the biodegradation of accessible pollutant fractions in a soil or sediment matrix, leaving a bound residue that is much more biologically unavailable and immobile (Luthy et al., 1997). The concept, however, is still in a developmental stage, and endpoints and appropriate measures of endpoints have not been defined. Very little research has been conducted on the applicability of biostabilization principles. It has been hypothesized that residual hydrocarbons remaining after biotreatment may represent an acceptable treatment endpoint. This is an important concept, but for wide acceptance, better understanding of the polycyclic aromatic hydrocarbon (PAH) release rates and mechanisms that bind or sequester recalcitrant compounds within contaminated material is required.

1.2 Relevance Recalcitrant compounds present one of the most pressing problems for biotreatment of contaminated soils or sediments. These compounds constitute a broad class of chemicals that appear as persistent contaminants in soils and sediments (NRC, 1997, 1994), including petroleum and fuel residues, tars, and creosotes. Such compounds have low solubility, low volatility, low intrinsic reactivity, and they typically exhibit very slow release rates from soil or sediment (Steinberg et al., 1987; Pignatello and Xing, 1996). These characteristics make biotreatment difficult in any geologic setting, necessitating extensive, site-specific testing. These characteristics also confound risk assessment, wherein the availability of recalcitrant compounds to organisms or the environment is unknown. In theory, soils and sediments contaminated with recalcitrant compounds may be treated utilizing various cleanup strategies. However, many proposed strategies have significant economic and feasibility problems. What is needed is an effective technology that supports the economics of disposal, eliminates adverse contaminant impacts, and supports the reuse of treated contaminated soils and sediments. In-place (in situ) biotreatment combined with biostabilization offers a strong possibility to achieve these goals. Regardless of whether the biotreatment system is passive (natural attenuation) or engineered (in-place treatment), a pragmatic solution is to focus active biotreatment on the available contaminant fraction and confirm that if residuals are released, the rate is sufficiently slow to allow consumption by the microbial community, i.e., the contaminant is biostabilized. Thus, an alternative environmental endpoint (cleanup level) may be appropriate,

Chapter one: Introduction to recalcitrant compounds

3

rather than basing decisions solely on total concentration of contaminant in the soil or sediment. It is also necessary to improve assessment of the toxicity and risk from residual hydrocarbons. Mechanisms that affect release rates and exposure need to be better defined. The potential benefits of such work could include reduced treatment costs, improved evaluation and design for cleanup technologies, greater regulatory and public acceptance of biotechnology, increases in the reuse/recovery opportunities for treated contaminated sediments, and potential application for in situ capping of contaminated soils and sediments.

1.3 Biodegradation and bioavailability Recalcitrant compounds in soils and sediments may be biodegraded by microorganisms to a residual concentration that no longer decreases with time or that decreases slowly over years with continued treatment (Thoma, 1994; Luthy et al., 1994; Loehr and Webster, 1997). Further reductions are believed to be limited by the availability of the recalcitrant compounds to microorganisms (Bosma et al., 1997; Erickson, 1993). Attempts have been made to increase this availability through the use of surfactants, but results have varied (Putcha and Domach, 1993; Bury and Miller, 1993; Chung et al., 1993; Edwards et al., 1994; Auger et al., 1995). Additionally, as contaminants age, they become less available compared to freshly contaminated material. As a consequence of binding with soils and sediments and subsequent slow release rates, residual recalcitrant compounds may be significantly less leachable by water and less toxic as measured by uptake tests (Alexander, 1995; GRI, 1995; Kelsey et al., 1997). Generally, contaminants can be degraded only when they exist in the aqueous phase and come into contact with the cell membrane of a microorganism. In this way, the contaminant serves as a substrate for the microorganism and is incorporated through membrane transport into the cell and utilized as an energy source in the cell’s principal metabolic pathways. However, physical or chemical phenomena can limit the bulk solution concentration of the contaminant and thus significantly reduce the ability of the microorganisms to assimilate the contaminant. Therefore, the bioavailability of the contaminant can control the overall biodegradation of these compounds. Another important factor relevant to biodegradation and bioavailability is the location and density of microorganisms. The majority of bacteria in the environment are attached to surfaces, and their distribution in and on soils/sediments is very patchy. The majority of these bacteria range in size from 0.5 to 1.0 μm, whereas micropores present in soils and sediments measure far less than 1 μm. It is generally believed that bacteria are attached predominantly to the surfaces of soils and sediments and not to the interior surfaces of the micropores. It has been estimated that more than 90 of the microorganism types present in geologic matrices accumulate on the surfaces

4

Bioremediation of Recalcitrant Compounds

of soils/sediments (Costerton et al., 1987). Therefore, the majority of contaminant–microbial interactions occur in the biofilm that develops within macropores on the surfaces of soils/sediments. This suggests that partitioning of an organic contaminant from the solid phase of the soil/sediment to the aqueous phase in the larger pore spaces controls soil/sediment bioavailability. These partitioning mechanisms may include chemical bonding, surface complex formations, electrostatic interactions, and hydrophobic effects (Schwarzenbach et al., 1993; Stumm, 1992). For hydrophobic contaminants such as recalcitrant compounds, sorption increases with the content of the organic matter in the soil/sediment and the degree of hydrophobicity of the specific PAH. Typically, the rate of desorption can be attributed to the mass transfer of the sorbate molecules from sorption sites on and in the soil/sediment. Active bacteria should correspond to the higher available PAH concentrations, which occurs where desorption is the most intense.

1.4 The sequestration of recalcitrant compounds As already discussed, the partitioning of recalcitrant compounds represents a significant part of the mass transfer process. However, describing this process is complicated by a combination of diffusion, dissolution, encapsulation, and adsorption phenomena. The combined effect of these factors can generally be defined as sequestration (Luthy et al., 1997). The complexities of sequestration can best be understood in terms of examining the different scales of observation associated with the heterogeneity of soils and sediments. Shown in Figure 1.1 are mineral surfaces in meso- and micropores (Luthy et al., 1997), both amorphous and condensed natural sorbent organic matter (SOM), and anthropogenic organic phase matter. The circled letters refer to different sorption mechanisms associated with specific domains within soils and sediments: Case A represents absorption into amorphous or soft natural organic matter (e.g., within an organic phase akin to solvent partitioning). Case B represents absorption into condensed or hard organic polymeric matter or combustion residue (e.g., into solid-like organic matter). Case C represents adsorption onto water-wet organic surfaces (e.g., soot). Case D represents adsorption to exposed water-wet mineral surfaces (e.g., quartz). Case E represents adsorption into microvoids or microporous minerals (e.g., zeolites) with porous surfaces at a water saturation of BB, BE, where BA, BB, BC, BD, and BE are the bioavailabilities for cases A, B, C, D, and E. These different domains within soils/sediments illustrate how structural and chemical heterogeneity can significantly affect how recalcitrant compounds behave. Although some correlation may be hypothesized with respect to bioavailability, it is difficult to specify the exact role each sorbent domain may have. For example, adherent or entrapped anthropogenic organic matter can also function as a sorbent (e.g., surfactants, soot, or NAPLs such as oils and tars) (Edwards et al., 1994; Gustafsson et al., 1997; Boyd and Sun, 1990). Also, there is a growing awareness that the affinity of nonpolar organics for SOM depends on the SOM’s origin and geologic

6

Bioremediation of Recalcitrant Compounds

history. Grathwohl (1990) showed that organic matter in unweathered shales and high-grade coals enhanced sorption by more than an order of magnitude when compared to more oxidized organic matter found in recent geologically young material and highly weathered SOM. Huang and Weber (1997) have suggested that changes in SOM oxygen-containing functional groups, e.g., changes in SOM O/C atomic ratios, can lead to a greater affinity of SOM for recalcitrant compounds. Gustafsson et al. (1997) differentiated two forms of SOM — humic acid and soot — for sediment by thermal treatment (Grathwohl and Reinhard, 1993). Currently, there are very few direct observational data revealing the microscale location or locations in which recalcitrant compounds accumulate when they associate with soils and sediments (Gillette et al., 1999; Ghosh et al., 2000). As a result, researchers must rely on inferences from macroscopic experimental observations that capture overall behavior and provide empirical evidence for deducing sorbent/sorbate mechanistic models (Wu and Gschwend, 1988; Nkedi-Kizza et al., 1989; Brusseau et al., 1991; Grathwohl and Reinhard, 1993; Werth and Reinhard, 1997; Huang and Weber, 1997; Gustafsson et al., 1997). There is no reason to presume that only one sorption mechanism dominates in any particular case. Indeed, in real systems, more than one process likely contributes to rate-limited sorption behavior. Current methods for assessing sorption and sequestration of recalcitrant compounds on soils and sediments do not provide a basic understanding of what is attainable by biostabilization or the bioavailability of recalcitrant compounds, or information to aid interpretation of results of ecotoxicological testing of residuals after biotreatment. Whether residual recalcitrant compounds remaining after biotreatment represent an acceptable endpoint level requires understanding of the mechanisms that bind contaminant recalcitrant compounds within soil or sediment. Research is needed that will assess the fundamental character of the binding of recalcitrant compounds at the microscale level in parallel with the development of bioslurry treatment and ecotoxicological testing, to show how the nature of PAH association with soils and sediments relates to biostabilization and achievable treatment endpoints. Research is needed to identify those factors affecting the bioavailability of recalcitrant compounds on soils and sediments and the development of the technical basis for enhancing natural recovery processes involved in the in situ biotreatment of soils and sediments contaminated with recalcitrant compounds. A special focus is needed on improving the mechanistic understanding of sequestration and bioavailability of recalcitrant compounds in soils and sediments. Such research could result in providing guidelines for the assessment and prediction of the bioavailability of recalcitrant compounds for in situ biotreatment.

 

Chapter one: Introduction to recalcitrant compounds

 

 

7

References Alexander, M. 1995. How toxic are toxic chemicals in soils? Environ. Sci. Technol. 29: 2713–2717. Auger, R.L., Jacobson, A.M., and Domach, M.M. 1995. Aqueous phase fluorescence quenching technique for measuring naphthalene partition coefficients. Environ. Sci. Technol. 29: 1273–1278. Bosma, R.M.P., Middeldorp, P.J.M., Schraa, G., and Zehnder, A.J.B. 1997. Mass transfer limitation of biotransformation: quantifying bioavailability. Environ. Sci. Technol. 31: 248–252. Boyd, S.A. and Sun, S. 1990. Residual petroleum and polychlorobiphenyl oils as sorptive phases for organic contaminants in soils.  Environ. Sci. Technol. 24: 142–144. Brusseau, M.C., Jessup, R.E., and Rao, P.S.C. 1991. Nonequilibrium sorption of organic chemicals: elucidation of rate-limiting processes. Environ. Sci. Technol. 25: 134–142. Bury, S.J. and Miller, C.A. 1993. Effect of micellar solubilization on biodegradation rates of hydrocarbons. Environ. Sci. Technol. 27: 104–110. Chung, G.Y., McCoy, B.J., and Scow, K.M. 1993. Criteria to assess when biodegradation is kinetically limited by intraparticle diffusion and sorption. Biotechnol. Bioeng. 41: 625–632. Costerton, J.W., Cheng, K.J., Geesey, G.G., Ladd, T.J., Nickel, J.C., Dasqupta, M., and Marrie, T. 1987. Bacterial biofilms in nature and disease. Annu. Rev. Microbiol. 41: 435–464. Edwards, D.A., Liu, Z., and Luthy, R.G.J. 1994. Surfactant solubilization of organic compounds in soil/aqueous systems. J. Environ. Eng. 120: 5–22. Environmental Protection Agency (EPA). 1994. Assessment and Remediation of Contaminate Sediments (ARCS) Program, Final Summary Report, EPA-905-S-94-001. EPA, Chicago. Erickson, D.C., Loehr, R.C., and Neuhauser, E.F. 1993. PAH loss during bioremediation of manufactured gas plant site soils. Water Res. 27: 911–919. Gas Research Institute (GRI). 1995. Environmentally Acceptable Endpoints in Soil: Risked-Based Approaches to Contaminated Site Management Based on the Availability of Chemicals in Soil, Draft Report of Workshop Proceedings. Gas Research Institute, Chicago. Ghosh, U., Luthy, R.G., Gillette, J.S., and Zare, R.N. 2000. Microscale location, characterization, and association of polycyclic aromatic hydrocarbons on harbor sediment particles.  Environ. Sci. Technol. 34: 1729–1736. Gillette, J.S., Luthy, R.G., Clemett, S.J., and Zare, R.N. 1999. Direct observation of polycyclic aromatic hydrocarbons on geosorbents at the subparticle scale. Environ. Sci. Technol. 33: 1185–1192. Grathwohl, P. 1990. Influence of organic matter from soils and sediments from various origins on the sorption of some chlorinated aliphatic hydrocarbons: implications on Koc correlations. Environ. Sci. Technol. 24: 1687–1693. Grathwohl, P. and Reinhard, M. 1993. Desorption of trichloroethylene in aquifer material: rate limitation at the grain scale. Environ. Sci. Technol. 27: 2360–2366. Gustafsson, O., Haghseta, F., Chan, C., MacFarlane, J., and Gschwend, P.M. 1997. Quantification of the dilute sedimentary soot phase: implications for PAH speciation and bioavailability.  Environ. Sci. Technol. 31: 203–209.

   

8

Bioremediation of Recalcitrant Compounds

Huang, W. and Weber, W.J., Jr. 1997. A distributed reactivity model for sorption by soils and sediments. 10. Relationships between desorption, hysteresis, and the chemical characteristics of organic domains.  Environ. Sci. Technol. 31: 2562–2569. Huang, W. and Weber, W.J., Jr. 1997. Thermodynamic considerations in the sorption of organic contaminants by soils and sediments. 1. The isosteric heat approach and its application to model inorganic sorbents.  Environ. Sci. Technol. 31: 3238–3243. Kelsey, J.W., Kottler, B.D., and Alexander, M. 1997. Selective chemical extractants to predict bioavailability of soil-aged organic chemicals. Environ. Sci. Technol. 31: 214–217. Linz, D.G. and Nakles, D.V., Eds. 1997. Environmentally Acceptable Endpoints in Soil. American Academy of Environmental Engineers, Annapolis, MD. Loehr, R.C. and Webster, M.T. 1997. Effects of treatment on contaminant availability, mobility, and toxicity. In Environmentally Acceptable Endpoints in Soil, D.G. Linz and D.V. Nakles, Eds. American Academy of Environmental Engineers, Annapolis, MD, chap. 2. Luthy, R.G., Aiken, G.R., Brusseau, M.L., Cunningham, S.D., Gschwend, P.M., Pignatello, J.J., Reinhard, M., Traina, S., Weber, W.W., Jr., and Westall, J.C. 1997. Sequestration of hydrophobic organic compounds by geosorbents. Environ. Sci. Technol. 31: 3341–3347. Luthy, R.G., Dzombak, D.A., Peters, C.A., Roy, S.B., Ramaswami, A., Nakles, D.V., and Nott, B.R. 1994. Remediating tar-contaminated soils at manufactured gas plant sites. Environ. Sci. Technol. 28: 266A–277A. Moore, J.N., Brook, E.J., and Johns, C. 1989. Grain size partitioning of metals in contaminated coarse-grained river flood plain sediment. Environ. Geol. Water Sci. 14: 107–115. National Research Council (NRC). 1994. Alternatives for Groundwater Cleanup, National Research Council Report. National Academy Press, Washington, DC. National Research Council (NRC). 1997. Contaminated Sediments in Ports and Waterways, Cleanup Strategies and Technologies, National Research Council Report. National Academy Press, Washington, DC. Nkedi-Kizza, P., Brusseau, M.L., Rao, P.S.C., and Homsby, A.G. 1989. Nonequilibrium sorption during displacement of hydrophobic organic chemicals and calcium-45 through soil columns with aqueous and mixed solvents. Environ. Sci. Technol. 23: 814–820. Oja, V. and Suuberg, E.M. 1997. Development of a nonisothermal Knudsen effusion method and application to PAH and cellulose tar vapor pressure measurement. Anal. Chem. 69: 4619–4626. Pignatello, J.J. and Xing, B. 1996. Mechanisms of slow sorption of organic chemicals to natural particles. Environ. Sci. Technol. 30: 1–11. Putcha, R.V. and Domach, M.M. 1993. Fluorescence monitoring of polycyclic aromatic hydrocarbon biodegradation and the effect of surfactants. Environ. Prog. 12: 81–85. Schwarzenbach, R.P., Gschwend, P.M., and Imboden, D.M. 1993. Environmental Organic Chemistry. John Wiley & Sons, New York. Steinberg, S.M., Pignatello, J.J., and Sawhney, B.L. 1987. Persistence of 1,2-dibromoethane in soils: entrapment in intraparticle micropores. Environ. Sci. Technol. 21: 1201–1208.

 

Chapter one: Introduction to recalcitrant compounds

9

Stumm, W. 1992. Adsorption. In Chemistry of the Solid-Water Interface. John Wiley & Sons, New York, chap. 4. Thoma, G. 1994. Summary of the Workshop on Contaminated Sediment Handling, Treatment Technologies, and Associated Costs, April 21–22, 1994. Background paper prepared for the Committee on Contaminated Sediments, Marine Board, National Research Council, Washington, DC. Werth, C.J. and Reinhard, M. 1997. Effects of temperature on trichloroethylene desorption from silica gel and natural sediments. 2. Kinetics. Environ. Sci. Technol. 31: 697–703. Wu, S. and Gschwend, P.M. 1988. Numerical modeling of sorption kinetics of organic compounds to soil and sediment particles. Water Resour. Res. 24: 1373–1383.

chapter two

Toxicological exposure of bound recalcitrant compounds Herbert Fredrickson, John S. Furey, and Jeffrey W. Talley Contents 2.1 Introduction .................................................................................................. 11 2.2 Bioavailability of recalcitrant compounds and environmental risk assessment..................................................................12 2.3 Equilibrium partitioning and sediment quality guidelines..................14 2.4 Recalcitrant compounds in sediments .....................................................15 2.5 Effects of diagenesis and weathering on recalcitrant compound geosorbents...............................................................................17 2.6 Koc-based predictions...................................................................................18 2.7 New protocols ..............................................................................................20 2.8 Microbial degradation recalcitrant compounds in sediment ...............22 2.9 Thermal desorption mass spectrometry of recalcitrant compounds...............................................................................23 2.10 Conclusions...................................................................................................26 References...............................................................................................................26

2.1 Introduction This chapter relates the importance of the toxicological exposure potential of recalcitrant compounds in sediments and dredged material to implementation of public laws and regulations governing environmental risk assessment; summarizes recent peer-reviewed literature on sediment recalcitrant 11

12

Bioremediation of Recalcitrant Compounds

compounds’ exposure potential in the context of microbial degradation and the sorbant quality of sediment organic carbon; and introduces the practical utility of thermal desorption mass spectrometry with respect to identification and quantification of recalcitrant compounds, measuring recalcitrant compounds’ release energy, and the compatibility of the development of field-portable direct-sampling analytical technologies.

2.2 Bioavailability of recalcitrant compounds and environmental risk assessment The Clean Water Act (Section 404 of PL 92-500) and the Marine Protection, Research, and Sanctuaries Act (also known as the Ocean Dumping Act, Section 103 of PL 92-532) require that sediment-associated contaminants be evaluated for their ability to accumulate in biota. Jointly, the U.S. Army Corps of Engineers (USACE) and the U.S. Environmental Protection Agency (EPA) adopted a tiered system to evaluate this bioaccumulation potential (Implementation Manual for Section 103, a.k.a. the Green Book, and Implementation Manual for Section 404, a.k.a. Inland Manual). Definitive bioaccumulation tests require that three different organisms be exposed to sediment for 28 days and then the recalcitrant compounds’ body burdens determined using standard analytical techniques. From a practical perspective, it is not feasible to test all sediments and dredged material the USACE must manage. It is also apparent that noncontaminated sediments do not warrant bioaccumulation testing, and some sediments are so contaminated that bioaccumulation is a foregone conclusion. The EPA/USACE testing manuals describe a screening level protocol termed thermodynamic bioaccumulation potential (TBP) (McFarland, 1995). TBP has been widely used in tier 2 evaluations to exclude from further testing sediments from both extremes of the contamination level continuum. TBP predicts the partitioning behavior of recalcitrant compounds between sediment organic carbon and benthic organism lipid. TBP is based on a thermodynamic model (Mackay, 1982) of the environment as a system composed of various compartments where contaminants have come to equilibrium though passive processes. At equilibrium, fugacity (i.e., escaping tendency) is equal in all sorptive and solution phases (Mackay, 1991). On the basis of fugacity, it is possible to predict the equilibrium distribution of a nonpolar contaminant between any two phases. The two most relevant phases with respect to the bioaccumulation of recalcitrant compounds from contaminated sediment are sediment organic matter and organism lipid. The sorption of recalcitrant compounds to sediments has been simply but elegantly described. Karickhoff et al. (1979) combined thermodynamic theory (i.e., fugacity) (Mackay, 1979) with empirical correlations to derive a systematic procedure for predicting contaminant sediment sorptive behavior. In spite of the “high degree of variability and complexity in sediment composition and large number of potential sorptive interactions,” Karickhoff

Chapter two: Toxicological exposure of bound recalcitrant compounds

13

intentionally developed a simple mathematical format that required a minimum of measured parameters. He felt a balance must be struck between a complex model few could afford to parameterize and the degree of accuracy and precision required in its application (Karickhoff et al., 1979). Karickhoff showed that for neutral hydrophobic contaminants (i.e., water solubilities less than 10–3 M), sorption isotherms in the low loading limit are linear and reversible. Their partition coefficients (Kp) were highly correlated to the organic carbon content of the soils/sediments in this data set. Referencing sorption to organic carbon content produced a partition coefficient to organic carbon (Koc) that was independent of other bulk sediment/soil parameters. Karickhoff’s (1981) “justifiable simplification” found even wider application when he showed that Koc could be directly derived from the contaminants’ octanol–water partition coefficients. Concurrently, Könemann and van Leeuwen (1980) showed a linear relationship between Koc and the partitioning of a series of chlorobenezenes from sediments to lipid normalized benthic infaunal biomass. McFarland (1984) synthesized information from Karickhoff and Könemann and van Leeuwen and derived a relationship for TBP (McFarland and Clarke, 1986): TBP = AF(CS/fOC)fL where AF = accumulation factor CS = recalcitrant compound concentration in whole sediment fOC = decimal fraction of organic carbon in sediment fL = decimal fraction of lipid in targeted organism Biota-sediment accumulation factors (BSAFs) have also been empirically determined and used to describe the distribution of recalcitrant compounds between lipid normalized biomass and organic carbon normalized sediment. BSAF = (CT/fL)/(CS/fOC) where CT/fL = the lipid normalized contaminant tissue concentration CS/fOC = the organic carbon normalized contaminant sediment concentration Initial TBP predictions, derived from an arbitrarily fixed AF of 4, were shown to consistently overestimate polycyclic aromatic hydrocarbon (PAH) bioaccumulation from contaminated sediments by factors ranging between 41 and 386 (McFarland, 1995). Precision and accuracy of TBP predictions were improved to a factor of 10 when empirically derived BSAFs from one field reference sediment contaminated with PAH were used to calculate TPB for a second field sediment contaminated with PAH (Clarke and McFarland,

14

Bioremediation of Recalcitrant Compounds

2000). That is, field reference–derived BSAFs were substituted for AFs in the original TBP equation. Clarke and McFarland (2000) concluded that TBP was a useful screening tool for eliminating sediments with negligible likelihood of causing unacceptable bioaccumulation from further testing and tended to generally overestimate recalcitrant compounds’ bioaccumulation from sediment.

2.3 Equilibrium partitioning and sediment quality guidelines In the absence of site-specific information, environmental managers must use the best available information, and this often entails the use of model predictions to support sediment management decisions. To this end, the logic of Karickhoff’s fugacity-based model of recalcitrant compounds–sediment partition coefficients Kp normalized to organic carbon content Koc has been combined with that of TBP to predict benthic marcofauna exposure levels (Figure 2.1). Predicted exposure levels are subsequently interpreted in the context of aquatic toxicity databases. This equilibrium partitioning (EqP)-based logic is the basis for deriving sediment quality criteria as proposed by DiToro et al. (1991). TBP predictions of bioaccumulation potentials and EqP estimates of exposure potentials are both derived from Koc. The accuracy, precision, and general applicability of predictions made on the basis of Koc-predicted recalcitrant compounds–sediment–pore water equilibrium partitioning have been debated in the technical literature since it was first proposed. The practical ecological and economic consequences of this issue have escalated as applications of the Koc model have expanded beyond that of a sediment screening

Biota

Sediment Carbon

Pore Water Koc

Figure 2.1 Equilibrium partitioning (EqP) as described by DiToro et al. (1991) is predicated on a model that assumes that equilibrium exists between the contaminant sorbed to sediment organic carbon, pore water, and lipid of benthic biota. The partitioning of recalcitrant compounds between sediment organic matter and pore water is predicted from Koc (Karickhoff, 1981).

Chapter two: Toxicological exposure of bound recalcitrant compounds

15

tool. Some suggest that EqP-derived predictions combined with aquatic toxicity databases can be used as stand-alone pass/fail predictors of sediment quality whose implementation can be modeled after EPA’s Water Quality Criteria. A review of this issue is beyond the scope of this chapter. The reader is referred to articles promoting the use of EqP model predictions in sediment management decisions (DiToro et al., 1991; Ankley et al., 1996; Burkhard, 2000) and those that argue for more limited use of the models in sediment management decisions (Iannuzzi et al., 1995; Driscoll and Landrum, 1997; O’Connor et al., 1998; van Beelen et al., 2001; Condor et al., 2002). Instead, we will focus the remainder of this discussion on technical issues relevant to two factors that can have major effects on the dosage of recalcitrant compounds realized by benthic macrofauna that are not addressed in Koc-based TBP or EqP models: recalcitrant compounds’ sequestration in sediment and microbial degradation of recalcitrant compounds in sediment.

2.4 Recalcitrant compounds in sediments Luthy et al. (1997) characterized matter in soils and sediments as geosorbents. Soils and sediments are heterogeneous at the scale of samples, aggregates, and particles. Structurally or chemically different constituents of sediments interact differently with recalcitrant compounds in terms of binding energies and associated rates of sorption and desorption. Complex assemblages of the components can cause complex mass transfer phenomena. The term sequestration refers to some combination of diffusion limitation, adsorption, and partitioning. Sorption and desorption rates for recalcitrant compounds in geosorbents occur on timescales ranging from fast (e.g., minutes to days) to slow (e.g., weeks to years). Although their relative proportions vary greatly, most recalcitrant compounds’ contaminated sediments to date have both rapidly and slowly desorbing recalcitrant compound fractions. Desorption rate differences are thought to be due to processes such as intra-aggregate diffusion, releases from micropores, or different forms of geosorbent organic matter. Two of these proposed geosorbant domains, soft amorphous organic matter and soot, have been shown to be particularly important when attempting to predict recalcitrant compounds’ equilibrium partitioning and, ultimately, exposure and toxicity. Decaying plant material (case A in Figure 2.2) is a major source of sediment organic matter and a major food source for detritivores. This low-density fraction of sediment organic matter from a New York–New Jersey estuary contained 10 times the level of PAH predicted by organic carbon normalized equilibrium partition coefficients (Koc) (Rockne et al., 2002). This fraction readily released PAH into the aqueous phase and was the controlling factor in whole-sediment PAH release. Drifting plant detritus has also been shown to be a major contributor to the total annual load of organochlorine contaminants (including polychlorinated biphenyl (PCB)) in the Detroit River (Lovett-Doust et al., 2002). These recent

16

Bioremediation of Recalcitrant Compounds Sediment (Organic Carbon)

Slow Pore Water Rap

Koc, rap BCF

Deposit-Feeder (Lipid)

Figure 2.2 Conceptual model of distribution of hydrophobic organic chemicals in sediment. rap = rapidly desorbing compartment; slow = slowly desorbing compartment; Koc,rap = partition coefficient between rapidly desorbing compartment and pore water (l/kg organic carbon); BCF = bioconcentration factor (l/kg lipid). (From Kraaij, R., Sequestration and Bioavailability of Hydrophobic Chemicals in Sediment, Ph.D. dissertation, University of Utrecht, Netherlands, http://www.library.uu.nl/digiarchief/dip/diss/1960191/inhoud.htm)

studies are especially important. They demonstrate that plant detritus, a major food source at the base of aquatic food webs, can be the major contributor to recalcitrant compounds’ theoretical maximium daily loads, and Koc significantly underestimates the bioaccumulation of recalcitrant compounds from this trophodynamically important geosorbant. In this common aquatic environmental situation, Koc-based environmental risk predictions (i.e., TBP, EqP, and sediment quality criteria) would not be protective. Koc-derived predictions of bioaccumulation and toxicity of sediments containing soot (case B in Figure 2.2) have also been shown to be inaccurate. Socha and Carpenter (1987) compared PAH-contaminated sediments from two sites within Puget Sound. Koc-predicted pore water PAH levels agreed with empirically determined pore water PAH levels (within a factor of 4) at a creosote-impacted site. However, no PAH was detected in sediment pore water from a site impacted by combustion products and natural PAH, even though detectable levels were predicted using Koc. McGroddy and Farrington (1995) published similar results on PAH-contaminated sediments in Boston Harbor. Pore water PAH levels were depleted relative to those predicted using Koc. Variances for individual PAHs varied, but only 0.2 to 5.0% of Koc-predicted phenanthrene was actually measured in sediment pore water. PAH associated with pyrogenically derived soot particles was suggested as the reason for the discrepancies (McGroddy et al., 1996). Paine et al. (1996) showed that heavily PAH-contaminated sediments (highest levels of 10,000 mg/kg and mean levels of 150 mg/kg) from Kitimat Arm, at the head of Douglas Channel in British Columbia, did not change benthic community structure, were not toxic to benthic fauna, and generally did not accumulate in the commercially important Dungeness crab. Most of the PAH in this sediment originated from the washout of a wet air scrubber from

Chapter two: Toxicological exposure of bound recalcitrant compounds

17

Sediment (Organic Carbon)

Slow

Pore Water

Rapid Koc, rapid

Microbial Degradation

CO2

Bioconcentration Factor

Deposit-Feeder (Lipid)

Figure 2.3 The recalcitrant compound pore water pool is probably the most biologically available. Kraaij’s (2002) conceptual model (modified above) equates recalcitrant compound in pore water to the rapidly desorbing fraction of recalcitrant compounds from sediment organic matter. Part of the pore water recalcitrant compounds can be taken up into benthic macrofaunal lipid. However, neither Kraaij’s conceptual model nor most of those currently proposed take into consideration the ability of sedimentary bacterial communities to mineralize recalcitrant compounds. The factors that determine this partitioning of the pore water recalcitrant compounds’ pool between macrofaunal lipid and microbial mineralization are not well understood.

aluminum smelter potlines. Aluminum smelter–derived PAH in sediments from Sunndalsfjord, Norway (Naes and Oug, 1998 and Naes et al., 1999) were present at lower levels (15 mg/kg) than Kitmat sediment but were likewise not biologically available because they were associated with soot particles. Song et al. (2002) showed that black carbon constituted between 18 and 41% of the total organic carbon of soil and sediment samples collected from Guangzhou, China. The percentage of soot in any particular series of sediment samples can be highly variable due to variability in air and water currents, which deposit them in aquatic systems, and sediment particle segregation, resuspension, redistribution, and transport by episodic water currents.

2.5 Effects of diagenesis and weathering on recalcitrant compound geosorbents In addition to the sources of sediment organic matter (e.g., vascular plants, algae), diagenesis and weathering affect sediment quality characteristics that are correlated to rates of recalcitrant compounds’ sorption and desorption (Luthy et al., 1997). Some diagenetically aged organic matter (e.g., coal and shale) exhibits a high degree of condensation, is reduced in the relative amount of oxygen-containing functional groups (reflected in H/O and O/ C atomic ratios), and contains more aromatic carbon rings (measure by ultraviolet (UV) and infrared (IR) absorbance). This reduced organic matter

18

Bioremediation of Recalcitrant Compounds

has been characterized as hard or glassy. Glassy organic matter strongly binds recalcitrant compounds (Brannon et al., 1998) and is characterized by slow mass transfer rates and nonlinear adsorption kinetics (Haitzer et al., 1999; LeBoeuf and Weber, 2000). Kerogen is another organic matter fraction that has undergone diagenic alterations that has been shown to have nonlinear recalcitrant compounds’ sorption isotherms and high capacity to bind recalcitrant compounds (Song et al., 2002). Little is known about the distribution of kerogen in surface sediments in the context of sequestering recalcitrant compounds. A portion of kerogen and other organic material can also be dissolved in pore water in a form that is not removed by filtration (Gauthier et al., 1987) and thus greatly affects the pore water recalcitrant compound concentrations and mechanisms of recalcitrant compounds’ bioaccumulation. This emerging technical information on the sorption behavior of recalcitrant compounds with respect to the quality of sediment organic carbon can directly affect the USACE management of dredged materials. Soxhlet-extractable PAH levels in dredged material from a confined disposal facility at Milwaukee Harbor (Table 2.1) averaged 115 mg/kg, but only 46 mg/kg (i.e., less than half) was readily biologically available for either microbial degradation (Ringelberg et al., 2001) or bioaccumulation in earthworms (Talley et al., 2002). The empirically determined BSAF for total PAH accumulation from Milwaukee Harbor–dredged material into the earthworm (Eisenia fetida) was 0.08. Five percent of the dry weight mass of Milwaukee Harbor-dredged material was coal/coke. Sixty percent of the total extractable PAHs were associated with this coal/coke fraction, and almost none of it was biologically available. As a consequence, the potential for bioremediation to reduce the total extractable PAH from this dredged material is limited (Myers et al., 2002).

2.6 Koc-based predictions Appropriate use and informed interpretation of the data derived from screening tools are essential for effective sediment management. Sediments have been described in which Koc-based predictions of bioaccumulation and toxicity have been inaccurate. Thus, the universal application of Koc-based predictions without reasoned judgment in interpretating the resulting predictions can lead to both significant under- and overassessments of environmental risk. Karickhoff’s “justifiable simplification” will be an even more useful screening tool when its limits of applicability are more fully understood and appreciated. The environmental distribution and relative abundance of organic matter that sequesters recalcitrant compounds in sediment, and the fate of recalcitrant compounds when desorbed from this material, are currently not known but warrant further study. To gain this perspective, we present and discuss additional tests and environmental parameters that will improve assessments of recalcitrant compound bioaccumulation potential from sediments.

Coal

Silt/Clay

Fraction of Milwaukee CDF-Dredged Material

5%

95%

% Dry Wt. of Dredged Material

10,000 >60%

80–100 85%. They also reported an approximate 81% reduction in NAPL saturation based on pre- and postflushing partitioning interwell tracer tests (PITTs). Falta et al. (1999) presented results from a second cosolvent flushing study at Hill AFB, wherein the remedial mechanisms were NAPL mobilization and enhanced dissolution. Their test cell was approximately 5 m long by 3 m wide, and the clay-confining unit was 9 m below grade. They injected 28,000 l of a ternary cosolvent mixture (80% tert-butanol, 15% n-hexanol, and 5% water) over a 7-day period. Reductions in target contaminant concentrations measured from pre- and postflushing soil cores were reported to range from 70

Chapter five: Solvent contaminated soils and groundwater

63

to >90%, and an 80% reduction in total NAPL content was reported based on pre-and postflushing PITTs. Brooks (2000) described a third cosolvent flushing field demonstration at the Dover National Test Site (DNTS) at Dover Air Force Base, DE. This test was conducted in an isolated test cell into which a known volume of PCE had been injected. The dimensions of the cell were approximately 5 by 3 by 12 m deep, with a clay layer forming the lower boundary of the cell. DNAPL was injected into the formation at approximately 10.5 m below the surface, creating a source zone of approximately 1.5 m thickness. A total of 41,700 l of 95% ethanol was used during this flushing demonstration. Upon extraction from the test cell, the cosolvent solution was treated on site to remove PCE and reinjected into the test cell. Because of this recycling system, a total of 69,700 l of ethanol was displaced through the test cell. Based on mass balance measurements, 71% of the resident PCE was removed by this flushing process. These results and others have indicated that PCE- and TCE-constituted DNAPLs are amenable to extraction-based mass reductions. However, although the potential for substantial mass removal has been demonstrated, the debate over the relative and absolute benefit of partial source removal, in terms of costs and risk reduction, is currently unresolved. It is clear that even under the best of extraction scenarios, residual amounts of contaminants will remain at levels that could preclude meeting regulatory requirements for site closure.

5.1.3 Biotransformation of chlorinated solvents and subsurface microbial ecology Heterotrophic organisms (humans and most bacteria) oxidize organic compounds to obtain energy. In this process, electrons or reducing equivalents from the oxidizable organic compound (substrate/electron donor) are transferred to, and ultimately reduce, an electron acceptor. The electron acceptor may be an organic or inorganic compound. During this electron transfer process, usable energy is recovered through a complex series of oxidation-reduction (redox) reactions by the formation of energy storage compounds or electrochemical gradients. The oxidation of organic compounds coupled to the reduction of molecular oxygen is termed aerobic heterotrophic respiration. When the oxidation of the organic compounds is linked to other electron acceptors (i.e., sulfate, nitrate, ferric iron, chloroethenes) by bacteria, it is referred to as anaerobic respiration. The oxidation of the organic matter, linked to oxygen reduction or other electron acceptors, to yield energy and carbon compounds for microbial growth is termed oxidative catabolism. Pristine (oxygenated) groundwater environments are electron donor limited. When readily degradable organic matter enters the subsurface in sufficient quantities, it can produce a variety of chronologically and spatially defined metabolic zones. These zones are dependent on the availability of electron acceptors. As organic matter enters an oxygenated aquifer, whether

64

Bioremediation of Recalcitrant Compounds

it is a human-produced (anthropogenic) contaminant or an influx of natural material, indigenous aerobic heterotrophic microorganisms metabolize the organic material and consume the available oxygen in the process. When this occurs, aerobic respiration slows and eventually stops. This allows for the development of anaerobic metabolic communities utilizing other electron acceptors. Under these conditions, the subsurface environment becomes electron acceptor limited and the rates of transformations of the introduced organics are controlled by the availability of the various electron acceptors. The introduction of the parent chloroethenes (PCE, TCE) does not supply bio-oxidizable electron donors to subsurface environments, and thus will not create anoxic conditions in the impacted groundwater. Chloroethenes, however, are often codisposed with other organics or may mobilize native or anthropogenic organics, which can serve as electron donors and can be biologically utilized for the reduction of dissolved oxygen, other anaerobic electron acceptors, or the chloroethenes. It is important to remember that the reductive dechlorination process is one of several competing electron sinks, and the amount of reducing equivalents that are utilized for dechlorination of chloroethenes is determined by the energetics and relative concentrations of all of the available electron acceptors. The parent chloroethenes (PCE, TCE) are resistant to oxidative catabolism. This means that in the presence of oxygen, the biodegradation potential of these compounds varies from not detected (for PCE) to low (for vinyl chloride (VC)). TCE and less chlorinated daughters can be degraded under cometabolic conditions. However, since this process is not directly beneficial or may even be detrimental to the microorganisms and requires the presence of an inducer, its relevance is limited. In the subsurface and other anaerobic environments, chloroethenes can be transformed by microorganisms through a process known as reductive dechlorination. This is a stepwise removal of chloride ions: Reductive chlorination of chloroethenes: R-Clx + H2

R-Clx–1 + H+ + Cl–

(5.1)

This process transforms the chlorinated ethenes to nonchlorinated products (ethene, ethane), which do not pose a threat to human health or the environment. Halorespiration is the term used to describe the reductive dechlorination process when specific microorganisms obtain energy for growth using halogenated compounds, such as PCE, as terminal electron acceptors. Unfortunately, the more mobile and toxic daughter products, such as dichloroethene (DCE) and vinyl chloride (VC), are intermediates in the halorespiration pathway (Figure 5.1). If the process “stalls,” as it often seems to in the subsurface, before reaching nonchlorinated end products, the reductive dechlorination process may increase the potential risks to humans and the environment. Thus, the reductive dechlorination process can exacerbate or attenuate the problems created by the release of chloroethenes to the

Chapter five: Solvent contaminated soils and groundwater PCE Biotransformations for Chloroethenes TCE

× ×

65

No Evidence but - ΔG

No Evidence but - ΔG

? 1, 1-DCE

trans-DCE

CO2

cis-DCE

Some Field Evidence

Reductive Transformation Oxidative Catabolism

Vinyl Chloride

CO2

Ethene

Ethane

CO2

Figure 5.1 Biotransformation pathways for chlorinated ethenes.

subsurface and groundwater environments. Under some conditions, the less chlorinated chloroethenes (VC and perhaps DCE) may undergo oxidative catabolism under anaerobic conditions. This process may be an important mechanism for mass removal of vinyl chloride in locations where redox potentials increase, such as leading edges of contaminant plumes or areas with high dispersion or oxygen diffusion.

5.1.4 Technology status: in situ bioremediation Bioremediation and natural attenuation are two in situ technologies often used for remediation of dissolved phase contaminants. Bioremediation is defined as using the metabolic processes of microorganisms to degrade, detoxify, or mineralize contaminants. Bioremediation in the context of contaminated subsurface environments (soil, sediments, groundwater) usually involves the injection of nutrients to stimulate the desired biological processes. This is also referred to as biostimulation. A less common, although growing, approach is to introduce or inject specific microorganisms, usually in combination with nutrients — a process referred to as bioaugmentation. Natural attenuation is a remedial option that takes advantage of the widely observed phenomenon that concentrations of contaminants in the subsurface tend to decrease over time. This is due to a combination of physical, chemical, and biological processes that degrade, immobilize, or transform dissolved contaminants. For most contaminants, the most important of these attenuation mechanisms is biodegradation by native microorganisms. The relative treatment costs of these technologies are low, and in

66

Bioremediation of Recalcitrant Compounds

most cases, they lead to destruction of the target contaminant; however, these processes are often very slow and thus may be impractical due to risk of exposure to the public from the potential of contaminants in groundwater reaching water wells or surface water. The selection and implementation of both bioremediation and natural attenuation also suffer from a lack of familiarization by remedial managers, typically civil and environmental engineers, and, to a degree, the uncertainties as to long-term performance, particularly in the case of natural attenuation. Beginning in the 1990s, a number of successful demonstrations of bioremediation technologies were reported for chlorinated solvent sites. These included active reductive dechlorination systems (Beeman Sewell/Pinellas), cometabolic systems (Savanah River, Moffet), and, later in the 1990s, a variety of natural attenuation demonstrations (St. Joe, Dover). While these demonstrations fulfilled the proof of concept and led to the development of defined protocols for chlorinated solvent sites, such as the Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in Ground Water (USEPA, 1998) and the Reductive Anaerobic Biological In Situ Treatment Technology (ESTCP, 1998), a number of limitations on the use of biological treatment approaches became clear. Some of these limitations were applicable to all in situ bioremediation systems, including the challenges associated with delivery and mixing of nutrients, and rates of cleanup were often limited by subsurface transport rates rather than biological reaction rates. However, some of the limitations were unique to DNAPL and chlorinated solvent sites. Active and passive bioremediation technologies for oxidizable contaminants such as fuels and petroleum proved to be metabolically robust and redundant. This means that the genetic capability for the degradation of a variety of compounds was found to be present at almost all sites where toxicity was not a major issue. This catabolic capacity was found to exist for multiple populations and redox conditions, such that a variety of electron acceptors could be used in sequence to support degradation (Suflita and Sewell, 1991). However, for chlorinated solvents, the conditions that supported transformation were more limited and the capacity, at least for complete dechlorination, does not appear to be universal. These limitations resulted in much slower rates of activity and potential instability in active systems. Another potential limitation is that the technologies are focused on dissolved phase contaminants and have not yet been shown to remediate areas contaminated with separate phase contaminants. It has been believed that the highly toxic nature of the separate phase residual contaminants found in source zone environments suppresses bioactivity. Extensive research demonstrating the biological transformations of dissolved phase chloroethenes has been conducted (Fathepure and Boyd, 1988; DiStefano et al., 1991; DeBruin et al., 1992; Tandoi et al., 1994; Magnuson et al., 1998). Recent research has shown that complete dechlorination can occur even with initial PCE concentrations as high as 91 mg/l (DiStefano, et al., 1991; Yang and McCarty, 2000). The slow transformation rates coupled to the low dissolution

Chapter five: Solvent contaminated soils and groundwater

67

rates for DNAPLs and their significant residual mass lead to both very large plumes for treatment and extensive treatment times. These designed treatment times for chlorinated solvent sites can range from decades for active systems to centuries for natural attenuation systems. Combining bioremediation with enhanced source removal, however, could be a mechanism for obtaining complete site restoration. With the bulk of the residual phase removed by enhanced source removal, in situ biotreatment could transform the remaining contaminants to nonhazardous compounds at a rate in excess of dissolution.

