EPA/600/R-94/011
March 1994
20TH ANNUAL RREL RESEARCH SYMPOSIUM
ABSTRACT PROCEEDINGS
Coordinated by:
Science Applications International Corporation
Ft. Washington, PA 19034
Contract No. 68-C2-0148
Work Assignment No. 2-7
Work Assignment Manager:
Emma Lou George
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
Printed on Recycled Paper
-------
FOREWORD
Today's rapidly developing technologies and industrial practices frequently carry
with them the increased generation of materials, that if improperly dealt with, can threaten
both public health and the environment. The U. S. Environmental Protection Agency is
charged by Congress with protecting the Nation's land, air, and water resources. Under
a mandate of national environmental laws, the Agency strives to formulate and implement
actions leading to a compatible balance between human activities and the ability of
natural systems to support and nurture life. These laws direct the EPA to perform
research to define our environmental problems, measure the impacts, and search for
solutions.
The Risk Reduction Engineering Laboratory is responsible for planning,
implementing, and managing research, development, and demonstration programs to
provide an authoritative, defensible engineering basis in support of the policies, programs,
and regulations of EPA with respect to drinking water, wastewater, pollution prevention,
solid and hazardous waste, and Superfund-related activities. This publication is one of
the products of that research and provides a vital communication link between
researchers and users.
These Abstract Proceedings from the 1994 Symposium provide the results of
projects recently completed by RREL and current information on projects presently
underway. Those wishing additional information on these projects are urged to contact
the author or the EPA Project Officer.
RREL sponsors a symposium each year in order to assure that the results of its
research efforts are rapidly transmitted to the user community.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
-------
NOTICE
The abstracts presented in these Proceedings (with the exception of those from
the Hazardous Substance Research Centers) have been reviewed in accordance with the
U.S. Environmental Protection Agency's peer and administrative review policies and
approved for presentation and publication. Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.
in
-------
ABSTRACT
The Twentieth Annual Risk Reduction Engineering Laboratory (RREL) Research
Symposium was held in Cincinnati, Ohio, March 15-17, 1994. The purpose of this
Symposium was to present the latest significant research findings from ongoing and
recently completed projects funded by the Risk Reduction Engineering Laboratory.
These Proceedings are organized into two sections. Part One contains extended
abstracts of the paper presentations. Part Two contains abstracts of the poster displays.
Subjects include pollution prevention demonstrations and life cycle analysis; remediation
technologies from the SITE Program, RREL technologies, and oil spills remediation
technologies; drinking water and wastewater technologies; municipal solid waste
technologies; and hazardous waste technologies.
IV
-------
ACKNOWLEDGEMENTS
The 20th Annual RREL Research Symposium was planned in partial fulfillment of
Contract No. 68-C2-0148, Work Assignment No. 2-7 by Science Applications International
Corporation (SAIC) under sponsorship of the U.S. Environmental Protection Agency.
Emma Lou George of the Risk Reduction Engineering Laboratory (RREL) was the Work
Assignment Manager responsible for coordinating this project. The conference program
and activities were planned by a committee consisting of the following individuals: Emma
Lou George, Lou Garcia, Franklin Alvarez, Randy Parker, and Walter Feige of RREL and
Lisa Kulujian of SAIC.