5.1.5 Flask to field Although the solvent flushing and bioremediation technologies have been shown in laboratory and limited field evaluations to enhance the remediation of contaminated aquifers, there has been no previous attempt to couple these technologies in the field. There are a number of factors that have not been fully evaluated for both the individual and coupled technologies. The high degree of spatial variability in the subsurface environment, the difficulty in obtaining in situ mixing of the remedial fluid, the changes in the hydraulic properties of the system as the DNAPL is removed, and the effects of high concentrations of remedial solvent mixtures on the microbial ecology are major factors that must be accounted for in the system design to ensure the success of implementation. Beginning in the late 1990s, Environmental Protection Agency (EPA) researchers began evaluating the effects of potential cosolvents and their breakdown products on the reductive dechlorination process (Gibson and Sewell, 1992; Gibson et al., 1994). Observations from the Hill AFB source control trials also indicated that native biological transformations were occurring during NAPL source removals. These early observations, laboratory experiments, and field tests supported the development of a process conceptual model and testable pilot system. The Solvent Extraction Residual Biotreatment (SERB) technology is a treatment train approach that combines both cosolvent extraction and in situ bioremediation to obtain complete site restoration. Advantages of this technology are: • Rapid removal of large masses of DNAPL in a very short time • In situ treatment train technique • Cosolvent injection/extraction overcomes transport and mixing limitations of standard bioremediation infiltration and injection techniques • DNAPL mass removal reduces toxicity to microbes for bioremediation • Continued removal of dissolved contaminants following cessation of pumping • Combination of an active process, solvent extraction, with an essentially passive process, residual biotreatment

68

Bioremediation of Recalcitrant Compounds

(a)

(b)

Figure 5.2 Schematic representing application of the Solvent Extraction Residual Biotreatment (SERB) technology. (a) Complex distribution of DNAPL in the subsurface. (b) Dissolution of DNAPL by cosolvent flushing technology. (c) Following cosolvent extraction, separate and mixed areas of DNAPL and cosolvent exist in the subsurface. (d) Development of bioactive zones conducive to the reductive dechlorination of DNAPL.

Chapter five: Solvent contaminated soils and groundwater

69

(c)

(d)

Figure 5.2 (Continued)

The overall SERB process is presented in Figure 5.2, which shows the initial characterization of the site (Figure 5.2a), the application of the cosolvent flushing technology (Figure 5.2b), the potential for mixing of electron donor and electron acceptor (Figure 5.2c), and the creation of reactive zones where bioremediation can occur (Figure 5.2d).

70

Bioremediation of Recalcitrant Compounds

5.1.6 Biotreatment (SERB) technology Site characterization activities are used to identify the location and extent of the source area and whether a separate phase DNAPL is present, but most often these procedures are not conducted to give the level of detail required for the successful application of many remediation technologies. Figure 5.2a shows the complex distribution of DNAPL that can exist in the subsurface and how the DNAPL moves through the aquifer due to gravity and pools on less permeable layers. Site characterization, groundwater flow characterization, and optimization modeling must be conducted to determine the number and placement of injection and recovery wells for the cosolvent extraction process. The selection of a cosolvent is based on the properties of contaminant to be removed, cosolvency properties, and regulatory constraints concerning injection of compounds. The cosolvent extraction process utilized in the SERB demonstration was based on the principle of enhanced dissolution of the contaminant into the cosolvent, which was then removed through the recovery/extraction wells. PCE was the contaminant at the SERB demonstration site, and ethanol was selected as the cosolvent for enhanced dissolution. Some in situ extraction technologies are also designed to mobilize the residual contaminant, which is also removed through the recovery/extraction wells. Mobilization of the contaminant requires excellent hydraulic control, especially in the case of DNAPL contaminants, to ensure the DNAPL is not allowed to contaminate the aquifer further. Upon selection of the cosolvent and determination of the injection/extraction design, the cosolvent flushing test is conducted (Figure 5.2b). Maximum contact and mixing of the cosolvent and residual contaminant is the most important criterion in the design of the injection/extraction system for successful application of the technology. Because of the complex nature of the subsurface environment, it is known that some portions of the DNAPL will not be contacted by the cosolvent, and the majority of the flow will be through layers of higher hydraulic conductivity. Following the application of the cosolvent extraction technology and cessation of active pumping, it is envisioned that there will be separate areas of DNAPL contamination and cosolvent remaining in the aquifer but that there will also be areas where the DNAPL and cosolvent are mixed (Figure 5.2c). The areas where the cosolvent (electron donor) and DNAPL (electron acceptor) are mixed contain the highest potential to stimulate bioremediation. The SERB technology application at the demonstration site was designed to enhance the reductive dechlorination of the residual DNAPL contaminant, PCE. Ethanol was selected as the cosolvent because it was both an acceptable cosolvent and an acceptable electron donor for the reductive dechlorination process. Stimulation of the bioremediation processes should result in subsurface bioactive zones that will remove the contaminant from the groundwater at a rate faster than it dissolves from the remaining residual DNAPL in the

Chapter five: Solvent contaminated soils and groundwater

71

subsurface (Figure 5.2d). If this is achieved, complete site restoration is possible. Mixing of the cosolvent and the contaminant for both dissolution and removal of the separate phase DNAPL, and the subsequent biodegradation of the remaining contaminant, is the most important criterion in the design of the injection/extraction system for the SERB technology. Delivery of the electron donor to the subsurface has been a problem for various in situ treatment technologies and is usually limited by transport considerations. By conducting the cosolvent flush, the concurrent exposure of the electron donor (cosolvent) to the electron acceptor (chlorinated solvent) is facilitated by the injection/extraction system and the cosolvency effect. Thus, subsequent biodegradation of the chlorinated solvent becomes a much more viable remediation strategy. For application of the SERB technology at the demonstration site, the reductive dechlorination of PCE was the bioremediation process that was to be stimulated. The pathways for the biotransformation of chloroethenes shown in Figure 5.1 demonstrate why PCE and TCE are extremely recalcitrant contaminants. There is no evidence for oxidative mechanisms for degradation of these compounds; thus, the reduction dechlorination pathway is the only known mechanism for breakdown to nontoxic constituents. There is evidence for both oxidative and reductive pathways for dechlorination of the other daughter products. The reductive dechlorination process is a stepwise removal of a chlorine atom. Production of cis-DCE rather than the other isomers is an indication that it is a product from the dechlorination of TCE. Monitoring of all daughter products of PCE gives an indication of the extent of dechlorination occurring in the system. The reductive dechlorination of chlorinated ethenes can be conducted by both halorespiration (catabolic) and cometabolic mechanisms. Different microbial populations are responsible for each of these processes. Because of the complexity and heterogeneity of the subsurface environment, it is hypothesized that there can be several competing mechanisms and organisms active at the same time but in different portions of the subsurface. This is displayed in Figure 5.3, where a consortium of microorganisms is shown to be responsible for different oxidation-reduction reactions. Various electron donors and electron acceptors are present in the subsurface and are utilized and degraded by these reactions. Evaluation of the concentrations of electron donors and electron acceptors in the site groundwater can be an indication of the types of reactions that may be occurring. Because most monitoring wells are screened over a large interval and water will move more quickly through the layers with a higher hydraulic conductivity, this type of analysis gives a general view of the screened interval.

5.2 Objectives The overall objective of this research, as for the entire Flask to Field (FTF) program, was to design, develop, and implement new biologically based

72

Bioremediation of Recalcitrant Compounds Electron Acceptors

Electron Donors

NO−3 Organic Compound

SO2− 4

(Ethanol, H2)

HCO−3

Fe3+ R-Clx N2(NH+4 )

Partially Oxidized Compound(s)

H2S

(Acetate) CO2

Fe2+ CH4 R-Clx−1

Figure 5.3 Complex oxidation-reduction reactions that may take place in the subsurface environment by microorganisms.

cleanup technologies to address environmental challenges faced by DOD, Department of Energy (DOE), EPA, and the Environmental Restoration Research community. These new technologies were to be evaluated in terms of both efficacy and performance. The specific objective in the chlorinated solvents area was to develop and test comprehensive site approaches, which addressed the unique nature of the environmental challenge presented by solvent/DNAPL sites. Where the FTF program differs from other government- and industry-driven programs is in the recognition of the value of fundamental research in directly supporting the design and testing and, more importantly, the evaluation of the performance to these novel remediation approaches. Because of this, the chlorinated solvents area project’s specific goal was to demonstrate the feasibility of the SERB technology for the comprehensive remediation of a DNAPL-contaminated site. This meant that the approach not only dealt with the dissolved/mobile phase contaminants but also addressed the need to remediate the separate phase material (DNAPL/residual) and limit the potential of the source area to serve as an extremely long-lived input to groundwater at the site. The project also focused on developing a coherent conceptual model of the interaction of the remedial approach, the subsurface microbial ecology, and the observed performance. To this end, significant effort was made to develop and test new molecular ecology tools and innovative site characterization tools and to support the overall test with concurrent laboratory research efforts. The regulatory goals of the SERB pilot test were to evaluate the potential and effectiveness of this technology for the Florida Department of Environmental Protection (DEP) Drycleaning Solvent Cleanup Program (DSCP) and the EPA’s DNAPL initiative. The specific goals in this regard were to demonstrate that the system configuration could achieve and maintain hydraulic containment while satisfying long-term injection permit requirements and

Chapter five: Solvent contaminated soils and groundwater

73

to collect sufficient data necessary for estimating the full-scale remedial technology design parameters for the selected site, and the postflush assessment of groundwater parameters to evaluate the potential for enhanced in situ microbial degradation of the chlorinated solvents with residual cosolvent from the extraction technology. Another objective was to evaluate the subsurface and environmental impact of large-scale ethanol injection. Ethanol was selected as the cosolvent for the solvent extraction demonstration because it has acceptable cosolvency properties and is a suitable electron donor for stimulation of in situ reductive dechlorination by native microorganisms. The specific objectives of the SERB demonstration were to: • Develop a greater understanding of halorespiration and evaluate its potential as a bioremediation technology for PCE and other halogenated compounds either by enhancement of native activity (biostimulation) or by introduction of active microorganisms (bioaugmentation) • Develop a detailed understanding of site geomorphology, contaminant distribution, and hydrology utilizing new and innovative tools and procedures • Develop molecular probes for the detection of known PCE-dechlorinating organisms • Conduct molecular analysis of site material for PCE-dechlorinating communities and isolates • Perform laboratory microcosm tests to detect and evaluate microbial processes • Evaluate cosolvent injection/extraction design and performance • Evaluate in situ bioprocess performance • Evaluate cost and performance • Predict long-term performance of the SERB approach • Design and produce an implementation manual for SERB technologies

5.3 Technical approach 5.3.1 Laboratory process and microbial ecology Attempts to achieve PCE dechlorination to ethene in the field have obtained variable success. This observation and previous work by researchers at the Center for Microbial Ecology at Michigan State University with materials from many different contaminated sites have suggested that this inconsistency may be due to site differences in microbial populations. Thus, one objective of the chlorinated solvents thrust area was to develop our understanding of the relationship between the extent of PCE dechlorination and particular microbial populations. To support this objective, the decision was made to develop a concerted effort focused on the analysis of microbial communities associated with successful PCE dechlorination, which was

74

Bioremediation of Recalcitrant Compounds

conducted in addition to and in support of the pilot test of the SERB technology. The subobjectives of the effort were as follows: • Identification and development of molecular probes for the detection of known PCE-dechlorinating organisms • Molecular analysis of PCE-dechlorinating communities and isolates • Enrichment of PCE-dechlorinating bacteria in materials from contaminated field sites • Testing of molecular probes for detection of chloroethene-dehalogenating communities in site materials • Testing of molecular probes for detection of chloroethene-dehalogenating communities in Sages/Jacksonville site material • Evaluating potential for ethanol for toxicity to native microorganisms

5.3.2 Field process and site evaluation 5.3.2.1 Site description The study was conducted at a former dry cleaner site in Jacksonville, FL, that was operational from 1968 to 1973 and from 1979 to 1989 and is currently abandoned. Prior to this, a gasoline service station was operated on the site from 1953 until the mid- to late-1960s. Although the PCE release history is unknown, dry-cleaning fluids were not stored on site and were replenished by a delivery truck where occasional releases occurred. A former floor drain sump was also used to collect drainage from the floor near the dry-cleaning machine, and high subsurface PCE concentrations were found in this area. Suspected sources of PCE contamination included the sump area, the area outside the back door of the building, where filter tubes were washed with PCE, and the area where a waste oil tank was removed that may have been used for both service station activities (parts cleaning) and dry-cleaning waste. The sump area was targeted for the cosolvent flushing pilot test (Figure 5.4).

5.3.2.2 Site/source zone characterization The subsurface site materials consisted primarily of fine-grained sands with a discontinuous clay layer at approximately 10.7 m below ground surface (bgs). The water table ranged from 2 to 2.6 m bgs during the 21/2 years of groundwater monitoring. The natural hydraulic gradient during this time ranged from 0.0025 to 0.006. High PCE concentrations were measured at locations in the sump area over a vertical interval of 7.9 to 9.6 m bgs. A high-frequency subsampling technique determined the PCE in this area to be in thin layers (5 to 8 cm thick) that were not necessarily continuous over the extent of the source area (Jawitz et al., 2000). Initial concentrations of PCE in groundwater from wells in the targeted area ranged from 42 to 90 mg/l. More detailed characterization data can be

Chapter five: Solvent contaminated soils and groundwater

75

N Former Fuel UST Locations Former Dry Cleaner Bldg Floor Drain Former Dry Cleaner Machine Location

Secondary DNAPL Source Area Former Waste Oil UST Boiler Room

DNAPL Source Area Target Zone Sump for Floor Drain

Drainage Canal

0

20

40

60

80 Ft.

Figure 5.4 Sage’s site map indicating former building and suspected contamination areas.

found in the reports prepared for the Florida Department of Environmental Protection by LFR Levine Fricke (1997a, 1997b, 1998, 2000).

5.3.2.3 Evaluation of preinjection natural attenuation potential Initial evaluation of the Sage’s site included an evaluation of natural attenuation. Groundwater parameter data from the existing wells were evaluated for the potential for biodegradation based on scores for the natural attenuation parameters developed according to the weighting guidelines provided in Wiedemeier et al. (1996). The detailed results of this evaluation are reported by LFR Levine Fricke (1997a, 1997b), in which the majority of the wells (five of six) indicated inadequate to limited evidence for natural attenuation of chlorinated solvents (Table 5.1). Reductive dechlorination was found not to be favored because of aerobic conditions, lack of sulfate reduction or methane production, and low levels of dechlorination products such as chloride, ethene, and ethane. Based on this assessment, the conclusion was that remediation by natural attenuation had limited potential as a significant component of the overall remedial strategy for effective cleanup of the site. It was also recognized that the DNAPLs present in the subsurface would serve as continuing sources of chemicals that would contribute as a long-term source to the dissolved groundwater plumes. The recommendation was to conduct the cosolvent flushing pilot test to study the effectiveness of alcohol flooding for DNAPL remediation.

5.3.2.4 Well description Three injection wells (IWs) surrounded by six recovery wells (RWs) were used for implementation of the cosolvent flushing test and groundwater monitoring. These wells were 10-cm-diameter polyvinyl choride (PVC) and

76

Bioremediation of Recalcitrant Compounds

Table 5.1a Natural Attenuation Parameters and Criteria Points

Analyte

Concentration Criteria

Interpretation

Criteria Points Awarded

Oxygen

1 mg/l

Vinyl chloride may be oxidized aerobically, but reductive dechlorination will not occur

–3

Nitrate

1 mg/l

Reductive pathway possible

3

Sulfate

1 mg/l

Reductive pathway possible

3

Methane

>0.1 mg/l

Ultimate reductive daughter product

2

Methane

>1 mg/l

Vinyl chloride accumulates

3

Methane

2° background

Daughter product of organic chlorine; compare chloride in plume to background conditions

—

Perchloroethene

Material released

2

Trichloroethene

Material released or daughter product of perchloroethene

2

Dichloroethene

Material released or daughter product of perchloroethene; if amount of cis-1,2-dichloroethene is greater than 80% of total dichloroethene, it is likely a daughter product of trichloroethene

2

Chapter five: Solvent contaminated soils and groundwater

77

Table 5.1a Natural Attenuation Parameters and Criteria Points (Continued)

Analyte

Concentration Criteria

Interpretation

Criteria Points Awarded

Vinyl chloride

Material released or daughter product of dichloroethenes

2

Ethene/ethane

Daughter product of vinyl chloride/ethene

>0.01 = 2; >0.1 = 3

Chloroethane

Daughter product of vinyl chloride under reducing conditions

2

1,1,1-Trichloroethane

Material released

—

1,1-Dichloroethene

Daughter product of trichloroethene or chemical reaction of 1,1,1-trichloroethane

—

Table 5.1b Interpretation of Natural Attenuation Scores Score

Interpretation

0 to 5

Inadequate evidence for biodegradation of chlorinated compounds

6 to 14

Limited evidence for biodegradation of chlorinated compounds

15 to 20

Adequate evidence for biodegradation of chlorinated compounds

>20

Strong evidence for biodegradation of chlorinated compounds

were installed with a 25-cm-diameter hollow-stem auger. A sand pack was used in the screened interval and a bentonite plug was placed immediately above the screened zone. The three IWs were screened from 7.6 to 9.9 m bgs, and the six RWs were screened from 7.9 to 9.6 m bgs. Monitoring wells (MWs) that were 5-cm-diameter PVC were also installed with a hollow-stem augor as described above and screened from 7.9 to 9.6 m bgs. A series of 2.5-cm PVC wells (C-wells) were installed with a cone penetrometer truck and screened from 8.3 to 9.1 m bgs. All wells were used for groundwater monitoring, and their locations are shown in Figure 5.5. A fully screened well (MW-516) was installed outside the contaminated area utilized for conducting flowmeter and aquifer pumping tests. The well was constructed of 5-cm schedule 40 PVC casing with machine 0.025-cm-slotted screen from approximately 3 to 10.7 m bgs. This well was installed with a sonic drilling rig, which pushed the casing into the subsurface. Seven multilevel samplers (MLSs) were installed within the source zone and the ring of RWs. The five sampling depths at each MLS location were 8.1, 8.7, 9.1, 9.4, and 9.9 m bgs for a total of 35 sampling locations. These locations were sampled by the University of Florida (UF), and data are reported elsewhere (Jawitz et al., 2000).

RW-001 25–35 feet 3 NA 2 3 2 3 0 NA NA 1 0 2 1 0 0 0 2 0 0 0 0 NA 0 19

Analyte

Oxygen Oxygen Nitrate Iron (II) Sulfate Sulfide Methane Methane Methane Oxidation reduction potential pH Total Organic Carbon Temperature Alkalinity Chloride Perchloroethene Trichloroethene Dichloroethene Vinyl chloride Ethene/Ethane Chloroethane 1,1,1-trichloroethane 1,1-dichloroethene

Total Points Awarded

7

NA –3 2 0 2 3 0 NA NA 0 0 2 1 0 0 0 0 0 0 0 0 NA 0

MW-001 10–20 feet

9

NA –3 2 0 2 3 0 NA NA 0 0 2 1 0 0 0 2 0 0 0 0 NA 0

MW-002 8–18 feet

Table 5.1c Natural Attenuation Scores for Wells at Sage’s Site

11

NA –3 2 3 2 0 2 NA NA 0 0 2 1 0 2 0 0 0 0 0 0 NA 0

MW-004 23–33 feet

5

NA –3 0 0 2 3 0 NA NA 0 0 2 1 0 0 0 0 0 0 0 0 NA 0

MW-009 2–10 feet

6

NA –3 2 0 2 3 0 NA NA 1 0 0 1 0 0 0 0 0 0 0 0 NA 0

MW-010 Not Recorded

78 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater

3

79

N

1 7 3

4 514

505

512

513

3

2

1

509

2

2

1 7

4 6

3

5 506

516

510 511

508 507 # Injection Well # Recovery Well # C-well (1 inch PVC) Monitoring Well (2 inch PVC)

2

Figure 5.5 Well locations at the former Sage’s Dry Cleaner site in Jacksonville, FL.

5.3.2.5 Groundwater flow characterization Characterization of the groundwater flow was initially conducted by LFR Levine Fricke (1997a, 1997b). Groundwater flow direction in the surficial aquifer was generally westward toward the St. Johns River. Groundwater elevations monitored over the project period showed similar results (Figure 5.6). The flow in the adjacent drainage canal is to the north, which may impact groundwater flow. The relationship of the drainage canal to groundwater flow in the surficial aquifer was unclear from the initial investigation. During periods following major rainfall events, the depth to water table would be expected to decrease, and the drainage canal could potentially impact local groundwater flow direction. The hydraulic gradient during the project period ranged from 0.0025 to 0.006. A slug test conducted by LFR in MW-4, which was screened from 7 to 10.1 m bgs, gave an average hydraulic conductivity of 3.2 m/day. Slug tests conducted in MW-1, MW-2, and MW-3, which were screened from approximately 3 to 6.1 m bgs, gave an estimated average hydraulic conductivity of 6.1 m/day. The results from this study were used in the design of the solvent extraction system. In July 2000, an electromagnet borehole flowmeter was used by the EPA to define the relative hydraulic conductivity distribution of aquifer materials screened by well MW-516 (Acree, 2000). The electromagnetic borehole flowmeter is a commercially available system manufactured by Tisco, Inc., and consists of a 1.25-cm-ID (inner diameter) downhole probe, a 2.5-cm-ID downhole probe, and an electronics module. The investigation was conducted using procedures based on the methods of Molz et al. (1994) and Young et al. (1998). Flow rate measurements using the electromagnetic borehole flowmeter were made under ambient and constant-rate pumping conditions at

80

Bioremediation of Recalcitrant Compounds

37.45

37.20

N

Ground Water Flow

20

5

0

37.4

37.20

Drainage Canal

.70 37

DNAPL Source Area Target Zone

40

60

80 Ft.

Figure 5.6 Groundwater elevations at the Sage’s site on September 22, 1999, showing groundwater flow to the west toward the drainage canal.

a measurement interval between 15 and 60 cm within the well screen. The 1.25-cm-ID probe was used for flow measurements under ambient conditions and the 2.5-cm probe was used for measurements under pumping conditions. The test was performed at a constant pumping rate of 8.8 l/min and repeated at a rate of 4.0 l/min. The rates were chosen to induce sufficient flow from each test interval with negligible head loss across the 2.5-cm-ID downhole probe. Measurements of vertical flow rates under ambient and constant-rate pumping (induced flow) conditions were analyzed using methods described by Young et al. (1998). The estimates of hydraulic conductivity for this test represent averages over the measurement interval, which was between 15 and 60 cm within the well screen, and are dependent upon the interval thickness. The data collected indicated that relatively conductive materials were present from the water table to a depth of approximately 7.9 m bgs at this location. Materials of lower hydraulic conductivity were present below this depth. Drawdown in adjacent monitoring wells was measured during the constant-flow-rate extraction test performed for the flowmeter survey, and these data were used for estimation of the hydraulic conductivity of the aquifer materials. The methods of Neuman (1975) were used to estimate the hydraulic conductivity using an aquifer thickness of 9.1 m, which corresponds approximately to the interval from the water table to the depth of the well screen. These estimates of aquifer hydraulic conductivity

Chapter five: Solvent contaminated soils and groundwater

81

−1 Water Table

−2

Top of Screen

−3 Depth (mBTOC)

−4 −5 −6 −7 −8 Flushing Zone

−9 −10

Bottom of Screen

−11 0

5 10 15 20 Hydraulic Conductivity (m/d)

25

Figure 5.7 Estimated hydraulic conductivity profile for materials screened by well MW-516.

ranged from 4.3 to 14.3 m/day using the drawdown data from five monitoring wells with an average of 9.1 m/day for the aquifer. This average value of 9.1 m/day was then used to estimate the hydraulic conductivity profile using the estimates of relative hydraulic conductivity for each interval measured during the flowmeter study (Figure 5.7). The estimates of hydraulic conductivity for the individual intervals with the flowmeter test show variability with depth as would be expected based on the site geology (Figure 5.7). The average hydraulic conductivity for the materials in the overlying sands from 3 to 7.6 m bgs is approximately 14.9 m/day, which is much higher than the value of 6 m/day estimated by LFR. The value for the lower sands from 7.6 to 10.7 m bgs of 3 m/day is significantly lower than the upper sands and is similar to the value of 3.2 m/day estimated by LFR. The site characterization showed that the highest concentrations of PCE contaminant in the aquifer sediment (6000 to 90,000 mg/kg) were from 8 to 9.2 m bgs (Jawitz et al., 2000). This corresponds approximately to the zone with the lowest measured hydraulic conductivities. Based on the properties of PCE, it has been shown that PCE will migrate downward through the aquifer to a zone of lower permeable materials, where it will be trapped in the pore space. Depending upon the volume of DNAPL moving through the aquifer and the thickness of the zones with lower hydraulic conductivities, it is possible for the DNAPL to break through these zones and move deeper into the aquifer.

82

Bioremediation of Recalcitrant Compounds

5.4 Materials and methods 5.4.1 Laboratory evaluations 5.4.1.1 Molecular ecology A number of chloroethene sites were used to provide source material for evaluation. Microcosm experiments were conducted to evaluate sediment, soil, and aquifer materials for their potential to reductively dechlorinate PCE and other chlorinated ethenes under anaerobic conditions and to assess the ability of various enhancing agents to stimulate PCE dechlorination. Eleven freshwater sediments, four marine sediments, five soils, and seven aquifer samples were used to construct microcosms containing 10 to 15% (v/v) sample under anaerobic conditions with either anaerobic mineral medium or phosphate-buffered saline (PBS) and amended with PCE. Dechlorination of PCE was monitored by gas chromatography (GC) under conditions that can resolve ethene and each of the chlorinated ethenes. The enrichments that completely dechlorinated PCE were further enriched on the less chlorinated ethenes, cis-DCE, and VC, separately. The resulting subcultures completely dechlorinated their respective chlorinated ethene and were maintained for more than 10 transfers in basal salt mineral medium. Subcultures were subjected to an analysis of the 16S rRNA genes by denaturing gradient gel electrophoresis (DGGE) and terminal restriction fragment length polymorphism (T-RFLP). The changes in community structure were evaluated. This technique is performed by polymerase chain reaction (PCR) amplification of 16S rRNA genes from DNA extracted from a microbial community in which one of the primers is labeled with a fluorescent molecule. The resulting fluorescently labeled PCR product is then subject to restriction enzyme digests. The fragments are resolved using an automated DNA sequencer, which allows for the detection of only the fluorescently labeled terminal fragments. The resolved terminal fragments provide a fingerprint of the community and an estimate of the number of ribotypes in a community. The Sage’s site material was subjected to direct and postamplification ribosomal gene probing with probes developed from known PCE-dechlorinating bacteria.

5.4.1.2 Ethanol toxicity Soil samples were collected from the former dry cleaner site in accordance with methods described in the published literature, ensuring sample integrity. Microcosms were prepared in triplicate in glass serum bottles fitted with Teflon™-coated rubber septa with approximately 50 g of soil. The amount of soil used for the microcosms was dictated by the quantity required for analyzing the phospholipid fatty acid (PLFA) profiles of the samples. The liquid phase in each microcosm consisted of Byrd’s Mill medium (BMM), ethanol, and other additives, such as KNO3, Na2SO4, or PCE, depending upon the electron acceptor condition selected for the particular experiment. Byrd’s Mill medium was made with Byrd’s Mill water (Byrd’s Mill Spring,

Chapter five: Solvent contaminated soils and groundwater

83

OK) at 500 ml/l, distilled water at 500 ml/l, and (NH4)2HPO4 at 1.32 g/l, and this supplied the basic nutrients to the sample’s microbial community. Different concentrations of ethanol, ranging from 0.05 to 70% v/v, were used for the study. Anaerobic microcosms were prepared in a glove box with a nitrogen–hydrogen atmosphere, and the aerobic microcosms were prepared in a laminar flow hood. Samples from each microcosm were withdrawn at regular intervals to determine the degradation of added ethanol. Degradation of ethanol was expressed in terms of attenuation rates of ethanol and was measured with different electron acceptors (O2, NO3–, SO42–, and PCE) and different stressor levels. Exposure time to the stressor for the different microcosms was also varied and can be divided into two groups: (1) long-term exposure and (2) transient exposure. Spent microcosms were frozen and sent for PLFA analysis.

5.4.2 Field studies 5.4.2.1 Groundwater sampling and analysis Groundwater samples were collected prior to the implementation of the cosolvent extraction technology to provide background conditions. Following the cosolvent extraction test, groundwater samples were collected and analyzed quarterly over a 2 1/2 -year period for nutrients, chloroethenes, ethanol, dissolved gases, and volatile fatty acids. Samples for ethanol and alcohol tracer analysis and for PCE and daughter product analysis were collected in 40-ml volatile organic analysis (VOA) vials without headspace. The alcohol compounds were separated by gas chromatography on a DB624 capillary column and detected with a flame ionization detector. Chlorinated compounds were analyzed by purge-and-trap gas chromatography, which utilized a 0.32 mm × 60 m RESTEK Rtx-Volatiles column for separation and a flame ionization detector for detection. Samples for chloride and sulfate analysis were collected in 100-ml polyethylene bottles and analyzed using Waters capillary electrophoresis method N-601. Samples for nitrate, phosphate, and dissolved organic carbon analyses were collected in 100-ml polyethylene bottles and preserved with concentrated sulfuric acid. Nitrate and phosphate were analyzed using a Lachat system. Dissolved organic carbon samples were analyzed with a Dohrman Carbon Analyzer. Subsamples for volatile fatty acid analysis were collected from the 100-ml nonacidified polyethylene bottle described above. These samples were analyzed by high-performance liquid chromatography (HPLC) with a Dionex system utilizing a Dionex ICE-AS1 IonPac column and AMMS-ICE MicroMembrane Suppressor. A Waters 431 conductivity detector was used for analyte detection. Dissolved gas samples were collected in 100-ml glass serum bottles without headspace and sealed with Teflon-lined septa and crimp caps.

84

Bioremediation of Recalcitrant Compounds

Samples were analyzed for methane, ethane, and ethene using an HP P200 Series Micro Gas Chromatograph and a thermal conductivity detector.

5.4.2.2 Core material collection Subsurface sediment samples were collected for direct molecular analysis (16S and PLFA) prior to the cosolvent flushing test and at approximately 1 and 2 years after the cosolvent flushing test. Samples were collected from approximately 7.9 to 9.4 m bgs to coincide with the zone targeted by the cosolvent flushing test. The results of the molecular analysis are not available at this time. Subsurface materials collected from these locations and locations where additional MLS, injection/extraction wells, and monitoring wells were installed have also been used as a source of material and microorganisms for laboratory biotransformation studies. Core materials collected for the molecular analysis prior to the cosolvent flushing test were put into sterile conical tubes and frozen by immersing into a liquid nitrogen canister for 2 to 5 min and stored frozen until analysis. Sterile spatulas were utilized to obtain samples of the undisturbed centers of the core. Between core sections, all materials used were sterilized with isopropyl alcohol. Core materials were collected from IW-3, as this well was installed, from the area between RW-6 and RW-7 (C6), and near MW-2 (C5). Additional materials were collected during installation of the C-wells (C1 to C4, C7). Locations of these samples are shown in Figure 5.8. In July 1999 and July 2000, core materials were collected near the original locations using the direct-push Simco rig. These materials were collected into polyethylene sleeves that were sectioned by sawing and then capped N

C4 C12,13 C34

C7 C20 C37

C3 C15 C33

C2 C14 C32

C1 C16,17,18 C31 C6 C10,19 C36

C5 C1 C35

Core Locations 1998 Samples 1999 Samples 2000 Samples

Figure 5.8 Locations of core materials collected for molecular analysis.

Chapter five: Solvent contaminated soils and groundwater

85

Table 5.2 Core Sample Locations for Direct Molecular Analysis (16S and PLFA)

Location DNAPL target area, between RW-6 and RW-7 Background, near MW-2 Upgradient, well C7 Downgradient, well C1 Downgradient, well C2 Downgradient, well C3 Downgradient, well C4

Pretest Core (1998)

1-Year Posttest Core (1999)

2-Year Posttest Core (2000)

C6

C10, C19

C36

C5 C7 C1 C2 C3 C4

C11 C20 C16, C17, C18 C14 C15 C12, C13

C35 C37 C31 C32 C33 C34

with plastic caps. Materials used to handle these cores were sterilized with isopropyl alcohol between core sections. Cores were immersed in a liquid nitrogen canister for 5 min and stored frozen until analysis. A compilation of the core locations and times of collection is given in Table 5.2.

5.4.2.3 Cosolvent flush and partitioning tracer tests The pilot-scale field test of in situ alcohol flushing was conducted in August 1998 as described by Jawitz et al. (2000). The alcohol flushing began on August 9, 1998, and ended on August 15, 1998. The injection/extraction system was designed so that the IWs were screened deeper than the RWs to promote an upward flow of the injected fluids. The configuration of injection/extraction wells was selected based on numerical model simulations conducted by Sillan (1999). A total flow of 15.1 l/min was distributed equally to the three IWs. The interior recovery wells (RW-3, RW-4, RW-6, and RW-7) had extraction rates of 5.9 l/min, and the outer two wells (RW-2 and RW-5) had extraction rates of 3.4 l/min, which resulted in a 2:1 extraction-to-injection ratio. Steady-state flow was maintained so that the water table position remained constant during the tests. The test zone was flushed with 34 kl (equivalent to 2 pore volumes) of a 95% ethanol/5% water mixture over the 3-day period. In the initial 10 hours the ethanol concentration was ramped from 0 to 95% to minimize density problems. Posttest hydraulic containment began on August 15, 1998, and continued until August 25, 1998, after the ethanol concentration in the treatment system influent dropped below the 10,000 mg/l termination criterion (LFR Levine Fricke, 1998). A partitioning tracer test conducted prior to the alcohol flushing was used to estimate approximately 68 l of PCE in the zone swept by the injection/recovery well system (Jawitz et al., 2000). A postflushing partitioning tracer test indicated that approximately 26 l of PCE remained in the swept zone, so approximately 42 l of PCE was removed during the cosolvent flush. A more detailed description of the tracers used and data evaluation is given in Jawitz et al. (2000).

86

Bioremediation of Recalcitrant Compounds

5.4.2.4 Hydrogen gas analysis In July 1999, hydrogen gas analysis was conducted on groundwater from selected monitoring wells. Monitoring wells were pumped at a rate of 250 to 300 ml/min through a 250-ml gas sampling bulb containing a gas pocket of approximately 20 ml volume. A 2-ml gas sample was collected from the gas pocket with a gas-tight syringe and injected into an RGA3 Reduction Gas Analyzer. Samples were collected and analyzed over time until equilibrium was reached at individual wells.

5.5 Accomplishments 5.5.1 Solvent extraction (SE) pilot test The cosolvent flushing pilot test activities were completed on September 10, 1998, with the collection of posttest groundwater samples. More detailed information concerning the cosolvent flushing test and an evaluation of the partitioning tracer data can be found in Jawitz et al. (2000) and LFR Levine Fricke (1998). Enhanced dissolution and solubilization of PCE was demonstrated as a result of cosolvent flushing. Analytical data from the recovery wells, except for RW-5, show that the peak PCE concentrations were 3 to 80 times larger than the initial PCE concentrations (Figure 5.9). According to the partitioning tracer test, RW-5 showed very low amounts of DNAPL present in the swept zone, and the precosolvent flush groundwater PCE concentration was 1 mg/l. During the cosolvent flush, the maximum concentration of PCE in this well was 2 mg/l, with a maximum ethanol concentration of 4% (Figure 5.9d). The partitioning tracer test indicated that the other recovery wells contained a significant volume of DNAPL in the individual swept zones. These were 3.2 l for RW-2, 11.6 l for RW-3, 4.6 l for RW-4, 7.2 l for RW-6, and 14 l for RW-7. RW-3 (Figure 5.9b), RW-4 (Figure 5.9c), and RW-6 (Figure 5.9e) showed peak PCE concentrations that were approximately 25 times the initial concentration. RW-7 showed the greatest increase in PCE concentration, which was more than 80 times the initial concentration (Figure 5.9f). The PCE concentrations in RW-2 were variable at the beginning of the cosolvent flush and reached a concentration of approximately two to three times the initial concentration (Figure 5.9a). The maximum ethanol concentration in these wells ranged from 18 to 45%. Groundwater concentrations of PCE in all of the recovery wells decreased to initial concentrations following the cosolvent flushing test. Comparison of average groundwater concentrations in the recovery wells prior to and post cosolvent flushing showed a high degree of variance and only 4% decrease in concentrations (Table 5.3). This same comparison for the three injection wells gave a 99% decrease in concentrations and a similarly large variance. Averaging the concentrations for both the recovery and injection wells showed a decrease in average concentration of 62% (Table 5.3).

PCE (mg/L)

0

10

20

30

Ethanol

PCE

40

40

1 0

3 2

5 4

0

200

400

600

800

0

30

50

60 50 40 30 20 10 0

30 PCE 25 20 Ethanol 15 10 5 0 100

Ethanol

PCE

1400 1200 1000 800 600 400 200 0 0

−10

500 400 300 200 100 0

Ethanol (%)

Volume (kL)

50

RW006

Volume (kL)

10

RW003 Ethanol (%)

Volume (kL)

20

RW005

Volume (kL)

1400 1200 1000 800 600 400 200 0 −10 30

Volume (kL)

50

RW007

Volume (kL)

10

RW004

Ethanol

PCE

Ethanol

PCE

50

35 30 25 20 15 10 5 0

70 60 50 40 30 20 10 0

Ethanol (%)

Ethanol (%)

Figure 5.9 Ethanol and PCE concentrations in recovery wells during the cosolvent flushing test. (a) Ethanol and PCE concentrations in RW-2. (b) Ethanol and PCE concentrations in RW-3. (c) Ethanol and PCE concentrations in RW-4. (d) Ethanol and PCE concentrations in RW-5. (e) Ethanol and PCE concentrations in RW-6. (f) Ethanol and PCE concentrations in RW-7.

2 0

0

24 21 18 15 12 9 6 3 0 Ethanol (%)

6 4

10 8

−40

60

Ethanol

PCE

Ethanol (%)

PCE (mg/L)

PCE (mg/L) PCE (mg/L)

RW002

PCE (mg/L) PCE (mg/L)

160

Chapter five: Solvent contaminated soils and groundwater 87

88

Bioremediation of Recalcitrant Compounds Table 5.3 PCE Removal Effectiveness Calculated by Different Techniques Technique RW (n = 6) IW (n = 3) RW and IW (n = 9) Core material Partitioning tracer

Preflushing (mg/l)

Postflushing (mg/l)

Removal Effectiveness

17.2 ± 25.9 54.5 ± 46.3 29.6 ± 36.1

16.5 ± 13.5 0.6 ± 0.5 11.2 ± 13.2

0.04 0.99 0.62 0.65 0.62–0.63

This calculation compares well with estimates of the amount of PCE removed during the cosolvent flushing test by two techniques explained in more detail in Jawitz et al. (2000). Comparison of core material collected prior to and following the cosolvent flushing test indicated a removal effectiveness of 0.65. The overall removal effectiveness for the recovery wells calculated from the partitioning tracer data was 0.62 to 0.63, but it varied significantly at individual recovery wells, depending upon the technique used for analysis. The hydraulic containment system was adequate to maintain capture within the testing zone. Groundwater elevations measured during the course of the pilot test showed an inward gradient from the outer monitoring wells (MW-505, MW-506, MW-507, and MW-509) toward the recovery wells. DNAPL PCE migration was not observed during the test, and ethanol was detected in only two of the outer monitoring wells following the test. In the first sampling event following the cosolvent flushing test, the ethanol concentration in MW-505 was 4620 mg/l, and in MW-506 was 0.74 mg/l. MW-506 is located upgradient from the targeted area, and MW-505 is located to the north. Neither well is located downgradient of the direction of groundwater flow. Analysis of the partitioning tracer data by Jawitz et al. (2000) estimated 42 l of PCE was removed by the cosolvent flushing test. Remaining PCE DNAPL was estimated to be 26 l. Approximately 92% of the 34 kl of injected alcohol was recovered at the end of the hydraulic containment so that 2.72 kl of ethanol remained in the subsurface as an electron donor for enhancement of the microbial processes. The cosolvent extraction demonstration at the Sage’s site was a successful demonstration of the removal of a significant mass of DNAPL contaminant from a selected zone. Improvement in removal effectiveness could be obtained by selection of a remedial cosolvent with a higher cosolvency property and by increased volume of cosolvent pumped through the system. This test demonstrated the capability of this technology to be utilized at other Florida DEP DSCP sites. Information obtained from this test will aid in the design of a full-scale technology implementation for the Sage’s site.

Chapter five: Solvent contaminated soils and groundwater

89

5.5.2 Residual biotreatment (RB) monitoring and assessment The postflush assessment of the ethanol-enhanced in situ microbial degradation of the chlorinated solvents is ongoing. This assessment of in situ reductive dechlorination, the residual biotreatment phase of SERB, is based on site monitoring of geochemistry and contaminant chemistry as well as laboratory and field evaluations of the impact of ethanol on site microbial communities. Groundwater samples from the monitoring wells at the site have been collected and analyzed quarterly for 21/2 years along with additional site characterization activities (Table 5.4). A database has been established that contains the measured results, including residual ethanol concentrations, its possible metabolic products, PCE and potential biotransformation intermediates, and other geochemical parameters. The compiled data were graphed by individual wells to show changes in concentration of the parameters over time. Relatively complete data sets were obtained for all MW- and C-wells. IWs and RWs were not sampled consistently throughout the field evaluation; thus, the bar charts of this data are somewhat limited. IW-1 to IW-3, RW-2, and RW-5 were not sampled after September 1999. Contour plots of the data were made to better visualize the changes in the parameters over time. Surfer™ was used to create these contour plots Table 5.4 Sampling Dates and Activities at the Sage’s Site Date July 16, 1998 August 11, 1998 August 9–15, 1998 August 15, 1998 August 25, 1998 September 8, 1998 October 20, 1998 December 3, 1998 January 26, 1999 May 26, 1999 July 14, 1999 September 23, 1999 March 14, 2000 June 14, 2000 July 13, 2000 September 23, 2000 December 29, 2000 March 20, 2000

Event Preethanol flush groundwater monitoring, site characterization, partitioning tracer test C-wells sampled (included with preethanol flush) Cosolvent flushing test Partitioning tracer test, hydraulic containment End hydraulic containment 1-month groundwater monitoring 2-month groundwater monitoring 3.5-month groundwater monitoring 5.5-month groundwater monitoring 9.5-month groundwater monitoring 11-month groundwater monitoring Hydrogen gas survey, core sampling 13.5-month groundwater monitoring 19-month groundwater monitoring 22-month groundwater monitoring 23-month groundwater sampling (partial) Core sampling 25-month groundwater monitoring 28-month groundwater monitoring 31-month groundwater monitoring

90

Bioremediation of Recalcitrant Compounds

with the data from the MW-, C-, RW-, and IW-wells. MW-1 to MW-3, which were screened over a shallower interval and were outside the contaminated plume, were not utilized in this exercise, nor was MW-508, which was screened over a deeper interval but showed high PCE concentrations. Changes in concentrations of the measured parameters give an indication of the change in subsurface geochemistry and the stimulation of microbial activity as a result of the SERB process. Ethanol concentrations stayed at approximately 10,000 mg/l in the area of the cosolvent extraction during the first year of groundwater monitoring and then decreased to less than 2000 mg/l after the second year (Figure 5.10). MW-505 and MW-509 were the only monitoring wells outside the injection/ extraction area to show significant concentrations of ethanol (>4000 mg/l) following the cosolvent flush. These wells bound the injection/extraction system on the north and west. Downgradient wells (MW-510, MW-512, MW-513, and MW-514) showed low concentrations of ethanol (90 mg/l), and it is suspected that these wells were completed in a previously undetected source

0

2000

4000

6000

8000

10000

12000

14000

RW-003

RW-004

(a)

Well Name

RW-006

Ethanol by Well

RW-007

31 Months

28 Months

25 Months

22 Months

19 Months

13.5 Months

11 Months

9.5 Months

5.5 Months

3.5 Months

2 Months

1 Month

Figure 5.10 Ethanol concentrations by well during groundwater monitoring. (a) Ethanol concentrations in the recovery wells. (b) Ethanol concentrations in the monitoring wells. (c) Ethanol concentrations in the C-wells.

Conc. (mg/L)

16000

Chapter five: Solvent contaminated soils and groundwater 91

0

1000

2000

3000

4000

5000

Ethanol by Well

(b)

Well Name

MW-505 MW-506 MW-507 MW-508 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514

Figure 5.10 (Continued)

Conc. (mg/L)

6000

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

92 Bioremediation of Recalcitrant Compounds

0

500

1000

1500

2000

2500

Figure 5.10 (Continued)

Conc. (mg/L)

3000

C1

C2 (c)

Well Name

C3

Ethanol by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 93

94

Bioremediation of Recalcitrant Compounds ~1 Month Post-Flush

~2 Months Post-Flush

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

16000 mg/L

MW-505

~3.5 Months Post-Flush

14000 mg/L 12000 mg/L

MW-506

MW-510MW-509 MW-511 MW-507

10000 mg/L 8000 mg/L 6000 mg/L

~5.5 Months Post-Flush

4000 mg/L 2000 mg/L C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-510 MW-509 MW-511 MW-507

MW-506

0 mg/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.11 Ethanol contour plots over the groundwater monitoring period (12,000 mg/l = 260 mM = 1.2%).

~9.5 Months Post-Flush

~13.5 Months Post-Flush

C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-510 MW-509 MW-511 MW-507

C3

MW-506

MW-510 MW-509 MW-511 MW-507

~19 Months Post-Flush

16000 mg/L 14000 mg/L 12000 mg/L

MW-505 MW-506

10000 mg/L 8000 mg/L 6000 mg/L

~22 Months Post-Flush

4000 mg/L 2000 mg/L C7 C4

C3

MW-514MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 C2

MW-505

MW-510 MW-509 MW-511 MW-507

MW-506

0 mg/L

C3 MW-512 C1

MW-505

MW-510 MW-509 MW-511 MW-507

MW-506

Figure 5.12 Ethanol contour plots over the groundwater monitoring period (12,000 mg/l = 260 mM = 1.2%).

area. C1 and C2 maintained relatively high concentrations, and C3 and C4 appeared to have increasing concentrations of PCE for the first 6 months of groundwater monitoring. Concentrations of PCE in all of these wells decreased to less than 10 mg/l after 2 years of monitoring. It is interesting to note that all wells showed a decrease in PCE concentration even though

Chapter five: Solvent contaminated soils and groundwater

95

~28 Months Post-Flush

~25 Months Post-Flush

C7

C7 C4

C3

MW-514 MW-513 C2

C4 MW-512

C1

MW-514MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-512 C1

16000 mg/L

MW-505

MW-510MW-509 MW-511 MW-507

MW-506

14000 mg/L 12000 mg/L 10000 mg/L 8000 mg/L 6000 mg/L

~31 Months Post-Flush

4000 mg/L 2000 mg/L C7

C4

0 mg/L

C3

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.13 Ethanol contour plots over the groundwater monitoring period (12,000 mg/l = 260 mM = 1.2%).

C1, C2, and C3 showed an increase in ethanol concentrations over this same period. The contour plots of the PCE data show dramatically decreasing concentrations in the groundwater over the monitoring period (Figure 5.15 to Figure 5.17). TCE and cis-DCE concentrations, which are daughter products that form by dechlorination of PCE, increased in the groundwater following the cosolvent flushing test. Prior to the introduction of ethanol to the subsurface, concentrations of TCE and cis-DCE were extremely low. TCE concentrations increased in all wells during the first 19 months of monitoring up to as high as 15 mg/l and then decreased to relatively low concentrations (10 mg/l) was detected in the farther downgradient MW- and C-wells (Figure 5.19). Contour plots of the TCE and cis-DCE concentrations over time give a good visualization of the daughter product formation at the site (Figure 5.20 to Figure 5.25). Groundwater samples were also analyzed for the isomers of cis-DCE, trans-1,2-DCE, and 1,1-DCE. These compounds were detected at low levels and made up less than 10% of the DCE analyzed. This ratio is indicative of biological transformations that seem to favor cis-DCE as the predominant product under most conditions. Vinyl chloride was not detected in any groundwater samples collected from the site during the monitoring period. Vinyl chloride is the

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

IW-001

IW-002 (a)

IW-003 RW-002 RW-003 RW-004 RW-005 RW-006 RW-007 Well Name

PCE for IW & RW’s

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.14 PCE concentrations by well during groundwater monitoring. (a) PCE concentrations in injection and recovery wells. (b) PCE concentrations in monitoring wells. (c) PCE concentrations in C-wells.

Conc. (ug/L)

100000

96 Bioremediation of Recalcitrant Compounds

0

10000

20000

30000

40000

50000

60000

Figure 5.14 (Continued)

Conc. (ug/L)

70000

M

05

-5 W M

06 -5 W M

07

-5 W M

08

-5 W M

10

-5 W

(b)

M Well Name

09

-5 W

PCE for MW’s

M

11

-5 W M

12

-5 W

M

13 -5 W

M

-5 W

14

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 97

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

Figure 5.14 (Continued)

Conc. (ug/L)

100000

C1

C2 (c)

Well Name

C3

PCE in C Wells

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

98 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater Pre-Ethanol Flush

99

~1 Month Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80000 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

70000 ug/L

MW-506

60000 ug/L 50000 ug/L 40000 ug/L

~2 Months Post-Flush

~3.5 Months Post-Flush

30000 ug/L 20000 ug/L C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

10000 ug/L 0 ug/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.15 PCE contour plots over the groundwater monitoring period (80,000 μg/ l = 480 μM). ~5.5 Months Post-Flush

~9.5 Months Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80000 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

70000 ug/L

MW-506

60000 ug/L 50000 ug/L 40000 ug/L

~13.5 Months Post-Flush

~19 Months Post-Flush

30000 ug/L 20000 ug/L

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

10000 ug/L 0 ug/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.16 PCE contour plots over the groundwater monitoring period (80,000 μg/ l = 480 μM).

next dechlorination product after DCE and is a known carcinogen and regulated contaminant. Interestingly, ethylene, which is the final dechlorination product in the biotransformation of PCE, was detected at noticeable levels within 6 months following the cosolvent flushing test (Figure 5.26). Ethylene concentrations were somewhat variable, which may be due more to the difficulty in sampling and analysis than to biological activity. The bar charts

100

Bioremediation of Recalcitrant Compounds ~22 Months Post-Flush

~25 Months Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-509 MW-510 MW-511

80000 ug/L

MW-505

70000 ug/L

MW-506

60000 ug/L

MW-507

50000 ug/L 40000 ug/L ~28 Months Post-Flush

~31 Months Post-Flush

30000 ug/L 20000 ug/L

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1 MW-509 MW-510 MW-511

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-505

MW-506 MW-507

MW-509 MW-510 MW-511

10000 ug/L 0 ug/L

C3

MW-506 MW-507

Figure 5.17 PCE contour plots over the groundwater monitoring period (80,000 μg/l = 480 μM).

of ethylene show a consistent trend of formation in the same wells that showed relatively high ethanol concentrations over time (MW-505, MW-509, C1 to C3, RW-3, and RW-7). The contour plots for this data indicate the area of main activity is from the injection/extraction area downgradient to well C2 (Figure 5.27 to Figure 5.29). The formation of the daughter products of PCE is an indication that in situ reductive dechlorination is beginning to occur. The process appears to have begun initially in the area where the cosolvent extraction was conducted and later in locations downgradient of the targeted source area. Production of chloride is concomitant with the reductive dechlorination of PCE, and because it is a relatively nonreactive anion, one would expect to see an increase in chloride concentrations over time. Background chloride concentration was less than 10 mg/l except in the injection/extraction zone, where concentrations were as high as 20 mg/l. Overall, there was a trend of increasing chloride concentrations over time (Figure 5.30). In particular, monitoring wells that showed the formation of cis-DCE also had an increase in chloride concentrations. Data from the RWs did not show this relationship and had more variable concentrations of both cis-DCE and chloride. Contour plots of chloride do show an increase in concentrations in the area immediately downgradient of the injection/extraction zone within 1 month of the cosolvent flushing test (Figure 5.31 to Figure 5.33). After approximately 1 year, this area of higher chloride concentrations appears to have moved farther downgradient to the area of C1 and C2. The maximum measured chloride concentration was approximately 2 mM, which is almost four times the maximum measured PCE concentration of 500 μM. This compares well to the complete dechlorination of PCE, which

0

1000

2000

3000

4000

5000

6000

7000

8000

RW-006

RW-007 (a)

Well Name

RW-003

RW-004

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.18 TCE concentrations by well during groundwater monitoring. (a) TCE concentrations in recovery wells. (b) TCE concentrations in monitoring wells. (c) TCE concentrations in C-wells.

Conc. (ug/L)

9000

TCE by Well

Chapter five: Solvent contaminated soils and groundwater 101

0

2000

4000

6000

8000

10000

12000

14000

TCE for MW’s

(b)

MW-505 MW-506 MW-507 MW-508 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514 Well Name

Figure 5.18 (Continued)

Conc. (ug/L)

16000

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

102 Bioremediation of Recalcitrant Compounds

0

2000

4000

6000

8000

10000

Figure 5.18 (Continued)

Conc. (ug/L)

12000

C1

C2 (c)

Well Name

C3

TCE by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 103

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

RW-003

RW-004 (a)

Well Name

RW-006

cis-DCE by Well

RW-007

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.19 cis-DCE concentrations by well during groundwater monitoring. (a) cis-DCE concentrations in recovery wells. (b) cis-DCE concentrations in monitoring wells. (c) cis-DCE concentrations in C-wells.

Conc. (ug/L)

10000

104 Bioremediation of Recalcitrant Compounds

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

cis-DCE for MW’s

(b)

Well Name

MW-505 MW-506 MW-507 MW-508 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514

Figure 5.19 (Continued)

Conc. (ug/L)

20000

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 105

0

10000

20000

30000

40000

50000

60000

Figure 5.19 (Continued)

Conc. (ug/L)

70000

C1

C2 (c)

C3 Well Name

cis-DCE by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

106 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater

107

~1 Month Post-Flush

Pre-Ethanol Flush

C7

C7 C4

C3

MW-514 MW-513 C2

C4 MW-512

C1

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-512 C1

14000 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

12000 ug/L

MW-506

10000 ug/L 8000 ug/L

~2 Months Post-Flush

6000 ug/L

~3.5 Months Post-Flush

4000 ug/L 2000 ug/L C7 C4

C3

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-514 MW-513 MW-512 C2 C1 MW-506

0 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.20 TCE contour plots over the groundwater monitoring period (14,000 μg/l = 107 μM).

~9.5 Months Post-Flush

~5.5 Months Post-Flush

C7

C7 C4

C3

C4

MW-514 MW-513 MW-512 C2 C1

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

14000 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

12000 ug/L

MW-506

10000 ug/L 8000 ug/L 6000 ug/L

~22 Months Post-Flush

~19 Months Post-Flush

4000 ug/L 2000 ug/L C7

C7 C4

C3

MW-514 MW-513 C2

C4

MW-514 MW-513 MW-512 C2 C1

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.21 TCE contour plots over the groundwater monitoring period (14,000 μg/ l = 107 μM).

108

Bioremediation of Recalcitrant Compounds ~25 Months Post-Flush

~28 Months Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

14000 ug/L

MW-505

MW-509 MW-510 MW-511 MW-507

12000 ug/L

MW-506

10000 ug/L 8000 ug/L 6000 ug/L

~31 Months Post-Flush

4000 ug/L 2000 ug/L C7 C4

0 ug/L

C3

MW-512 MW-514 MW-513 C2 C1 MW-509 MW-510 MW-511

MW-505 MW-506 MW-507

Figure 5.22 TCE contour plots over the groundwater monitoring period (14,000 μg/l = 107 μM).

Pre-Ethanol Flush

~1 Month Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 MW-509 MW-510 MW-511

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

MW-505 MW-506

MW-509 MW-510 MW-511

MW-507

16000 ug/L MW-505

14000 ug/L MW-506

12000 ug/L

MW-507

10000 ug/L 8000 ug/L 6000 ug/L

~3.5 Months Post-Flush

~2 Months Post-Flush

4000 ug/L 2000 ug/L C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1 MW-509 MW-510 MW-511

C4

MW-514 MW-513 MW-512 C2 C1

MW-505 MW-506 MW-507

0 ug/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.23 cis-DCE contour plots over the groundwater monitoring period (16,000 μg/l = 165 μM).

Chapter five: Solvent contaminated soils and groundwater ~5.5 Months Post-Flush

109

~9.5 Months Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

MW-505 MW-506

MW-509 MW-510 MW-511 MW-507

MW-509 MW-510 MW-511

16000 ug/L MW-505

14000 ug/L MW-506

12000 ug/L 10000 ug/L

MW-507

8000 ug/L ~13.5 Months Post-Flush

6000 ug/L

~19 Months Post-Flush

4000 ug/L 2000 ug/L

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0 ug/L

C3

MW-509 MW-510 MW-511

MW-505 MW-506 MW-507

Figure 5.24 cis-DCE contour plots over the groundwater monitoring period (16,000 μg/l = 165 μM).

~22 Months Post-Flush

~25 Months Post-Flush

C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-509 MW-510 MW-511

16000 ug/L MW-505

14000 ug/L MW-506

12000 ug/L

MW-507

10000 ug/L 8000 ug/L 6000 ug/L

~31 Months Post-Flush

~28 Months Post-Flush

4000 ug/L 2000 ug/L C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1 MW-509 MW-510 MW-511

C4

MW-514 MW-513 MW-512 C2 C1

MW-505 MW-506 MW-507

0 ug/L

C3

MW-509 MW-510 MW-511

MW-505 MW-506 MW-507

Figure 5.25 cis-DCE contour plots over the groundwater monitoring period (16,000 μg/l = 165 μM).

0

0.002

0.004

0.006

0.008

0.01

0.012

RW-003 (a)

Well Name

RW-007

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.26 Ethylene concentrations by well during groundwater monitoring. (a) Ethylene concentrations in RW-3 and RW-7. (b) Ethylene concentrations in monitoring wells. (c) Ethylene concentrations in C-wells.

Conc. (mg/L)

0.014

Ethylene by Well

110 Bioremediation of Recalcitrant Compounds

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

Ethylene by Well

(b)

MW-505 MW-506 MW-507 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514 Well Name

Figure 5.26 (Continued)

Conc. (mg/L)

0.008

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 111

0

0.002

0.004

0.006

0.008

0.01

Figure 5.26 (Continued)

Conc. (mg/L)

0.012

C1

C2 (c)

C3 Well Name

Ethylene by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

112 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater

113

~2 Months Post-Flush

~1 Month Post-Flush

C7

C7 C4

C3

MW-514 MW-513 C2

C4 MW-512

C1

MW-509 MW-510 MW-511

C3

MW-514 MW-513 C2

MW-505 MW-506

MW-512 C1

MW-509 MW-510 MW-511

MW-507

MW-505

0.0125 mg/L MW-506

0.0100 mg/L MW-507

0.0075 mg/L 0.0050 mg/L

~5.5 Months Post-Flush

~3.5 Months Post-Flush

0.0025 mg/L C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1 MW-509 MW-510 MW-511

C4

MW-514 MW-513 C2

MW-512 MW-506

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-507

0.0000 mg/L

C3

MW-506

Figure 5.27 Ethylene contour plots over the groundwater monitoring period (12.5 μg/l = 446 μM).

~13.5 Months Post-Flush

~9.5 Months Post-Flush

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-505

MW-509 MW-510 MW-511 MW-507

0.0125 mg/L MW-506

0.0100 mg/L 0.0075 mg/L 0.0050 mg/L

~22 Months Post-Flush

~19 Months Post-Flush

0.0025 mg/L C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0.0000 mg/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.28 Ethylene contour plots over the groundwater monitoring period (12.5 μg/l = 446 μM).

114

Bioremediation of Recalcitrant Compounds ~25 Months Post-Flush

~31 Months Post-Flush 0.0125 mg/L C7

C4

C7

C3

C4

MW-512 MW-514 MW-513 C2 C1

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0.0100 mg/L

C3

0.0075 mg/L MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0.0050 mg/L 0.0025 mg/L 0.0000 mg/L

Figure 5.29 Ethylene contour plots over the groundwater monitoring period (12.5 μg/l = 446 μM).

would produce four chloride ions. It is not expected that we see this relationship in an open system, and these were the highest concentrations measured over the entire sampling period. The targeted zone does consist of materials with a relatively low hydraulic conductivity, and the range in hydraulic gradient at the site suggests that overall movement of groundwater through the site could be from 15 to 33 m/year. Partial oxidation of ethanol will result in the formation of 1 mole of acetic acid and 2 moles of hydrogen gas (Equation 5.2), whereas complete oxidation of ethanol will result in the formation of 2 moles of carbon dioxide and 6 moles of hydrogen gas (Equation 5.3). Acetic acid (>200 μM) was observed in all wells except for MW-506, MW-507, MW-511, and C4 (Figure 5.34). Acetic acid was not measured in MW-506 (upgradient), MW-507 (south of injection/extraction zone), and MW-511 (farthest well southwest of monitored area), and these same wells never had detectable ethanol concentrations during the monitoring period. C4 (downgradient and farthest well northwest of monitored area) had measured concentrations of acetic acid from 200 to 600 μM after 2 years, even though ethanol was never detected at this location. Partial oxidation of ethanol to acetic acid and hydrogen gas: CH3CH2OH + H2O

CH3COOH + 2H2

(5.2)

Complete oxidation of ethanol to carbon dioxide and hydrogen gas: CH3CH2OH + 3H2

2CO2 + 6H2

(5.3)

Initially, the contour plots of acetic acid show increasing concentrations in the area of the cosolvent flushing test (Figure 5.35 to Figure 5.37). This trend continues for approximately 19 months, after which the overall measured concentrations decrease, and the area of highest concentration is farther downgradient in the area of MW-510 and MW-513. This coincides with the decrease in ethanol concentrations after 19 months, as would be expected,

0

10

20

30

40

50

RW-003 Well Name (a)

RW-007

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months

Figure 5.30 Chloride concentrations by well during groundwater monitoring. (a) Chloride concentrations of RW-3 and RW-7. (b) Chloride concentrations of the monitoring wells. (c) Chloride concentrations of the C-wells.

Conc. (mg/L)

60

Chloride by Well

Chapter five: Solvent contaminated soils and groundwater 115

0

10

20

30

40

50

60

70

80

MW-505

Figure 5.30 (Continued)

Conc. (mg/L)

90

MW-506

MW-507

MW-509 (b)

MW-510 MW-511 Well Name

Chloride by Well

MW-512

MW-513

MW-514

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

116 Bioremediation of Recalcitrant Compounds

0

10

20

30

40

50

60

Figure 5.30 (Continued)

Conc. (mg/L)

70

C1

C2 (c)

C3 Well Name

Chloride by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 117

118

Bioremediation of Recalcitrant Compounds ~1 Month Post-Flush

Pre-Ethanol Flush

C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80 mg/L

MW-505

70 mg/L

MW-509 MW-510 MW-511 MW-507

MW-506

60 mg/L 50 mg/L 40 mg/L

~2 Months Post-Flush

30 mg/L

~3.5 Months Post-Flush

20 mg/L 10 mg/L C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.31 Chloride contour plots over the groundwater monitoring (80 mg/l = 2.26 mM).

~9.5 Months Post-Flush

~5.5 Months Post-Flush

C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4

MW-514 MW-513 MW-512 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80 mg/L

MW-505

70 mg/L

MW-509 MW-510 MW-511 MW-507

MW-506

60 mg/L 50 mg/L 40 mg/L 30 mg/L

~19 Months Post-Flush

~13.5 Months Post-Flush

20 mg/L 10 mg/L C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1

C4 MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-514 MW-513 MW-512 C2 C1 MW-506

MW-509 MW-510 MW-511

0 mg/L

MW-505 MW-506 MW-507

Figure 5.32 Chloride contour plots over the groundwater monitoring (80 mg/l = 2.26 mM).

Chapter five: Solvent contaminated soils and groundwater ~22 Months Post-Flush

119

~25 Months Post-Flush

C7 C4

C3 MW-512

MW-514 MW-513 C2

C7 C4

C1

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80 mg/L

MW-512 C1

MW-509 MW-510 MW-511

MW-505

70 mg/L MW-506

60 mg/L

MW-507

50 mg/L 40 mg/L 30 mg/L

~31 Months Post-Flush

~28 Months Post-Flush

20 mg/L 10 mg/L C7 C4

C3

MW-514 MW-513 C2

C7 C4

MW-512 C1

MW-509 MW-510 MW-511

MW-514 MW-513 C2

MW-505 MW-506 MW-507

C3

0 mg/L

MW-512 C1

MW-509 MW-510 MW-511

MW-505 MW-506 MW-507

Figure 5.33 Chloride contour plots over the groundwater monitoring (80 mg/l = 2.26 mM).

since ethanol is the parent product (Equation 5.2). After the significant decrease in ethanol, it is expected that the microbial population would shift to using acetic acid as the electron donor so that there would be a decrease in these concentrations as well. The acetic acid contour plots show this trend (Figure 5.35 to Figure 5.37). The highest measured concentration of acetic acid was approximately 2.5 mM, which is 100 times less than the maximum ethanol concentration of 260 mM. This would indicate that a significant amount of the ethanol is being completely degraded to carbon dioxide and hydrogen gas. The hydrogen gas survey conducted approximately 1 year after the cosolvent flushing test indicated that high levels of hydrogen were being produced (Figure 5.38). Concentrations of hydrogen gas in wells near the injection/extraction area were more than 2 μM, and downgradient locations measured from 5 to 30 nM. Hydrogen gas is produced from the oxidation of ethanol and can be utilized by microorganisms responsible for the catabolic and cometabolic reductive dechlorination of PCE and its daughter products. Background concentrations of sulfate at the site were approximately 40 mg/l. All locations showed a decrease in sulfate concentrations over the monitoring period (Figure 5.39). Sulfate was removed from the area immediately downgradient and including the injection/extraction zone within 3 months of the cosolvent flushing test (Figure 5.40 to Figure 5.42). After approximately 14 months, the area of sulfate removal expanded to include most of the downgradient wells. Even MW-506, which was located upgradient of the flushing area, showed a decrease in sulfate concentrations. The contour plots of this data depict the sulfate removal and give a strong

0

500

1000

1500

2000

RW-003 (a)

Well Name

RW-007

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.34 Acetic acid concentrations by well during groundwater monitoring. (a) Acetic acid concentrations in RW-3 and RW-7. (b) Acetic acid concentrations in the monitoring wells. (c) Acetic acid concentrations in the C-wells.

Conc. (mg/L)

2500

Acetic Acid for RW’s

120 Bioremediation of Recalcitrant Compounds

0

200

400

600

800

1000

1200

Acetic Acid by Well

(b)

MW-505 MW-506 MW-507 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514 Well Name

Figure 5.34 (Continued)

Conc. (mg/L)

1400

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 121

0

100

200

300

400

500

600

700

800

900

Figure 5.34 (Continued)

Conc. (mg/L)

1000

C1

C2 (c)

Well Name

C3

Acetic Acid by Well

C4

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

122 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater

123

~2 Months Post-Flush

~1 Months Post-Flush

C7

C7 C4

C3

C4

MW-512 MW-514 MW-513 C2 C1

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

160 mg/L

MW-505

140 mg/L

MW-509 MW-510 MW-511 MW-507

MW-506

120 mg/L 100 mg/L 80 mg/L

~3.5 Months Post-Flush

60 mg/L

~5.5 Months Post-Flush

40 mg/L 20 mg/L C7 C4

C3

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0 mg/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.35 Acetic acid contour plots over the groundwater monitoring period (120 mg/l = 2.0 mM).

~9.5 Months Post-Flush

~13.5 Months Post-Flush

C7 C4

C3

MW-514 MW-513 C2

C7 C4

MW-512 C1

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-512 C1

160 mg/L

MW-505

140 mg/L

MW-509 MW-510 MW-511 MW-507

MW-506

120 mg/L 100 mg/L 80 mg/L

~19 Months Post-Flush

60 mg/L

~22 Months Post-Flush

40 mg/L 20 mg/L C7 C4

C3

MW-514 MW-513 C2

C7 C4

MW-514 MW-513 MW-512 C2 C1

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

0 mg/L

C3 MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.36 Acetic acid contour plots over the groundwater monitoring period (120 mg/l = 2.0 mM).

124

Bioremediation of Recalcitrant Compounds ~25 Months Post-Flush

~28 Months Post-Flush

C7 C4

C7

C3

MW-514 MW-513 C2

C4 MW-512

MW-514 MW-513 C2

MW-505

C1

MW-509 MW-510 MW-511 MW-507

C3

MW-506

160 mg/L

MW-512 MW-505

C1

140 mg/L

MW-509 MW-510 MW-511 MW-507

MW-506

120 mg/L 100 mg/L 80 mg/L 60 mg/L

~31 Months Post-Flush

40 mg/L 20 mg/L 0 mg/L

C7 C4

C3

MW-514 MW-513 C2

MW-512 MW-505

C1

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.37 Acetic acid contour plots over the groundwater monitoring period (120 mg/l = 2.0 mM).

N MW-512

MW-513

5 nM

18 nM

MW-505

156 nM

2 uM

MW-510 MW-511

DNAPL Source Area Target Zone

RW-007

33 nM

MW-507

24.8 uM

4.7 uM 0

20

40

60

80 ft.

Figure 5.38 Hydrogen gas concentrations measured 1 year following the cosolvent flushing test.

indication that sulfate-reducing bacteria are becoming active and creating conditions conducive to the reductive dechlorination process. Along with the removal of sulfate, there was a production of methane, which indicates that methanogenesis has been enhanced. Methane production began after approximately 6 months and was measured in all wells (Figure 5.43). As with the sulfate removal, methane production began in the area near the cosolvent flushing test and concentrations downgradient increased after 14 to 19 months (Figure 5.44 to Figure 5.46).

0

5

10

15

20

25

30

35

40

45

RW-003 Well Name (a)

RW-007

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.39 Sulfate concentrations by well during groundwater monitoring. (a) Sulfate concentrations in RW-3 and RW-7. (b) Sulfate concentrations in the monitoring wells. (c) Sulfate concentrations in the monitoring wells.

Conc. (mg/L)

50

Sulfate by Well

Chapter five: Solvent contaminated soils and groundwater 125

0

10

20

30

40

50

60

70

80

MW-505

Figure 5.39 (Continued)

Conc. (mg/L)

90

MW-506

MW-507

MW-509 (b)

MW-510 MW-511 Well Name

Sulfate by Well

MW-512

MW-513

MW-514

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

126 Bioremediation of Recalcitrant Compounds

0

5

10

15

20

25

30

35

40

45

Figure 5.39 (Continued)

Conc. (mg/L)

50

C1

C2 (c)

C3 Well Name

Sulfate by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 127

128

Bioremediation of Recalcitrant Compounds ~1 Month Post-Flush

Pre-Ethanol Flush

C7

C7 C4

C4

C3

MW-514 MW-513 MW-512 C2 C1

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

80 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

70 mg/L

MW-506

60 mg/L 50 mg/L 40 mg/L

~2 Months Post-Flush

30 mg/L

~3.5 Months Post-Flush

20 mg/L 10 mg/L C7

C7 C4

C3

MW-514 MW-513 C2

C4

MW-514 MW-513 MW-512 C2 C1

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.40 Sulfate contour plots over the groundwater monitoring period (80 mg/l = 833 μM).

~5.5 Months Post-Flush

~9.5 Months Post-Flush

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1 MW-509 MW-510 MW-511

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

MW-505 MW-506

MW-509 MW-510 MW-511 MW-507

MW-507

80 mg/L

MW-505

70 mg/L

MW-506

60 mg/L 50 mg/L 40 mg/L 30 mg/L

~19 Months Post-Flush

~13.5 Months Post-Flush

20 mg/L 10 mg/L C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C4

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.41 Sulfate contour plots over the groundwater monitoring period (80 mg/l = 833 μM).

Chapter five: Solvent contaminated soils and groundwater

129

~25 Months Post-Flush

~22 Months Post-Flush

C7 C4

C3

MW-514 MW-513 C2

C7 C4

MW-512 C1

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-512 C1

80 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

70 mg/L

MW-506

60 mg/L 50 mg/L 40 mg/L

~31 Months Post-Flush

~28 Months Post-Flush

30 mg/L 20 mg/L 10 mg/L

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.42 Sulfate contour plots over the groundwater monitoring period (80 mg/l = 833 μM).

5.5.3 Microbial ecology assessment The only known pure culture that is capable of complete reductive dechlorination of PCE to ethene is the obligately hydrogenotrophic organism Dehalococcoides ethenogenes. Hydrogen is generally considered the ultimate electron donor to stimulate the reductive dechlorination of chloroethenes. Dsulfuromonas sp. strain BB1 is a dechlorinator that utilizes acetate as an electron donor to support the reductive dechlorination of PCE and TCE. Löffler et al. (2000) developed 16S rDNA-based PCR methods for detection of PCE-dechlorinating Dehalococcoides sp. and Dsulfuromonas sp. strain BB1 and evaluated their presence in environmental samples. Aquifer materials collected during installation of MW-510 and IW-001 were utilized in their study along with environmental samples from other sites. The MW-510 sample resulted in visible amplification in the initial PCR with the universal primers. Nested PCR with the Dehalococcoides-targeted primers yielded a positive signal with the MW-510 sample. Microcosm studies conducted with the aquifer materials confirmed the results obtained with the molecular approach. MW-510 microcosms were the only aquifer material–based microcosms that indicated the presence of a hydrogenothrophic PCE-dechlorinating population. Vinyl chloride and ethene accumulated in the hydrogen-fed microcosms, whereas acetate-amended cultures showed only negligible dechlorination (Löffler et al., 2000). The results of the molecular analysis and microcosm studies support the data from the Sage’s site that indicates complete reductive dechlorination of PCE to ethene. Additional samples collected prior to the cosolvent extraction

0

2

4

6

8

10

12

RW-003 Well Name (a)

RW-007

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Figure 5.43 Methane concentrations by well during groundwater monitoring. (a) Methane concentrations of RW-3 and RW-7. (b) Methane concentrations in the monitoring wells. (c) Methane concentrations in the C-wells.

Conc. (mg/L)

14

Methane by Well

130 Bioremediation of Recalcitrant Compounds

0

2

4

6

8

10

12

14

16

18

MW-505

Figure 5.43 (Continued)

Conc. (mg/L)

20

MW-506

MW-507

MW-509 (b)

MW-510 MW-511 Well Name

Methane by Well

MW-512

MW-513

MW-514

1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 11 Months 13.5 Months 19 Months 22 Months 25 Months 28 Months 31 Months

Chapter five: Solvent contaminated soils and groundwater 131

0

1

2

3

4

5

6

7

Figure 5.43 (Continued)

Conc. (mg/L)

8

C1

C2

C3 Well Name (c)

Methane by Well

C4

C7

Pre-Flush 1 Month 2 Months 3.5 Months 5.5 Months 9.5 Months 13.5 Months 19 Months 22 Months 23 Months 25 Months 28 Months 31 Months

132 Bioremediation of Recalcitrant Compounds

Chapter five: Solvent contaminated soils and groundwater

133

~2 Months Post-Flush

~1 Month Post-Flush

C7

C7 C4

C3

MW-514 MW-513 MW-512 C2 C1 MW-509 MW-510 MW-511

C4

C3

MW-514 MW-513 MW-512 C2 C1

MW-505 MW-506

18 mg/L MW-505

MW-509 MW-510 MW-511 MW-507

MW-507

16 mg/L 14 mg/L

MW-506

12 mg/L 10 mg/L

~3.5 Months Post-Flush

8 mg/L 6 mg/L

~5.5 Months Post-Flush

4 mg/L 2 mg/L C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-512 MW-514 MW-513 C2 C1

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.44 Methane contour plots over the groundwater monitoring period (18 mg/l =1.12 mM).

~13.5 Months Post-Flush

~9.5 Months Post-Flush

C7

C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C4

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

18 mg/L

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

16 mg/L 14 mg/L 12 mg/L

MW-506

10 mg/L 8 mg/L ~19 Months Post-Flush

6 mg/L

~22 Months Post-Flush

4 mg/L 2 mg/L C7 C4

C3

MW-512 MW-514 MW-513 C2 C1

C7 C4

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

0 mg/L

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.45 Methane contour plots over the groundwater monitoring period (18 mg/ l =1.12 mM).

134

Bioremediation of Recalcitrant Compounds ~28 Months Post-Flush

~25 Months Post-Flush

C7 C4

C7

C3

MW-514 MW-513 C2

C4 MW-512

C1

MW-514 MW-513 C2

MW-505

MW-509 MW-510 MW-511 MW-507

C3

MW-506

MW-512 C1

MW-509 MW-510 MW-511

18 mg/L 16 mg/L

MW-505 MW-506 MW-507

14 mg/L 12 mg/L 10 mg/L 8 mg/L 6 mg/L

~31 Months Post-Flush

4 mg/L 2 mg/L C7 C4

C3

MW-514 MW-513 C2

0 mg/L

MW-512 C1

MW-505

MW-509 MW-510 MW-511 MW-507

MW-506

Figure 5.46 Methane contour plots over the groundwater monitoring period (18 mg/l = 1.12 mM).

test and 1 and 2 years after the test are being analyzed to determine the changes in microbial ecology that may have occurred. These samples are currently being analyzed and the data are not available at this time.

5.5.4 Ethanol toxicity assessment The effects of continuous and transient exposure of ethanol on subsurface microbial populations were evaluated in laboratory microcosms and in terms of available field data. Ethanol oxidation, as evidenced by concentration decreases, was observed at up to 1% (v/v) concentrations for transient exposure and 0.1% in continuous exposure microcosms (Table 5.5). Analysis of field ethanol concentration data yielded a first-order degradation rate of 0.3/ year. Laboratory rates appeared to be much faster, ranging from 13 to 90/ year (Table 5.6). We feel this is a reflection of the toxic effect of higher ethanol concentrations in the source area and that the observed field rates are impacted by limitations on the size and location of bioactive zones. The observed acetate and methane formation rates are insufficient to account for ethanol losses and indicate the presence of other electron sinks, such as dechlorination and sulfate reduction. Examination of PFLA profiles in laboratory microcosms suggests that higher concentrations of ethanol (at or above 5%) reduced microbial populations and that the decrease in biomass was influenced by exposure time (data not shown).

Chapter five: Solvent contaminated soils and groundwater

135

Table 5.5 Estimates of PCE and Ethanol Mass Remaining in Subsurface by Various Methods PCE Ethanol Ratio Volume Mass Volume Mass Moles Ethanol: (l) (moles) (l) (moles) Moles PCE Extraction wells and partitioning tracers Dissolved phase — after solvent extraction Dissolved phase — after residual biotreatment (2.5 years) Total system PCE, ethanol from extraction well concentration Total system PCE, ethanol from dissolved phase — after solvent extraction a b c

26a

254a

2720b

47,234

186:1

1.1c

11c

295c

5119c

465:1

0.4c

4c

110c

1926c

482:1

265

47,234

178:1

265

5119

19:1

Estimated by partitioning tracer data (Jawitz et al., 2000). Estimated from recovery well concentrations (Jawitz et al., 2000). Calculated from Surfer contour plots.

Table 5.6 Rate Constants for cis-DCE Formation for Individual Wells Well Name C1 C2 C3 C4 MW-505 MW-506 MW-507 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514 RW-3 RW-4 RW-6 RW-7 Average of RW-3, RW-6, and RW-7

k μg/l/day) (μ 14.65 57.29 10.27 4.67 0.24 0.0008 0.003 11.64 1.09 0.14 0.54 6.29 10.86 6.32 2.89 4.27 6.01 5.54

Correlation Coefficient 0.25 0.59 0.68 0.27 0.19 0.20 0.28 0.59 0.49 –1.47 0.25 0.54 0.63 0.79 0.08 0.50 0.66 0.64

136

Bioremediation of Recalcitrant Compounds

5.5.5 Evaluation of system performance 5.5.5.1 Electron donor evaluation The production of acetate and hydrogen gas is an indication that the biodegradation of ethanol is occurring at the site. The hydrogen produced by the oxidation of ethanol can be used for the dechlorination of PCE (Equation 5.1). With incomplete oxidation of ethanol (Equation 5.2), 2 moles of ethanol would be required to produce enough hydrogen for complete dechlorination of PCE. With complete oxidation of ethanol (Equation 5.3), only 1 mole of ethanol would be required. The Surfer contour plots of PCE and ethanol groundwater concentrations were used to estimate the mass of each compound in the dissolved phase. Calculations were made for the groundwater concentration data collected immediately following the solvent extraction test and after 21/2 years of groundwater monitoring (Table 5.5). The estimated residual PCE DNAPL was obtained from the partitioning tracer test data. Ethanol mass was also estimated based on concentrations measured in the recovery wells during the solvent extraction test, which was subtracted from the total mass injected (Table 5.5). These estimates of compound mass were used to calculate the ratio of ethanol to PCE to evaluate the relative amount of electron donor available for the reductive declorination of PCE (Table 5.5). The estimate of ethanol mass from the recovery well data (47,234 moles) is 10-fold higher than that from the measured dissolved phase concentrations (5119 moles). The total mass of PCE in the dissolved phase (11 moles) is also much lower than the estimate of the DNAPL residual from the partitioning tracer test (254 moles), as would be expected. The data indicate that excess electron donors (ethanol) are present to drive the reductive dechlorination of PCE to completion. The ratio of ethanol to PCE is approximately 180:1 when the estimate of ethanol from the recovery wells is used, compared to 19:1 when the estimate of ethanol from the groundwater concentrations is used. If we assume it takes 2 moles of ethanol for complete dechlorination and there are no competing reactions, then there is 9 to 90 times the amount of ethanol to remove the 265 moles of PCE remaining in the subsurface. The dissolved phase concentrations were used to estimate the masses of ethanol and PCE immediately following the cosolvent flush and after 2 1/2 years of groundwater monitoring. Although these estimates do not take into account the residual PCE DNAPL that is present in the system, they do show that excess ethanol remains in the dissolved phase. The ratios from these estimates range from 465:1 immediately following the cosolvent extraction to 482:1 after 2 1/2 years of groundwater monitoring. If we assume there are no competing reactions and there is complete oxidation of ethanol, these ratios indicate that there is more than 200 times the amount of ethanol present in solution available for the complete dechlorination of PCE. Even though the concentrations of both ethanol and PCE have decreased over time, their ratio has remained constant.

Chapter five: Solvent contaminated soils and groundwater

137

Although all of these estimates assume no competing terminal oxidation processes, such as methanogenesis or sulfate reduction, they show that enough ethanol is present in the subsurface that, at even a low efficiency, the theoretical demand could be met. Methane production and sulfate reduction indicate that competing processes are active at the site and will consume some of the ethanol in the system.

5.5.5.2 Estimates of transformation rates As with most multicomponent, ecologically competitive subsurface microbial processes, mathematical descriptions of the fate of PCE and ethanol at the Sage’s site are problematic. Typically zero- and first-order degradation rates are used as a predictive tool for evaluating system performance. We have chosen to derive both zero- and first-order rates, based on data types, to provide values in a usable and generally accepted format for comparisons to other projects and literature values. Conceptually, a first-order rate should be more appropriate for ethanol oxidation, because it is a catabolic process most likely controlled by a single enzyme (per species). Reductive dechlorination, on the other hand, is thought to be limited by available reducing equivalents (hydrogen), which is controlled by a complex set of interactions between competing populations and processes. Thus, the dechlorinating populations seem to have a defined/fixed capacity for transformation, and the behavior of the dechlorination process often seems to fit the zero-order model better (Gibson et al., 1994). However, both approaches are bounded by assumptions such as limited growth (probably a good assumption under anaerobic conditions), single reactive species or fixed ratio of species (a questionable assumption), and conditions that do not vary spatially over the site (a poor assumption).

5.5.5.3 Single-well rate constants The accumulation of cis-DCE in the groundwater from selected wells is plotted in Figure 5.47. The total range of rates of formation of cis-DCE for the site are shown in Table 5.6, which includes the rate constant in μg/l/ day and correlation coefficient for the equation fit through zero. The swept volume for the solvent extraction test was 17.3 kl, and the average rate of cis-DCE production for the RW-wells was 5.54 μg/l/day. If we assume that the total mass of dissolved phase PCE measured during the last sampling event (659 g) was contained in this volume, we calculate that it would take approximately 19 years to consume the PCE in solution. Using the lowest rate of 1.09 μg/l/day and the highest rate of 57.29 μg/l/day, we calculate that it would take 96 to 2 years, respectively, for complete consumption of the dissolved phase PCE. These calculations show the potential for removal of PCE but do not take into account any residual DNAPL PCE contained in the aquifer that would serve as a long-term source of contamination.

138

Bioremediation of Recalcitrant Compounds cis-DCE Formation 70000

Concentration (ug/L)

60000 50000

y = 52.702x 2 R = 0.6131

40000 30000 20000 10000 0 −200

y = 1.0906x R2 = 0.489 0

200

400 600 Time (Days)

C2

MW-510

RW-007

C3

MW-514

Linear (C2)

800

1000

1200

Linear (MW-510)

Figure 5.47 Production of cis-DCE from PCE by microbial reductive dechlorination supported by the residual ethanol from the cosolvent flush at the Sage’s site.