-------
PART ONE
Using Life-Cycle Assessment to Evaluate Pollution Prevention Options in the Printing
Industry
Mary Ann Curran, U.S. EPA, RREL 1
Chemical Ranking for Potential Health and Environmental Effects
Gary Davis, University of Tennessee 6
NMP-Based Coatings Remover at Tooele Army Depot
Johnny Springer, Jr., U.S. EPA, RREL 7
Selected Technology Assessments and Evaluations Under the WREAFS Program
Kenneth R. Stone, U.S. EPA, RREL 10
Recycling Nickel Electroplating Rinse Waters by Low Temperature Evaporation and
Reverse Osmosis
Paul M. Randall, U.S. EPA, RREL , . 13
Source Reduction in the Pulp and Paper Industry
Thomas J. Holdsworth, U.S. EPA, RREL 18
Pesticide Treatability Data Base, Version 2.0
T. David Ferguson, U.S. EPA, RREL 20
The Eco Logic Gas-Phase Chemical Reduction Process
Gerard W. Sudell, Foster Wheeler Enviresponse, Inc 24
Vitrification of Superfund Site Soils and Sludges
Emilio D. Spinosa, Ferro Corporate Research 28
Photothermal Destruction of Off-Gas from SVE and Thermal Desorption Treatment
Processes
Chien T. Chen, U.S. EPA, RREL 29
Potential Application and Limitations of TiO2 Photocatalytic Oxidation for the
Inactivation of Microorganisms
James Owens, U.S. EPA, RREL , 35
SITE Demonstration of the Colloid Polishing Filter Method (CPFM)
Annette M. Gatchett, U.S. EPA, RREL 38
Controlling Copper, Lead and Iron in Small Systems
Darren A. Lytle, U.S. EPA, RREL 41
VI
-------
Ultrafiltration Membrane Research for Small Drinking Water Systems
James A. Goodrich, U.S. EPA, RREL 46
Point of Use Treatment for Arsenic Removal
Kim R. Fox, U.S. EPA, RREL 52
The Effectiveness of Sorbents for Solid-Bed Metal Capture in an Incinerator:
Screening Tests at the Incineration Research Facility
Gregory J. Carroll, U.S. EPA, RREL . 55
Pilot-Scale Incineration Tests of UDMH and Nitrogen Tetroxide - Former Soviet Union
Liquid Ballistic Missile Propellants
Donald A. Oberacker, U.S. EPA, RREL 60
Evaluation of a Rotary Kiln Incinerator as a Thermal Desorber
Justice A. Manning, U.S. EPA, CERI 64
Results of the MITE Program's Material Recovery Facility (MRF) Evaluations
Lynnann Hitchens, U.S. EEPA, RREL 71
Full-Scale Leachate-Recirculating MSW Landfill Bioreactor Assessments
David A. Carson, U.S. EPA, RREL . 75
Protocol - A Computerized Solid Waste Quantity and Composition Estimation System
Albert J. Klee, U.S. EPA, RREL 80
Price Information and the Hazardous Waste Remediation Industry
Gordon M. Evans, U.S. EPA, RREL 84
Using Fourier Transform Infra-red Spectroscopy (FT-IR) to Monitor the Progress of
Plant Based Bioremediation Efforts
Lawrence C. Davis, Kansas State University 88
Field Scale Evaluation of Grass-Enhanced Bioremediation of PAH Contaminated Soils
Darwin L. Sorensen, Utah State University 92
Isolation of Pollutants Using a Biobarrier Technology
J. William Costerton, Montana State University 95
**
Kinetics of Biodegradation, Sorption and Desorption of Phenol, Substituted Phenols
and Polycyclic Aromatic Hydrocarbons in Soil Slurry Systems
Henry H. Tabak, U.S. EPA, RREL 102
VII
-------
Membrane Biofiltration in Anaerobic/Aerobic Applications
Amit Pundit, U.S. EPA, University of Cincinnati ....... , ............ 109
Recent Advances in Biofiltration
Rakesh Govind, University of Cincinnati ......................... 115
Remediation of Contaminants with Biological Activated Carbon Systems
Thomas C. Voice, Michigan State University . . .................... 122
Soil Slurry Bioreactors Bench Scale Studies
John A. Glaser, U.S. EPA, RREL ....................... ....... 127
Use of Composting Techniques to Remediate Contaminated Soils and Sludges
James H. Johnson, Jr., Howard University ............... ........