If we include the mass of residual DNAPL PCE (26 l) that was estimated to remain in the swept volume and use the average rate of cis-DCE production of 5.54 μg/l/day, it would take more than 1000 years for complete removal. This calculation assumes that there are no mass transfer limitations. Similar analysis of PCE removal rates for the RW-wells gives an average rate of –37.9 μg/l/day, which is more than six times the rate of cis-DCE formation (Figure 5.48). Table 5.7 is a compilation of the rates of PCE removal and correlation coefficients for the individual wells. Using the average rate of removal for the RW-wells, it would take approximately 3 years for complete removal of the dissolved phase PCE and approximately 200 years for complete removal of the residual DNAPL PCE. Again, it is assumed that there are no mass transfer limitations.

5.5.5.4 Total mass removal The total masses of cis-DCE and PCE were calculated for each sampling event based on the Surfer contour plots. The masses were put on a relative basis by dividing by the initial mass measured before the cosolvent flushing test. Rate constants for each compound were obtained by fitting a linear regression to the data (Figure 5.49). The rate constant for PCE was –0.56/year and for cis-DCE was 0.81/ year for the site. Converting these rate constants to half-lives gives values of 1.2 and 0.9 years, respectively. These values indicate that the activity

Chapter five: Solvent contaminated soils and groundwater

139

PCE Removal 100000

Concentration (ug/L)

90000 80000 70000 60000 50000 40000 30000 20000 10000 0 −200

0

200

400 600 Time (Days)

RW-003

RW-006

800

1000

1200

RW-007

Figure 5.48 Removal of PCE by microbial reductive dechlorination supported by the residual ethanol from the cosolvent flush at the Sage’s site (Y = –37.9X + 44,778).

Table 5.7 Rate Constants for PCE Removal for Individual Wells Well Name C1 C2 C3 C4 MW-505 MW-506 MW-507 MW-509 MW-510 MW-511 MW-512 MW-513 MW-514 RW-3 RW-4 RW-6 RW-7 Average of RW-3, RW-6, and RW-7

k μg/l/day) (μ –110.8 –102.46 –11.48 –18.87 –0.60 –0.0021 –0.062X –11.94 –2.17 –0.70 –6.65 –22.84 –47.98 –34.59 –7.70 –25.77 –50.47 –37.89

R2 0.90 0.92 0.25 0.82 0.03 0.01 0.12 0.19 0.58 0.18 0.41 0.36 0.51 0.37 0.10 0.31 0.44 0.31

140

Bioremediation of Recalcitrant Compounds PCE and cis-DCE Rate Constants Based on Total Mass 2.5 y = 0.8102x − 0.0844 R2 = 0.8243

2 1.5 ln (C/Co)

1 0.5 0 −0.5 −0.5 −1 −1.5

0

0.5

1

1.5

2

2.5

3

y = −0.5608x + 0.0766 R2 = 0.8198

−2 Time (years) ln (PCE C/Co)

Linear (ln (PCE C/Co))

ln (cis-DCE C/Co)

Linear (ln (cis-DCE C/Co))

Figure 5.49 Rate constants for cis-DCE and PCE calculated using total mass estimated from contour plots of data for individual sampling events.

occurring at the Sage’s site in the dissolved phase is quite rapid; the dissolved PCE plume could meet the regulatory limit of 5 μg/l within 10 to 15 years.

5.6 Conclusions on utility in remediation The goal of the chlorinated solvents thrust area was to develop and demonstrate biological remediation technologies for chlorinated solvents. This project focused on one such technological approach for source area treatment, a subarea for which current technologies are limited but which must be addressed if viable, cost-effective approaches are to be implemented at solvent-contaminated sites. The SERB technology has several advantages for remediation of free-phase DNAPLs in the subsurface. These include: • Rapid removal of large masses of DNAPL in a very short time • In situ treatment train technique • Cosolvent injection/extraction overcomes transport and mixing limitations of standard bioremediation infiltration and injection techniques • DNAPL mass removal reduces toxicity to microbes for bioremediation • Continued removal of dissolved contaminants following cessation of pumping • Combination of an active process, solvent extraction, with an essentially passive process, residual biotreatment

Chapter five: Solvent contaminated soils and groundwater

141

The ability to remove large masses of contaminant in a relatively short time is attractive to both regulators and stakeholders, who might be reluctant to support an extremely long natural attenuation option for the duration of natural DNAPL dissolution (assuming that natural attenuation processes are functioning in a sufficiently protective manner). It is also attractive to responsible or potentially responsible parties who might otherwise have no options other than expensive long-term and energy-intensive pumping-and-treating operations for containment. The in situ treatment train approach improves on either natural attenuation or enhanced bioremediation in the DNAPL scenario by both removing large masses of DNAPL and reducing toxicity for long-term biological processes to operate on the dissolved contaminants. There are also disadvantages to the SERB technology, as there are for any remediation scheme: • The technique is innovative and remains to be proven at sites having a variety of characteristics. • Both cosolvent and surfactant in situ floods have the potential and are implemented to mobilize or dissolve the contaminants. If hydrodynamic control is inadequate, the potential exists for contaminating larger volumes of the aquifer and the groundwater. • The cosolvent flood itself will probably not reduce contaminant concentrations to the needed regulatory requirements; hence, a successful residual biotreatment phase is very important. • The choice of cosolvent is critical for both the solvent extraction and residual biotreatment phases of the technique, and success or failure can depend upon this choice. Laboratory studies must be considered before going to field scale. • There is considerable regulatory resistance to the injection of foreign materials (e.g., cosolvents) into the groundwater, which may require a significant effort and large amounts of documentation and persuasion to overcome. • The residual biotreatment phase of SERB might still take a very long time to come to completion following the cosolvent flooding phase, depending upon the conditions created and the mass of contaminant remaining. Recommendations for transitional research include: • Additional pilot-scale and full-scale demonstrations • Further characterization of shifts in microbial ecology in response to SERB • Evaluation of mixtures of optimal cosolvent/electron donor solutions • Modeling of impact of source removal on long-term economics for site-remediation costs

142

Bioremediation of Recalcitrant Compounds

5.7 Technology transfer The environmental challenge posed by separate phase chlorinated solvents in the subsurface is a problem found at a large number of both military and industrial sites. The application of the SERB technology could have a significant effect on cleanup efforts for both DOD and private industry. Subsurface restoration practitioners should be able to make direct use of the results of the laboratory and field investigation. The results are also directly applicable to the needs of the regulatory community and should promote the acceptance of innovative technologies for subsurface remediation. The SERB project was focused on a controlled field test, which demonstrated the capabilities of a treatment train technology that incorporated both active and passive remediation. The reports and presentations from this work will provide technical guidance as to the feasibility for implementation at full scale. The field demonstration was conducted in conjunction with an ongoing collaborative field research project, which involved the State of Florida DEP, the University of Florida, LFR Levine Fricke (site operations contractor), the EPA Technical Innovation Office, and the EPA National Risk Management Research Laboratory (NRMRL). The involvement of this diverse group of collaborators should facilitate technology transfer efforts. The EPA NRMRL’s Groundwater Ecosystems and Restoration Division, located in Ada, OK, houses the Superfund Technology Support Center (TSC), which provides a mechanism for technical assistance and technology transfer to move research results to the private sector. The TSC has provided technical assistance on more than 300 Superfund sites since 1987 and has conducted numerous technology transfer seminars for EPA regional personnel, state personnel, and private contractors who are responsible for subsurface remediation at hazardous waste sites. The TSC provides a very effective means for transferring the results of this research to the user community. An implementation manual for the SERB technology is currently in preparation. This manual will discuss the considerations that should be made before application of the technology as well as the results from the SERB demonstration project. This manual will be produced as an EPA report that will be widely distributed and available to government personnel and private industry.

5.8 Notice The work upon which this report is based was supported by the U.S. Environmental Protection Agency (EPA) through its Office of Research and Development with funding provided by the Strategic Environmental Research and Development Program (SERDP), a collaborative effort involving the U.S. Environmental Protection Agency, the U.S. Department of Energy (DOE), and the U.S. Department of Defence (DOD). It has not been subjected to Agency review and, therefore, does not necessarily reflect the views of the Agency and no official endorsement should be inferred. Mention of trade

Chapter five: Solvent contaminated soils and groundwater

143

names or commercial products does not constitute endorsement or recommendation for use.

References Acree, S.D. 2000. Electromagnetic Borehole Flowmeter Survey Results, Sages Dry Cleaner Site. Summary report submitted to USEPA/ORD/NRMRL/SPRD/SRB. Augustijn, D.C.M., Dai, D., Rao, P.S.C., and Wood, A.L. 1994a. Solvent flushing dynamics in contaminated soils. In Transport and Reactive Processes in Aquifers, T. Dracos and F. Stauffer, Eds. A.A. Balkema, Rotterdam, The Netherlands, pp. 557–562. Augustijn, D.C.M., Jessup, R.E., Rao, P.S.C., and Wood, A.L. 1994b. Remediation of contaminated soils by solvent flushing. J. Environ. Eng. 120: 42–57. Augustijn, D.C.M., Lee, L.S., Jessup, R.E., Rao, P.S.C., Annable, M.D., and Wood, A.L. 1997. Remediation of soils and aquifers contaminated with hydrophobic organic chemicals: theoretical basis for the use of cosolvents. In Subsurface Restoration Handbook, C.H. Ward, J.A. Cherry, and M.R. Scalf, Eds. Ann Arbor Press, Chelsea, MI, pp. 231–250. Brandes, D. and Farley, K. 1993. Importance of phase behavior on the removal of residual DNAPLs from porous media by alcohol flooding. Water Environ. Res. 65: 869–878. Brooks, M.C. 2000. Characterization and Remediation of DNAPL Resulting from a Controlled Release. Ph.D. dissertation, University of Florida, Gainesville. De Bruin, W.P., Kotterman, M.J.J., Posthumus, M.A., Schraa, G., and Zehnder, A.J.B. 1992. Complete biological reductive transformation of tetrachloroethene to ethane. Appl. Environ. Microbiol. 58: 1996–2000. DiStefano, T.D., Gossett, J.M., and Zinder, S.H. 1991. Reductive dechlorination of high concentrations of tetrachloroethene to ethene by an anaerobic enrichment culture in the absence of methanogenesis. Appl. Environ. Microbiol. 57: 2287–2292. ESTCP. 1998. A Treatability Test for Evaluating the Potential Applicability of the Reductive Anaerobic Biological In Situ Treatment Technology (RABITT) to Remediate Chloroethenes. Available at: http://www.estcp.org/documents/ techdocs/Rabitt_Protocol.pdf Falta, R.W., Brame, S.E., and Roeder, E. 1997. The application of density controlled flooding solutions for remediation. Geological Society of America, Abstract Programs, abstract 51134. Falta, R.W., Lee, C.M., Brame, S.E., Roeder, E., Coates, J.T., Wright, C., Wood, A.L., and Enfield, C.G. 1999. Field test of high molecular weight alcohol flushing for subsurface nonaqueous phase liquid remediation. Water Resour. Res. 35: 2095–2108. Fathepure, B.Z. and Boyd, S.A. 1988. Reductive dechlorination of perchloroethylene and the role of methanogens. FEMS Microbiol. Lett. 49: 149–156. Fu, J.K. and Luthy, R.G. 1986. Aromatic compound solubility in solvent/water mixtures. J. Environ. Eng. 112: 328–345. Gibson, S.A., Roberson, D.S., Russell, H.H., and Sewell, G.W. 1994. Effects of three concentrations of mixed fatty acids on dechlorination of tetrachloroethene in aquifer microcosms. Environ. Toxicol. Chem. 13: 453–460.

144

Bioremediation of Recalcitrant Compounds

Gibson, S.A. and Sewell, G.W. 1992. Stimulation of reductive dechlorination of tetrachloroethene in anaerobic aquifer microcosms by addition of short-chain organic acids or alcohols. Appl. Environ. Microbiol. 58: 1392–1393. Imhoff, P.T., Gleyzer, S.N., McBride, J.F., Vancho, L.A., Okuda, I., and Miller, C.T. 1995. Cosolvent-enhanced remediation of residual dense nonaqueous phase liquids: experimental investigation. Environ. Sci. Technol. 29: 1966–1976. Jawitz, J.W., Sillan, R.K., Annable, M.D., Rao, P.S.C., and Warner, K. 2000. In-situ alcohol flushing of a DNAPL source zone at a dry cleaner site. Environ. Sci. Technol. 34: 3722–3729. LFR Levine Fricke. 1997a. Contamination Assessment Report, Former Sages Dry Cleaner. Report submitted to Florida Department of Environmental Protection (December 23, 1997), Tallahassee, FL. LFR Levine Fricke. 1997b. Second Groundwater Sampling Event and Preliminary Evaluation of Natural Attenuation. Report submitted to Department of Environmental Protection (December 24, 1997), Tallahassee, FL. LFR Levine Fricke. 1998. Cosolvent Flushing Pilot Test Report, Former Sages Dry Cleaner. Report submitted to Florida Department of Environmental Protection (December 4, 1998), Tallahassee, FL. LFR Levine Fricke. 2000. Addendum, Contamination Assessment Report, Former Sages Dry Cleaner. Report submitted to Florida Department of Environmental Protection (February 28, 2000), Tallahassee, FL. Löffler, F.E., Sun, Q., Li, J., and Tiedje, J.M. 2000. 16S rRNA gene-based detection of tetrachloroethene-dechlorinating Desulfuromonas and Dehalococcoides species. Appl. Environ. Microbiol. 66: 1369–1374. Lowe, D.F., Oubre, C.L., and Ward, C.H., Eds. 1999. Surfactants and Cosolvents for NAPL Remediation: A Technology Practices Manual. Lewis Publishers, Boca Raton, FL. Lunn, S.R.D. and Kueper, B.H. 1996. Removal of DNAPL pools using upward ethanol floods. In Non-Aqueous Phase Liquids (NAPLs) in Subsurface Environment: Assessment and Remediation, ASCE Proceedings, L.N. Reddi, Ed., Washington, DC, November 12–14, 1996, pp. 345–356. Lunn, S.R.D. and Kueper, B.H. 1997. Removal of pooled dense, nonaqueous phase liquid from saturated porous media using upward gradient alcohol floods. Water Resour. Res. 33: 2207–2219. Lunn, S.R.D. and Kueper, B.H. 1999. Manipulaton of density and viscosity for the optimization of DNAPL recovery by alcohol flooding. J. Contaminant Hydrol. 38: 427–445. Luthy, R.G., Dzombak, D.A., Peters, D.A., Ali, M.A., and Roy, S.B. 1992. Solvent Extraction for Coal Tar, Report EPRI TR-101845. Electr. Power Res. Inst., Palo Alto, CA. MacKay, D.M. and Cherry, J.A. 1989. Groundwater contamination: pump-and-treat remediation. Environ. Sci. Technol. 23: 630–636. Magnuson, J.K., Stern, R.V., Gossett, J.M., Zinder, S.H., and Burris, D.R. 1998. Reductive dechlorination of tetrachloroethene to ethene by a two-component enzyme pathway. Appl. Environ. Microbiol. 64: 1270–1275. Milazzo, J.T. 1993. Remediation of DNAPL-Contaminated Groundwater Using Alcohol Flooding: Separate Phase of Displacement vs. Enhanced Dissolution. M.S. thesis, Clemson University, Clemson, SC. Molz, F.J., Boman, G.K., Young, S.C., and Waldrop, W.R. 1994. Borehole flowmeters: field application and data analysis. J. Hydrol. 163: 347–371.

Chapter five: Solvent contaminated soils and groundwater

145

Morris, K.R., Abramowitz, R., Pinal, R., Davis, P., and Yalkowsky, S.H. 1988. Solubility of aromatic pollutants in mixed solvents. Chemosphere 17: 285–298. Neuman, S.P. 1975. Analysis of pumping test data from anisotropic unconfined aquifers considering delayed yield. Water Resour. Res. 11: 329–342. Peters, C.A. and Luthy, R.G. 1993. Coal tar dissolution in water miscible solvents: experimental evaluation. Environ. Sci. Technol. 27: 2831–2843. Peters, C.A. and Luthy, R.G. 1994. Semiempirical thermodynamic modeling of liquid – liquid phase equilibria: Coal tar dissolution in water-miscible solvents. Environ. Sci. Tech. 28(7): 1331–1340. Rao, P.C.S., Annable, J.D., Sillan, R.K., Dai, D., Hatfield, K., and Graham, W.D. 1997. Field-scale evaluation of in situ cosolvent flushing for enhanced aquifer remediation. Water Resour. Res. 33: 2673–2686. Roeder, E., Brame, S.E., and Falta, R.W. 1996. Swelling of DNAPL by cosolvent to allow its removal as an LNAPL. In Non-Aqueous Phase Liquids (NAPLs) in Subsurface Environment: Assessment and Remediation, ASCE Proceedings, L.N. Reddi, Ed., Washington, DC, November 12–14, 1996, pp. 333–344. Roeder, E. and Falta, R.W. 1998. Phase density difference reversal during horizontal cosolvent flooding of tetrachloroethylene. In: Nonaqueous-Phase Liquids Remediation of Chlorinated and Recalcitrant Compounds. G.B. Wickramanayake and R.E. Hinchee, Eds., Battelle Press, Proceedings of the First International Conference on Remediation of Chlorinated and Recalcitrant Compounds, May 18–21, 1998, 67–72. Roy, S.B., Dzombak, D.A., and Ali, M.A. 1995. Assessment of in situ solvent extraction for remediation of coal tar sites: column studies. Water Environ. Res. 67: 4–14. Sillan, R.K. 1999. Evaluation of In-Situ DNAPL Remediation and Innovative Site Characterization Techniques. Ph.D. dissertation, University of Florida, Gainesville. Sillan, R.K., Annable, M.D., Rao, P.S.C., Dai, D., Hatfield, K., Graham, W.D., Wood, A.L., and Enfield, C.G. 1998. Evaluation of in-situ cosolvent flushing dynamics using a network of spatially distributed multilevel samplers. Water Resour. Res. 34: 2191–2202. Suflita, J.M. and Sewell, G.W. 1991. Anaerobic Biotransformation of Contaminants in the Subsurface, Environmental Research Brief EPA/600/M-90/024. USEPA, Ada, OK. Tandoi, V., DiStefano, T.D., Bowser, P.A., Gossett, J.M., and Zinder, S.H. 1994. Reductive dehalogenation of chlorinated ethenes and halogenated ethanes by a high-rate anaerobic enrichment culture. Environ. Sci. Technol. 28: 973–979. USEPA. 1998. Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in Ground Water (EPA/600/R-98/128), USEPA/ORD/NRMRL, Cincinnati, OH. Wiedemeier, T.H., Swanson, M.A., Moutoux, D.E., Wilson, J.T., Kampbell, D.H., Hansen, J.E., and Hass, P. 1996. Overview of the technical protocol for natural attenuation of chlorinated aliphatic hydrocarbons in groundwater under development for the U.S. Air Force Center for Excellence. In Symposium on Natural Attenuation of Chlorinated Organics in Groundwater, EPA/540/R-96/ 509. USEPA, Washington, DC, pp. 335–359. Yalkowsky, S.H. and Roseman, T. 1981. Solubilization of drugs by cosolvents. In Techniques of Solubilization of Drugs, S.H. Yalkowsky, Ed. Marcel Dekker, New York, pp. 91–134.

146

Bioremediation of Recalcitrant Compounds

Yang, Y. and McCarty, P.L. 2000. Biologically enhanced dissolution of tetrachloroethene DNAPL. Environ. Sci. Technol. 34: 2979–2984. Young, S.C., Julian, H.E., Pearson, H.S., Molz, F.J., and Boman, G.K. 1998. Application of the Electromagnetic Borehole Flowmeter, EPA/600/R-98/058. USEPA, Ada, OK.

chapter six

Enhancing PCB bioremediation James M. Tiedje, Tamara V. Tsoi, Kurt D. Pennell, Lance D. Hansen, Altaf Wani, and Desirée P. Howell

Contents 6.1 Project background and rationale ...........................................................148 6.2 Objectives ....................................................................................................151 6.2.1 Overall objectives ..........................................................................151 6.2.2 Research objectives to design PCB-growing GEMs ................152 6.2.3 Research objectives to enhance PCB remediation ...................153 6.2.4 Field-Test Phase Objectives .........................................................154 6.3 Technical approach ....................................................................................155 6.3.1 Summary.........................................................................................155 6.3.2 FeSO4 amendment.........................................................................157 6.3.3 Sequential inoculations ................................................................157 6.3.4 Surfactant Amendments...............................................................161 6.4 Accomplishments of the flask evaluation..............................................162 6.4.1 Designing and testing PCB-growing GEMs .............................162 6.4.1.1 Characterization of aerobic PCB metabolism by biphenyl-degrading organisms ....................................162 6.4.1.2 Conceptual proof of designing PCB growth pathway..............................................................167 6.4.1.3 Developing gene transfer system for G+/G– PCB-degrading bacteria................................... 170 6.4.1.4 Degradative capabilities of the recombinant RHA1(fcb).........................................................................171 6.4.1.5 Survival and activity of GEM RHA1(fcb) in nonsterile soil ..................................................................173

147

148

Bioremediation of Recalcitrant Compounds

6.4.1.6 Developing and testing molecular tracking recombinant organisms in situ .....................................175 6.4.1.7 Construction of multiple ortho-PCB dechlorinator LB400(ohb)...............................................177 6.4.1.8 Growth on defined PCB mixtures ...............................182 6.4.1.9 Validation of PCB remediation strategy in soil (microcosm studies) ................................................182 6.4.1.10 Developing protocol for inoculum delivery ..............184 6.4.1.11 Recommendations for inoculum delivery during pilot test ..............................................................187 6.4.1.12 Compatibility of anaerobic and aerobic phases in remediation process .......................................................187 6.4.2 Microbial-surfactant compatibility experiments ......................188 6.4.3 Plasmid stability studies ..............................................................189 6.4.4 PCB-surfactant solubilization experiments...............................191 6.4.5 Mathematical modeling ...............................................................192 6.4.6 PCB transformation experiments ...............................................193 6.5 Accomplishments of the pilot evaluation..............................................196 6.5.1 Site consideration for field test ...................................................196 6.5.2 Site description ..............................................................................200 6.5.3 Pilot-scale demonstration.............................................................201 6.5.4 Sampling schedule ........................................................................204 6.5.5 Analytical methods .......................................................................206 6.6 Conclusions.................................................................................................208 6.7 Recommendations for further transitional research ............................208 References.............................................................................................................209

6.1 Project background and rationale PCBs remain among the most expensive hazardous waste cleanup problems facing the country. Through the use of existing technology, principally incineration, the cleanup cost is estimated to exceed $20 billion. If PCB concentrations could be reduced in situ using bioremediation approaches, these costs could be substantially reduced. The research on microbial degradation of PCBs has a 20-year history (Ahmed and Focht, 1973), and many field trials of PCB bioremediation have taken place. This research has shown that bioremediation requires a more sophisticated technology than the simplistic attempts that have been tried thus far. The 20 years of PCB research, however, defined the barriers that must be overcome to achieve successful bioremediation, and the discoveries in basic biochemistry and molecular biology have now provided feasible approaches to overcome these barriers. The fundamental barriers to bioremediation of PCBs are: • The absence from nature of organisms that will grow on PCBs

Chapter six: Enhancing PCB bioremediation

149

• The slow rate of reductive dechlorination and the fact that it is usually incomplete • The low solubility of PCBs, and hence their poor bioavailability • Practical barriers, including a microbial delivery technology that ensures high survivability of introduced microorganisms in soils • Appropriate field-scale remediation technologies. Polychlorinated biphenyls represent a class of chlorinated compounds, the general structure for which is given in Figure 6.1. Each of the numbered positions may or may not be chlorinated, resulting in 209 different congeners. For industrial purposes, PCBs were manufactured as complex mixtures containing from 60 to 90 congeners by the catalytic chlorination of biphenyl (Shulz et al., 1989). Depending on the amount of chlorine added, these mixtures were mobile oils, viscous liquids, or sticky resins, but all were nonflammable, thermally stabile, chemically inert, and excellent electrical insulators. Because of these properties, they were widely used as dielectric fluids in electrical capacitors and transformers and as plasticizers. Smaller but still significant amounts were used as lubricants, hydraulic fluids, heat transfer fluids, cutting oils, extenders in waxes, pesticides, and inks, and in carbonless copy paper (Hutzinger et al., 1974). In the United States and Great Britain, nearly all PCBs were manufactured by Monsanto under the trade name Aroclor and given a four-digit numerical designation. The first two digits in the numerical designation indicate either PCBs (12), polychlorinated terphenyls (PCTs) (54), or mixtures of PCBs and PCTs (25 or 44), and the last two digits indicated the percent chlorine by weight. Thus, Aroclor 1242, for example, is a PCB mixture that is 42% chlorine by weight and averages 3.1 chlorines per molecule. Aroclor 1260 contains 60% chlorine and averages about six chlorines per molecule. Aroclor 1016 is an exception to this scheme; it contains 41% chlorine by weight and appears to be a fractional distillation product from Aroclor 1242, with a marked reduction in the amount of congeners with five or more chlorines. In other countries, PCB mixtures were manufactured under the trade names Fenclor, Pheneclor, Pyralene, Clophen, and Kanechlor, to name a few (Hutzinger et al., 1974). 3

2

6′ 1

4

5

5′

1′

6

4′

2′

3′

Biphenyl

Figure 6.1 The general structure of polychlorinated biphenyl compounds.

150

Bioremediation of Recalcitrant Compounds

As the result of manufacturing processes and spills, several hundred million pounds of PCBs have been released into the environment (Hutzinger and Veerkamp, 1981), and the same properties that made them so industrially useful make them environmentally persistent. Because they are sparingly soluble in water, they have a limited potential for migration through soil, and even the bulk of PCBs deposited in sediments may remain in place for decades. They are also lipid soluble and therefore bioaccumulate, increasing risks associated with exposure through the food chain. A variety of adverse biological effects have been ascribed to them. Perhaps the most notable ecotoxicological effect of PCBs concerns poor reproductive success and deformities in some fish and fish-eating birds (Ludwig et al., 1993). Also, PCBs are suspected carcinogens, and there is epidemiological evidence that they can cause abnormal neurological development in infants and children and alter immunological responses (ATSDR, 2000). Thus, they are recognized as one of the most problematic and persistent environmental contaminants. The remediation of PCB-contaminated soils and sediments typically involves excavation of the contaminated material followed by landfill disposal or incineration. The high costs, long-term liability, and regulatory issues associated with this approach have reduced the attractiveness of excavation as an ultimate remediation option. In addition, excavation and off-site transport of PCB-contaminated wastes may actually increase the potential for human exposure. Recognition of the potential economic and health implications associated with traditional PCB treatment methods has led to a renewed interest in the development of in situ and on-site treatment technologies, including enhanced bioremediation processes. Due to their low solubility in water (or hydrophobicity) and low vapor pressures, PCB congeners are not effectively removed from soil/sediment systems by conventional abiotic remediation technologies such as soil vapor extraction or solvent flushing. Thus, the current state-of-the-art for PCB remediation typically involves the excavation of PCB-contaminated soil/sediment, followed by incineration. Estimated costs for incineration are on the order of $300 to $600/ton of soil, including transportation and excavation costs. As noted above, this remediation method frequently involves increased risk of human exposure, due to the excavation and transport of PCB-contaminated soils. The primary routes of exposure for this scenario are inhalation and dermal contact with soil particles containing sorbed-phase PCBs. A competing in situ technology that is currently under development involves thermal desorption and oxidation of PCB-contaminated surface soils (Iben et al., 1996). The technique involves the use of a thermal blanket containing resistive tubular heaters spaced at 8-cm intervals. The thermal blanket is placed over the soil surface and covered with a layer of insulation (vermiculite or ceramic fiber) and an impermeable sheet of fiberglass-reinforced silicon rubber. Off-gases are extracted through a central tube and passed through a thermal oxidizer operated at about 900˚C. A pilot-scale study has been conducted at an abandoned racetrack where PCB-containing oil was applied to the soil surface to reduce dust. PCB concentrations

Chapter six: Enhancing PCB bioremediation

151

averaged approximately 680 mg/kg from 0 to 7.5 cm (0 to 3 in.) depth and 100 mg/kg for 7.5 to 15.0 cm (3 to 6 in.) depth. Below 15 cm (6 in.), PCB concentrations were below the 2 mg/kg target level. The thermal blanket was heated to about 900˚C for 21 h, with temperatures at the 15 cm depth maintained at 250˚C for about 29 h after the heaters were turned off. In most cases, soil PCB concentrations were reduced to well below the 2 mg/kg target. The costs for thermal blanket remediation were estimated at $150 to $200/ton for a larger site (>6 ha) for a treatment depth of 15 cm (6 in.). These costs may be viewed as rather optimistic because they were based on a scaled-up application (not the actual test case) and the depth of treatment was only 6 in. Limitations and concerns of this technology include: • Small depth of treatment • Potential for downward migration of mobilized PCBs • Potential formation of undesirable low-temperature thermal products near the edge of the treated zone

6.2 Objectives 6.2.1 Overall objectives The primary objectives of the research described herein was to: • Develop genetically engineered organisms that will grow on PCBs • Evaluate surfactants and FeSO4 to enhance PCB dechlorination • Implement and test PCB bioremediation in pilot-scale reactors The goal of the first objective was to construct pathways for PCB degradation that would result in bacteria capable of using PCB congeners as a growth substrate, and to use these organisms to remove products of anaerobic reductive PCB dechlorination (i.e., the less chlorinated mono-, di-, and trichlorobiphenyls, predominantly ortho- and ortho+para-chlorinated congeners). To achieve this goal, two metabolic capabilities were combined in the same organism, (a) cometabolism of PCBs to chlorobenzoates and dechlorination and (b) mineralization of chlorobenzoates as a growth substrate. Our research activities have focused on several biphenyl-degrading, PCB-cometabolic bacterial strains studied in our laboratory and the dechlorination genes found, isolated, and studied as part of the Great Lakes and Mid-Atlantic Hazardous Suibstance Research Center (GLMAC HSRC) project. Once those strains were constructed, the ability of the designed organisms to enhance PCB degradation in soils was evaluated. The practical effectiveness of constructed bacteria was tested at the U.S. Army Engineer Waterways Experiment Station (WES). The second objective of the project was designed to identify and evaluate surfactants capable of enhancing the bioremediation of PCBs in soils and sediments. Research activities specifically focused on the selection of

152

Bioremediation of Recalcitrant Compounds

surfactants that are compatible with the engineered bacteria discussed below, including Rhodococcus erythreus NY05, Rhodococcus RHA1, and Comamonas testosteroni VP44. The experimental approach involved a systematic screening of selected surfactants in microbial batch systems for toxicity or inhibitory effects, prior to the addition of engineered bacteria and surfactant to a contaminated soil or sediment system. The ideal surfactant candidate would not be readily utilized as a growth substrate by the bacteria, possibly serving as a preferential growth substrate over PCBs or reducing the selective pressure for PCB growth genes. In addition to biological compatibility, the selected surfactants were also tested for sorptive losses to several natural soils, capacity to solubilize PCB congeners, coupled solubilization and microbial transformation, and effects on plasmid stability. The third objective focused on pilot-scale implementation of PCB bioremediation using soil collected from Lake Ontario Ordinance Works (LOOW) Picatinny Arsenal and General Electric at a site located in Rome, GA. The practical effectiveness of a two-phase anaerobic–aerobic bioremediation system was evaluated in cooperation with engineers in the Strategic Environmental Research and Development Program (SERDP) bioconsortium at Georgia Tech and the Waterways Experiment Station (WES).

6.2.2 Research objectives to design PCB-growing GEMs • Develop gene cloning and chromosomal integration technique in Rhodococcus strains. • Design genetically enhanced para- and ortho-PCB-growing gram-negative and Rhodococcus strains. • Evaluate the fate of the designed organisms and their effect on PCBs in soils. • Develop methods allowing the tracking of introduced organisms and genes in situ. • Evaluate effect of anaerobic-aerobic shift and FeSO4 and FeS on survivability and PCB degradative activity of the designed organisms in soils. • Develop suitable protocol for soil inoculation with the engineered microorganisms. • Enhance anaerobic reductive PCB dechlorination in PCB-contaminated soil. • Evaluate the recombinant PCB remediation two-phase technology on the pilot scale using PCB-contaminated soils. • Evaluate the feasibility of anaerobic PCB dechlorination in contaminated soils to enrich the congeners that would be accessible for degradation by aerobic genetically enhanced microorganisms. • Establish methods allowing rapid and quantitative detection of genetically engineered PCB-growing bacteria in situ.

Chapter six: Enhancing PCB bioremediation

153

• Characterize the PCB-growing variants of strains Rhodococcus sp. RHA1 and Burkholderia xenovorans LB400 possessing hydrolytic para-dechlorination (fcb) and oxygenolytic ortho-chlorobenzoate (ohb) dechlorination genes, respectively, for their substrate range. • Evaluate the survivability of the recombinant PCB-growing organisms RHA1(fcb) and LB400(ohb) and their impact on PCBs in soil microcosms. • Evaluate the effect of FeSO4 and FeS on the fate and activity of the recombinant strains LB400(ohb) and RHA1(fcb) in soil microcosms. • Evaluate the survivability of introduced biphenyl-degrading bacteria in soils. • Isolate and evaluate the feasibility of genetic enhancement of indigenous biphenyl-degrading organisms from the PCB-contaminated soil to increase chances for successful PCB remediation in this environment, which has a complex contamination profile.

6.2.3 Research objectives to enhance PCB remediation • Investigate the growth of Comamonas testosteroni VP44 and Rhodococcus erythreus NY05 on biphenyl and 4-chlorobiphenyl (4-CBP) in the presence of selected nonionic surfactants (Tween 80, Tergitol NP-15, and Witconol SN-120). • Develop analytical methods (high-performance liquid chromatography (HPLC) and gas chromatography (GC)) to measure the concentration of both surfactant and PCB congeners in the aqueous phase. • Measure the rates of micellar solubilization and equilibrium solubilization capacity of the selected surfactant for specific PCB congeners. • Measure the sorption and desorption of surfactants on natural soils possessing a range of organic carbon (OC) contents (Wurtsmith aquifer material, 0.02% OC; Appling soil, 0.7% OC; and Webster soil, 3.5% OC) and the influence of surfactant sorption on PCB sorption by the solid phase. • Quantify the aerobic degradation of both surfactant and PCB congeners in aqueous systems by recombinant variants of Comamonas testosteroni VP44 and Rhodococcus erythreus NY05. • Conduct microcosm experiments (500-ml reactors) to assess the desorption and aerobic dechlorination of sorbed-phase PCB congeners in the presence of Tergitol NP-15 and Tween 80. • Develop and test a mathematical model to describe the coupled sorption/desorption, micellar solubilization, and transformation of PCB congeners under sequential anaerobic and aerobic conditions. • Investigate the growth of Comamonas testosteroni VP44 and Rhodococcus erythreus NY05 on biphenyl and 4-CBP in the presence of selected nonionic surfactants (Tween 80, Tergitol NP-15, and Witconol SN-120).

154

Bioremediation of Recalcitrant Compounds • Design and test mixing capacity of low-water-content bioreactor (~1/ 5 scale) for the treatment of PCB-contaminated soils. • Perform economic analysis of bioreactor system and compare costs to competing technologies based on site-specific information. • Optimize anaerobic–aerobic reactor treatment scheme, including materials handling; the addition of bulking agents, biocarriers, and nutrients; and disposal procedures. • Conduct surfactant performance tests using PCB-contaminated soils to test the potential for surfactant sorption losses and PCB phase distribution and PCB desorption rates. • Assist in scale-up of the bioreactor and implementation of pilot-scale tests. • Assist in regulatory compliance with regard to PCBs and genetically engineered microorganisms (GEMs) handling and disposal procedures. • Coordinate interaction among PCB thrust area PIs (J. Tiedje at MSU, K. Pennel at Georgia Tech, and L. Hansen at WES) and incorporate laboratory results into field-scale reactor design and operation. • Evaluate mixing performance and materials addition of full-scale reactor during operation.

6.2.4 Field-Test Phase Objectives • Evaluate the effects of vermiculite and Fe(II) on survival and activity of GEMs in soil. • Evaluate survivability and PCB degradative activity of GEMs in contaminated soils. • Develop protocol for preparation and delivery of GEMs in pilot-scale reactors. • Conduct laboratory-scale experiments to evaluate the effects of surfactant additions on PCB desorption and biodegradation under mixing regimes similar to those utilized in the pilot-scale reactors. • Develop and evaluate mathematical models designed to simulate PCB desorption and biodegradation in the presence and absence of surfactants. • Validate the design and strategy for two-phase anaerobic–aerobic bioremediation of PCB-contaminated soils using laboratory-scale soil microcosms. • Evaluate the use of PCB-growing GEMs in combination with enhanced anaerobic dechlorination and surfactants for bioremediating PCB-contaminated soils. • Design pilot-scale treatment systems, including slurry and land-farming reactors, to test PCB remediation technologies. • Develop cost estimates for pilot-scale treatment systems and perform economic analysis of full-scale implementation with comparisons to conventional PCB treatment technologies.

Chapter six: Enhancing PCB bioremediation

155

• Evaluate the effects of different solids’ loading rates (i.e., water contents) on the application of amendments and GEMs and the bioremediation of PCBs. • Determine the maximum solids’ loading rate for optimum activity of GEMs in order to offset subsequent dewatering costs for the disposal/reuse of stabilized soils. • Identify PCB-contaminated sites available for the field test and evaluate feasibility of their PCB treatment. • Conduct ex situ two-phase anaerobic–aerobic PCB remediation pilot-scale tests and evaluate efficiency of the PCB removal and performance of GEMs.

6.3 Technical approach 6.3.1 Summary This project addresses key barriers to bioremediating PCBs, which are to: • Develop microorganisms that will grow on the major congeners produced by anaerobic dechlorination of PCBs • Improve bioavailability of PCBs through use of surfactants • Optimize field delivery of anaerobic–aerobic PCB bioremediation technology • Validate the new two-phase anaerobic–aerobic bioremediation strategy in a pilot-scale test Three central components of the project were: the use of a combination of genetically engineered organisms that will grow on PCBs, the use of surfactants to enhance the bioavailability, and bioslurry experiments as a first stage in a flask-to-field transfer technology (Figure 6.2). The lack of organisms that grow on PCBs results from the fact that organisms with PCB (biphenyl moiety)-cometabolizing activity do not seem to have the ability to use the chlorobenzoate product for growth. Moreover, those few organisms that have this ability would attempt to metabolize it via the chlorocatechol pathway, forming an acyl halide, which is immediately toxic. Hence, if any organism in nature does have the capacity to grow on PCBs, it would be suicidal. Through the use of dechlorination genes, we have devised a scheme to remove chlorines before chlorocatechols are formed, thereby providing an energy (growth) product and avoiding toxicity. This approach should be a more desired solution for PCB remediation because it would avoid the need to manage cometabolism, which can be difficult to achieve in situ. One of the biological barriers to effective PCB remediation is slow rates of intrinsic anaerobic dechlorination. The first step in optimizing a sequential anaerobic-aerobic biotreatment process for PCBs is to maximize the extent of dechlorination. Several dechlorination processes, generally believed to be

156

Bioremediation of Recalcitrant Compounds General Scheme Anaerobic Reductive Dechlorination FeSO4 - Rates/Conc.

Seed Inoculum - Hudson river

Tween 80 - Sorption/Availability Aerobic PCB Dechlorination Genes ohb, fcb G+/− Cassettes

Organisms - Burkholderia LB400 (G−) - Rhodococcus RHA1 (G+) Inocula Technology - Vermiculite Low, Medium and High Solids Reactors

Figure 6.2 Sequential anaerobic–aerobic PCB remediation strategy.

due to the actions of different species of microorganisms, have been recognized on the basis of their congener specificities (Table 6.1). In Table 6.1, dechlorination process M, for example, removes both flanked and unflanked meta-chlorines. A flanked meta-chlorine is one that is adjacent to another (ortho- or para-) chlorine. An unflanked chlorine has no other chlorine next to it. Not fully captured in Table 6.1 is the fact that these processes also vary in their abilities to attack more heavily chlorinated Table 6.1 Regiospecificities of the Various Described PCB Dechlorination Processes Dechlorination Activity M Q H H’ P N T LP

Susceptible Chlorines Flanked and unflanked meta Flanked and unflanked para Flanked para Doubly flanked meta Flanked para Meta of 2,3- and 2,3,4- groups Flanked para Flanked meta Meta of hepta- and octa-CBs Unflanked para

Chapter six: Enhancing PCB bioremediation

157

congeners. Processes M and Q, for example, are less effective in dechlorinating Aroclor 1260 than is process N. Thus, the extent of PCB dechlorination that occurs is dependent on which microorganisms are present and active, and more extensive dechlorination can be achieved by combining dechlorination processes with complementary activities. This has occurred naturally in the Upper Hudson River, where PCB dechlorination was first recognized. In several locations the combined activities of at least the first four processes listed in Table 6.1 have removed most of the meta- and para-chlorines from Aroclor 1242, with the result that about 90% of the remaining PCBs are monoand dichlorinated congeners substituted only in the ortho-positions. These congeners constitute less than 8% of Aroclor 1242. In laboratory experiments, PCB dechlorination by complementary processes was achieved using two different approaches. First, the addition of FeSO4 fosters simultaneous dechlorination by processes M and Q (Zwiernik et al., 1998), and second, enhanced dechlorination of Aroclor 1260 was realized using sequential inoculations of process N and process Q microorganisms (Quensen et al., 1990b).

6.3.2 FeSO4 amendment Prior experience indicates that meta-dechlorination is more readily achieved than para-dechlorination and that process Q dechlorination is not reliably achievable. While investigating the use of ferrous sulfate as a way of precipitating heavy metals, which can inhibit dechlorination, a means to rescuing process Q activity was discovered. Aroclor 1242 dechlorination was more extensive in microcosms to which 10 mM FeSO4 was added, but dechlorination occurred only subsequent to FeSO4 depletion. Our thinking is that at least some of the dechlorinating microorganisms are sulfate reducers (i.e., they dechlorinate PCBs in the absence of sulfate) and that the addition of a small amount of sulfate allowed them to increase in numbers. We also believe that the iron precipitated the sulfide formed and that the para-dechlorinating microorganisms are sensitive to this sulfide. Evidence for this last point comes from the fact that precipitation of sulfides with FeCl2 or PbCl2 also enhanced dechlorination but to a lesser extent (Zwiernik et al., 1998). Adding FeSO4 to the PCB-dechlorinating microcosms altered the dechlorination pattern achieved. Without the addition of FeSO4, pattern M resulting from meta-dechlorination alone was achieved. With ferrous sulfate, pattern C resulting from the combined meta- and para-dechlorinating activities of processes M and Q was achieved. Thus, the addition of FeSO4 is one way of achieving enhanced PCB dechlorination through the combined activities of complementary dechlorination processes.

6.3.3 Sequential inoculations A second way of obtaining a combination of complementary dechlorination activities is to use inocula from different sources. The results from such an

158

Bioremediation of Recalcitrant Compounds

Average meta & para Chlorines

5

4

3 Control Silver Lake Inoculum Hudson River Inoculum Hudson River & Silver Lake Inocula Silver Lake, then Hudson River Inocula

2

1 0

10

20 30 40 Incubation Time (Weeks)

50

60

Figure 6.3 Extent of Aroclor 1260 dechlorination obtained using mixed and sequential inoculations with Hudson River and Silver Lake microorganisms.

experiment are depicted in Figure 6.3. The progress of Aroclor 1260 dechlorination in various treatments is compared by plotting the average number of chlorines remaining in the meta- or para-positions vs. time. The most extensive dechlorination occurred with sequential inoculations with Silver Lake sediments, followed by Hudson River microorganisms. Process N microorganisms in Silver Lake sediments are much more effective in dechlorinating the heavily chlorinated congeners in Aroclor 1260 than are the microorganisms in Hudson River sediments, which are mainly process M and Q microorganisms. The extent of dechlorination achievable by Silver Lake microorganisms alone is limited because process N removes only flanked meta-chlorines. Hudson River microorganisms (M and Q together), however, have the potential to remove all meta- and para-chlorines, whether or not there is an adjacent chlorine. Combined activities could not be achieved by simply mixing microorganisms from the two sediments apparently because of incompatibility, but we were able to achieve enhanced activity through sequential inoculations, first with Hudson River microorganisms and then with Silver Lake microorganisms. A possible explanation can be gleaned from examining the chromatographic profiles of the PCBs generated in the various treatments (Figure 6.4). The profiles for Hudson River microorganisms alone and mixed with Silver Lake microorganisms are similar, and indicate that dechlorination was limited to the meta-positions. In other words, process Q or para-dechlorination activity was lost, and even N activity was diminished. The profile for Silver Lake microorganisms alone indicates that high levels of ortho- and para-substituted congeners, notably 24-26-CB (peak 21), 24-24-CB (peak 26), and 246-24-CB (peak 34), were formed. These congeners could serve as a substrate for process Q microorganisms in the subsequent inoculation with Hudson River microorganisms. It is possible that the accumulation of these congeners even selected for, or primed, this para-dechlorination activity. In

Mole Percent

60

70

80

90

20

30 40 50 60 Peak Number

70

80

90

0

10

0 0

5

5

15

20

25

10

Hudson River Inoculum Hudson River & Silver Lake Inocula

10

15

20

25

50

0

40

0 30

5

5 20

10

10

10

15

15

0

20

25

20

Control

0

0

10

10

20

20

50

60

70

80

70

80

Silver Lake Inoculum followed by Hudson River Inoculum

40

30 40 50 60 Peak Number

30

Silver Lake Inoculum

90

90

Figure 6.4 Aroclor 1260 dechlorination patterns obtained using mixed and sequential inoculations with Hudson River and Silver Lake microorganisms.

Mole Percent

25

Chapter six: Enhancing PCB bioremediation 159

160

Bioremediation of Recalcitrant Compounds

any event, the result was that ortho-only-substituted congeners increased from less than 1% in Aroclor 1260 to 39%, with sequential inoculations at the end of the experiment. Another biological barrier to PCB bioremediation is that there are no bacteria that have been found in nature that can grow on important PCB congeners as a growth substrate. Anaerobic cometabolic reductive dechlorination of highly chlorinated PCBs produces less chlorinated congeners (mono-, di-, and trichlorinated biphenyls), preferentially ortho-chlorobiphenyls and ortho- + para-derivatives. Aerobic biphenyl-degrading bacteria can nonspecifically co-oxidize less chlorinated PCBs, but this process requires biphenyl or related other sources as a growth substrate. The major product of this dead-end cometabolism is the respective chlorobenzoate. If the barrier of growth on PCBs can be overcome, then a natural enrichment by growth on PCBs should occur, increasing the rate of PCB removal. This approach should be a more desired solution for PCB remediation, because it would avoid the need to manage cometabolism, which can be difficult and costly. Previously, we obtained a collection of bacterial isolates that aerobically degrade biphenyl and cometabolize certain PCB congeners. We have cloned and characterized the genes specifying anoxic hydrolytic para-dechlorination (fcb operon from Arthrobacter globiformis KZT1) and oxygenolytic ortho-dechlorination (ohb operon from Pseudomonas aeruginosa 142). These genes encode enzymes that remove chlorine from chlorobenzoates, funneling nonchlorinated products into common degradative pathways for nonchlorinated aromatics. Combining these dehalogenase genes and biphenyl oxidation pathways should result in engineered pathways for ortho-, para-, and ortho- + para-chlorinated PCB congeners that are of a major concern in the proposed anaerobic–aerobic bioremediation scheme (Figure 6.5). In contrast to approaches employed by other groups, using the specific chlorobenzoate dehalogenases not only allows the choice of a desirable host (for example, PCB-tolerant bacteria, gram positive or gram negative) but also prevents accumulation of toxic-aromatic-ring meta-cleavage products that would be produced from using broad specificity benzoate oxygenases that produce chlorocatechol from chlorobenzoate. Another potential problem we envision for in situ bioremediation is that combining the anaerobic phase for reductive dechlorination of highly chlorinated PCBs and the aerobic phase for oxidation of less chlorinated PCB congeners in the same remediation scheme might be too complicated to manage if the common aerobe, such as gram-negative bacteria, were used for the aerobic phase. To overcome this barrier, we propose to use not only gram-negative but also gram-positive bacteria for construction of PCB degradation pathways. The well-known ability of gram-positive microaerophilic bacteria, such as bacilli, corynebacteria, and rhodococcaceae, to persist in harsh environments and survive anaerobic conditions should reduce remediation costs and increase chances of success.

Chapter six: Enhancing PCB bioremediation

161

H

Figure 6.5 Engineering PCB pathways.

6.3.4 Surfactant Amendments Although microbial transformation of PCBs has been the subject of intense study over the past 25 years, it is now apparent that the use of simplistic remediation approaches, based on conventional bioreactor and land-farming strategies, will not be successful. One of the primary barriers to effective PCB bioremediation is the limited availability of PCBs to microbial populations. PCBs are extremely hydrophobic compounds, which results in their low-equilibrium solubilities and slow rates of desorption from solid phases. These physical and chemical barriers may contribute to incomplete bioremediation of PCB-contaminated sites and the inability to reach target PCB concentrations. To overcome such limitations, we have proposed the use of surfactants to increase the equilibrium solubility and mass transfer rate of PCBs into the aqueous phase. Three commercially available surfactants were selected for study: Tween 80, Witconol SN-120, and Tergitol NP-15. These surfactants cost approximately $1.00/lb and, thus, surfactant costs for a 5000 mg/l solution would be approximately $3.78/ton of soil, assuming a water-filled porosity of 0.3.

162

Bioremediation of Recalcitrant Compounds

6.4 Accomplishments of the flask evaluation 6.4.1 Designing and testing PCB-growing GEMs 6.4.1.1 Characterization of aerobic PCB metabolism by biphenyl-degrading organisms Analysis of the dechlorination patterns in anaerobic sediments resulted in identification of eight ortho- and ortho- + para-chlorinated PCB congeners, which account for up to 80% of the total Aroclor 1242 products in anaerobically dechlorinated sediments and are primary targets for the aerobic phase of the PCB bioremediation scheme. Correspondingly, we have concentrated on studying aerobic metabolism of these eight PCB congeners, both individual and given as defined mixtures M and C (Figure 6.6). We characterized PCB cometabolism by biphenyl-degrading gram-positive Rhodococcus erythreus NY05 and Rhodococcus sp. RHA1, and gram-negative Comamonas testosteroni VP44 and Burkholderia xenovorans LB400. Metabolism of PCBs by these strains has been studied in detail for the range of substrates, rates of PCB oxidation, intermediates, dead-end products, and specificity of biphenyl ring oxidation (summarized in Maltseva et al., 1999). The 2-, 4-, and 2,4-CBs were easily degraded via preferential oxidation of the nonchlorinated ring of these CBs. Incomplete transformation of 2-CB by 50 40 30 20 10 0

A: Aroclor 1242

Mole Percent

0

10

20

50 40 30 20 10 0

30

40

50

60

B: Pattern M

· 0

10

50 40 30 20 10 0

20

30

40

50

60

C: Pattern C

0

10

20 30 40 Peak Number

50

60

Figure 6.6 Pattern M and C profiles compared to Aroclor 1242. See Quensen et al., 1990b for cogeners represented by each peak number.

Chapter six: Enhancing PCB bioremediation

163

RHA1 was indicated by nonstoichiometric yield of 2-chlorobenzoic acid (2-CBA) and the appearance of yellow color with a maximum absorbance at a wavelength of 394 nm, 2-hydroxy-6-oxo-6-phenylhexa-2,4-dienoic acid (HOPDA). Some fading of the yellow color and a shift toward shorter wavelengths were detected during the next two days of incubation, indicating further transformation of the meta-cleavage product; however, no increase of 2-CBA concentration was noted, indicating no complete transformation. Accumulation of the 2-CB meta-cleavage product was also detected during incubation of 2-CB with NY05, but in contrast to strain RHA1, this intermediate completely disappeared after 24 h of incubation followed by equimolar production of 2-CBA. In a difference from the other three strains, the dynamics of accumulation of 2-CBA from 2-CB by LB400 suggested that it could further degrade this acid (40 to 60%) with no metabolites of 2-CBA degradation found (HPLC). 2,6-CB proved to be the congener most resistant to microbial attack, with less than 5% depletion by NY05 and VP44, and only 5 to 10% degradation by strains LB400 and RHA1, yielding up to 5% as 2,6-CBA. Accumulation of HOPDA during incubation of RHA1 with 4-CB, 2,4-CB, and 2,6-CB indicated incomplete transformation of some of the (Cl)HOPDAs formed from these congeners into the corresponding chlorobenzoic acids and pentadienes. Congeners with one of the rings containing chlorine in the para-position (2,4′ and 2,4,4′-CB) were efficiently degraded by LB400 and RHA1 via preferential oxidation of their ortho-substituted rings, in agreement with the previously reported transformation of 2,4,4′-CB by LB400 (Seeger et al., 1995). Interestingly, following a decrease in absorption of HOPDA (434 nm), during a longer (24 h) incubation of LB400 with 2,4,4′-CB, a minor peak of HOPDA with a maximum of 398 nm was observed, suggesting that the 4′-chlorinated ring of 2,4,4′-CB could also be oxidized. This was confirmed by GC/MS (mass spectrometry) verification of minor 2,4-CBA production among the degradation products of 2,4,4′-CB (Maltseva et al., 1999). Complete depletion of 2-4′- and 2,4,4′-CB was also observed with NY05 and VP44; however, amounts of chlorobenzoates formed did not exceed 10 to 12% of the expected value, in agreement with formation of high amounts of (Cl)HOPDA. The metabolite of 2,4-CB was characterized by molecular ion at m/z 430 (mass to charge ratio), low abundance of M ± 15 and M ± 35 ions, and a prominent ion M ± 117 arising from the loss of CH 3, Cl, and COO-TMS (trimethylsilane) from the molecular ion, respectively. The same fragmentation pattern was obtained for the metabolite of 2,4-4′-CB (molecular ion 464). The mass spectral features of these compounds are in good agreement with previously published mass spectra of the TMS derivatives of chlorobiphenyl meta-cleavage products (HOPDAs) produced by other bacterial strains (Furukawa et al., 1979a, 1979b; Masse et al., 1989). Absorption maxima of these HOPDAs occurred at 398 nm indicated that they have the chlorine substituent at the ortho-position (Seeger et al., 1995) and therefore were formed by oxidation of the para-chlorinated ring. Although both strains produced some 2,4-CBA from 2,4-4′-CB, only NY05 produced traces of 2-CBA from 2,4′-CB as transformation product. No significant decrease of these HOPDAs, nor a

164

Bioremediation of Recalcitrant Compounds

corresponding increase of the chlorobenzoic acids, was detected during another 72 h of incubation, suggesting that meta-cleavage products formed by NY05 and VP44 were stable. Recovery of about 5% of 2,4′-CB as 4-CBA by both strains showed that they could also oxidize the ortho-chlorinated ring. Advantageously to NY05 and VP44, strains LB400 and RHA1 efficiently degraded 80 to 100% of 2,2′- and 2,4,2′-CB (chlorine in ortho-position on each ring of the biphenyl moiety). Both strains transformed 2,4,2′-CB into equimolar amounts of 2,4-CBA, and there was a 60 to 80% transformation rate of 2,2′-CB into 2-CBA. In the latter case, RHA1 accumulated HOPDA (maximum of 394 nm), accounting for the difference in the depleted CB and accumulated 2-CBA, whereas LB400 partially depleted 2-CBA, similar to 2-CB oxidation. NY05 and VP44, again, were much less active, depleting only 5 to 30% of 2,2′- and 2,4,2′-CB, with equimolar yields of 2-CBA and 2,4-CBA. It should be noted that LB400 advantageously exhibited only transient appearance of (Cl)HOPDAs, whereas strain RHA1 accumulated HOPDA from 2-CB and 2,2′-CB. This difference between two strains with similar PCB degradative ability might be due to the different substrate specificity of their respective HOPDA hydrolases, which have only about 30% sequence identity (Hofer et al., 1993; Masai et al., 1997). Poor turnover of HOPDAs can exclude some CBs from further productive metabolism. Two principal and complementary modes of PCB metabolism (preferential oxygenation of ortho- or para-chlorinated ring of biphenyl moiety) were demonstrated by Rhodococcus strains NY05 and RHA1 and have been found most attractive for use in bioremediation. Consumption rates and products of metabolism of the eight PCB congeners by strains NY05, VP44, and LB400 were determined. These findings were summarized in Pellizari et al. (1996), Hrywna et al. (1999), and Maltseva et al. (1999). We have studied metabolism of defined PCB mixtures simulating most extensive (pattern C) and average (pattern M) anaerobic dechlorination products accumulating in PCB-contaminated soils and sediments. Strains NY05, RHA1, and LB400 were used in resting cell assays and showed from 40 to 80% depletion of mix M after 24 h of incubation, with 65 to 95% of the expected recovery as chlorobenzoates, mostly 2-CBA and 4-CBA (Figure 6.7). Similarly, the bacteria depleted 60 to 75% of mix C after 24 h of incubation, with 70 to 90% recovery as chlorobenzoates, predominantly 2-CBA (Figure 6.8). 2,2′-CB, 2,6-CB, 2,4,2′-CB, and 2,4,4′-CB were the most resistant to microbial attack when supplied in mixtures M and C. Accumulation of yellow color was detected during degradation of mixtures M and C by Rhodococcus strains NY05 and RHA1, in accordance with HOPDA production found in transformation of individual congeners. Not surprisingly, ortho-directed RHA1 and LB400 exhibited higher activity toward mixtures M (65 and 80%, respectively) and C (70 and 75%, respectively). Degradation rates of PCB congeners containing chlorine on both biphenyl rings decreased when they were supplied in mixtures, possibly due to competition from easily degradable congeners with one nonchlorinated

Chapter six: Enhancing PCB bioremediation

165

A Mix M

CB, uM

200 150 100 50 0

1

2

3

7

B 200 CBA, uM

CB, uM

6

150 100 50 0

2

3

4

5

6

50

CBA, uM

50 2

3

4

5

6

100 50

50

2

3

4

5

6

0

6

12 18 Time, h

24

LB400

150 100 50 0

1

24

100

200 CBA, uM

150

12 18 Time, h RAH1

7

Mix M + LB400

200

6

150

0 1

0

200

Mix M + RHA1

100

0

100

7

150

0

NY05

150

0 1

200 CB, uM

5

Mix M + NY05

200

CB, uM

4

0

6

7

12 18 Time, h

24

Figure 6.7 Degradation of mix M by three BP degraders.

ring. From 70 to 95% of depleted chlorobiphenyls were recovered as chlorobenzoates. In the only comparable investigation, albeit using much lower concentration of PCBs, Alcaligenes eutrophus strain H850, also an ortho-directed strain (Bedard et al., 1987a), was reported to deplete 81% of Aroclor 1242 anaerobic dechlorination products in Hudson River sediments (10 ppm, pattern C). The production of chlorobenzoates was not monitored in that work (Bedard et al., 1987b). These experiments showed that degradation of several congeners is affected when PCBs supplied in complex mixtures, primarily less efficient degradation of 2,4,4-CB, 2,2-CB, and 2,6-CB. Strains LB400 and RHA1 with preference toward hydroxylation of ortho-chlorinated biphenyl ring appeared more efficient PCB degraders compared to NY05 with preference toward para-chlorinated ring. We have identified several intermediate products of aerobic PCB oxidation with potential biotoxic effects, among them meta-cleavage products; chlorinated HOPDAs and dihydrodiols and monoand dihydroxybiphenyls. The amount of (Cl)HOPDA produced by

166

Bioremediation of Recalcitrant Compounds A Mix C

CB, uM

300 200 100 0

1

2

3

4

100

2

3

4

5

6

Mix C + RHA1

100

2

3

4

5

6

100

2

3

4

5

6

0

6

12 18 Time, h

24

LB400

200 100 0

1

24

RHA1

300 CBA, uM

200

12 18 Time, h

100

7

Mix C + LB400

300

6

200

0 1

0

300 CBA, uM

100

0

200

7

200

NY05

300

0 1

300

0

B

CBA, uM

CB, uM

7

200

0

CB, uM

6

Mix C + NY05

300

CB, uM

5

7

0

6

12 18 Time, h

24

Figure 6.8 Degradation of mix C by BP degraders.

para-directed strains NY05 and VP44 was especially significant from 2,4-CB and 2,4,4-CB. Biochemical assay suggested that these chlorinated intermediates reversibly inhibit hydrolase activity, preventing complete transformation of the HOPDAs to the respective chlorobenzoates (Maltseva et al., 1999). Aerobic degradation of PCBs by biphenyl-growing bacteria is usually cometabolic and results in partial degradation, particularly the accumulation of chlorobenzoates (Abramowicz, 1990; Unterman, 1996). Introduction of genes for dehalogenases that control removal of chlorine from ortho- and para-chlorinated benzoates into PCB-cometabolizing strains should result in growth of the recombinant bacteria on the targeted ortho- and ortho- + para-substituted congeners. The ideal host for constructing genetically modified microorganisms that would grow on anaerobic Aroclor dechlorination products should:

Chapter six: Enhancing PCB bioremediation

167

• Completely deplete targeted congeners • Accumulate no intermediates of CB degradation • Not funnel produced chlorobenzoates into unproductive or toxic pathways Based on the results described above, we proposed that although none of the investigated bacteria completely met all these criteria, the ortho-directed strains RHA1 and LB400 catalyzed more efficient oxidation of a wider range of CBs, with higher yield of chlorobenzoates, than did the para-directed strains VP44 and NY05. Thus, the strains preferentially oxidizing ortho-chlorinated CB rings are still more suitable for genetically engineering bacteria capable of growth on Aroclor anaerobic dechlorination products.

6.4.1.2 Conceptual proof of designing PCB growth pathway In contrast to in vivo construction, the use of sequenced and well-characterized genes permits targeting of specific compounds for degradation and better prediction of expected intermediates and products with minimal effects on existing cellular metabolism. In a two-phase anaerobic–aerobic PCB bioremediation scheme, a limited number of ortho-, 2,6-, 2,4′-, and 2,4-chlorobiphenyls are major targets for the aerobic degradation of PCBs (Quensen et al., 1990a, 1990b). These congeners, upon oxidation by BP pathway enzymes, produce the respective ortho-, para-, and ortho- + para-chlorobenzoates. Therefore, introduction of specific ortho- and para-dechlorination genes into biphenyl-degrading bacteria should result in growth on, and mineralization of, the targeted PCB congeners by the recombinant organism. We have published (Hrywna et al., 1999) the first results on construction of the recombinant variants using the PCB-cometabolizing Comamonas testosteroni strain VP44 as the host (Figure 6.5). The dehalogenation genes used were the fcb operon for hydrolytic dechlorination of para-chlorobenzoate and ortho- + para-chlorinated biphenyls, such as 2-, 2,2′-, 2,4,2′-, 2,4,4′-, which we previously isolated from Arthrobacter globiformis strain KZT1 (Plotnikova et al., 1991; Tsoi et al., 1991), and the ohb operon for oxygenolytic dechlorination of ortho-halobenzoate, which we isolated from Pseudomonas aeruginosa strain 142 under the related HSRC project (Tsoi et al., 1999). The ohb DNA region contains structural genes ohbAB encoding small and large subunits, respectively, of the terminal oxygenase of the three-component ortho-halobenzoate 1,2-dioxygenase, potential regulatory gene ohbR, and ohbC, the gene of unknown function that overlaps the ohbB gene in an opposite transcriptional direction. Prior results showed that expression of the ohb genes in Escherichia coli and Pseudomonas putida cells results in dechlorination and transformation of 2-CBA to catechol and, in the case of P. putida, a recombinant pathway for growth on the chlorobenzoate. Plasmid pE43 (Tsoi et al., 1999), composed of the ohb operon cloned in a broad-host-range vector pSP329, was introduced in strain VP44, and resulting transformants

168

Bioremediation of Recalcitrant Compounds

grown on Luria-broth (LB) medium containing tetracycline were then transferred onto solidified mineral medium K1 plates with 1.25 and 2.5 mM 2-CBA as a sole source of carbon. After an extended 8-week incubation, colonies for containing plasmid pE43 were fully grown and confirmed for 2-CBA-specific growth by transfer in liquid K1 + 2-CBA media. 4-CBA-growing recombinant strain VP44(pPC3) was constructed by introduction of the recombinant plasmid pPC3 carrying a 4.36-Kb fcb DNA fragment from A. globiformis KZT1. The fcb DNA fragment contains structural genes fcbABC encoding a denosine triphosphate (ATP)–dependent coenzyme A (CoA) ligase, hydratase/dehalogenase, and thioesterase, respectively. Prior results showed that expression of the fcb operon in E. coli and P. putida results in dechlorination and conversion of 4-CBA to 4-hydroxybenzoate (4-HBA) and, in the case of P. putida, a recombinant pathway for mineralization of 4-CBA (Plotnikova et al., 1991; Tsoi et al., 1991). Upon transformation of VP44 with plasmid pPC3, transformants that grew on LB400 medium with tetracycline were transferred to K1 plates with 4-CBA (5 mM) as a sole source of carbon. Transformant VP44(pPC3) formed fully grown colonies on 4-CBA plates after less than 1 week of incubation. Batches of recombinant VP44(pE43) and VP44(pPC3) grew efficiently on respective chlorobenzoates and chlorobiphenyls as a sole source of carbon and completely mineralized up to 10 mM of the chloroaromatics within 2 days, as confirmed optical density, chloride release, substrate disappearance, and protein yield. Growth of VP44(pPC3) containing the fcb genes on 4-CBA was similar to that of VP44(pE43::ohb) on 2-CBA, except for a shorter lag period, in accordance with plate growth tendencies, perhaps due to higher toxicity of 2-CBA. Protein yield was proportional to the concentration of substrate and similar to that for benzoate-grown cultures. When grown on VP44(pPC3) and VP44(pE43), recombinant strains 2-CB and 4-CB, respectively, exhibited only transient accumulation of the corresponding chlorobenzoates. Strain 4-CB exhibited rapid growth on 44(pPC3), with only transient production of yellow color (HOPDA) and 4-CBA (maximal accumulation of the 4-CBA was about 50 μM, or about 5% of the original concentration) in the culture supernatant during log phase, with concomitant release of inorganic chloride. Growth of strain 2-CB on 44(pE43) was slightly slower, with a greater accumulation of 2-CBA, 125 μM or 12.5% of original substrate concentration, which persisted slightly longer. The parent strain VP44 was capable of growth on low concentrations of both 2- and 4-monochlorobiphenyls, with accumulation of stoichiometric amounts of the corresponding chlorobenzoates in the culture supernatant. Recombinant VP44(pE43) yielded approximately the same amount of protein per mole of 2-CB, as compared to growth on biphenyl, and released stoichiometric amounts of chloride. It also grew on 4-CB, however, releasing no chloride and only about half as much protein compared to the growth on 2-CB and biphenyl, with a stoichiometric amount of 4-CBA accumulated. While recombinant VP44(pPC3) growth on 4-CB was marked by stoichiometric chloride release and protein yield comparable to that observed on

Chapter six: Enhancing PCB bioremediation

169

biphenyl, its growth on 2-CB yielded no dechlorination, only about half as much protein compared to its growth on 4-CB and biphenyl, and stoichiometric amounts of 2-CBA released in the medium. These results confirm the ability of the parent strain VP44 to utilize pentadiene derived from the nonchlorinated ring of both the para- and ortho-chlorobiphenyl molecule for growth. Thus, the introduction of the dehalogenation genes resulted in recombinant organisms that are capable of complete mineralization of orthoand para-chlorobiphenyls. Growth on biphenyls via the BP pathway for biphenyl degradation yields a benzoate and pentadiene as key intermediates, either or both of which may be chlorinated, depending on the PCB congener (Hofer et al., 1993). Introduction of the ohb (pE43) and fcb (pPC3) operons controlling ortho- and para-dechlorination of the aromatic ring, respectively, resulted in a recombinant pathway for degradation of chlorobiphenyls via dechlorination of chlorobenzoates. Both recombinant variants VP44(pE43) and VP44(pPC3) compared favorably to other chlorobiphenyl-degrading recombinant strains previously constructed via intergenic mating. Both of our strains were capable of growth on chlorobenzoates and chlorobiphenyls at high concentrations of up to 10 mM. By comparison, strain M3GY (McCullar et al., 1994) was grown on 0.89 mM 3-4′-CB with a doubling time of approximately 20 days. Strain JHR22 (Havel and Reineke, 1991) was reported to grow on several chlorobiphenyls, including 2-CB and 4-CB, and exhibited doubling times of approximately 16 and 10 h. Another strain, UCR2 (Hickey et al., 1992), was reported to mineralize both 2-CB and 2,5-CB, with doubling times of 20 and 48 h. Introduction of specific dechlorination genes can permit growth on otherwise recalcitrant substrates and may be more easily accomplished than methods previously used to generate recombinant PCB degraders. Each recombinant strain was also tested for growth on other PCB congeners. Strain VP44(pE43) grew on plates with 2,2′- and 2,4′-dichlorobiphenyls, and VP44(pPC3) grew on plates with 4,4′- and 2,4′-chlorobiphenyls, the congeners expected to yield either 2-CBA or 4-CBA as intermediates (Pellizari et al., 1996). To our knowledge, none of the previously constructed PCB-growing strains grew on congeners with both substituted rings. No growth was observed when parent VP44 cultures were amended with chlorobiphenyls containing halogen atoms in both rings of biphenyl moiety, such as 2,2′- and 2,4′-CB, suggesting the strain does not possess a pathway for efficient oxidation of chlorinated pentadiene. Apparently, this is a common characteristic for natural biphenyl-degrading bacteria that can grow on monochlorobiphenyls via oxidation of a nonchlorinated ring producing chlorobenzoate as a final product but do not grow on PCB congeners with both chlorinated aromatic rings (Ahmed and Focht, 1973; Brenner et al., 1994). One of the exceptions is Pseudomonas strain MB86 (Barton and Crawford, 1988), which was isolated on 4-CBA and grew poorly on 4-CB, probably due to toxicity of 4-chloroacetophenone, which is formed as an intermediate from 4-CB. Chlorinated pentadienes presumably generated

170

Bioremediation of Recalcitrant Compounds

from the degradation of PCBs, based on analogy to the biphenyl pathway (Omori et al., 1986), are uncharacterized, and therefore their fate remains unknown, although it was suggested that chlorinated pentadiene can be metabolized through formation of chloroacetate followed by dehalogenation to acetate (Brenner et al., 1994). Strain VP44 did not grow on chloroacetate. In summary, the ohb and fcb operons for ortho- and para-dechlorination of chlorobenzoates, respectively, were successfully expressed in C. testosteroni strain VP44. Although the parent VP44 was incapable of either growth on or dechlorination of chlorobenzoates, the introduced genes encoding dehalogenation of these compounds permitted both. The resulting transgenic strains were capable of growth on and dechlorination of the respective chlorobenzoates and chlorobiphenyls. Degradation of ortho-chlorinated PCB congeners is especially significant given their predominance among the products of anaerobic PCB dechlorination. Up to 80% molar of the PCBs present following anaerobic dechlorination of Aroclor 1242 consists strictly of ortho-chlorinated congeners; 2-CB alone may constitute as much as 40% molar of the total PCBs in anaerobically dechlorinated sediments (Bedard and Quensen, 1995). These results were summarized in a recent paper by Hrywna et al. (1999) and demonstrated an alternative approach for the construction of PCB-degrading bacteria, i.e., using genes encoding peripheral enzymatic activities for modification of xenobiotics into substrates for the central metabolic pathway for degradation of aromatic compounds. The introduction of specific dechlorination genes and their expression in the biphenyl-degrading bacterium C. testosteroni strain VP44 demonstrated the efficacy of this method for extending the substrate range for PCB degradation by biphenyl-degrading bacteria.

6.4.1.3 Developing gene transfer system for G+/G– PCB-degrading bacteria Because of inefficient rates of degradation of important PCB congeners such as 2,2′-, 2,4′-, and 2,4,2′-CBs by strain VP44, other BP degraders, particularly gram-positive RHA1 and NY05, as well as the most active strain LB400, were chosen for subsequent design of PCB-growing GEMs. Gram-positive bacteria, especially Rhodococcus strains, offer a number of advantages for environmental use, including higher growth yields on biphenyl, the presence of multiple PCB metabolic systems allowing co-oxidation of a wider range of PCB congeners (Masai et al., 1997; Seto et al., 1995), and more tolerance to environmental stresses such as drought or exposure to toxic compounds (Warhust and Fewson, 1994; Tsoi et al., 1991). Genetic engineering of catabolic pathways in Rhodococcus, however, is not well developed. We constructed a broad-host-range shuttle vector pRT1 suitable for transferring (dehalogenase) gene cassettes into Rhodococcus as well as gram-negative strains, based on RP4/RK2 derivative pSP329 and Rhodococcus-specific replicon pRC1 (Rodrigues et al., 2001). We then cloned the gene cassette carrying 4-CBA degradation operon fcbABC into pRT1, yielding

Chapter six: Enhancing PCB bioremediation

171

Growth of the recombinant RHA1 in 4-CBA 2.5

8

2

6

1.5

4

1

2

0.5

OD600

Chloride ion release (mM)

10

0

20

40

60 Time (h)

80

100

120

Figure 6.9 Growth of RHA1(fcb) on 4-CBA.

plasmid pRHD34, and showed that the E. coli lac promoter contained in pRT1 enhanced expression of the fcb genes in both Rhodococcus strains NY05 and RHA1. Although the parent RHA1 did not oxidize 4-CBA, its recombinant derivative RHA1(pRHD34) grew exponentially in medium containing 4-CBA as the sole carbon source, concomitantly releasing stoichiometric amounts of chloride (Figure 6.9). The fcb operon in strain RHA1 appeared to be stable under nonselective conditions, as verified by polymerase chain reaction (PCR) amplification using fcbA- and fcbB-specific primers (Rodrigues et al., 2001).

6.4.1.4 Degradative capabilities of the recombinant RHA1(fcb) Similar to VP44, Rhodococcus strain RHA1(pRHD34::fcb) grew on and completely mineralized 4-CB, releasing nearly stoichiometric amounts of chloride. The molar growth yield of the recombinant strain on biphenyl and 4-CB was 177 ± 6 and 189 ± 9 g dry weight of cells/mol of substrate, respectively, which is similar to the theoretical value of 173 g dry weight of cells/mol for complete 4-CB oxidation. No transient formation of 4-CBA was detected during the growth, and only slight transient formation of HOPDA was detected in the log growth phase. We noted that the 4-CBA-grown RHA1(pRHD34) inoculum was imperative for growth on 4-CB, possibly due to the rapid turnover of 4-CB with accumulation of intermediate compounds such as HOPDA when the BP

172

Bioremediation of Recalcitrant Compounds

pathway is fully induced. Seto et al. (1996) have suggested that chlorinated HOPDA or its metabolites inhibited growth of wild-type RHA1. We observed that with biphenyl-grown inoculum there was no 4-CB growth, even though formation of HOPDA was rapid and intense. The PCB degradation range of the RHA1(pRHD34::fcb) was examined in biphenyl-grown resting cell assays. We observed complete degradation of 2-, 4-, and 2,4-CB after 24 h of incubation. Accumulation of equimolar amounts of 2- and 2,4-CBA was observed; however, no 4-CBA was detected, whereas the parent RHA1 accumulated stoichiometric quantities of 4-CBA from 4-CB. This suggested activity of the enzymes encoded by the fcb operon in the recombinant strain. Because the wild-type strain RHA1 could degrade the major products from pattern M anaerobic dechlorination with a theoretical recovery of 50% of para-chlorobenzoates (Maltseva et al., 1999), we tested the activity of the recombinant strain with mix M. Similar degradation rates for mixture M congeners by the wild-type RHA1 and its recombinant were observed, with 60% of mix M PCBs degraded. However, whereas the parent RHA1 accumulated 2-, 4-, and 2,4-CBAs, only trace concentrations of 4-CBA were found in recombinant RHA1(pRHD34::fcb) (Figure 6.10). Notably, HOPDA concentrations were 38% less for the recombinant strain, possibly due to partial removal of the bottleneck presented by CBAs. Final concentrations of 2- and 2,4-CBA were similar for both wild-type and transformant strain, with no utilization of these metabolites in either case (Rodrigues et al., 2001). Chlorobipheny1, uM

A

Mix M

150 300

Chlorobenzoate, uM

B

250

100

200 150

50

100 50 0 150

A

0

Mix M + RHA1 (+fcb)

0

4

8

12

16

20

24

Time, h 2-CBA

4-CBA

2, 4-CBA

100

Degradation of Mix M by the resting cells of Rhodococcus sp. RHA1 (+fcb) (A) and accumulation of a chlorobenzoates (B).

50

0 2-CB

4-CB

2-2′/2, 6-CB

2, 4-CB

2-4′ -CB

2, 4-2′ -CB

Figure 6.10 Degradation of mix M by RHA1(fcb).

2, 4-4′ -CB

Chapter six: Enhancing PCB bioremediation

173

Schematic Representation of a Microcosm Experiment

PCB contaminated soil (20 g)

Serial dilutions

Luria-Bertani Medium Extraction and Analysis

Luria-Bertani Medium with Rifampicin 50 μg/ml

Whole cell PCR for the fcb genes

Figure 6.11 Microcosm experiment.

6.4.1.5 Survival and activity of GEM RHA1(fcb) in nonsterile soil We investigated the survival and metabolic activity of the recombinant PCB degrader RHA1(pRHD34::fcb) in nonsterile soil microcosms (Figure 6.11). The loamy sand soil (84.6% sand, 12.1% silt, and 3.4% clay) with 3.5% organic matter, pH 7.6, was passed through a 4-mm mesh sieve and stored at 4ºC until used; it originated from a noncontaminated area adjacent to PCB-contaminated soils at Picatinny Arsenal, NJ (28 miles northwest of New York City). Prior to inoculation, indigenous bacterial populations were assessed by staining with 5-(4,6-dichlorotriazine-2-yl) aminofluoroscein (DTAF) followed by epifluorescence microscopy (Bloem, 1995). Total bacterial counts of the soil before inoculation averaged to 9.9 × 108. Because we could not distinguish our strain by color and colony morphology in comparison to the indigenous bacteria, we selected a rifampicin-resistant mutant for tracking our recombinant strain. No indigenous bacteria grew on this medium. Spontaneous rifampicin (Rif+) mutants of the RHA1(pRHD34::fcb) were isolated according to Smith and Tiedje (1980). RHA1(pRHD34::fcb)Rif cells were grown on 4-CBA and, after washing twice (50 mM phosphate buffer, pH 7.0), were resuspended in mineral K1 medium (Tsoi et al., 1999) and inoculated into 20 g of soil amended with 100 ppm 4-CB, at a final density of 104 cells/g soil and 30% moisture content. Results of this study have been summarized in Rodrigues et al. (2001). As shown in Figure 6.12, recombinant organisms grew in both sterile and

174

Bioremediation of Recalcitrant Compounds 120 50/50

50/50

8

50/50 100

50/50

50/50

80

50/50

60

6 50/50 50/50

% 4-CB remaining

Log CFU/g soil

7

40 50/50

50/50 5

49/50

50/50

20

50/50 4

0

10

20

30 Time (days)

40

50

0 60

Figure 6.12 Survival and degradation of 4-CB in soil by RHA1(fcb). (Redrawn from Rodrigues et al., Environ. Sci. Technol., 35, 663–668, 2001.)

nonsterile soil concurrent with 4-CB removal but did not grow in the noncontaminated, nonsterile control soil, the latter in agreement with growth being 4-CB specific. Most significantly, the growth of the recombinant strain also occurred in the nonsterile 4-CB-contaminated soil with concurrent 4-CB removal. This implicated competitiveness of the RHA1 (fcb), which are within the indigenous soil community. The noninoculated soil showed no removal of PCB, providing further evidence that the inoculated strain was responsible for PCB degradation. The RHA1 (fcb) appeared stable in both the sterile and nonsterile soils, and in 60 days only one of the 700 colonies examined did not yield a fcbB amplicon when PCR probed with fcb gene-specific primers (Figure 6.12) (Rodrigues et al., 2001). The long-term effect of carrying exogenous DNA sequences or expression of new pathways on the fitness of the recombinant strain was not studied (Lenski et al., 1994); however, the period in which aerobic PCB treatment is needed in the field would not normally be much longer than 60 days.

Chapter six: Enhancing PCB bioremediation

175

6.4.1.6 Developing and testing molecular tracking recombinant organisms in situ The success of any bioremediation process depends upon establishing favorable conditions for microorganisms to degrade the target contaminant. Failures in bioremediation more often result from a lack of knowledge about ecological constraints than from information about the genetic capabilities of the organism. Good data on population dynamics in bioremediation are important for providing insights into better methods of inoculation, maintenance, and management of contaminated sites. Although numerous molecular markers have been developed and used for detection of microorganisms (Burlage et al., 1990; Recorbet et al., 1992; Hwang and Farrand, 1997), including successful applications based on the green flourescent protein (GFP) gene (Bloemberg et al., 1997), the addition or alteration of genetic traits may have deleterious effects on the environmental fitness of microorganisms (Lenski, 1991). Other molecular tracking methods, such as amplified ribosomal DNA restriction analysis (ARDRA) (Nüsslein and Tiedje, 1998), single-strand conformation polymorphism (SSCP) (Lee et al., 1996; Schwieger and Tebbe, 1998), denaturing gradient gel electrophoresis (DGGE) (Recorbet et al., 1992; Muyzer et al., 1993), and terminal restriction fragment length polymorphism (T-RFLP) (Muyzer et al., 1998), lack a quantitative component. In situ hybridization with fluorescently labeled oligonucleotide probes (FISH) targeted to the rRNA sequences provides for quantitation and has proven useful in aqueous environments (DeLong et al., 1989), but it is labor intensive, does not offer species-to-strain level specificity, and proved difficult to use in soil. Real-time PCR (RTm-PCR) was originally developed for measuring the copy number of a targeted gene (Dölken et al., 1998; Heid et al., 1996; Higuchi et al., 1993) and presented a promising tool for the detection and quantification of microorganisms. In this method, a double-labeled probe is used to measure the accumulation of fluorescence of a released reporter dye during PCR. This fluorescence is then correlated to the amount of product formed in real time during amplification when PCR is more quantitative (Higuchi et al., 1993). The method has been successful in clinical studies with Listeria monocytogenes (Bassler et al., 1995), Yersinia pestis (Higgins et al., 1998), Mycobacterium tuberculosis (Desjardin et al., 1998), and Borrelia burgdorferi (Pahl et al., 1999), as well as for assessing ecology of denitrifying nitrite reductase (Grüntzig et al., 2001). We applied this technique to study the population dynamics in soil of a recombinant Rhodococcus RHA1(pRHD34::fcb) genetically engineered to grow on a chlorobiphenyl by quantifying both the 16S rRNA gene and the introduced 4-chlorobenzoate fcb degradation operon (Rodrigues et al., 2002). Species-specific 16S rRNA regions were identified for Rhodococcus sp. strain RHA1 using the phylogeny of these strain and close relatives. All candidate primers included more than one species of the genus Rhodococcus because of the inherent difficulty in designing a 16S rRNA-based strain-specific probe. The

176

Bioremediation of Recalcitrant Compounds

RHA1 16S rRNA gene sequence corresponding to E. coli positions 599 to 619 was chosen for designing the TaqMan-16S rDNA probe based on the presence of only two other perfect matches for this primer, R. marinonascens and R. globerulus. The probe contained FAM (6-carboxy-fluorescein) as a reporter fluorochrome on the 5′ end and TAMRA (N,N,N′,N′-tetramethyl-6-carboxy-rhodamine) as a quencher on the 3′ end of the nucleotide sequence. A flask trial was performed to test the specificity of the designed probe, and fluorescence during PCR amplification was observed for Rhodococcus strains only. Similarly, primers and a suitable probe targeting the fcbB gene (chlorobenzoyl–CoA hydrolase/dehalogenase) were designed and tested. In this case, among some closely related Rhodococcus species, quantitative amplification of the sequence was possible only for RHA1(pRHD34::fcb). To test whether the recombinant organism could be detected within the soil community, a soil microcosm experiment was conducted, similar to that described in the above section, and the results of this study were summarized in Rodrigues et al. (2002). After 30 days of incubation, the estimated numbers of RHA1(pRHD34::fcb) cells per gram of soil measured by the two TaqMan probes were similar to the values obtained for culturable rifampicin-resistant RHA1(fcb) cells (colony forming units , CFUs/g soil) throughout the growth cycle. Growth was expected on the particular congeners in the PCB mixture, and the absence of increased counts may have been due to protozoan grazing, although this was not investigated. The method did allow measurement of population dynamics in soil and an assessment of the stability of the engineered gene (Figure 6.13). Molecular Tracking of RHA 1 (fcbB) in Picatinny Arsenal Soil Using Real Time PCR 40 R2 = 0.9816

35

RHA 1 in Soil (CFUs) Plasmid pRHD34 (copy #)

30

RHA1 Whole Cell DNA (copy #)

Ct

25 20 15 R2 = 0.9984

10

R2 = 0.9964 5 0 −4

1

6 11 16 21 ln CFUS or Copy Number

26

31

Figure 6.13 Molecular tracking GEMs in soil. (Redrawn from Rodrigues et al., J. Microbiol. Meth., 51, 181–189, 2002.)

Chapter six: Enhancing PCB bioremediation

177

Detection limits for PCR-based methods depend upon numerous factors, among them the type and composition of the matrix, the type of target organism, the number and diversity of bacteria in the sample, the Taq polymerase used, and the bias introduced by the DNA extraction protocol (Löffler et al., 2000). Our experiments suggested that the target sequence can also affect detection, as seen by the different standard curves for fcb-targeted gene vs. the 16S rRNA-targeted gene. One explanation for different sensitivities might be the size of the strain RHA1 genome (3.0 Mb) (Tonso, 1997) in comparison to the plasmid pRHD34 size (14.4 Kb), limiting the ability of the TaqMan-16S rDNA probe to find its target. However, chromosomal and plasmid targets of RHA1 16S rRNA gene gave the same count. The lower count values observed when the TaqMan-fcb probe was used are probably due to the probe efficiency, and not to the difficulty of genome strand separation when targeting the 16S rRNA genes. Penalty score analysis for the two TaqMan probes indicated differences between the two probes; values of 250 for the TaqMan-16S rDNA and 49 for TaqMan-fcb (Perkin Elmer Applied Biosystems, 1998), suggest that the probe targeting the fcb operon met most of the criteria required for RTm-PCR, while the 1.6S rDNA end did not. Use of rDNA phylogenetic probes is hampered by a limited number of specific regions within the hypervariable sequence of the 16S rRNA gene (Stackebrandt and Rainey, 1995). Thus, the TaqMan-16S probe design had to be constructed from a predetermined position, whereas the TaqMan-fcb probe was chosen from many different possibilities within the entire fcb operon. In summary, a real-time PCR assay using fluorescently labeled oligonucleotides (TaqMan probes) was developed and tested on a model system of the recombinant Rhodococcus sp. strain RHA1(pRHD34::fcb) in nonsterile soil. One primer and probe set targeted the hypervariable region of the 16S rRNA gene, and the other set targeted the recombinant 4-cB degradation (fcb) operon. The 16S rDNA probe detected RHA1(pRHD34::fcb) and phylogenetically related Rhodococcus species, whereas the fcb probe was specific for the recombinant strain. The method had a 6-log dynamic range of detection (102 to 107) for both probes in batch cultures but a lower sensitivity in soil. The estimated number of cells in soil by real-time PCR measurement corresponded to the number of RHA1(pRHD34::fcb) CFUs recovered from soil. The real-time PCR method is easy to perform, has high throughput, and was found to be reliable for targeting organism in nonsterile soil.