Use of Chemical Dispersants for Oil Spills in Marine Waters
Daniel Sullivan, U.S. EPA, RREL .................... . ......... ''35
Nutrient Application Strategies for Oil Spill Bioremediation in the Field
Albert D. Venosa, U.S. EPA, RREL ............................ 139
Bioventing of a Jet Fuel Spill in a Cold Climate with Soil Warming: A Field Evaluation
Gregory D. Sayles, U.S. EPA, RREL ........................... 144
SITE Demonstration of Pneumatic Fracturing and Hot Gas Injection
Uwe Frank, U.S. EPA, RREL .................... . ............ 150
Case Study of the Application of Soil Vapor Extraction-Air Sparging Technology to
Leaking UST Site
Chi-Yuan Fan, U.S. EPA, RREL ............................... 156
Particle Separation (Soil Washing) Process for the Treatment of Contaminated Soils
Peter Wood, Warren Spring Laboratory ......................... 161
SITE Program Demonstration of In Situ Steam Enhanced Recovery Process at the
Rainbow Disposal Site in Huntington Beach, California
Paul de Percin, U.S. EPA, RREL ..... ......................... 165
Electrokinetic Soil Remediation - A Pjlot Scale Study
Yalcin B. Acar, Louisiana State University ........................ 168
COGNIS Terramet™ Lead Extraction Process
William E. Fristad, COGNIS, Inc ....................... ---- ---- 173
VIII
-------
Acid Extraction Treatment System (AETS) for Treatment of Metal Contaminated Soils
Stephen W. Paff, CHMR 178
Treatment of Organic Wastes in Aqueous Matrices by X-Ray
Esperanza Piano Renard, U.S. EPA^ RREL 183
Recovery of Lead by a Chelation/Electromembrane Process
Ronald J. Turner, U.S. EPA, RREL 191
Ultraviolet Light Degradation of Polychlorinated Biphenyls
Marilyn Barger, Hofstra University 196
Remediation of Metal-Contaminated Soil by Electric Fields
Ronald F. Probstein, Massachusetts Institute of Technology .. . 201
Fate of Terpenes in the Activated Sludge Process
Franklin R. Alvarez, U.S. EPA, RREL 206
Experimental and Detailed Modeling Studies of Pyrolytic and Oxidative Processing of
Chlorocarbon/Hydrocarbon Systems
Robert B. Barat, New Jersey Institute of Technology 208
Release of Chlorinated Organic Compounds from a Contaminated Estuarine Sediment
Spyros G. Pavlostathis, Georgia Institute of Technology 212
Flume Studies on the Detachment of Kaolinite Clay and Associated Contaminants
from a Coarse Sediment
T.W. Sturm, Georgia Institute of Technology 216
PART TWO
The U.S. EPA Incineration Research Facility
Robert Thurnau, U.S. EPA, RREL 221
Four Year Study of Asbestos in New Jersey Public Schools
Thomas J. Powers, U.S. EPA, RREL 222
Update of the EPA Developed Full Scale Debris Washing System
Naomi P. Barkley, U.S. EPA, RREL 224
IX
-------
An Evaluation of Drinking Water Samples Prepared Using Ultrafiltration and UV-TiO2
Technologies
Kathleen S. Patterson, U.S. EPA, RREL 225
RREL Site Remediation Technical Support Program
Benjamin L. Blaney, U.S. EPA, RREL 226
The Environmental Protection Agency's Innovative Technology Program
Norman J. Kulujian, U.S. EPA, Region III 227
Evaluation of In-vessel Composter Designs for Hazardous Waste Treatment
John A. Glaser, U.S. EPA, RREL 229
Acoustic Location of Leaks in Pressurized Underground Petroleum Pipelines
Robert W. Hillger, U.S. EPA, RREL 230
Plunging Water Jets: Evaluating an Innovative, High Current, Diversionary Oil Boom
John S. Farlow, U.S. EPA, RREL 231
Application of the Electron Beam Treatment Process to Multi-Source Hazardous
Waste Leachate Treatment
William J. Cooper, High Voltage Environmental Applications, Inc 234
Photolysis/Biodegradation of PCB Contaminated Soils
Ed Alperin, IT Corporation 235
USEPA - DOE Joint Assessment Program
Emma Lou George, U.S. EPA, RREL 236
Decision-Support Software for Soil Vapor Extraction Technology Application:
Hyperventilate
Chi-Yuan Fan, U.S. EPA, RREL 237
Removal of Organic Compounds from Drinking Water Using Membrane Technology
Carol Ann Fronk, U.S. EPA, RREL 238
Packaged Water Treatment Plant Operation and Field Data Documentation Project
Susan Campbell, U.S. EPA, RREL 239
Risk Reduction Engineering Laboratory (RREL) Drinking Water Technology Activities -
U.S. EPA
Clois J. Slocum, U.S. EPA, RREL 240
-------
Pollution Prevention Research Branch Program
Emma Lou George, U.S. EPA, RREL . .