6.4.1.7 Construction of multiple ortho-PCB dechlorinator LB400(ohb) To design organisms capable of growing on the most environmentally relevant PCB cogeners, we have recruited the P. aeruginosa strain 142 oxygenolytic ohb operon for ortho-dechlorination of mono-, di-, and trichlorobenzoates. Plasmid pRO41 (Figure 6.14) contains ohbABCR genes coding for the iron-sulphur protein, terminal oxidoreductase (ISP-OHB), potential ATP binding cassette (ABC) transporter, and putative transcriptional regulator,

oriT oriV

PISIHIII

XI

Plac

pRT1 10.8 kb G+/G−

KICIRI

ohb clc

pOCC1 18.6 kb G+/G−

XIBIMIKICIRI

BamHI

PISIHIII

BamHI

pRC, 2.6 kb

XbaI + PstI

Figure 6.14 Construction of plasmid pRO41.

PstI

pK4 5.2 kb

pRC

XbaI

α pSP329 Plac 8.2 kb tetR bhr tetA trfA

BamHI

pUC18 M. C. S.

Plac RI

EcoRI

BamHI

PISIHIII

XI

3.5 kb

ohbRohbABC

KpnI + XbaI

KpnI

4.3 EcoRI-fragment clc (pDC100)

KpnI + XbaI

Construction of pRT1, pRO41, and pOCC1

ohbRABC

pRO41 14.3 kb G+/G −

KICIRI

Plac

XbaI

pE43 pE43 ohbABCR

178 Bioremediation of Recalcitrant Compounds

Chapter six: Enhancing PCB bioremediation

179

respectively, as we previously reported (Tsoi et al., 1999). Plasmid pOCC1, in addition, contains the clc genes for the chlorocatechol pathway from D-plasmid pAC27 (Ghosal and You, 1989) under control of the lac promoter. The plasmids may conveniently be used for transformation of laboratory or indigenous biphenyl-degrading organisms, which customarily belong to either the Rhodococcus genus or a variety of gram-negative bacteria. Addition of the clc genes presumably would allow recombinant PCB degraders to grow on congeners that contain di- or trichlorinated rings, yielding respective di- or trichlorobenzoates that would be ortho-dechlorinated by the ohb genes to the respective mono- or dichlorocatechols. Plasmids pRO41 and pOCC1 were introduced in four different biphenyl-degrading strains: VP44 and LB400 (gram negative), and Rhodococcus NY05 and RHA1 (gram positive). Recipient VP44 served as a control because we reported previously expression of the ohb genes in this host (Hrywna et al., 1999). Controls were the same recipients transformed with vector pRT1. After 8 (VP44) to 12 (LB400 and NY05) weeks of incubation, fully grown single colonies (a few per each plating) grew on 2-CBA in the case of plasmids pRO41 and pOCC1. No growth was detected for pRT1, as expected. No growth was detected for pOCC1 on dichlorobenzoates (dCBA). No growth has been detected in the case of Rhodococcus RHA1. From each original transformant, a few randomly chosen colonies grown on 2-CBA were serially transferred in 2-CBA and BP batches and shown to fully grow on 2 mM 2-CBA in 2 to 5 days, depending on the recipient (fastest, VP44; slowest, NY05). The best-growing GEM was LB400(pRO41), which was used in further studies. In our previous experiments (see above sections), resting cells of LB400 partially oxidized 2-CBA; this correlated with a less than expected yield of 2-CBA from oxidation of 2-CBP and 2,2′-CBP. However, control variant LB400(pRT1) did not grow on 2-CBA, thus implicating the cometabolic nature of the 2-CBA oxidation by LB400. Recombinant LB400(pRO41) (Figure 6.15) efficiently grew on 2-CBA with stoichiometric release of Cl–. It appears that many biphenyl degraders can normally grow on 4-CB and 2-CBP via oxidation of the nonchlorinated ring (pentadiene) (Hrywna et al., 1999). Control strain LB400(pRT1) was grown on BP and inoculated in K1 medium with 1 mM 2-CBP. LB400(pRT1) grew on 1 mM 2-CBP, accumulating 2-CBA and Cl– in culture liquid. The residual 2-CBA and Cl– total to approximately 1.2 mM, as measured, indicating complete consumption of 2-CB. Again, less than stoichiometric production of 2-CBA indicated its partial oxidation. The recombinant LB400(pRO41) was grown on BP and inoculated in 1 mM 2-CBP (Figure 6.16). The GEM completely oxidized both 2-CBP and 2-CBA, yielding more biomass than the control LB400(pRT1), with an estimated 1.2 mM Cl– released. Haddock et al. (1995) purified biphenyl dioxygenase from LB400 and showed that 97% of 2-CBP is dihydroxylated at the 2,3 positions of the nonchlorinated ring and only 3% is dihydroxylated/dehalogenated in the 2,3 positions of the chlorinated ring. Therefore, most of the 0.7 mM Cl– released by control strain LB400(pRT1) supposedly came from dechlorination of

180

Bioremediation of Recalcitrant Compounds Growth of LB400 (pRO41) on 2CBA 0.5

OD, 3 mM 2CBA OD, 2 mM 2CBA OD, 1 mM 2CBA 2CBA, 3 mM

3.5 3.0 2.5

0.3

2.0 0.2

1.5 1.0

0.1

2CBA, Cl−(mM)

OD600

0.4

4.0

2CBA, 2 mM 2CBA, 1 mM Cl−, 3 mM 2CBA Cl−, 2 mM 2CBA Cl−, 1 mM 2CBA

0.5 0.0

0

20

40

60 80 Time, hours

100

120

140

Growth of control LB400 (pRT1) on 2CBA 0.5 OD, 2 mM 2CBA OD, 1 mM 2CBA 2CBA, 2 mM 2CBA, 1 mM Time, h vs Col 16 Time, h vs Col 17

0.3 0.2 0.1 0.0 0

20

40 60 Time, hours

80

100

2.50 2.25 2.00 1.75 1.50 1.25 1.00 0.75 0.50 0.25 0.00

2CBA, Cl− (mM)

OD600

0.4

0.0

Figure 6.15 Growth of LB400(pRO41::ohb) and LB400 on 2-CBA.

2-CBA. In conjunction with data by Bedard et al. (1986) on co-oxidation of 2-CBA (without growth), it could be that the biphenyl dioxygenase (BDO) from LB400 does have 2,3-dihydroxylation/dehalogenation activity on both 2-CBA and 3-CBA, unless LB400 possesses another chlorobenzoate-dehalogenating enzyme. In this case, no catechol would be produced, in accordance with the inability of LB400 to grow on 2- and 3-CBA. To verify this, we PCR-amplified and cloned the BDO genes in E. coli, and the bphA1A2A3A4B genes appeared to be responsible for the partial dechlorination of 2-CBA. However, no product of this dechlorination process has been identified. Strain LB400 is one of only a few organisms known to efficiently oxidize an environmentally important 2-2′-CB. Haddock et al. (1995) and Arnett et al. (2000) showed with purified BDO that the initial dihydroxylation of 2,2′-CBP occurs in 2,3 positions with the release of chlorine. Consequently, the dechlorinated ring is further oxidized to pentadiene, which can be consumed for growth, whereas the second ring converts to 2-CBA. LB400(pRO41) grew on 1 mM 2,2′-CBP and released 2.2 mM Cl–, as

Chapter six: Enhancing PCB bioremediation Growth of LB400 (pRO41) and control LB400 (pRT1) on 2-CB (1 mM)

0.35

1.5

0.25 0.20

1.0

−

0.15

2CBA, Cl , (mM)

OD, pRO41 OD, pRT1 Cl−, pRO41 Cl−, pRT1 2CBA, pRO41 2CBA, pRT1

0.30

OD600

181

0.10

0.5

0.05 0.00 0

20

40 Time, hours

60

80

0.0

Growth of LB400 (pRO41) and control LB400 (pRT1) on 2-2′-CB (1 mM) 0.25 0.20 0.15

2.5 2.0 1.5

−

0.10

1.0

0.05 0.00

2CBA, Cl , (mM)

OD600

3.0

OD, pRO41 OD, pRT1 Cl−, pRO41 Cl−, pRT1 2CBA, pRO41 2CBA, pRT1

0.5 0

20

40

60 80 100 Time, hours

120

0.0 140

Figure 6.16 Growth of LB400(pRO41::ohb) and LB400 on ortho-PCBs.

measured, indicating complete mineralization of 2,2′-CB. Therefore, simultaneous expression of the two ortho-dehalogenating dioxygenases, BDO and ISP-OHB, resulted in complete mineralization of 2,2′-CB. In addition, LB400 grew on 3-CBA due to the indigenous chlorocatechol pathway. Consequently, LB400(pRO41) efficiently grew on 2,3- and 2,5-dichloribenzoates and on dichlorobenzoates 2,4- and 2,6-dCBAs, although the latter two are poor growth substrates. We showed that efficiency of degradation of 2-CBP and 2,2′-CBP by recombinant LB400(pRO41) was dependent on preparation of inoculum. Whereas complete mineralization of 2-CBP was not affected by whether the inoculum was grown on BP or 2-CBA, a dramatic difference was found with a higher-chlorinated 2,2′-CB. In the latter case, complete mineralization was achieved only with inoculum grown on 2-CBA, whereas only partial degradation was observed with BP-grown inoculum. Indeed, BP-initiated culture LB400(pRO41) behaved similarly to control strain LB400(pRT1), except that no 2-CBA accumulated in the former, indicating that the incomplete mineralization could not be attributed to plasmid instability (loss of the ohb genes).

182

Bioremediation of Recalcitrant Compounds

The initial ratio of participating dioxygenases (BDO and ISP-OHB) therefore appeared crucial for more difficult substrates when inhibitory effects of reaction products were suspected (such as the case of 2,2′-CBP). Consequently, LB400(pRO41) inocula for bioaugmentation of PCB-contaminated soils should be grown on CBA prior to delivery in soil.

6.4.1.8 Growth on defined PCB mixtures Control strain LB400(pRT1) was grown on BP and inoculated in K1 medium with 1 mM mix C. Mix C represents the best scenario that would be seen in anaerobically dechlorinated sediments initially contaminated with Aroclor 1242, 1248, and perhaps 1256. There were no previous reports on growth of LB400 on any artificial mixes or naturally occurring dechlorination products. As shown in Figure 6.17, the control strain grew on mix C, and the sum of released Cl– and yielded 2-CBA, 4-CBA, and 2,4-CBA totaled approximately 1.4 mM Cl–, compared to the theoretically expected 1.5 mM. Of the total 1.4 mM chlorine released, 0.65 mM was released as Cl– and 0.7 mM accumulated as 2-CBA in accordance with growth of the LB400(pRT1) on individual 2and 2,2′-CBP, and with only partial consumption of 2,2′-CBP. In contrast, recombinant LB400(pRO41) released 1.5 mM Cl– and only negligible amounts of 4- and 2,4-CBA and chlorocatechol, indicating nearly total mineralization. Further flask experiments were done to characterize growth and degradation of products expected from the anaerobic phase of Aroclor remediation. In these experiments, the inoculum of GEM LB400(pRO41::ohb) was prepared on 2,5-CBA, whereas control strain LB400 was inoculated from BP-grown culture. Individual flasks in triplicates represented time points. Production of biomass (optical density, OD, at 600 nm), degradation of PCBs (GC), accumulation of CBAs (HPLC), and release of Cl– (Bergmann and Sanik method, 1957) were measured. The recombinant LB400(pRO41::ohb) grew efficiently on both lighter mix C (Figure 6.17) and heavier mix M (Rodrigues et al., 2005, submitted). Importantly, we observed significantly higher biomass production and Cl– release, greatly diminished accumulation of CBAs, and potentially toxic intermediates such as (Cl)HOPDAs by the recombinant LB400(pRO41::ohb). It also well sustained the exceptionally high rates of PCB oxidation characteristic for the parent LB400. These results suggested that the GEM is superior to its wild-type parent for bioremediation purposes.

6.4.1.9 Validation of PCB remediation strategy in soil (microcosm studies) In preparation for a field test, we conducted a series of laboratory experiments both in flasks and in soil microcosms to validate applicability of the proposed two-phase anaerobic–aerobic PCB remediation employing GEMs. Based on growth in flasks on defined PCB mixtures M and C, on Aroclor toxicity testing, and on survival of different GEMs in soil microcosm experiments, we have concluded that using a combination of two GEMs, the longer-surviving gram-positive Rhodococcus RHA1(pRHD34::fcb) and the

Chapter six: Enhancing PCB bioremediation

183

Growth on 1 mM Mix C − Cl , mM 2CBA, mM 4CBA, mM 2, 4CBA, mM

0

20

40 Time, hours

60

80

1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

CBA, Cl−, mM

OD600

Growth of LB400 (pRT1) on Mix C (1 mM) 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

2.0 Col 7 vs Col 8 Col 7 vs Col 9 Col 7 vs Col 10 Col 7 vs Col 11 Col 7 vs Col 12

1.5

1.0

CBA, Cl−, mM

OD600

Growth of LB400 (pRO41) on Mix C (1 mM) 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

0.5

0.0 0

20

40

60

80

100

120

Time, hours

Figure 6.17 Growth of LB400(pRO41::ohb) and LB400 on mix C.

more active and PCB tolerant gram-negative Burkholderia xenovorans LB400(pRO41::ohb), is the most effective for achieving maximum PCB degradation. Sediment samples were obtained from the Red Cedar River (Michigan) and contaminated with Aroclor 1242. The sediment was inoculated with anaerobic-dechlorinating microorganisms eluted from the River Raisin (Michigan) sediment and submitted to anaerobic conditions for 1 year. The anaerobic treatment was followed by inoculation with the aerobic GEMs, rifampicin-resistant recombinants RHA1(fcb) and LB400(ohb). The aerobic inoculums were grown on 3 mM (nominal concentration) medium containing BP and 2,5-CBA, respectively. Cells were added to 1 g of contaminated sediment (50% solids microreactors) to give a density of 104 (low-density treatment) or 106 (high-density treatment) cells g–1 of sediment for each recombinant strain. Noncontaminated sediment as well as noninoculated contaminated sediment submitted to the same conditions were used as controls. Flasks were continually shaken at 150 rpm and incubated in duplicates at 30˚C for 30 days. Sediment samples were stored at –20˚ for soil DNA and PCB extractions. Samples were taken periodically and immediately diluted. Population dynamics of recombinant RHA1(fcb) and LB400(ohb) strains were

184

Bioremediation of Recalcitrant Compounds

tracked by plating on rifampicin-containing Luria agar plates (Figure 6.18). Direct bacterial counts of the sediment before inoculation averaged 4.82 × 108 cells g–1 sediment. Indigenous bacteria could not be isolated below our limit of detection. Bacterial counts increased for both high- and low-density inoculation treatments. In the high inoculation treatment, RHA(fcb) and LB400(ohb) cell numbers increased to 7.8 × 106 and 1.4 × 107 cells/g sediment, respectively. Increases in cell numbers were also observed in the low-density treatment at day 15. LB400(ohb) colonies were not observed on agar plates for the noncontaminated treatment, whereas RHA1(fcb) colonies increased from 1.0 × 104 to 2.2 × 105 cells/g soil. The calculated CFUs/g sediment, using real-time PCR, for RHA1(fcb) and LB400(ohb) with TaqMan-fcb and -ohb probes were 5.3 × 106 and 4.4 × 106 cells/g sediment, respectively, at day 15. Sediment samples taken from noninoculated controls did not show fluorescence above the threshold level. The fcb and ohb operons in strains RHA1 and LB400, respectively, appeared to be stable, as verified by PCR amplification with the ohb and fcb gene-specific primers. Similar degradation patterns for Aroclor 1242 were observed for high and low cell densities (Figure 6.19) by day 30 of the experiment. PCB removal for high and low cell densities was 57 and 54%, respectively, and 4% of PCBs were degraded in the noninoculated treatment, indicating that the inoculated strains were actively degrading the products of the anaerobic dechlorination of Aroclor 1242. It should be noted that degradation of PCBs in these 50% solid microreactors was significantly lower than the results of flask experiments. To investigate whether these rates were affected by water content, we conducted a similar test in 20% solids (wt/v) microslurries. The slurries were aerated at room temperature for 15 days and sacrificed for PCB extraction and analyses. Under these conditions, the Aroclor 1242 dechlorinated products (60 ppm) were degraded to a final concentration of 16 ppm. In conclusion, using a model system of River Raisin sediment historically contaminated with Aroclor 1242, we validated the proposed two-phase anaerobic–aerobic PCB bioremediation scheme. In these experiments, enhanced anaerobic PCB dechlorination was achieved using Hudson River inoculum and resulted in a shift in the congener profile from highly to lower-chlorinated PCB congeners. The anaerobic treatment was followed by inoculation with aerobic PCB-growing GEMs. This resulted in 78% removal of PCBs. Whereas the aerobic incubation caused a slight degree of intrinsic PCB degradation, inoculation with the PCB-growing GEMs was essential to achieve rapid PCB removal.

6.4.1.10

Developing protocol for inoculum delivery

To find out if we could enhance the delivery of inoculum to the soil, we compared two methods of inoculation: addition of cells directly to the soil and addition of cells to vermiculite, which was then added to the soil. We chose vermiculite over other possible inoculum carriers such as rice hulls, perlite, and calcium alginate beads because vermiculite is inexpensive, readily available, and graded for size uniformity.

Real Time PCR

DNA extraction

0.25 g of soil

104 cells/g

Figure 6.18 PCB degradation in 50% solids (wt/v) slurries.

PCB extraction and analysis

1 g of soil

106 cells/g

Serial dilution

Non-contaminated

Whole cell PCR for isolated colonies (fcb and ohb primers)

Luria-Bertani Medium with Rifampicin 50 μg/ml

1 g of soil

Non-inoculated

Schematic Representation of the Sediment Experiment

Chapter six: Enhancing PCB bioremediation 185

186

Bioremediation of Recalcitrant Compounds

100

% PCB remaining

80

60

40

20

0 0

10

Non-inoculated control 104 cells . g−1 of sediment 106 cells . g−1 of sediment

15

30

Days

Figure 6.19 PCB degradation in 50% slurry microreactors.

In our trials, two cell densities were used, a low density of ~104 cells/g and a high density of ~106 cells/g. Rifampicin-resistant cells were grown in liquid K1 mineral salt media with 20 mM biphenyl and then transferred to K1 with 2 mM 2-CBA (LB400(ohb)) or 4-CBA (RHA1(fcb)) to log phase. Cells were then added to fine-grade vermiculite. We used vermiculite at a rate of 3% w/w vermiculite to soil. We diluted our inoculum with nonsterile tap water so that we added a total volume of 1 ml of dilute inoculum to 1 ml of vermiculite. We homogenized the inoculum/vermiculite mixture and then added it to soil. In the nonvermiculite treatment, we diluted the inoculum to a volume that would adjust the soil water potential to –0.5 bars. This volume is dependent on the water potential of the soil and must be adjusted for each new case. Prior to the addition of the inoculum, 2-CBP or 4-CBP was added from a 1 M stock to 4 ml of acetone, which was dribbled into the soil with constant stirring. The soil was then vigorously mixed in a sealed metal container for 30 min and allowed to stand overnight at room temperature for volatilization of acetone from the soil. Two grams of inoculated, chlorobiphenyl-contaminated soil was dispensed into each of 24 10-ml serum vials per treatment, and these were then sealed with a Teflon-coated rubber septum and crimp top. Samples were taken at random, in triplicate over the course of the experiment. Microcosms were incubated at 25˚C and sampled on days 0, 2, 5, 10, 15, 20, 30, and 60. Sampling was done by injecting 5 ml of 10 mM

Chapter six: Enhancing PCB bioremediation

187

potassium phosphate buffer (pH 7) into each vial. The six vials were vortexed horizontally at one time for 10 min at the maximum speed and allowed to stand for 3 min to permit sedimentation of soil particles from the water column. Serial dilutions were prepared and spread onto Luria agar plates containing rifampicin. Immediately following cell extraction, the microcosms were frozen at –20˚C. PCBs were then extracted from these same vials and analyzed using a GC. Survival of Rhodococcus sp. strain RHA1(pRHD34::fcb) and Burkholderia xenovorans strain LB400(pRO41::ohb) was greatly improved when vermiculite was used as an inoculum carrier. Growth did occur in the presence of chlorobiphenyl with and without vermiculite. However, the growth was an order of magnitude greater when vermiculite was used as a carrier.

6.4.1.11

Recommendations for inoculum delivery during pilot test

Recommendations for inoculum delivery during pilot test:

1.

Inoculation density a. Organisms will be inoculated at 106 cells/g. 2. Vermiculite a. Use 3 g vermiculite/100 g of soil (for slurry, use weight of soil + water) (3% w/w). b. Use a medium- to fine-grade vermiculite for greater surface-to-volume (mass) ratio. 3. Vermiculite impregnation a. Cell inoculum is diluted to bring the total water content of the vermiculite to a ratio of 1 ml to 1 g. b. Diluted cells are mixed into vermiculite immediately after they are harvested until homogenous. 4. Addition of impregnated vermiculite to soil a. Impregnated vermiculite is added to soil immediately following addition of cells to vermiculite. b. Homogeneity of vermiculite and soil in high solids (land farming) and medium solids treatments must be reached.

6.4.1.12

Compatibility of anaerobic and aerobic phases in remediation process

In the anaerobic-aerobic bioremediation system to be employed, FeSO4 will be added to the soil to enhance anaerobic reductive dechlorination. To find out if residual FeSO4 or the resulting FeS from this process will have a negative influence on the efficiency of aerobic metabolism, we grew LB400 and RHA1 in the presence of these Fe(II) compounds (Figure 6.20).

188

Bioremediation of Recalcitrant Compounds Growth of RHA1 (fcb) on 4-CBA in the Presence of Fe(II) Compounds 5 1e + 8 1e + 7 3 1e + 6 2

1e + 5

1 0

cfu/ml

4-CBA (mM)

4

1e + 4

0

10

20 Time (hr)

cfu/ml w/10 mM FeSO4 & 10 mM FeS cfu/ml w/20 mM FeS cfu/ml 4-cba Only Control

30

1e + 3 40

10 mM FeSO4 & 10 mM FeS [4-cba] w/20 mM FeS [4-cba] in 4-cba Only Control

Figure 6.20 Effect of Fe(II) on growth of RHA1(fcb).

LB400(ohb)Rif+ and RHA1(fcb)Rif+ freshly grown cells were added to a final density of 104 cells/ml in K1 media with 2-CBA or 4-CBA, or 2 mM 2-CB or 2 mM 4-CB with either 20 mM FeSO4, 10 mM FeSO4 + 10 mM FeS, 10 mM FeS, or an Fe-free control. The pH of each treatment was adjusted to ~7. The results of studies using CBAs as growth substrates indicate that FeSO4 and FeS do not appear to adversely affect either organism. Given these results, the anaerobic-stage FeSO4 treatments used to stimulate anaerobic dechlorination should not affect the activities of either aerobic organism in soil.

6.4.2 Microbial-surfactant compatibility experiments A number of research groups have evaluated the use of surfactants to enhance the bioremediation of hydrophobic organic compounds (HOCs), including phenanthrene and naphthalene (Laha and Luthy, 1991; Tsomides et al., 1995). However, these studies have yielded mixed results, showing either enhancement, reduction, or no measurable effect on contaminant biodegradation following surfactant addition (Rouse et al., 1994). These conflicting results are due, in large part, to the complex interactions that may occur between surfactants, microorganisms, the solid phase, and the contaminant(s) of interest. A series of batch reactor experiments was first conducted to determine the effects of surfactants on the growth and survival of two biphenyl-degrading strains of bacteria, gram-positive Rhodococcus erythreus strain NY05 and

Chapter six: Enhancing PCB bioremediation

189

gram-negative Comamonas testosteroni strain VP44. Three commercially available nonionic surfactants (Witconol SN-120, Tergitol NP-15, and Tween 80) were selected for study based on their solubilization potential and susceptibility to microbial degradation (Pennell et al., 1997; Yeh et al., 1998). Of the three surfactants evaluated, the linear alcohol ethoxylate (Witconol SN-120) completely inhibited microbial growth of both bacterial strains. In contrast, both microorganisms achieved substantial growth in solutions containing polyoxyethylene (20) sorbitan monooleate (Tween 80) as the sole carbon source, as well as in solutions containing Tween 80 with either biphenyl or BP a monochlorinated biphenyl (4-CBP). The nonylphenol ethoxylate (Tergitol NP-15) did not inhibit microbial growth when present in solutions containing either BP or 4-CBP. In addition, Tergitol NP-15 did not support microbial growth when present in solution as the sole carbon source. Examples of representative NYO5 and VP44 growth curves in the presence of Tween 80 and 4-CBP are shown in Figure 6.21. A summary of the experimental results obtained from the growth experiments for both microbial strains is presented in Table 6.2. Based on these results, Witconol SN-120 was eliminated from consideration as a surfactant in subsequent experiments. Tergitol NP-150 was attractive because it did not appear to negatively impact growth of NY05 and VP44 and would not be utilized as an alternative carbon source. However, growing concerns over the toxicity of nonylphenol ethoxylates in aquatic systems eliminated Tergitol NP-15 from use in the bioreactor studies (Ahel et al., 1994).

6.4.3 Plasmid stability studies A matrix of experiments was performed to ascertain whether microorganisms engineered to grow on a particular PCB congener or pathway would lose that ability if allowed to grow on other substrates, including surfactants. 10

1 1185 ppm (2.1 mM 4CB) 990 ppm (0.7 mM 4CB) 1309 ppm (0.5 mM 4CB) 1309 ppm (0.4 mM 4CB) BP, 38 mM (ref )

0.1

0.01

Optical Density

Optical Density

10

1 1961 ppm (0.9 mM 4CB) 1839 ppm (0.6 mM 4CB) 1405 ppm (3.9 mM 4CB) 1098 ppm (0.61 mM 4CB) 1071 ppm (0.044 mM 4CB) 814 ppm (0.044 mM 4CB) BP, 6.6 mM (ref )

0.1

0.01 0

50

100 Time, hour

150

0

50

100 Time, hour

Figure 6.21 Examples of growth of NY05 and VP44 on Tween 80 + 4-CB.

150

190

Bioremediation of Recalcitrant Compounds Table 6.2 Summary of Results Obtained from NY05 and VP44 Growth Studies

Surfactant alone Surfactant + biphenyl Surfactant + 4-CB

Witconol SN-120

Tween 80

Tergitol NP-15

No growth No growth No growth

Growth Growth Growth

No growth Growth Growth

Initially, solutions of 4-CBA and 2-CBA were prepared by dissolving each acid in 1 N KOH. The pH of the solution was brought below 7.0 by addition of 2 N H2SO4. A growth flask was then prepared with approximately 100 ml of K1 nutrient solution, and 2-CBA or 4-CBA was added to reach a concentration of approximately 2 mM. The flasks were inoculated with engineered strains of RHA1, NY05, and VP44 provided by investigators at Michigan State University (MSU), to which 2 ml of 100 mM CBA solution was added each day. Once the solution in each flask reached an optical density of 1.0 or greater, 10 ml of the solution was used to inoculate a second flask containing the same K1 and CBA solution, and the process was repeated. After the second flask reached an optical density of 1.0, the solution was centrifuged and the supernatant was discarded. The biomass was resuspended in K1 media and centrifuged for a second time. This process was repeated two more times in an attempt to remove all of the CBA from solution. After the final centrifugation, the biomass was resuspended to a concentration of 0.2 g biomass/ml. This solution was equally divided and used to inoculate three growth flasks, the first containing a solution of K1 and BP, the second containing a solution of Tergitol NP-15 and BP, and the third containing a solution of Tween 80. A sequential plating method was used, with five plates being produced for each instance with the hope that the fifth plate would produce individual colonies. These plates were labeled, sealed with parafilm, and forwarded to laboratories for PCR analysis. A summary of the PCR analysis for RHA1(fcb) is given in Table 6.3. Here, note that under all of the growth conditions evaluated, the plasmid gene was detected by PCR analysis. Table 6.3 Representative Results of Plasmid Stability Tests Substrate

Plasmid Gene

Initial Substrate

PCR Analysis

Biphenyl + K1 Tween 80 + K1 Tergitol NP-15 + K1

RHA1 + fcb RHA1 + fcb RHA1 + fcb

4-CBA 4-CBA 4-CBA

5/5 5/5 5/5

Chapter six: Enhancing PCB bioremediation

191

6.4.4 PCB-surfactant solubilization experiments Surfactants have the capacity to greatly increase the overall aqueous solubility of hydrophobic organic compounds, including chlorinated biphenyls, at surfactant concentrations above the critical micelle concentration (CMC), the point at which individual surfactant monomers aggregate to form micelles in solution (e.g., Pennell et al., 1997). The aqueous solubility of 4-CBP in solutions of Tween 80 and Tergitol NP-15 was determined from batch experiments conducted in 26-ml glass serum tubes. The lower portions of the tubes were coated with 4-CBP that was transferred in a solution of hexane. The hexane was allowed to evaporate, leaving a thin film of 4-CBP of sufficient mass to exceed the solubility limit. Surfactant solutions were prepared over a concentration range of 50 to 2000 mg/l. Approximately 20 ml of surfactant solution was added to each serum tube; these tubes were placed on an orbital shaker and mixed at 150 rpm for 3 weeks. The solubility experiments were conducted at 22˚C (room temperature) and 30˚C (temperature of microbial growth studies). After equilibrium was reached, 0.4-ml samples were taken from the serum tubes and mixed with 1.2 ml of isopropanol in 1.6-ml glass autosampler vials. The concentration of 4-CBP was determined using an Agilent 6890 Gas Chromatograph equipped with an autosampler and electron capture detector. To improve detection for lower concentrations of 4-CBP, samples from the serum tubes were transferred as 10- to 15-ml aliquots to clean serum tubes, into which 3 to 5 ml of hexane was added. These serum tubes were mixed by hand and centrifuged at 3000 rpm for 20 min to separate the aqueous and organic phases. Samples were collected from the hexane phase and transferred to autosample vials for subsequent GC/ECD (electron capture detection) analysis. Results of representative 4-CBP solubility experiments, conducted for Tween 80 and Tergitol NP-15, are presented in Figure 6.22. The aqueous solubility of 4-CBP is approximately 2 mg/l (0.011 mM) at 22˚C. A linear enhancement in 4-CBP solubility was observed for both surfactants above the CMC, which is consistent with the conceptual model of HOC partitioning into the hydrophobic core of surfactant micelles (Rosen, 1992). The micellar solubilization capacity of a surfactant for a particular HOC can be described in terms of a molar solubilization ratio (MSR): MSR =

C o − C o ,cmc C s − C s ,cmc

(6.1)

where Co is the molar concentration of the organic species, Co,cmc is the molar concentration of the organic species at the CMC (i.e., the aqueous solubility), Cs is the molar concentration of surfactant in solution, and Cs,cmc is the molar concentration of surfactant at the CMC. The resulting MSR values for 4-CBP in micellar solutions of Tween 80 and Tergitol NP-15 at 22˚C were 0.33 and

192

Bioremediation of Recalcitrant Compounds

MSR = 0.38 (30°C) 0.35

MSR = 0.33 (22°C) R2 = 0.993 2

R = 0.994

0.2

MSR = 0.14 (22°C)

0.3

0.15

4-CB, mM

4-CB, mM

0.3 0.25

MSR = 0.20 (30°C)

0.35

0.25

R2 = 0.997

0.2 0.15

0.1

0.1

0.05

0.05

0

R2 = 0.998

0 0

0.5 Tween 80, mM

1

0

1 2 Tergitol NP-15, mM

3

Figure 6.22 Solubility enhancements of 4-CBP in solutions of Tween 80 and Tergitol NP-15 at 22 and 30˚C.

0.14, respectively. These findings indicate that Tween 80 has more than twice the capacity to solubilize 4-CBP than does Tergitol NP-15. The MSR values obtained above may be expressed as a micelle–water partition coefficient (Kmw) defined as K mw =

Xm Xa

(6.2)

where Xm is the mole fraction of organic species in the micellar phase and Xa is the mole of organic in the aqueous phase. The corresponding log Kmw values for 4-CBP in solutions of Tween 80 and Tergitol NP-15 were 6.10 and 5.85, respectively (for calculation details, see Pennell et al., 1997; Laha and Luthy, 1992). The Tergitol NP-15 value is consistent with existing log Kow–log Kmw correlations for alkylphenol ethoxylates, which yielded a log Kmw value of 5.7 (see Pennell et al., 1997).

6.4.5 Mathematical modeling To predict the effect of surfactant addition on the distribution of PCB congeners in a solid–liquid system, it is necessary to account for the potential impact of sorbed-phase and micellar surfactants on PCB phase distributions. Surfactant micelles will act to increase the amount of PCB in solution; however, sorbed-phase surfactant may also increase partitioning of the PCB to the solid phase. This effect will be a function of the surfactant critical micelle concentration (CMC), the soil sorption capacity for the surfactant (Sm), and the partitioning of the PCB congener among the aqueous, micellar, and solid phases. The overall or apparent solubility of a compound in the presence of surfactant can be represented as the amount of solute associated with

Chapter six: Enhancing PCB bioremediation

193

surfactant monomers plus the amount associated with surfactant micelles, which can be expressed as (Pennell et al., 2001; Sun et al., 1995) C o* = 1 + C s ,mn K mn + C s ,mc K mw Co

(6.3)

where Co* is the apparent solubility of organic species in the aqueous surfactant solution, Co is the solubility of organic species in water (micelle-free), Cs,mn is the concentration of surfactant monomers, Cs,mc is the concentration of surfactant micelles, Kmn is the organic species distribution coefficient between surfactant monomers and water, and Kmw is the organic species distribution coefficient between the micelles and water. This approach can be extended to include the effect of the sorbed-phase surfactant on the distribution of solute between the solid and aqueous phases:

K D* =

(

K D 1 + C s/om K s/om

(1 + C

o , mn

)

K mn + C o ,mc K mw

)

(6.4)

where KD* is the apparent soil–water distribution coefficient for the organic species, KD is the soil–water distribution coefficient in the absence of surfactant, Cs/om is the concentration of sorbed surfactant per unit mass of native soil organic matter, Ks/om is the solute distribution coefficient between sorbed-phase surfactant and the native organic matter (Ks/Kom), and Ks is the solute distribution coefficient between the sorbed-phase surfactant and water. Equation 6.4 was incorporated into a macro-based spreadsheet to predict the apparent distribution coefficient (KD*) as a function of surfactant concentration. Based on data obtained in laboratory experiments, the sensitivity of KD* to different values of Ks/om was investigated over a surfactant concentration range of 0 to 800 mg/l. An example of the effect of differences in the Ks/om value on the overall partitioning of PCB between the solid and liquid phases is given in Figure 6.23. Note that once the CMC of the surfactant was exceeded, the effect of surfactant sorption became minimal; that is, KD* decreases dramatically and approaches zero as the surfactant concentration is further increased. However, at low concentrations, partitioning of PCB congeners into the sorbed-phase surfactant may have a substantial impact on the distribution of PCBs within the system. The mathematical model was further adapted to account for rate-limited sorption and desorption of both the surfactant and the PCB congener(s) of interest.

6.4.6 PCB transformation experiments Microbial growth experiments confirmed the ability of biphenyl-degrading microorganisms to achieve substantial growth in solutions of Tween 80 and

194

Bioremediation of Recalcitrant Compounds 1600

K* (L/kg)

1400 1200

Ks/om = 0.5

1000

Ks/om = 1.0

800

Ks/om = 1.5

600

Ks/om = 2.0

400

Ks/om = 2.5

200 0 0

500 Tergitol NP-15 Concentration (mg/L)

1000

Figure 6.23 Effect of variations in values of Ks/om on the overall distribution coefficient.

Tergitol NP-15 that contained monochlorinated biphenyl (4-CBP). To address the fate of both the surfactants and the chlorinated biphenyls, a series of batch reactor experiments was performed to monitor the disappearance of 4-CBP from liquid batch cultures. The transformation experiments were conducted in 300-ml glass culture flasks that were sterilized and coated with 4-CBP. An aqueous nutrient solution (K1 media) or solution of surfactant, either Tergitol NP-15 or Tween 80, was added to the prepared flasks and allowed to mix at 150 rpm for 10 to 13 h at 30˚C. Resting cells of strain NY05, which had been rinsed three times to remove residual biphenyl, were suspended in K1 media at a biomass concentration of 0.2 g/ml. Samples were transferred from the flasks to 26-ml glass serum tubes, to which was added sulfuric acid (2 N H2SO4) to lower the pH to approximately 3.5 in order to halt microbial activity. Additional sample preparation included hexane extraction for treatments that did not contain surfactant solutions and centrifugation to remove biomass, after which the samples were transferred to glass autosampler vials for analysis by GC/ECD. In order to confirm that the disappearance of 4-CBP was microbially mediated, controls were included in the transformation experiments, in which flasks, both with and without surfactant solutions, were not inoculated with NY05 resting cells. These controls enabled detection of possible abiotic influences, such as volatilization and adsorption losses, which would appear as 4-CB degradation. Typical results of the abiotic controls, for solutions of both surfactants, are presented in Figure 6.24 with the respective surfactant solution. The transformation of 4-CBP by strain NY05 has been shown to result in the production of stoichiometric amounts of the corresponding chlorinated benzoic acid, 4-CBA. The detection of this degradation metabolite provided additional confirmation of the microbially mediated disappearance of 4-CB. The appearance of 4-CBA over time for a representative degradation

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0

6

Abiotic Control, 4CB (0.4 mM) NP (1720 ppm), 4CB (2.1 mM) NP (1133 ppm), 4CB (1.7 mM) NP (1720 ppm), 4CB (0.8 mM)

2 4 Time, hours

0.00

0.10

0.20

0.30

0.40

0.50

0 Abiotic Control, 4CB (0.5 mM) T80 (1273 ppm), 4CB (0.5 mM) T80 (1752 ppm), 4CB (0.5 mM) T80 (1752 ppm), 4CB (0.5 mM)

Time, hours

5

4-CB in Solutions of Tween 80

4-CB (mM)

4-CB (mM)

0

5 Time, hours Abiotic Control, 4CB (0.01 mM) 4CB (0.01)

0

0.002

0.004

0.006

0.008

0.01

4-CB in K1 Solution

Figure 6.24 Transformation of 4-CB by NY05 in solutions of Tergitol NP-15, Tween 80, and medium K1.

4-CB (mM)

0.35

4-CB in Solutions of Tergitol NP-15

Chapter six: Enhancing PCB bioremediation 195

196

Bioremediation of Recalcitrant Compounds Tween 80, 4680 ppm

Molar Concentration, mM

0.6 4-CB Disappearance

0.5

4-CBA Appearance

0.4 0.3 0.2

Co (4CB) = 1.4 mM Co (4CB) = 5.2 mM Co (4CB) = 5.8 mM

0.1 0.0 0

20

40

60

80 100 Time, hour

120

140

160

180

Figure 6.25 Disappearance of 4-CB and production of 4-CBA by NY05 in solution of Tween 80 (4680 mg/l).

experiment conducted in a solution of Tween 80 is presented in Figure 6.25. In this experiment, the flasks were inoculated before the system reached equilibrium (i.e., 4-CBP had not reached equilibrium solubility in the surfactant solution). As a result, a greater amount of 4-CBA was produced, on an equivalent molar basis, than was detected at time zero. These data indicate that the rate of 4-CB biotransformation to 4-CBA was faster than the rate of micellar solubilization of 4-CB into the micellar solution. Results from representative degradation experiments are presented in Figure 6.25 for surfactant and K1 solutions. Additional transformation experiments were performed to include a minimum mixing period of 3 weeks prior to inoculation. These extended mixing intervals allowed the 4-CB and surfactant systems to reach equilibrium before the microbial transformation was initiated. In addition, similar amounts of 4-CB mass were added to each system, both with and without surfactant solution, in order to facilitate comparison of the degradation efficiencies of the different systems. Results of the latter transformation experiments indicated that the presence of Tween 80 increased both the aqueous solubility and transformation of 4-CB when compared to systems absent of surfactant.

6.5 Accomplishments of the pilot evaluation 6.5.1 Site consideration for field test The ideal soil for the demonstration project should contain one of the less chlorinated Aroclors (1242 or 1248) at a concentration of 500 to 800 μg/g

Chapter six: Enhancing PCB bioremediation

197

and should be free of other contaminants, such as oils and heavy metals. Although more heavily chlorinated PCB mixtures, such as Aroclor 1260, could be treated, sequential inoculations would be necessary to obtain extensive enough dechlorination, and this would add to the time required for the anaerobic phase. Oils can be expected to slow both the anaerobic and aerobic steps by serving as a phase, to which the PCBs can partition, thus decreasing their availability to the microorganisms. The anaerobic step is relatively immune to heavy metals because under anaerobic conditions, they are precipitated as sulfides and thus not bioavailable. However, they may be toxic when conditions are made aerobic for the second stage treatment. Also, soil should have a moderate (2 to 3%) amount of organic carbon to provide substrates for the microorganisms. If the organic carbon level is much higher, sorption of PCBs to the organic matter may begin to limit PCB bioavailability, and growth of non-PCB-dechlorinating and -degrading microorganisms could be favored (both anaerobic and aerobic stages). A deficiency of organic carbon can easily be made up by the addition, preferably, of a simple carbon source, such as ethanol or trypticase soy broth, that does not bind PCBs. Many PCB-contaminated sites (Table 6.4) were evaluated. However, for various reasons these sites were found unsuitable for our field test. Finally, a GE site, located in Rome, GA, was found suitable for collecting the field samples. Several samples were analyzed for PCB profile and contamination levels (Table 6.5). The Aroclor 1242–contaminated soil from Rome, GA, met all of the criteria mentioned above and was used for the pilot-scale evaluation of the two-phase PCB bioremediation. The pilot-scale reactor treatment sequence involves an anaerobic phase, designed to dechlorinate the higher-chlorinated congeners, followed by an aerobic phase, to treat the accumulated lower-chlorinated congeners. To enhance the anaerobic treatment process, river sediment inoculum, ferrous sulfate, and surfactants were added to the reactors. During the aerobic treatment phase, genetically engineered microorganisms were added using a vermiculite carrier along with a nutrient/surfactant solution (Figure 6.26). Traditionally, low solids slurry reactors are used for creating the reduced environment for PCB dechlorination. In these reactors mixing is needed for maintaining the uniform moisture level throughout the system as well as for uniformly distributing the nutrient amendments. Achieving a well-mixed system is dependent on operational capacity of mixing systems, which in turn is dependent on the slurry concentration. Most of the traditional mixing systems can mix only slurries with consistencies up to 10 to 20% (solids:water, w/w). Once the treatment goals are achieved, the dewatering costs become prohibitive for the disposal/reuse of the solid material. To overcome these operational limitations, this study focuses on the economic feasibility and cost–benefit analysis of three bioremediation options with different solids loadings.

PA20241 PA20242 PA20243 PA20244

136 122 102 105

1260 1260 1260 1260

Picatinny Arsenal Site, Picatinny, NJ Sample Concentration PCB ID (μg/g) Aroclor RS1-1 RS2-1 RS3-1 RS4-1 RS5-1 RS6-1 RS7-1 RS8-1 RS9-1 RS10-1 RS11-1 RS12-1 0.7 0.7 0.4 0.5 0.2 0.3 0.8 1.8 0.5 0.1 0.5 0.6 1254 1254 1254 1254 1254 1254 1254 1254 1254 1254 1254 1254

Rosenberg Superfund Site, Cortland, NY Sample Concentration PCB ID (μg/g) Aroclor

Table 6.4 PCB Profile and Contamination Levels for Different Sites

1 2 3 4 5 6 7 8 9 10

4.6 10.6 84.4 72.4 69.0 35.6 29.7 22.6 13.9 29.5

1248 1248 1248 1248 1248 1248 1248 1248 1248 1248

Richardson Hill Superfund Site, Oneonta, NY Sample Concentration PCB ID (μg/g) Aroclor

198 Bioremediation of Recalcitrant Compounds

Chapter six: Enhancing PCB bioremediation

199

Table 6.5 PCB Profile and Contamination Level in Rome, GA, Soil Sample ID GT-1 GT-2 GT-3 GT-4 GT-5 GT-6 GT-7 GT-8 GT-9 a

Boreholea

Depth (ft)

1 1 1 1 1 2 2 2 1

1 1–1.5 1–1.5 1.5–2 1–1.5 2 1.5 1.5 2

Sample A Concentration (μg/g)

Sample B Concentration (μg/g)

Sample C Concentration (μg/g)

2412.9 653.1 505.3 852.2 807.6 155.2 1.6 158.7 353.9

765.6 597.7 526.6 687.9 435.9 154.0 114.0 208.1 100.9

590.5 940.6 254.7 798.5 577.3 294.4 13.6 214.4 201.8

Borehole 1 located approximately 0.5 ft from SB9B-5. Borehole 2 located approximately 1.5 ft from SB9B-5. The PCB contamination is predominantly Aroclor 1242.

Low Solids Reactor (8.3% solids, w/w) Msoil = 98 kg Mwater = 1087 kg

Medium Solids Reactor (53.9% solids, w/w) Msoil = 240 kg Mwater = 205 kg

High Solids Reactor (72.6% solids, w/w) Msoil = 215 kg Mwater = 82 kg

Water FeSO4 (20 mM) Ethanol (0.1% wt) Tween 80 (5 g/L)

Anaerobic Phase 6–8 Months

Anaerobic Inoculum Hudson River, NY (1.0% wt)

Tween + FeSO4 + EtOH

FeSO4 + EtOH

Control

Water Ammonium Phosphate

Aerobic Phase 2–4 Months Continuous Mixing

GEMs: LB400 + RHA1 Vermiculite Carrier (3.0% wt)

End Pilot Study Final Sample Analysis 8–10 Month Total Treatment

Figure 6.26 Flowchart illustrating process streams for pilot-scale biotreatment of Rome, GA, soil.

200

Bioremediation of Recalcitrant Compounds The main objectives of this pilot-scale tests were to: • Validate the usefulness of the two-phase bioremediation scheme that combines enhanced anaerobic dechlorination of Aroclor and genetically engineered aerobic PCB-growing microorganisms (GEMs) for bioremediating PCB-contaminated soils • Evaluate the effects of different solids loading rates (or moisture contents) on the bioremediation of PCBs and application of GEMs • Develop critical design information on the application of GEMs for full-scale bioremediation of PCB-contaminated soils

A pilot-scale demonstration study consisting of side-by-side comparisons of three bioremediation processes was performed at the Environmental Laboratory, U.S. Army ERDC, Vicksburg, MS. Three parallel bioremediation processes with different solids loading rates were used to evaluate the effects of solids loading on soil mixing, aeration, addition, and uniform distribution of nutrient amendments/GEMs, and final disposal/use of treated material. Each bioremediation process has three separate reactors (treatments) to evaluate the effects of ferrous sulfate and surfactants on PCB dechlorination (Table 6.6).

6.5.2 Site description PCB-contaminated soil was collected from the GE site in Rome, GA. The PCB concentration in this soil ranges from 500 to 1000 ppm (Table 6.5). The soil is clayey with little sand (87.4% fines, 11.6% sand, and 1% gravel). The soil has a liquid limit (LL) of 57 and plastic limit (PL) of 23. Chemical characteristics of the Rome, GA, soil are summarized in Table 6.7. Soil was collected from 6 ft below the ground surface. After excavating, the soil was cleaned of any pebbles/debris that was more than 1 in. in size. The collected soil was sieved through 1 in. mesh screen, thoroughly mixed, and homogenized at the site. About 2240 kg of homogenized soil was shipped to ERDC in 10 55-gallon drums. The soil was used to load the triplicate treatment reactors for each bioremediation process and for the initial physical and chemical characterizations of the collected soil. Table 6.6 Experimental Design Reactor

No. of Treatments

Low solids

3

Medium solids

3

High solids

3

Treatments Surfactant + iron + Iron + ethanol No amendments ethanol Surfactant + iron + Iron + ethanol No amendments ethanol Surfactant + iron + Iron + ethanol No amendments ethanol

Chapter six: Enhancing PCB bioremediation

201

Table 6.7 Rome, GA, Soil Chemical Characterization Nutrients TKN TP TOC

a

Concentration (mg/g) 0.51 0.58 6.27

Metals B As Cd Cr Co Cu Pb Mn Mo Ni Se Zn K Na Mg Ca Fe Al

Concentrationa (μg/g) 21.27 26.99 6.17 90.25 15.97 20.70 47.77 427.67 6.87 70.44 457.47 287.37 281.65 8.16 30.82 261.84 1.79 1057.46

Metal concentration is on a dry basis.

6.5.3 Pilot-scale demonstration The reactor setup consists of three reactors (treatments) for each solids loading under three bioremediation options. The reactors used for three different solids loadings are illustrated in Figures 6.27 through 6.29. The low solids reactor (Figure 6.27) consisted of a stainless steel 430-gallon tank with dimensions of 56 in. high and 47∫ in. internal diameter. The reactor has a cone-shaped bottom for convenient unloading. The reactor was equipped with a variable-speed agitator fitted with two propellers for mixing and aerating the slurry. The reactor was similar to the traditional bioslurry reactor commonly used for the ex situ treatment of contaminated soils and sediments. This reactor uses excessive water (90 to 95%) to create anaerobic conditions, to maximize the distribution of microbial amendments, and to enhance the mass transfer. Inherently, the treatment times are considerably shorter in such types of systems; however, the postremediation handling of excessive water generated from dewatering the treated material is a major disadvantage of these systems. The medium solids reactor shown in Figure 6.28 is a state-of-the-art shaftless screw reactor capable of mixing soil slurries with solids content up to 70% w/w. The 140-gallon medium solids reactor has a footprint of 16.5 ft2. The reactor was fitted with three shaftless screws to mix and convey the high solids slurries. The medium solids reactor is a novel idea in bulk material mixing and conveying. The shaftless screws are 1.93 m in length

202

Bioremediation of Recalcitrant Compounds

Figure 6.27 Low solids reactor 1.

and 12.7 cm in diameter. The screws rotate at variable speed in U-shaped troughs constructed within a rectangular metal housing measuring 1.93 m × 0.79 m × 0.56 m. The screws simply slide on the wear- and chemical-resistant lining. Each screw is operated by a 0.75-hp motor equipped with a gear control having a gear ratio of 40. Electrical inverter controllers (Baldor 15H) regulate individual screw speed and rotation. The side troughs are inclined, making a 4˚ angle with the bottom of the reactor. The middle trough is also inclined at the same angle but in the opposite direction of the side troughs. The middle screw rotates at twice the speed of the side screws in order to transport the material back to the lower end of the side screws for continuous mixing. The shaftless screw reactor significantly reduces the dewatering costs associated with low solids slurries. The shaftless screw reactor enhances the treatment efficiency by reducing the mass transfer limitations in high solids slurries and increases the soil microbial activity by facilitating the addition of nutrient amendments in high solids slurries, thereby reducing the treatment time and cost. The high solids reactor shown in Figure 6.29 simulates the contaminant treatment in in situ land farming. The reactor consists of a galvanized

Chapter six: Enhancing PCB bioremediation

203

Figure 6.28 Medium solids reactor 1.

oval-shaped iron pan with an open top. The top is covered with an airtight Plexiglas lid. The total volume of this reactor is 360 L, with dimensions of 122 m × 0.61 m × 0.61 m. Contaminated soil at field moisture content was treated in this high solids reactor. The reactor was equipped with a leachate collection system at the bottom, consisting of a 10 cm gravel layer covered with a geomembrane. A perforated aluminum sheet covered the geomembrane for uniform distribution of soil in the reactor. The process needs minimal maintenance; however, the treatment efficiency is low, with prolonged treatment times. The major advantage of this bioremediation option is that the treated material does not require any postremediation dewatering for disposal/reuse. During the aerobic phase, soil was aerated by tilling/ mixing the soil with a low-power rotary cultivator. Each treatment system underwent a two-phase cycle: an anaerobic dechlorination phase followed by an aerobic degradation phase. The anaerobic phase trigger was 6 to 8 months or a predetermined percent dechlorination, whichever comes first. The aerobic phase followed the anaerobic phase and lasted for about 2 to 4 months. All three reactors in each bioremediation process were seeded with anaerobic inoculum obtained from Hudson River sediment. Anaerobic inoculum was applied at 0.8% w/w (dry solid:dry soil). In two of the reactors for each bioremediation process (Table 6.9), iron (20 mM FeSO4 in pore water)

204

Bioremediation of Recalcitrant Compounds

Figure 6.29 High solids reactor.

and organic carbon (0.1% w/w; ethanol:slurry) were added along with the anaerobic inoculum. Surfactant (0.5% w/w; Tween 80:pore water) was added in one of the reactors for each bioremediation process. The quantities of soil, water, inoculum, and nutrient amendments in each of the reactors for the three bioremediation processes are given in Table 6.8. After completion of the anaerobic phase, the reactors were inoculated with GEMs (LB400(ohb) and RHAI(fcb)). Vermiculite impregnated with GEMs were applied at the rate of 3% (w/w; vermiculite:dry solids) of contaminated soil in each reactor. Vermiculite was used because of its beneficial effect on the long-term survival of the GEMs and to improve distribution of the inoculum during mixing.

6.5.4 Sampling schedule The sampling analysis schedule for different parameters is given in Table 6.9. The material in each reactor was analyzed at the beginning for PCB concentration, microbial biomass phospholipid fatty acid (PFLA), moisture content, pH, temperature, nutrient levels (sodium, potassium, nitrogen

215 215 215

High solids reactor 1 2 3

82 82 82

205 205 205

1087 1087 1087

Water (kg)

72.58 72.58 72.58

53.90 53.90 53.90

8.27 8.27 8.27

Solids Content (%)

b

a

Soil mass is on a dry basis. Ferrous sulfate was applied as FeSO4·7H2O. c Surfactant was applied as Tween 80. d Inoculum was applied as wet Hudson River sediment.

240 240 240

98 98 98

Medium solids reactor 1 2 3

Low solids reactor 1 2 3

Soila (kg)

Table 6.8 Reactor Loading Conditions

0.45 0.45 0.00

1.15 1.15 0.00

6.05 6.05 0.00

Ferrous Sulfateb (kg)

0.31 0.31 0.00

0.47 0.47 0.00

1.25 1.25 0.00

Ethanol (kg)

0.41 0.00 0.00

1.03 0.00 0.00

5.43 0.00 0.00

Surfactantc (kg)

3.91 3.91 3.91

4.33 4.33 4.33

1.78 1.78 1.78

Inoculumd (kg)

Chapter six: Enhancing PCB bioremediation 205

206

Bioremediation of Recalcitrant Compounds

(NKN), total phosphorous (TP), and total organic carbon (TOC)), and headspace oxygen concentration. Additionally, initial characterization included particle size distribution (PSD), Atterberg properties, metals concentration, particle surface area, and soil leachability. These characteristics were analyzed again at the end of the experiment. Routine soil/slurry samples were collected every month for PCB concentration, microbial biomass, nutrient analysis, soil moisture, and pH. Soil/slurry temperature and headspace oxygen concentration were monitored biweekly to check anaerobic conditions in the reactors. To ensure quality control of analytical results, samples were analyzed in composite replicates of seven for high solids, five for medium solids, and three for low solids reactors. Higher replicates were used in high solids reactors because of the level of heterogeneity and variability, compared to the well-mixed low solids reactor.

6.5.5 Analytical methods Particle size distribution (PSD) was measured by using a Coulter LS 100Q particle counter according to instrument specifications. The Atterberg limits test was performed according to the Corps of Engineers Laboratory Testing Manual, EM-1110-2-1906, “Appendix III: Liquid and Plastic Limits.” Bulk density of the contaminated soil was measured by weighing a known volume of soil at field conditions. Soil pH was determined by a soil-distilled water slurry (1:10, w/w) using a Cole-Parmer® pH meter and was reported as pH in water (pHw). PCB concentrations in soil samples were determined by high-resolution gas chromatography with electron capture detection (ECD). Total Kjeldahl nitrogen (TKN) and total phosphorous (TP) analyses were performed by a Lachat 8000 Flow Injection Analyzer (FIA). The preparation methods were modified versions of EPA-600/4-79-020 (1983 revision) 365.1 and 351.2, respectively. Total organic carbon (TOC) was determined by the Zellweger Analytic TOC analyzer, according to the instrument protocol. Moisture content of soil samples was analyzed by oven drying at 105˚C for 24 h. Oxygen concentration was measured in the headspace by using an LMSx Multigas Analyzer®. Surfactant analysis for Tween 80 concentrations in aqueous samples was done by a Hewlett Packard 1100 series. Detection of Tween 80 was achieved using a diode array selection HPLC controlled by a HP Vectro Pentium computer. Soil leachability was assessed by two alternative leachability methods. The Sequential Batch Leaching Test (SBLT) consists of four repeat extractions of the same sample using deionized water in a 4:1 (water:solid) ratio. The slurry was mixed for 24 h, centrifuged, and filtered, and the filtrate was analyzed for contaminants. The Synthetic Precipitate Leaching Procedure (SPLP) was performed according to SW846, EPA Method 1312. Being a nonvolatile extraction, a “bottle extraction vessel” such as an amber jar (1 l) with sufficient capacity to hold the sample and the extraction fluid will be

a

X X X X X X X X

X X X X X X X X X X X X X X X X X X

X X

X X X X X X X X X X

X X

X X X X X X X X X X

X X

X X X X X X X X X X

X X

X X X X X X X X X X

X X

X X X X X X X X X X

X X

X X

X X X X X X X X

X X

X X

X X X X X X X X

X X X X X X X X

t=1 15 t = 2 t = 3 t = 4 t = 5 t = 6 t = 7 t = 8 t = 9 t = 10 t = 11 t = 12 t = 13a t = 14 t = 15 t = 16 15 30 45 60 75 90 105 120 135 150 165 180 195 210 225 240

t =13 is the tentative start of the aerobic phase.

PSD Atterberg limits Bulk density Surface area Leachability Metals concentration Temperature Oxygen concentration Moisture content pH PCB concentration Microbial biomass TOC TKN TP Surfactant

Day

t=0 0

Table 6.9 Sample Analysis Schedule

Chapter six: Enhancing PCB bioremediation 207

208

Bioremediation of Recalcitrant Compounds

used. The method consists of a single extraction using a dilute acid solution. Extraction fluid used to determine the leachability of soil from a site that is in the east of the Mississippi River consists of a 60/40 weight percent mixture of sulfuric and nitric acids diluted with reagent water to a pH of 4.20 ± 0.05.

6.6 Conclusions Bioremediation can potentially result in dechlorination of PCBs and possibly even in mineralization of the contaminant. Energy costs are lower than other forms of treatment. Slurry phase treatment usually requires less time than solid phase biological treatment due to the increased rates of contaminant mass transfer. Furthermore, it is relatively simple to maintain either aerobic or anaerobic conditions in the reactor and to switch between these two conditions. As noted previously, this characteristic process is especially beneficial when treating PCBs. Anaerobic conditions can be maintained to allow for reductive dechlorination of the highly chlorinated congeners. After this stage is complete, aerobic conditions can be established to allow the resulting lightly chlorinated congeners to be metabolized aerobically. Principal drawbacks to preexisting bioremediation schemes are slow rates of the anaerobic Aroclor dechlorination in sediments and the inability of naturally occurring biphenyl-degrading organisms to dechlorinate and grow on lower-chlorinated PCB congeners resulting from the anaerobic phase. We have addressed these principal barriers and developed processes that resolve the above problems via enhancing rates of anaerobic dechlorination and allowing aerobes to grow on the products from anaerobic Aroclor dechlorination. Although results of the currently ongoing pilot test are yet to be determined, flask- and laboratory-scale soil remediation experiments indicate that the designed two-phase enhanced anaerobic dechlorination of Aroclor coupled with GEM-based enhanced aerobic degradation/mineralization of lower-chlorinated PCBs could be very beneficial as a remediation technology.

6.7 Recommendations for further transitional research The results of our research indicated that the developed bioremediation scheme could be beneficial in future technologies. However, much more research is needed to resolve remaining questions as well as improve current results by gaining knowledge on how the degradative processes affect and are being affected within the given metabolic networks in microbial communities. Recent developments in genomic sequencing and DNA microarray technology create opportunities for transferring these approaches to community and populational studies. The latter would allow for rapid advances in our understanding of the complex processes involved in successful biodegradation within a given microbial population, naturally occurring or engineered by the introduction of foreign members. Success in this direction

Chapter six: Enhancing PCB bioremediation

209

will determine the practicality of the emerging recombinant technologies for remediation of environmental hazards. The success of biodegradation may also be hampered by low bioavailability of the contaminant, mass transfer limitations of the electron acceptor, and low temperature. In addition, most feed soils will require some degree of size reduction, usually to a greater extent than that required with incineration. Particles must be small enough that they can be suspended by the reactor’s agitators, and gravel may jam agitator blades. Soils with large amounts of oily or greasy waste have been found to be problematic in slurry treatments. Because these wastes are hydrophobic, dispersants must be used to keep agglomerations of the waste from forming. Even so, the oils are often in large droplets with fairly low surface area with respect to their volume, so that microbial attack is hindered. In addition, the oils will adhere to the sides of the reactors and to the agitation equipment, often blocking air nozzles. Further physicochemical and engineering research is needed to evaluate and improve or better design technological processes.

References Abramowicz, D.A. 1990. Aerobic and anaerobic biodegradation of PCBs: a review. Crit. Rev. Biotechnol. 10: 241–248. Ahel, M., Giger, W., and Schaffner, C. 1994. Behavior of alkylphenol polyethoxylate surfactants in the aquatic environment. 2. Occurrence and transformation in rivers. Water Res. 28: 1143–1152. Ahmed, M. and Focht, D.D. 1973. Degradation of polychlorinated biphenyls by two species of Achromobacter. Can. J. Microbiol. 19: 47–52. Arnett, C.M., Parales, J.V., and Haddock, J.D. 2000. Influence of chlorine substituents on rates of oxidation of chlorinated biphenyls by the biphenyl dioxygenase of Burkholderia sp. strain LB400. Appl. Environ. Microbiol. 66: 2928–2933. ATSDR. 2000. Toxicological Profile of Polychlorinated Biphenyls (PCBs). U.S. Department of Health and Human Services, Public Health Services, Agency for Toxic Substances and Disease Registry, Atlanta, GA. Barton, M.R. and Crawford, R.L. 1988. Novel biotransformations of 4-chlorobiphenyl by a Pseudomonas sp. Appl. Environ. Microbiol. 54: 594–595. Bassler, H.I., Flood, S.J.A., Livak, K.J., Marmaro, J., Knorr, R., and Batt, C.A. 1995. Use of fluorogenic probe in a PCR-based assay for the detection of Listeria monocytogenes. Appl. Environ. Microbiol. 61: 3724–3728. Bedard, D.L., Haberl, M.L., May, R.J., and Brennan, M.J. 1987a. Evidence for novel mechanisms of polychlorinated biphenyl metabolism in Alcaligenes eutrophus H850. Appl. Environ. Microbiol. 53: 1103–1112. Bedard, D.L. and Quensen, J.F., III. 1995. Microbial reductive dechlorination of polychlorinated biphenyls. In Microbial Transformation and Degradation of Toxic Organic Chemicals, L.Y. Young and C. Cerniglia, Eds. Wiley-Liss Division, John Wiley & Sons, New York, 127–216. Bedard, D.L., Unterman, R., Bopp, L.H., Brennan, M.J., Haberl, M.L., and Johnson, C. 1986. Rapid assay for screening and characterizing microorganisms for the ability to degrade polychlorinated biphenyls. Appl. Environ. Microbiol. 51: 761–768.

210

Bioremediation of Recalcitrant Compounds

Bedard, D.L., Wagner, R.E., Brennan, M.J., Haberl, M.L., and Brown, J.F., Jr. 1987b. Extensive degradation of Aroclors and environmentally transformed polychlorinated biphenyls by Alcaligenes eutrophus H850. Appl. Environ. Microbiol. 53: 1094–1102. Bergmann, J.G. and Sanik, Jr., J. 1957. Determination of trace amounts of chloride in naphtha. Anal. Chem. 29: 241–243. Bloem, J. 1995. Fluorescent staining of microbes for total direct counts. In A.D.L. Akkermans, Ed., Molecular Microbial Ecology Manual. Kluwer Academic Publishers, Dordrecht, The Netherlands, p. 4.1.8:1–4.1.8:12. Bloemberg, G.V., O’Toole, G.A., Lugtenberg, B.J.J., and Kolter, R. 1997. Green fluorescent protein as a marker for Pseudomonas spp. Appl. Environ. Microbiol. 63: 4543–4551. Brenner, V., Arensdorf, J.J., and Focht, D.D. 1994. Genetic construction of PCB degraders. Biodegradation 5: 359–377. Burlage, R.S., Sayler, G.S., and Larimer, F. 1990. Monitoring of naphthalene catabolism by bioluminescence with nah-lus transcriptional fusions. J. Bacteriol. 172: 4749–4757. DeLong, E.F., Wickham, G.S., and Pace, N.R. 1989. Phylogenetic stains: ribosomal RNA-based probes for the identification of single cells. Science 243: 1360–1363. Desjardin, L.E., Chen, Y., Perkins, M.D., Teixeira, L., Cave, M.D., and Eisennach, K.D. 1998. Comparison of the ABI 7700 system (TaqMan) and competitive PCR for quantification of IS6110 DNA in sputum during treatment of tuberculosis. J. Clin. Microbiol. 36: 1964–1968. Dölken, L., Schüler, F., and Dölken, G. 1998. Quantitative detection of t(14;18)-positive cell by real-time quantitative PCR using fluorogenic probes. Biotechniques 6: 1058–1064. Furukawa, K., Tomizuka, N., and Kamibayashi, A. 1979a. Effect of chlorine substitution on the bacterial metabolism of various polychlorinated biphenyls. Appl. Environ. Microbiol. 38: 301–310. Furukawa, K., Tonomura, K., and Kamibayashi, A. 1978. Effect of chlorine substitution on the biodegradability of polychlorinated biphenyls. Appl. Environ. Microbiol. 35: 223–227. Furukawa, K., Tonomura, K., and Kamibayashi, A. 1979b. Metabolism of 2,4-4′-trichlorobiphenyl by Acinetobacter sp. P6. Agric. Biol. Chem. 43: 1577–1583. Ghosal, D. and You, I.S. 1989. Operon structure and nucleotide homology of the chlorocatechol oxidation genes of plasmids pJP4 and pAC27. Gene 83: 225–232. Grüntzig, V., Nold, S.C., Zhou, J., and Tiedje, J.M. 2001. Pseudomonas stutzeri nitrite reductase gene abundance in environmental samples measured by real-time PCR. Appl. Environ. Microbiol. 67: 760–768. Haddock, J.D., Horton, J.R., and Gibson, D.T. 1995. Dihydroxylation and dechlorination of chlorinated biphenyls by purified biphenyl 2,3-dioxygenase from Pseudomonas sp. strain LB400. J. Bacteriol. 177: 20–26. Havel, J. and Reineke, W. 1991. Total degradation of various chlorobiphenyls by cocultures and in vivo constructed hybrid pseudomonads. FEMS Microbiol. Lett. 78: 163–170. Heid, C.A., Stevens, J., Livak, K.J., and Willians, P.M. 1996. Real time quantitative PCR. Genome Res. 6: 986–994.

Chapter six: Enhancing PCB bioremediation

211

Hickey, W.J., Brenner, V., and Focht, D.D. 1992. Mineralization of 2-chloro- and 2,5-dichlorobiphenyl by Pseudomonas sp. strain UCR2. FEMS Microbiol. Lett. 98: 175–180. Higuchi, R., Fockler, C., Dollinger, G., and Watson, R. 1993. Kinetic PCR analysis: real-time monitoring of DNA amplification reactions. Biotechnology 11: 1026–1030. Higgins, J.A., Ezzell, J., Hinnebusch, B.J., Shipley, M., Henchal, E.A., and Ibrahim, M.S. 1998. 5′ nuclease PCR assay to detect Yersinia pestis. J. Clin. Microbiol. 36: 2284–2288. Hofer, B., Eltis, L.D., Dowling, D.N., and Timmis, K.N. 1993. Genetic analysis of a Pseudomonas locus encoding a pathway for biphenyl/polychlorinated biphenyl degradation. Gene 130: 47–55. Hrywna, Y., Tsoi, T.V., Maltseva, O.V., Quensen, J.F., III, and Tiedje, J.M. 1999. Construction and characterization of two recombinant bacteria that grow on ortho- and para-substituted chlorobiphenyls. Appl. Environ. Microbiol. 65: 2163–2169. Hutzinger, O., Safe, S., and Zitko, V. 1974. The Chemistry of PCBs. Robert E. Krieger Publishing Company, Malabar, FL. Hutzinger, O. and Veerkamp, W. 1981. Xenobiotic chemicals with pollution potential. In Microbial Degradation of Xenobiotics and Recalcitrant Compounds, T. Leisinger, R. Hutter, A.M. Cook, and J. Neusch, Eds. Academic Press, New York, pp. 3–45. Hwang, I. and Farrand, S.K. 1997. Detection and enumeration of a tagged Pseudomonas fluorescens strain by using soil with markers associated with an engineered catabolic pathway. Appl. Environ. Microbiol. 63: 602–608. Iben, I.E., Edelstein, W.A., Sheldon, R.B., Shapiro, A.P., Uzgiris, E.E., Scatena, C.R., Blaha, S.R., Silverstein, W.B., Brown, G.R., Stegemeier, G.L., and Vinegar, H.J. 1996. Thermal blanket for in-situ remediation of surficial contamination: a pilot test. Environ. Sci. Technol. 30: 3144–3154. Laha, S. and Luthy, R.G. 1991. Inhibition of phenanthrene mineralization by nonionic surfactants in soil-water systems. Environ. Sci. Technol. 25: 1920–1930. Lee, D.-H., Zo, Y.-G., and Kim, S.-J. 1996. Nonradioactive method to study genetic profiles of natural bacterial communities by PCR-single-strand-conformation polymorphism. Appl. Environ. Microbiol. 62: 3112–3120. Lenski, R.E. 1991. Quantifying fitness and gene stability in microorganisms. In Assessing Ecological Risks of Biotechnology, L.R. Ginzburg, Ed., Butterworth-Heinemann, Boston, MA, pp. 173–192. Lenski, R.E., Simpson, S.C., and Nguyen, T.T. 1994. Genetic analysis of a plasmid-encoded, host genotype-specific enhancement of bacterial fitness. J. Bacteriol. 176: 3140–3147. Löffler, F.E., Sun, Q., Li, J., and Tiedje, J.M. 2000. 16S rRNA gene-based detection of tetrachloroethene-dechlorinating Desulforomonas and Dehalococcoides species. Appl. Environ. Microbiol. 66: 1369–1374. Ludwig, J.P., Giesy, J.P., Summer, C.L., Bowerman, W., Aulerich, R., Bursian, S., Auman, H.J., Jones, P.D., Williams, L.L., Yillitt, D.E., and Gilbertson, M. 1993. A comparison of water quality criterion for the Great Lakes based on human and wildlife health. J. Great Lakes Res. 19: 789–807. Maltseva, O.V., Tsoi, T.V., Quensen, J.F., III, Fukuda, M., and Tiedje, J.M. 1999. Degradation of anaerobic reductive dechlorination products of Aroclor 1242 by four aerobic bacteria. Biodegradation 10: 363–371.

212

Bioremediation of Recalcitrant Compounds

Masai, E., Sugiyama, K., Iwashita, N., Shimizu, S., Hauschild, J.E., Hatta, T., Kimbara, K., Yano, K., and Fukuda, M. 1997. The bphDEF meta-cleavage pathway genes involved in biphenyl/polychlorobiphenyl degradation are located on a linear plasmid and separated from the initial bphACB genes in Rhodococcus sp. strain RHA1. Gene 187: 141–149. Masse, R., Messier, F., Ayotte, C., Levesque, M.-F., and Sylvestre, M. 1989. A comprehensive gas chromatographic/mass spectrometric analysis of 4-chlorobiphenyl bacterial degradation products. Biomed. Environ. Mass Spectrom. 18: 27–47. McCullar, M.V., Brenner, V., Adams, R.H., and Focht, D.D. 1994. Construction of a novel polychlorinated biphenyl-degrading bacterium: utilization of 3,4′-dichlorobiphenyl by Pseudomonas acidovorans M3GY. Appl. Environ. Microbiol. 60: 3833–3839. Muyzer, G., Brinkhoff, T., Nübel, U., Santegoeds, C., Schäffer, H., and Wawer, C. 1998. Denaturing gradient gel electrophoresis (DGGE) microbial ecology. In Molecular Microbial Ecology Manual, 3.4.4, 2nd ed., A.D.L. Akkermans, J.D.V. Elsas, and F.J.D. Bruijn, Eds. Kluwer, Dordrecht, Netherlands, pp. 1–27. Muyzer, G., de Waal, E.C., and Uitterlinden, A.G. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59: 695–700. Nüsslein, K. and Tiedje, J.M. 1998. Characterization of the dominant and rare members of a young Hawaiian soil bacterial community with small-subunit ribosomal DNA amplified from DNA fractionated on the basis of its guanine and cytosine composition. Appl. Environ. Microbiol. 64: 1283–1289. Omori, T., Sigimura, K., and Minoda, Y. 1986. Purification and some properties of a 2-hydroxy-6-oxo-6-phenyl-2,4-dienoic acid hydrolyzing enzyme from Pseudomonas criciviae S93B1 involved in the degradation of biphenyl. Agric. Biol. Chem. 50: 1513–1518. Pahl, A., Kühlbrandt, U., Brune, K., Röllinghoff, M., and Gressner, A. 1999. Quantitative detection of Borrelia burgdorferi by real-time PCR. J. Clin. Microbiol. 37: 1958–1963. Pellizari, V.H., Bezborodnikov, S., Quensen, J.F., III, and Tiedje, J.M. 1996. Evaluation of strains isolated by growth on naphthalene and biphenyl for hybridization of genes to dioxygenase probes and polychlorinated biphenyl-degrading ability. Appl. Environ. Microbiol. 62: 2053–2058. Pennell, K.D., Adinolfi, A.M., Abriola, L.M., and Diallo, M.S. 1997. Solubilization of dodecane, tetrachloroethylene and 1,2-dichlorobenzene in micellar solutions of ethoxylated nonionic surfactants. Environ. Sci. Technol. 31: 1382–1389. Pennell, K.D., Pavlostathis, S.G., Karagunduz, A., and Yeh, D.H. 2001. Influence of nonionic surfactants on the bioavailability of hexachlorobenzene to microbial reductive dechlorination. In Chemicals in the Environment: Fate, Impacts and Remediation, R.L. Lipnik, R.P. Mason, M.L. Phillips, and C.U. Pittman Jr., Eds. American Chemical Society, Washington, DC, chap. 27, pp. 449–466. Perkin Elmer Applied Biosystems. 1998. Taqman® Universal PCR Master Mix Protocol, P/N 4304449. Perkin Elmer Applied Biosystems, Foster City, CA. Plotnikova, E.G., Tsoi, T.V., Grischenkov, V.G., Zaitsev, G.M., Nagaeva, M.V., and Boronin, A.M. 1991. Cloning of the Arthrobacter globiformis fcbA gene for dehalogenase and construction of a hybrid pathway of 4-chlorobenzoic acid degradation in Pseudomonas putida. Genetika 27: 589–597 (in Russian).

Chapter six: Enhancing PCB bioremediation

213

Quensen, J.F., III, Boyd, S.A., and Tiedje, J.M. 1990b. Dechlorination of four commercial polychlorinated biphenyl mixtures (Aroclors) by anaerobic microorganisms from sediments. Appl. Environ. Microbiol. 56: 2360–2369. Quensen, J.F., III, Ye, D., Griffith, G.D., Tiedje, J.M., and Boyd, S.A. 1990a. Reductive dechlorination of Aroclors by anaerobic microorganisms. In General Electric Company Research and Development Program for the Destruction of PCBs, Ninth Progress Report, H.L. Finkbeiner and S.B. Hamilton, Eds. General Electric Company Corporate Research and Development, Schenectady, NY, pp. 15–24. Recorbet, G., Givaudan, A., Steinberg, C., Bally, R., Normand, P., and Faurie, G. 1992. Tn5 to assess soil fate of genetically marked bacteria: screening for aminoglycoside-resistance advantage and labelling specificity. FEMS Microbiol. Ecol. 86: 187–194. Rodrigues, J.L.M., Aiello, M.R., Urbance, J.W., Tsoi, T.V., and Tiedje, J.M. 2002. Use of both 16S rRNA and engineered functional genes with real-time PCR to quantify an engineered, PCB-degrading Rhodococcus in soil. J. Microbiol. Meth. 51: 181–189. Rodrigues, J.L.M., Maltseva, O.V., Tsoi, T.V., Helton, R.R., Quensen, J.F., III, Fukuda, M., and Tiedje, J.M. 2001. Development of a Rhodococcus recombinant strain for degradation of products from anaerobic dechlorination of PCBs. Environ. Sci. Technol. 35: 663–668. Rosen, M.J. 1992. Surfactants and Interfacial Phenomena, 2nd ed. John Wiley & Sons, New York. Rouse, J.D., Sabatini, D.A., Suflita, J.M., and Harwell, J.H. 1994. Influence of surfactants on microbial degradation of organic compounds. CRC Crit. Rev. Environ. Sci. Technol. 24: 325–370. Schulz, D.E., Petrick, G., and Duinker, J.C. 1989. Complete characterization of polychlorinated biphenyl congeners in commercial Aroclor and Clophen mixtures by multidimensional gas chromatography. Environ. Sci. Technol. 23: 852–859. Schwieger, F. and Tebbe, C.C. 1998. A new approach to utilize PCR-single-strand-conformation polymorphism for 16S rRNA gene-based microbial community analysis. Appl. Environ. Microbiol. 35: 4870–4876. Seeger, M., Timmis, K.N., and Hofer, B. 1995. Conversion of chlorobiphenyls into phenylhexadienoates and benzoates by the enzymes of the upper pathway for polychlorobiphenyl degradation encoded by the bph locus of Pseudomonas sp. strain LB400. Appl. Environ. Microbiol. 61: 2654–2658. Seto, M., Kimbara, K., Shimura, M., Hatta, T., Fukuda, M., and Yano, K. 1995. A novel transformation of polychlorinated biphenyls by Rhodococcus sp. strain RHA1. Appl. Environ. Microbiol. 61: 3353–3358. Seto, M., Okita, N., Sugiyama, K., Masai, E., and Fukuda, M. 1996. Growth inhibition of Rhodococcus sp. strain RHA1 in the course of PCB transformation. Biotechnol. Lett. 18: 1193–1198. Smith, M.S. and Tiedje, J.M. 1980. Growth and survival of antibiotic-resistant denitrifier strains in soil. Can. J. Microbiol. 26: 854–856. Stackebrandt, E. and Rainey, F.A. 1995. Partial and complete 16S rDNA sequences, their use in generation of 16S rDNA phylogenetic trees and their implications in molecular ecological studies. In Molecular Microbial Ecology Manual, 3.1.1, 2nd ed., A.D.L. Akkermans, J.D.V. Elsas, and F.J.D. Bruijn, Eds. Kluwer, Dordrecht, Netherlands, pp. 1–17.

214

Bioremediation of Recalcitrant Compounds

Sun, S., Inskeep, W.P., and Boyd, S.A. 1995. Sorption of nonionic organic compounds in soil-water systems containing micelle forming surfactant. Environ. Sci. Technol. 29: 903–913. Tonso, N.L. 1997. Genome conformation and genetic diversity: a closer look at several species of Rhodococcus isolated from contaminated soils. In Microbial Diversity and Genetics of Biodegradation, K. Horikoshi, M. Fukuda, and T. Kudo, Eds. Japan Scientific Societies Press, Tokyo, pp. 75–80. Tsoi, T.V., Plotnikova, E.G., Cole, J.R., Guerin, W.F., Bagdasarian, M., and Tiedje, J.M. 1999. Cloning, expression and nucleotide sequence of the Pseudomonas aeruginosa strain 142 ohb genes coding for oxygenolytic ortho-dehalogenation of halobenzoates. Appl. Environ. Microbiol. 65: 2151–2162. Tsoi, T.V., Zaitsev, G.M., Plotnikova, E.G., Kosheleva, I.A., and Boronin, A.M. 1991. Cloning and expression of the Arthrobacter globiformis KZT1 fcbA gene encoding dehlaogenase (4-chlorobenzoate-4-hydroxylase) in Escherichia coli. FEMS Microbiol. Lett. 81: 165–170. Tsomides, H.J., Hughes, J.B., Thomas, J.M., and Ward, C.H. 1995. Effect of surfactant addition on phenanthrene biodegradation in sediments. Environ. Toxicol. Chem. 14: 953–959. Unterman, R. 1996. A history of PCB biodegradation. In Bioremediation: Principles and Applications, R.L. Crawford and D.L. Crawford, Eds. Cambridge University Press, Cambridge, U.K., pp. 209–253. Warhust, A.W. and Fewson, C.A. 1994. Biotransformations catalyzed by the genus Rhodococcus. Crit. Rev. Biotechnol. 14: 29–73. Yeh, D.H., Pennell, K.D., and Pavlostathis, S.G. 1998. Toxicity and biodegradability screening of nonionic surfactants using sediment-derived methanogenic consortia. Water Sci. Technol. 38: 55–62. Zwiernik, M.J., Quensen, J.F., III, and Boyd, S.A. 1998. FeSO4 amendments stimulate extensive anaerobic PCB dechlorination. Environ. Sci. Technol. 32: 3360–3365.

chapter seven

Polycyclic aromatic hydrocarbons (PAHs): improved land treatment with bioaugmentation Hap Prichard, Joanne Jones-Meehan, Cathy Nestler, Lance D. Hansen, William Straube, William Jones, John Hind, and Jeffrey W. Talley Contents 7.1 Land-farming background .......................................................................217 7.1.1 Polycyclic aromatic hydrocarbons..............................................217 7.1.1.1 Chemical structure and source of contamination..................................................................217 7.1.1.2 Toxicity and benzo(a)pyrene toxic equivalent factors ...........................................................217 7.1.1.3 PAH bioavailability ........................................................220 7.1.1.4 Problem summary ..........................................................221 7.1.2 Available treatment options ........................................................222 7.1.3 Thrust area: early studies ............................................................222 7.1.3.1 Solid phase treatments...................................................224 7.1.3.2 Slurry phase treatment ..................................................226 7.1.3.3 Performance comparison...............................................226 7.1.3.4 The flask-to-field selected treatment option (land farming) .................................................................228 7.1.3.5 Microbiological studies..................................................228 7.2 Objectives ....................................................................................................241 7.3 Technical approach ....................................................................................242 7.3.1 Site description ..............................................................................242 215

216

Bioremediation of Recalcitrant Compounds 7.3.2 7.3.3 7.3.4

Soil characterization......................................................................243 Flask (bench-scale) experimental design...................................244 Flask (bench-scale) materials.......................................................244 7.3.4.1 Bacteria .............................................................................244 7.3.4.2 Amendments ...................................................................244 7.3.5 Flask (bench-scale) methods........................................................245 7.3.5.1 Isolation and characterization of PAH-degrading bacteria ................................................245 7.3.5.2 Biosurfactant production...............................................247 7.3.5.3 Bioaugmentation.............................................................248 7.3.5.4 Carrier technology development .................................249 7.3.5.5 Biostimulation .................................................................251 7.3.5.6 Microcosm preparation..................................................252 7.3.5.7 Microbial analysis...........................................................252 7.3.5.8 Chemical analysis ...........................................................252 7.3.6 Pilot studies: experimental design .............................................253 7.3.6.1 LTU objectives .................................................................255 7.3.6.2 Trough study objectives.................................................255 7.3.7 Land treatment units: assembly .................................................255 7.3.8 Trough study: assembly ...............................................................255 7.3.9 Pilot-scale materials ......................................................................257 7.3.10 Pilot-scale methods .......................................................................259 7.3.10.1 Sampling design .............................................................259 7.3.10.2 Physical analysis .............................................................260 7.3.10.3 Chemical analysis ...........................................................260 7.3.10.4 Microbial analysis...........................................................261 7.3.10.5 Metabolic analysis ..........................................................261 7.3.10.6 Statistical analysis...........................................................261 7.4 Accomplishments.......................................................................................262 7.4.1 Flask studies...................................................................................262 7.4.1.1 PAH removal ...................................................................262 7.4.1.2 Biosurfactant production...............................................266 7.4.1.3 Nutrient amendment .....................................................267 7.4.1.4 Vermiculite carrier technology .....................................268 7.4.2 LTU pilot project ...........................................................................270 7.4.2.1 Chemical characteristics ................................................270 7.4.2.2 PAH removal ...................................................................271 7.4.2.3 Microbial characterization.............................................273 7.4.2.4 Soil respiration ................................................................276 7.4.3 Trough pilot project ......................................................................276 7.4.3.1 Chemical characterization .............................................276 7.4.3.2 PAH removal ...................................................................276 7.4.3.3 Microbial characterization.............................................282 7.4.3.4 Metabolic analysis: trough soil respiration ................283 7.4.4 Comparison of LTUs and troughs..............................................284 7.5 Conclusions on utility in remediation....................................................285

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

217

7.5.1 Conclusions from flask studies...................................................285 7.5.2 Conclusions from LTUs................................................................286 7.5.3 Conclusions from trough study..................................................286 7.5.4 Utility to remediation of highly contaminated soil.................287 7.6 Recommendations for transitional research ..........................................288 7.6.1 PAH availability in soil and regulatory cleanup levels..........288 7.6.2 Phytoremediation of PAHs..........................................................288 7.7 Technology transfer ...................................................................................289 References.............................................................................................................290

7.1 Land-farming background 7.1.1 Polycyclic aromatic hydrocarbons 7.1.1.1 Chemical structure and source of contamination Polycyclic aromatic hydrocarbons (PAHs) are multiringed, organic compounds, characteristically nonpolar, neutral, and hydrophobic. PAHs have two or more fused benzene rings in a linear, stepped, or cluster arrangement. Although there are more than 100 known PAHs, Table 7.1 provides the chemical structure, abbreviated name, and molecular weight for the 15 PAHs that were analyzed in this study. PAHs occur naturally as components of incompletely burned fossil fuels, and they are also manufactured. Several of these manufactured homologues are used in medicines, dyes, and pesticides, but most are found in coal tar, roofing tar, and creosote, a commonly used wood preservative. PAHs are major chemical constituents of a wide variety of contaminants found at Department of Defense (DOD) installations. They are found in burning pits and as spills of creosote, fungicides, heavy oils, Bunker C fuels, and other petroleum-based products. The higher-molecular-weight (HMW) homologues are particularly recalcitrant and toxic. Some lower-molecular-weight PAHs are volatile, readily evaporating into the air. Others will undergo photolysis. Because they are hydrophobic and neutral in charge, PAHs are strongly adsorbed onto soil particles, especially clays. Park et al. (1990) studied the degradation of 14 PAHs in two soils. They found air phase transfer (volatilization) an important means of contaminant reduction only for naphthalene and 1-methylnaphthalene (the two-ring compounds). Abiotic mechanisms accounted for up to 20% of the total reduction but involved only two-and three-ring compounds. Biotic mechanisms were responsible for the removal of PAHs over three rings. The persistence of PAHs in the environment, coupled with their hydrophobicity, gives them a high potential for bioaccumulation.