Using Data Quality Objectives (DQOs) to Search for Hot Spots
Esperanza Piano Renard, U.S. EPA, RREL
241
242
XI
-------
-------
: USING LIFE-CYCLE ASSESSMENT TO EVALUATE POLLUTION PREVENTION
OPTIONS IN THE PRINTING INDUSTRY
Mary Ann Curran
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7837
Duane To "Me
Battel1e
505 King Avenue
Columbus, Ohio 43201-2693
(614) 424-7591
INTRODUCTION
The Risk Reduction Engineering Laboratory is evaluating the use of the
Life-Cycle Assessment (LCA) approach in selecting pollution prevention (P2)
options. Typically, Pz options which are available to a process operator for
"improving" the process are assessed using limited information with the final
selection often largely based on cost. The idea behind this study was to
expand this decision-making process to include life-cycle, cradle-to-grave
considerations leading to a better decision. A framework method was developed
for calculating a matrix scoring system. A general framework was developed
that lean be applied to many different industries, however, only a subset of
these criteria would be selected for evaluating P2 activities for a specific
industry. The lithographic printing industry was selected as a case study for
illustrating the application of the framework.
METHODOLOGY
i
: The preliminary framework methodology has been developed so that it can
use industry average data for an entire industry or site-specific data for an
individual company to determine which P activities result in the greatest
environmental improvement. This information can be used along with other
factors, such as cost, manufacturability, and performance, to break a tie
between similar Pd activities. It is expected that users of this methodology
include both industry and government.
The P2/LCA methodology has some limitations. The alternatives are
expressed in the form of fractions, where the denominator is the score for a
criterion before application of a specific P2 activity and the numerator is
the s,core for the same criterion after implementation of that P2 activity. A
specific P activity in a given industry can be compared on a relative basis
with :others calculated for the same industry to see which activity provides
the greatest, relative to the original process, environmental improvement.
However, the final result should not be used to claim that the process is good
or bad for the environment. The "scores" are only for selected criteria in
specific life-cycle stages that are expected to change due to the P2 activity.
The results are not based on all possible activities and do not represent all
possible impacts in all life-cycle stages, and thus do not represent a full
LCA. i
-------
A master list of impact categories was developed based on the work by
SETAC. As defined by SETAC , stressors are conditions that may lead to human
or ecological impairment or resource depletion. Stressor/impact chains can be
developed by considering the energy, water, and raw material inputs to each
life-cycle stage, as well as air, water, and solid waste emission outputs from
each life-cycle stage. The inputs and outputs can then be compared against a
list of potential impacts in order to develop stressor/impact chains. The
development of stressor/impact chains is designed to focus the P2 evaluation
only on those stressors and associated impacts that are expected to change in
one or more life-cycle stages as a result of implementation of a P2 activity.
During the calculation of a specific P2 activity for a particular industry,
only a subset of the criteria will be relevant. Furthermore, the
stressor/impact chains discussed above may indicate that significant changes
resulting from implementation of a P2 activity are only likely to occur during
selected stages for a certain criterion. In addition, some criteria are
relevant from an industry-average standpoint (i.e., when an entire industry is
evaluated), and other criteria may only be appropriate when the P2 factor is
determined from a site-specific standpoint (i.e., a single company at one
location).
RESULTS
To demonstrate the usefulness of this preliminary framework, the
lithographic printing industry was selected for a case study. Many different
P2 activities have been implemented by one or more of the seven lithographic
printers that agreed to provide information for this project. The selection
of a solvent substitute for blanket or press wash was selected for
demonstration and is modelled after a specific printer that has made two
changes in their blanket and press wash over a five-year period. Each change
was designed to incorporate the use of reduced volatile organic compound (VOC)
releasing solvents. The company started with 543 Type Cleaner in 1988,
changed to Ultra Fast Blanket Wash 2215 in 1990, and made another switch to
1044 Press Wash in 1993.* Records were kept of the quantity of solvent used
each year and the total sales increased by 61% over the five year period.