7.1.1.2 Toxicity and benzo(a)pyrene toxic equivalent factors The 15 compounds examined in this study (Table 7.2) are grouped together because (1) more information is available on them and (2) they are suspected

218

Bioremediation of Recalcitrant Compounds

Table 7.1 Molecular Weights and Structures for the 15 PAH Homologues Analyzed PAH Homologue

Abbreviation

Molecular Weight (amu)

Naphthalene

NAPHTH

128

Acenaphthylene

ACENAY

152

Acenaphthene

ACENAP

154

Fluorene

FLUORE

166

Anthracene

ANTHR

178

Phenanthrene

PHENAN

178

Fluoranthene

FLA

202

Pyrene

PYR

202

Benzo(a)anthracenea

BAANTHR

228

Chrysenea

CHRYS

228

Benzo(k)fluoranthenea

BKFLANT

252

Benzo(a)pyrenea

BaP

252

Benzo(g,h,i)perylenea

BGHIPY

276

Indeno(1,2,3-c,d)pyrenea

I123PY

276

Dibenzo(a,h)anthracene

DBAHANT

278

a

Structure

Indicates a BaP toxic equivalent compound. See Table 7.2 for values.

Table 7.2 Toxic Equivalency Factors for the Seven PAHs of Greatest Environmental Significance Compound (Abbreviation) Benzo(a)anthracene (BAANTHR) Chrysene (CHRYSE) Benzo(b)fluoranthene (BBFLANT) Benzo(k)fluoranthene (BKFLANT) Benzo(a)pyrene (BaP) Indeno(1,2,3-c,d)pyrene (I123PYR) Dibenzo(a,h,)anthracene (DBAHANT)

Toxic Equivalent Factor (by Nisbet and LaGoy, 1992) 0.1 0.01 0.1 0.1 1.0 0.1 1.0

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

219

to be the most harmful of the PAHs. The effects they exhibit in animal and human studies are representative of the class as a whole. In addition, these are the PAHs to which the public is most commonly exposed. Also, they are found in highest concentration on National Priority List hazardous waste sites (ATSDR, 1995a and 1995b). As a class of compounds, PAHs have been classified as carcinogens, mutagens, and immunosuppressants. Even slight differences in PAH chemical structure and activity result in different toxic potencies and different health effects from the individual PAHs. The importance of PAH chemical structure as an indicator of potential carcinogenicity has been reviewed in Pitot and Dragan (1996). Some PAHs have been classified as carcinogens only in laboratory animals. Others, including benzo(a)pyrene (BaP) and benzo(a)anthracene, have been identified as human carcinogens. Still others are possible carcinogens or not classifiable because the testing is incomplete. Tumors usually occur at the point of entry into the body (i.e., the skin, lungs, eyes, intestines). However, metabolism of these compounds can result in an increase in their toxic potency and tumor formation in secondary organs (i.e., bladder, colon, liver). Metabolites of these compounds can also be carried into cells where they form adducts with DNA through covalent bonding. The best-studied mutation is in the 12th codon of the Hras codon The PAHs elicit multiple responses from the body’s immune system due to their effects on humoral and cell-mediated immunity as well as host resistance (Burns et al., 1996). The mechanisms of PAH immunosuppression have been reviewed by White et al. (1994). BaP is often used as an indicator for risk assessment of human exposure because it is highly carcinogenic, persistent in the environment, and toxicologically well understood. This breadth of knowledge does not exist for most of the other PAH compounds. Because PAHs occur as mixtures of different concentrations of different homologues, toxic equivalency factors (TEFs) were proposed, similar to those used in the risk assessment of mixtures of polychlorinated biphenyls (PCBs). The Environmental Protection Agency (EPA, 1984) took the first step by separating PAHs into carcinogenic and noncarcinogenic compounds. All of the PAHs were rated, using BaP as a reference and giving it a value of 1.00. However, this method led to an overestimation of exposure risk because the carcinogenicity of most of the compounds was unknown and the interactions between compounds in mixtures had not been determined. In an attempt to overcome this liability, Nisbet and LaGoy (1992) developed a new method based on the compounds’ response while testing one or more PAHs concurrently with BaP in the same assay system (usually lung or skin cell carcinoma). BaP remained the reference carcinogen, assigned the value of 1.00. Sixteen other PAHs were ranked in comparison to BaP carcinogenicity. This system was tested by Petry et al. (1996), who assessed the health risk of PAHs to coke plant workers. There are drawbacks to any system that uses equivalency factors. The uncertainties in this case arise primarily from dealing with inconsistent mixtures. Carcinogenic potency could be affected by differences in bioavailability, a competition for binding sites, cocarcinogenic

220

Bioremediation of Recalcitrant Compounds

action, or the effects of metabolism. Nevertheless, Petry et al. (1996) found that the BaP equivalents developed by Nisbet and LaGoy (1992) were valid markers for PAH health risk assessment. Environmental risk assessment, in slight contrast to human health risk, looks at the PAHs that usually occur in contaminated environmental systems and that have the highest TEFs (by the Nisbet and LaGoy (1992) system). The seven PAHs listed in Table 7.2 have the highest environmental risk: benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, indeno(1,2,3-c,d)pyrene, and dibenzo(a,h)anthracene.

7.1.1.3 PAH bioavailability As indicated earlier, contaminants react chemically and physically with different kinds of soil particles, which change the physical and chemical natures of both components. Biological availability, or bioavailability, is used to describe both the amount of toxin available in soil to harm organisms (humans, other animals, plants) and, in the case of bioremediation, the amount of toxin available to be metabolized by microorganisms after contaminant–soil interactions. In situ bioremediation is a managed or spontaneous process in which microbiological processes are used to degrade or transform contaminants to less toxic or nontoxic forms, thereby remedying or eliminating environmental contamination. Although these microbiological processes may decrease contaminant concentrations to levels that no longer pose an unacceptable risk to the environment or human health, the contaminants that remain in treated soils still might not meet stringent regulatory levels, even if they represent site-specific, environmentally acceptable endpoints (National Research Council, 1997). PAHs in soils may be biodegraded by microorganisms to a residual concentration that no longer decreases with time or that decreases slowly over years with continued treatment (Thoma, 1994; Luthy et al., 1994; Loehr and Webster, 1997). Further reductions are limited by the availability of the PAHs to microorganisms (Bosma et al., 1997; Erickson et al., 1993). Additionally, as contaminants age they become less available than freshly contaminated material. The adherence and slow release of PAHs from soils are other obstacles to remediation (National Research Council, 1994; Moore et al., 1989). Because they bind with soils and suffer subsequent slow-release rates, residual PAHs may be significantly less leachable by water and less toxic as measured by uptake tests (Gas Research Institute, 1995; Alexander, 1995; Kelsey et al., 1997). Generally, contaminants can only be degraded when they exist in the aqueous phase and in contact with the cell membrane of a microorganism (Fletcher, 1991). The contaminant serves as a growth substrate for the microorganism and is incorporated into the cell through membrane transport and utilized as an energy source in the cell’s principal metabolic pathways. However, physical or chemical phenomena can limit the bulk solution concentration of the contaminant and thus significantly reduce the ability of the microorganism to assimilate the contaminant. Therefore, the availability of the contaminant can control the overall biodegradation of these compounds.

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

221

Other important factors relevant to biodegradation and bioavailability are the location and density of microorganisms. The majority of bacteria in the environment are attached to surfaces, and their distribution in and on soils is very patchy. The majority of these bacteria range in size from 0.5 to 1.0 μm, whereas micropores present in soils measure far less than 1 μm. It is generally believed that bacteria are attached predominantly to the surface of soil particles and not to the interior surfaces of the micropores. It has been estimated that more than 90% of the microorganisms present in geologic matrices accumulate on the surfaces of soil (Costerton et al., 1987). Therefore, the majority of contaminant–microbial interactions occur in the biofilm that develops within macropores on the surfaces of soils. This suggests that partitioning of an organic contaminant from the solid phase of the soil to the aqueous phase in the larger pore spaces controls soil biovailability. These partitioning mechanisms may include chemical bonding, surface complex formations, electrostatic interactions, and hydrophobic effects (Schwarzenbach et al., 1993; Stumm, 1992). For hydrophobic contaminants such as PAHs, sorption increases with the content of the organic matter in the soil/sediment and the degree of hydrophobicity of the specific PAH. Typically, the rate of desorption can be attributed to the mass transfer of the sorbate molecules from sorption sites on and in the soil. Active bacteria should correspond to the higher available PAH concentrations, which occur where desorption is the most intense. Current methods for assessing sorption and sequestration of PAHs on soils do not provide a basic understanding of the bioavailability of recalcitrant PAHs. They also lack information to aid interpretation of results of ecotoxicological testing of residuals after biotreatment. Whether residual PAHs remaining after biotreatment represent an acceptable cleanup endpoint requires understanding of the mechanisms that bind contaminant PAHs within soil or sediment. Research is needed that will assess the fundamental character of the binding of PAHs in parallel with the development of biotreatment and ecotoxicity testing, to show how the nature of PAH association with soils relates to bioavailability and achievable treatment endpoints.

7.1.1.4 Problem summary PAHs are large, multi-ring compounds, many of which are toxic to humans and the environment. Whereas the lighter-molecular-weight homologues may be removed by volatilization, the higher-molecular-weight compounds are increasingly more toxic and more resistant to both chemical and biological degradation. PAHs are tightly bound to the humic fraction of the soil. The binding strength increases with exposure time, making aged soils more difficult to remediate. Research was needed to: • Increase the availability of the PAHs for biological degradation • Establish remediation endpoints that maintain public health and safety and are realistically achievable

222

Bioremediation of Recalcitrant Compounds

The Flask to Field PAH project (Flask, for short) focused on the first of these objectives, increasing the availability of PAH compounds for biological degradation and increasing the overall biodegradation of the high-molecular-weight homologues.

7.1.2 Available treatment options Treatment of PAH-contaminated soil can be performed either ex situ or in situ, and each of these has both abiotic and biotic technologies available. Of the ex situ treatments, the abiotic choice is a destructive technology — incineration. Biotic options include slurry bioreactors and compost reactors. The available abiotic in situ treatments include soil flushing and stabilization. Electrokinetic (E-K) separation is in preliminary development. Biotic treatments performed in situ include bioventing, phytoremediation (on soils with low PAH concentrations), and land farming. These options generally separate into either of two treatment approaches: highly engineered solutions such as solid phase or slurry phase treatment, and minimally engineered in situ treatment. Examples of each strategy are presented in Table 7.3, along with a summary of the inherent benefits and limitations associated with each technology. The difficulties associated with the use of biotreatments have been analyzed in Talley and Sleeper (1997), with reviews of pertinent technologies. Detailed information including cost summaries and case studies can be obtained at http://www.frtr.gov

7.1.3 Thrust area: early studies The approach taken during the Flask studies separated the objective into two broad tasks: (1) isolating and characterizing a microorganism, or consortium of microorganisms capable of degrading the higher-molecular-weight PAHs, and (2) selecting a means of releasing the PAHs from the soil into the surrounding soil pore spaces where it would, presumably, be available to degradation. Each of these tasks was further separated into smaller research areas. In order to find an organism that would degrade higher-molecular-weight PAHs, a new isolation method was developed. New and existing strains were characterized and metabolic pathways have been described. It became necessary to explore the potential of cometabolism for the degradation of these PAHs. At the same time, the effects of chemical surfactants on soil–PAH binding were studied. Bacteria that are natural surfactant producers were isolated and the biosurfactant activities were compared to the chemical surfactants. When a decision for bioaugmentation had been made, a method had to be found to deliver the chosen microorganisms into the contaminated soil. Several different technologies contributed to the Flask portion of the study, as illustrated in Figure 7.1: slurry reactors, land farming, and composting. The project began with an examination of available treatment options. Three promising technologies were selected: a high-technology system of slurry bioreactors and two low-technology soil treatment systems — composting and land

Land farming Composting Engineered soil cell Soil treatment

Aqueous reactors Soil slurry reactors

Biosparging Bioventing GW circulation In situ bioreactors

Solid phase

Bioreactors

In situ

Material input requires physical removal Relatively high capital costs Physicochemical control Extended treatment time Monitoring progress and effectiveness

Same as solid phase Chemical solubility LNAPL/DNAPL present Geological factors Regulatory aspects for groundwater

Catabolic capabilities of indigenous microflora Presence of metals and other organics pH, temperature, moisture control Biodegradability Same as solid phase Toxicity of amendments Toxic concentrations of contaminants

Space requirements Extended treatment time Control of abiotic loss Mass transfer Bioavailability

Cost efficient Modestly effective for HMW PAHs Low O&M costs Perform on site, in place

Most rapid degradation Controlled conditions Enhanced mass transfer Use of surfactants and inoculants Least cost Noninvasive Complements natural attenuation processes Soil and water treated simultaneously

Factors to Consider

Limitations

Benefits

Source: Modified from Mueller, J. et al., Antonie Leeuwenhoek, 71, 329–343, 1997.

Examples

Technology

Table 7.3 Summary of Existing Treatment Options for PAH-Contaminated Soil

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs) 223

224

Bioremediation of Recalcitrant Compounds Remediation of Soil PAHs

Bioslurry

Nutrient amendments Traditional landfarming bioaugmentation

Bulking agents mixing

Composting ineffective

Too expensive Incomplete degradation

Modified Land farming nutrient amendment, bulking agent, bioaugmentation and limited tilling

Figure 7.1 Project history leading to the selection of land farming as a PAH treatment technology.

farming. Research in each of these technologies provided insight into areas of PAH bioremediation useful to the final treatment selection: isolation and characterization of PAH-degrading bacteria and elucidation of the degradation pathways, surfactant chemistry, bioaugmentation, and microbial carrier technology. This produced a final treatment with aspects of all three technologies. For field studies, land farming with bioaugmentation and biostimulation was selected as the treatment option for PAH-contaminated soils. The treatments and the technologies that support them are discussed in the following sections.

7.1.3.1 Solid phase treatments Solid phase treatments, commonly known as land farming and composting, are two of the most commonly applied technologies for the remediation of PAH-contaminated soil (Gray et al., 2000; Harmsen, 1991; Mueller et al., 1991a, 1991b; Mueller-Hurtig et al., 1993; Yare, 1991). Most PAH-contaminated soils contain a significant number of PAH degraders that have been enriched because of the presence of the PAHs, but they are often constrained in their degradation capability because of some limiting factor. Common limiting factors include inadequate aeration, poor contact of the microorganisms with the PAHs due to the adherence of the PAHs to surfaces and nonaqueous phase liquid (NAPL) materials, and the absence of sufficient nitrogen to sustain extensive mineralization of the contaminant carbon. Any engineering activity that reduces these limitations brings the native degraders into action. Advantages of solid phase treatment are that large quantities of contaminated soil can be treated at the same time and that operation and maintenance activities (costs) are minimal. In general, contaminated soil is placed in aboveground treatment areas that are designed for proper effluent collection, and then the soil is handled in specific ways to enhance indigenous microbial activity. Composting usually involves the addition of readily

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

225

degradable organic matter (bulking) and fertilizer. Bulking refers to the addition of inexpensive, readily available materials (straw, manure, sewage sludge, wood chips, rice hulls, etc.) that enhance aeration of the soil and improve soil texture. The bulking agents also dilute the soil contaminant, reducing the concentration of toxic chemicals. The resulting mineralization of the added organic matter also aids in the degradation of PAHs. Some control of aeration and temperature is usually required (Potter et al., 1997). Kastner et al. (1998) found that the addition of compost enhanced the degradation of PAHs in soil due to the presence of the organic solids in the compost. Composting was not selected as our treatment option because the loss of PAHs was due to soil binding, which did not reduce the toxicity of the soil (Johnson, 1998). Land-farming remediation of PAH-contaminated soils, based on the degradative activities of natural microbial communities, is a well-used and generally reliable technology. It has been applied to a variety of soils and contaminant types (i.e., creosote, coal tar, and petroleum) and is generally preferred over nonbiological approaches such as stabilization, chemical oxidation, and incineration (Mueller et al., 1995). The focus of land farming is to stimulate the degradation capabilities of natural microbial communities by providing oxygen and nitrogen and to use physical mixing of the soil to distribute the contaminants over a greater surface area and bring them into contact with the microorganisms. Traditional agricultural procedures are used to provide mixing (tilling, bulking), moisture (irrigation), and nitrogen (fertilizer). Fertilizer can be applied as commercial formulations or as organic waste material (commonly manure). Land farming is often, but not necessarily, limited to the treatment of 6 to 12 in. of soil at a time (normal depth of tilling), but as one layer (lift) is successfully treated, successive lifts can be applied on top. The disadvantages of traditional land farming are the time required, the operation and maintenance costs, and the space required for the lifts. The degradation that results from solid phase remediation often follows a two-stage process involving an initial rapid phase with extensive PAH degradation (a few months) and a second slower phase of degradation (months to years) with relatively little further change in PAH concentration (Brown et al., 1995; Cornelissen et al., 1998; Mueller et al., 1998; Pollard et al., 1994). Differences in desorption rates from soil particles into the interstitial water are frequently cited as the cause of the two-stage process. Initially, rapid desorption occurs, and degradation is limited by microbial factors. At some point, desorption slows and degradation rates decrease accordingly. The transition from the fast to the slow phase is often critical. If it occurs before cleanup criteria are met, treatment times are greatly extended and costs increase substantially. Cleanup criteria are often based on the concentration of benzo(a)pyrene equivalents. If concentrations cannot be reduced below a previously established cleanup value, then the soil is not considered clean and must be either disposed of as a hazardous waste or further “engineered” with another treatment technique, such as chemical oxidation, the

226

Bioremediation of Recalcitrant Compounds

addition of surfactants, or physical restructuring, in an attempt to encourage additional degradation. Treatment times are greatly extended and costs increase substantially. The treatment options then become: (1) disposal, (2) short-term treatment, or (3) impoundment. Impounding the treated soils and allowing natural degradation processes to eventually degrade the high-molecular-weight (HMW) PAHs can be considered. With the impoundment option, leachable hydrocarbons, mainly the low-molecular-weight PAHs, are not likely to cause environmental problems because they have been largely removed during the initial degradation phase. In essence, these soils can be considered biostabilized (Luthy et al., 1997; Talley et al., 2000); that is, the very slow leaching of the residual HMW PAHs is counterbalanced by the degradation capabilities of the indigenous microbial communities. However, this requires that the impounded soil be carefully managed and monitored over periods of years, a task that can add considerable long-term cost. Therefore, the problem became one of identifying the causes of the rapid-to-slow transition of degradation and learning what could be done to control it.

7.1.3.2 Slurry phase treatment Slurry phase treatment involves the processing of contaminated soil using a contained system, or bioreactor, in which the contaminated soil receives the maximum amount of mixing and aeration. A variety of reactors can be used: fixed film, plug flow, and slurry reactors. Degradation rates in a bioreactor are usually considerably faster than solid phase systems (Pinelli et al., 1997). This is because intimate contact between the PAHs and the microorganisms is provided and optimal conditions for microbial growth and degradation are maintained with considerable uniformity. Because of the contained nature of the system, inoculation with selected organisms (bioaugmentation) or the addition of surfactants is reasonable. However, slurry phase treatment is expensive to set up, operate, and maintain, especially as only relatively small quantities of soil can be treated at a time. As the size of the reactor increases, optimal conditions will be compromised due to the physical nature of the systems. Costs can often can be reduced by using existing facilities, such as lined lagoons and basins, for the treatment.

7.1.3.3 Performance comparison From a comparison of the performance of slurry phase and solid phase treatments, it can be seen that solid phase treatment is slower than slurry phase treatment (Figure 7.2). Mueller et al. (1991a, 1991b) divided PAHs into three groups based on the number of rings and added the mixtures to soil in both slurry and solid treatments. Group 1 consisted of two-ring PAHs (naphthylene, methyl and dimethyl naphthalenes, and biphenyl). Group 2 consisted of three-ring PAHs (acenaphthylene, acenaphthene, phenanthrene, anthracene, and methyl-anthracene). Group 3 consisted of various four-, five-, and six-ring PAHs (fluoranthene, pyrene, benzo(b)fluorene, chrysene,

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

227

100 A

Group 1 PAHs Group 2 PAHs Group 3 PAHs

% Biodegradation

80

60

40

20

0 0

2

4

6 Weeks

8

10

12

100 B

% Biodegradation

80

60

40 Group 1 PAHs Group 2 PAHs Group 3 PAHs

20

0 0

5

10

15 Days

20

25

30

Figure 7.2 Biodegradation of PAHs during solid phase (A) and slurry phase (B) treatment of contaminated surface soils amended with nutrients. Group 1 was 2-ring PAHs (naphthylene, methyl and dimethyl naphthalenes, biphenyl), Group 2 was 3-ring PAHs (acenaphthylene, acenaphthene, phenanthrene, anthracene, methyl-anthracene), and Group 3 was 4-, 5-, and 6-ring PAHs (fluoranthene, pyrene, benzo(b)fluorine, chrysene, benzo(a)pyrene, benzo(a)anthracene, indenopyrene, benzo(b,k)fluoranthene).

benzo(a)pyrene, benzo(a)anthracene, indenopyrene, and benzo(b,k)fluoranthene). For the group 3 PAHs, 40% removal occurred in 30 days in the slurry reactor and in 12 weeks in the solid phase reactors. However, degradation

228

Bioremediation of Recalcitrant Compounds

of these higher-molecular-weight PAHs appeared to plateau in the slurry reactor, whereas degradation was still occurring at the end of the solid phase reactor experiment. Thus, the added time of treatment may be rewarded by a greater extent of degradation.

7.1.3.4 The flask-to-field selected treatment option (land farming) In order to avoid the possibility of long-term passive treatment or the use of additional active treatment technologies, the initial, rapid phase of PAH degradation must be expanded and the second, slower phase enhanced. This will provide a greater and more predictable degradation of HMW PAHs. Improvements in the degradation of HMW PAHs result from the enhancement of the metabolic capabilities of natural microbial communities and the increased availability of the PAHs to the microorganisms. These two considerations appear to have the greatest effect on the second phase of degradation. Enhancing metabolic capabilities of microbial communities will require an understanding of where the deficiencies originate, including an appreciation for the metabolic pathways used by these organisms. Enhancing bioavailability will require knowledge of the interactions between the degrading microorganisms and the availability of the PAHs from the bound state. To achieve these enhancements, chemicals and microorganisms can be added to soil. Any amendment additions must generate enough cost savings in the end to pay for the additions, an aspect that was considered carefully in this work. Each of these enhancement strategies is presented in detail.

7.1.3.5 Microbiological studies 7.1.3.5.1 Isolation/characterization of PAH-degrading bacteria. Bacterial isolates that have been enriched for their ability to grow on low-molecular-weight PAHs (i.e., naphthalene, phenanthrene, fluorene, and, to some extent, indan, acenaphthene, and anthracene) were studied. The common bacterial genera encountered include Pseudomonas, Alicaligenes, Mycobacterium, Rhodococcus, Comamonas, and Sphingomonas. This is a relatively small range of genera considering the prevalence of PAHs in the environment; however, it does show that the ability to degrade low-molecular-weight PAHs is common. Recent studies have emphasized the potential importance of Mycobacterium and Sphingomonas species (Bastiaens et al., 2000; Bouchez et al., 1995; Fredrickson et al., 1995; Givindaswami et al., 1995; Kastner et al., 1994; Meyer et al., 1999), with some indication that Sphingomonas strains are more likely to degrade aqueous phase PAHs and Mycobacterium are more likely to degrade solid PAHs (crystals) because of their hydrophobic cell surfaces (Bastiaens et al., 2000). Quantitative polymerase chain reaction (PCR) has been used to identify microorganisms able to degrade PAHs in soil. PCR is a technique that increases the number of copies of a specific region of DNA. From a single microorganism, enough DNA can be produced to allow precise identification of that species. In some soil samples where PAH degradation

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

229

is actively occurring, the responsible organism appears to be a Bulkholderia-type organism, which is difficult to grow in the laboratory (Laurie and Lloyd-Jones, 2000). Again, it must be emphasized that these studies find bacteria that seem to grow well on phenanthrene and fluorene but with a limited ability to grow on or cometabolize HMW PAHs. A number of studies have shown that bacteria are able to grow on the four-ring PAHs, specifically fluoranthene (Boldrin et al., 1993; Dagher et al., 1996; Kastner et al., 1994; Lloyd-Jones and Hunter, 1997; Mueller et al., 1990; Mueller et al., 1994; Weissenfels et al., 1991) and pyrene (Boldrin et al., 1993; Churchill et al., 1999; Dean-Ross and Cerniglia, 1996; Fritzsche, 1994; Grosser et al., 1991; Heitkamp et al., 1988a, 1988b; Jimenez and Bartha, 1996; Kastner et al., 1994; Lloyd-Jones and Hunter, 1997; Rehmann et al., 1998; Schneider et al., 1996). Mycobacterium, Rhodococcus, Alcaligenes, and Sphingomonas are the genera commonly encountered. Results from a phylogenic study with a variety of Pseudomonas and Sphingomonas strains isolated from PAH-contaminated soils show that these two groups can be separated, to some extent, based on their physiological characteristics (Biolog® Identification System) and fatty acid compositions. The Biolog results are shown in Figure 7.3, modified from Mueller et al. (1997). The symbols represent different soil samples used for isolation, and the numbers refer to strain designations. The cluster in the upper left (7–10, 15–17) consisted entirely of phenanthrene 21 0.67

28

16

10 8 9

3 15

2

0.21 17

P3

7

14 20 13 26

−0.26 12

18

11

27

4 1 6

23 −0.72 0.92

19

25

24 5 29

22

1.79 0.86

0.15 P1

−0.06

−0.62

P2

−0.98 −1.40

Figure 7.3 Balloon plot of principal components analysis of Biolog responses from PAH-degrading isolates. (Modified from Mueller, J.G. et al., Environ. Sci. Technol., 25, 1055–1061, 1991a; Mueller, J.G. et al., Environ. Sci. Technol., 25, 1045–1055, 1991b.)

230

Bioremediation of Recalcitrant Compounds

degraders (predominantly Pseudomonas species). The cluster in the lower right (1, 4–6, 13, 24–26, 29) consisted of fluoranthene degraders, predominantly Sphingomonas species. These results suggest that certain catabolic characteristics are associated with specific genera. That is, Sphingomonas and Mycobacterium could be the primary genera that are able to attack HMW PAHs. This may be related to characteristics of their cell membranes that allow these PAHs to diffuse inside the cells, or it may be related to the presence of key membrane-associated dioxygenases. Interestingly, most of the pure cultures that have been isolated for their ability to grow on pyrene have been Mycobacterium. We are aware of only one case in which a Gram-negative microorganism, a Pseudomonas strain, was able to grow on pyrene (Thibault et al., 1996). Numerous fungal species are also able to partially degrade PAHs, both low and high molecular weight (Baldrian et al., 2000; Boonchan et al., 2000; Mueller et al., 1997). The initial reactions of PAH degradation by fungi are usually ascribed to their extracellular lignolytic enzymes (usually the laccases and perioxidases), and these organisms may be involved in PAH turnover in unpolluted soils. The effectiveness of a coculture of a fungus (Penicillium) with the bacterium Stenotrophomonas and a mixed bacterial population in degrading five-ring PAHs suggests that the initial oxidation products produced by the fungi are then further degraded by the bacteria. There was no indication that this coculture affects the bioavailability of the HMW PAHs. It may not be a particularly useful technique for bioremediation, but it is interesting in terms of the natural way in which PAHs might be degraded in the environment. The most important observation is that there are a number of PAH degraders that have relatively broad cometabolic capabilities, especially for HMW PAHs (Dagher et al., 1996; Mahaffey et al., 1988; Mueller et al., 1997; Schneider et al., 1996; Schocken and Gibson, 1984). The term cometabolism, as used in this study, is the ability to transform (oxidize) a particular PAH, but without growth on that PAH. Presumably, partial degradation products are generated, which cannot be further metabolized to produce carbon and energy. There have been several reports of PAH-cometabolizing Sphingomonas species (Dagher et al., 1996; Fredrickson et al., 1995; Kastner et al., 1994; Mueller et al., 1990). Sphingomonas paucimobilis strain EPA 505 has been shown to have a substantial cometabolic capability for HMW PAHs (Mueller et al., 1990; Ye et al., 1996). Sphingomonas strain B1 (formally Beijerinckia B1), isolated originally for its ability to grow on biphenyl, has been shown to cometabolize benzo(a)anthracene to acid metabolites (Gibson et al., 1973; Mahaffey et al., 1988). Whether sphingomonads are commonly associated with PAH degradation is yet to be assessed, but a study examining the diversity of bacteria able to degrade PAHs (Ye et al., 1996) showed that Sphingomonas species tended to be the isolates capable of degrading fluoranthene, whereas the bacteria able to degrade phenanthrene were more commonly associated with Pseudomonas strains. Dagher et al. (1996) compared three PAH-degrading Pseudomonas sp. with a Sphingomonas sp. and found

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

231

the latter to be the most efficient PAH degrader, with the ability to possibly grow on fluoranthene (FLA). With these previous studies as a basis, other fluoranthene degraders were isolated and evaluated with respect to their taxonomic and metabolic characteristics. Little is known concerning the bacterial genera that may be responsible for microbial degradation of HMW PAHs in PAH-contaminated sites; however, the results reported here suggest that Sphingomonas sp. are commonly isolated from PAH-contaminated soils and their metabolic diversity may vary considerably (Kim et al., 1996). Therefore, our efforts focused on sphingomonads and on S. paucimobilis, in particular. 7.1.3.5.2 Cometabolism to enhance metabolic capabilities. The cometabolic capabilities of S. paucimobilis strain EPA 505 are shown in Table 7.4 and Figure 7.4. After a 28-day incubation of a PAH mixture with EPA 505 and a chemical surfactant, there was considerable reduction in the higher-molecular-weight PAHs. Ye et al. (1996), using resting cell suspensions of EPA 505 and individual PAHs, in place of a mixture, showed a similar pattern of degradation of HMW PAHs, but to a greater extent. This is shown in Figure 7.4. Error bars for the standard deviation of triplicates are not shown in this figure, but the error was approximately 10%. The PAHs in this experiment were not growth substrates for EPA 505. This cometabolic capability is not typical of all PAH-degrading strains studied. Table 7.4 shows the effectiveness of several isolates in their ability to degrade different PAH fractions of creosote. The CRE strains were isolated for their ability to grow on phenanthrene, the UNP/NP strains for their ability to grow on fluoranthene, and the PYR1 strain for its ability to grow Table 7.4 Amount of PAHs Remaining Following Biodegradation of PAH Fractions by Selected Strains of PAH Degraders Isolate/ Strain Initial Uninoculated CRE 7a N3P2b UN1P1b EPA 505b PRY-1c

Two-Ring PAHs

Three-Ring PAHs

149.5 71.3 1.7 0.5 0.5 0.3 4.8

168.1 107.0 14.8 4.9 3.4 2.2 70.6

(0)d (NA) (98) (99) (99) (100) (93)

(0) (NA) (86) (95) (97) (98) (34)

Four-, Five-, and Six-Ring PAHs 60.0 55.9 47.2 46.1 43.0 18.6 40.3

(0) (NA) (16) (18) (23) (67) (28)

Total PAHs 377.6 (NA) 234.1 (NA) 63.7 (73) 51.5 (78) 46.9 (80) 21.1 (91) 115.7 (51)

Phenanthrene degrader. Fluoranthene degraders. c Pyrene cometabolizer. d Numbers in parentheses indicate percent removed versus the uninoculated control. Source: Modified from J. Mueller et al., 1997. Antonie Leeuwenhoek 71, 329–343. a

b

232

Bioremediation of Recalcitrant Compounds

100

% Degradation

80

Dibenzo(a,l)pyrene Dibenz(a,h)anthracene Benzo(b)fluoranthene Chrysene Benzo(a)pyrene 1-nitropyrene Benz(a)anthracene Pyrene

60

40

20

0 0

2

4

6

8 10 Hours

12

14

16

18

Figure 7.4 Cometabolism of PAHs by EPA 505.

on pyrene. It is clear that of the strains studied, EPA 505 was the most effective in terms of removing the four-, five-, and six-ring PAHs (67% removal compared to 28% removal for next best strain, Mycobacterium strain PYR-1) (Heitkamp et al., 1988a, 1988b) and in producing the greatest removal of PAHs overall (91% removal). Thus, just because a strain grows well on a particular PAH does not mean that it will have broad cometabolic capabilities. The next question investigated was whether this metabolic capability could be expected in most soils. 7.1.3.5.3 Metabolic characteristics. In preliminary mechanistic studies performed with pyrene, a nongrowth PAH for EPA 505, it appears that at least two different partial degradation products of pyrene were produced. Both degradation products represent the opening of an aromatic ring, and both indicate an inability to cleave carbon moieties to use for growth. The proposed pathway for the cometabolism of pyrene by strain EPA 505 is shown in Figure 7.5. The pathway represents possible attacks on the pyrene rings based on products detected by, and inferred from, gas chromatography mass spectrometry (GC-MS) analysis. The products suggest that pyrene was dihydroxylated at both the 2,3 position and the 9,10 position. There was no evidence to suggest two separate dioxygenase systems and, thus, it was assumed that a single enzyme system is able to recognize both positions. Clearly, both of the oxidized rings were then opened, one by meta-cleavage and one by ortho-cleavage. This is consistent with the mechanism by which the strain attacks fluoranthene (see below). It is also consistent with the reported 1,2-dioxygenation of pyrene and subsequent meta-cleavage (Heitkamp et al., 1988a, 1988b; Walter et al., 1991) and the 4,5-dioxygenation followed by ortho-cleavage (Rehmann et al., 1998), both seen with Mycobacterium and Rhodococcus sp. that are able to

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs) 6

5

7

4

8

3 O2

233

9

2 1

O2

10

OH

Pyrene

OH

OH OH

O2

O2

OH COOH COOH COOH CO2

O

9-Hydroxy-perinaphthylidiene pyruvate (product E) Thermal decomposition in GC analysis OH

OH

COOH

10-Hydroxy-1-Phenanthoric Acid (product Ga)

CH2 O

O

(product G)

Thermal decomposition in GC analysis

Figure 7.5 Proposed pathway for the cometabolism of pyrene by S. paucimobilis strain EPA 505.

grow on pyrene. In the Sphingomonas species, however, the specificity of the enzymes involved in the cleavage of either a three-carbon fragment (meta-cleavage) or a two-carbon fragment (ortho-cleavage) is apparently too narrow to allow further metabolism of the pyrene products. It is interesting to speculate that this may be true of Sphingomonas species in general because, at the present time, there are no known Sphingomonas species that are able to grow on pyrene. Mycobacterium strains, on the other hand, can be readily isolated for their ability to grow on pyrene. In preliminary studies, we have observed transient pyrene degradation products from selected Mycobacterium strains that have the same spectral characteristics as the substituted perinaphthalene product produced by EPA 505. If this is the case, then initial degradation of pyrene may be similar in both Mycobacterium and Sphingomonas species. Recent evidence demonstrates that nah-like genes can be found in Mycobacterium and Rhodococcus species (Hamann et al., 1999). However, the Mycobacterium strains clearly have the ability to further metabolize the intermediates produced from a naphthalene-like attack on pyrene. This was verified by the work of Rehmann et al. (1998) and Heitkamp et al. (1988a, 1988b), who observed further degradation products. Why Sphingomonas strains have not acquired the necessary enzymes for this transformation

234

Bioremediation of Recalcitrant Compounds

is unclear. Perhaps it is related to the organic matter that each species normally utilizes in nature. It is assumed that EPA 505 will hydroxylate other HMW PAHs in the same manner and produce ring-opened intermediates that cannot be further metabolized. Thus, this strain appears to have a remarkable breadth of cometabolic capability due to very loose specificity of the dioxygenase enzymes that are responsible for metabolizing phenanthrene and fluoranthene. 7.1.3.5.4 Biodegradation pathways. As microorganisms in the environment are confronted with PAHs of increasing numbers of aromatic rings, the biochemical strategy for removing carbon fragments that can be oxidized through the intermediary metabolism of the organisms becomes more complex. In many cases, the activity of PAH-degrading enzymes is constitutive, yet the presence of certain growth PAHs seems to increase enzyme activity (Aitken et al., 1998). Stringfellow and Aitken (1995) and Stringfellow et al. (1995) have shown that the growth substrate phenanthrene (and salicylate) was able to induce the cometabolism of fluoranthene and pyrene, both of which were not growth substrates in Pseudomonas saccharophilia strain P-15, an isolate from PAH-contaminated soil. Studies with the same organism also revealed that chrysene and benzo(a)anthracene could be mineralized and the mineralization was stimulated by pregrowth on phenanthrene (Chen and Aitken, 1999). Chrysene was not a metabolic inducer. Benzo(a)pyrene was not mineralized significantly regardless of whether cells were pregrown on phenanthrene or not. In soils, it is possible that an inducer such as salicylate could be added directly to the soil to induce greater PAH degradation activity, but it is probably impractical in the final analysis (Chen and Aitken, 1999). The more likely scenario is to make the HMW PAHs as available as possible during the time when the inducing PAHs, most likely phenanthrene, are actively being degraded. The potential of sequential degradation of the PAHs complicates issues of cometabolic induction. Phenanthrene is usually in high concentration in PAH-contaminated soil. If it is competing with the cometabolized PAH for the same active site of an enzyme, cometabolism is likely to be slowed by phenanthrene, assuming it is the preferred substrate for the enzymes. In studies examining the effect of PAH mixtures on cometabolism, Luning Prak and Pritchard (2002a) have shown that phenanthrene does inhibit the cometabolism of pyrene in strain EPA 505. These experiments were carried out in the presence of the chemical surfactant Tween 80 in order to produce concentrations of PAHs that could be readily followed by HPLC analysis. In these studies, the rate of phenanthrene degradation is the same in the presence and absence of pyrene, suggesting that if only one enzyme system is responsible for the initial PAH metabolism, then phenanthrene is clearly the preferred PAH substrate. The presence of phenanthrene does not totally preclude pyrene cometabolism (i.e., there is some decay over time), but pyrene metabolism is much slower in the absence of phenanthrene.

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

235

However, once phenanthrene was degraded, pyrene metabolism commenced at a faster rate than that seen with no preexposure to phenanthrene. These results have important implications for land-farming treatment. If the bioavailability of the PAHs can be enhanced by the addition of surfactants, then once the readily degradable PAHs are removed, one can assume that, for a short period, the microbial communities will be fully induced and this will be the time in which maximum cometabolism can be expected. Thus, if organisms with broad cometabolic capabilities are to be added to soil during land-farming treatment, timing is important. They must be added during a time early enough in treatment when they will be induced by growth on low-molecular-weight PAHs, yet not so late that they are outcompeted for these low-molecular-weight PAHs by the indigenous microbial communities. 7.1.3.5.5 Surfactants to enhance bioavailability. 7.1.3.5.5.1 Chemical surfactants. One limitation to degradation of HMW PAHs is their low solubility in water and high affinity for surfaces. Bioavailability of HMW PAHs in soil has been extensively studied, and it is clear that slow desorption of the PAHs from soil particles plays a key role in the ability of bacteria to degrade these PAHs. As a contaminated soil ages, PAHs tend to move into the deeper recesses of soil particles, soil aggregates, and the organic matter sorbed to soil particle surfaces (Cornelissen et al., 1998; Jixin et al., 1998; Pignatello and Xing, 1996; Zhang et al., 1998). Consequently, desorption is usually described as a rapid initial release of PAHs that are close to the surface and a very slow release of PAHs that are more deeply sorbed. Although considerable amounts of sorbed PAHs will eventually leach out over years, this time frame is usually too long for shorter-term remediation techniques, such as land-farming treatment (months), to be effective. If strategic modifications of bioremediation techniques can be made to increase desorption rates over the shorter treatment term, then the added amount of degradation may mean meeting cleanup goals in a reasonable time. Because most microorganisms take in PAHs from the aqueous phase, their degradation rate is often limited by the mass transfer from the sorbed or nonaqueous liquid phase into the aqueous phase. One method to enhance PAH transfer rates into the aqueous phase is to add surfactants to increase micellar solubilization of the PAHs (Luning Prak and Pritchard, 2002b). Surfactants have been found to enhance the degradation rates of individual PAHs in pure and mixed cultures (Grimberg et al., 1996; Guha and Jaffe, 1996; Guha et al., 1998; Liu et al., 1995; Madsen and Kristensen, 1997; Tiehm, 1994; Volkering et al., 1995; Willumsen et al., 1998). Success is controlled by the type and concentration of surfactant utilized and the type of organisms tested. Many surfactants can be toxic to the microorganisms used. In the work of Willumsen et al. (1998), Tween 80 (0.24 mM) had a stimulatory effect on the mineralization of fluoranthene by both Sphingomonas and Mycobacterium sp., but Triton X-100 (0.48 mM) was quite toxic to most PAH degraders,

236

Bioremediation of Recalcitrant Compounds

as determined by their ability to mineralize glucose (Figure 7.6). Fan 9 and VF1 are both Mycobacterium strains. EPA 505 and FLA 10-1 are both Sphingomonas strains. Interestingly, one of the strains tested in these experiments was able to recover from the effects of Triton X-100 after extended incubation. This suggests that the surfactant interacts with the cell membrane, perhaps allowing for greater transport of PAHs inside, and that the extent of this interaction determines the eventual toxicity.

E

B

F

C

G

Strain EPA505

60 50 40 30 20 10 0 60 50 40 30 20 10 0

D

H

Strain FAn9

A

70 60 50 40 30 20 10 0

Strain VF1

40

Accumulated 14CO2 (%)

Mineralization of 14C-fluoranthene

Strain 10-1

Mineralization of 14C-glucose

30 20 10 0

0

10

20 30 hours

40

50

0

50

100 hours

150

Figure 7.6 Mineralization of fluoranthene and glucose by four fluoranthene-degrading strains in the presence (squares and triangles) and absence (circles) of surfactants. Surfactants were Tween 80 (squares) and Triton X-100 (triangles). (From Ho, Y. et al., J. Ind. Microbiol. Biotechnol., 24, 100–112, 2000.)

Chapter seven: Polycyclic aromatic hydrocarbons (PAHs)

237

Table 7.5 Mass Balance of Radioactivity and Fluoranthene after 48 h Incubation of EPA 505 Cells with Different Concentrations of Triton X-100

Triton 0% 0.005% 0.03% 0.1% 2.0% Killed cells

Carbon FLA Dioxide Residual Recovery of Soluble Cell Mass (mg) Radioactivity Radioactivity Radioactivitya Radioactivity 81 100 95 95 96 100

4 4 9 12 26