Based on the stressor/impact chain for this particular process, eight
scoring criteria were selected in three life cycle stages; raw materials
acquisition stage, materials manufacture (petroleum refining), and product
fabrication (printing). These criteria (see Table 1) were selected from a
larger list of scoring criteria because specific stressors are expected to
change due to the decreased volatility of the blanket/press wash mixture. A
ninth scoring criterion is available as a fall-back option for scoring
stressors that cannot be readily quantified with available data. Since not
all of the data are available for accurate scoring of every criterion in this
example, hypothetical scores have been developed to demonstrate the
calculation process for the original solvent wash and each substitute. The
scores of all three solvents are shown in Table 2.
* Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
TABLE 1. SCORING CRITERIA FOR CALCULATION OF SOLVENT SUBSTITUTION
Raw Material Acquisition Stage
Resource Renewabilitv
9
7
5
3
: i
*
Renewability
Renewability
Nonrenewable,
Nonrenewable,
Nonrenewable,
Insufficient
< 1 year
1-10 years
sustainability
sustainability
sustainability
data
> 500 years
50 - 500 years
< 50 years
Material Manufacturing Stage
Petroleum Refining - Energy Usage
9
7
5
3
1
*
< 5,000 BTU/lb
5,000 - 10,000 BTU/lb
10,000 - 20,000 BTU/lb
20,000 - 30,000 BTU/lb
> 30,000 BTU/lb
Insufficient data
Petroleum Refining - Airborne Emissions
9 < 20 mg/lb
20 - 200 mg/lb
200 - 500 mg/lb
500 - 2,000 mg/lb
> 2,000 mg/lb
Insufficient data
7
5
3
1
*
Petroleum Refining - Waterborne Effluents
: 9 < 1,000 mg/lb
7 1,000 - 2,000 mg/lb
5 2,000 - 3,000 mg/lb
3 3,000 - 5,000 mg/lb
1 > 5,000 mg/lb
: *
Insufficient data
Product Fabrication Stage
Printing - VOC Emissions
9 < 20 mg/lb
7 20 -200 mg/lb
5 200 - 500 mg/lb
3 5,00 - 2,000 mg/lb
1 > 2,000 mg/lb
* Insufficient data
-------
Printing - Ozone Depleting Potential Relative to CFC-11
9 < 0.01 (all non-halogenated chemicals, HFC-125,
HFC-134a, HFC-143a, HCFC-152a)
7 0.01 - 0.39 (HCFC-22, HCFC-123, HCFC-124, HCFC-1415,
HCFC-142b)
5 0.40 - 0.69 (CFC-115)
3 0.70 - 0.99 (CFC-114)
1 >1.00 (CFC-11, CFC-12, CFC-113, Carbon Tetrachloride,
Halon-1301)
* Insufficient data
Printing - Global Warming Potential
Relative to CO, over 100 Years
9 <1 (H-2401, H-2311)
7 1-99 (CO,, HCFC-123, H-1211, H-1202, H-2402, H-1201,
Methane, HCFC-141b)
5 100 - 499 (Nitrous Oxide, HCFC-124, HFC-152a, Methyl
Chloroform)
3 500 - 4,999 (CFC-11, CFC-113, Carbon Tetrachloride,
HCFC-22, HFC-125, HFC-134a, HCFC-142b, HFC-143a)
1 >5,000 (CFC-12, CFC-114, CFC-115)
* Insufficient data
Printing - Inhalation Toxicitv
9
7
5
3
1
*
NOAEL > 1,000 mg/m in air
NOAEL 10 - 1,000 mg/m3 in air
NOAEL 0.1 - 10 mg/m3 in air
NOAEL 0.01 - 0.1 mg/m3 in air
NOAEL < 0.01 mg/m3 in air
Insufficient data
Fall-Back Option where Data are Unavailable
Estimated Status Relative to Industry Norm
9 Much better (> 50% better)
7 A little better (25 - 50% better)
5 Equal to norm (±25%)
3 A little worse (25 - 50% worse)
1 Much worse (< 50% worse)
-------
TABLE 2. HYPOTHETICAL EXAMPLE FOR SOLVENT SUBSTITUTIONS
AT A LITHOGRAPHIC PRINTER
LCA STAGE/SCORING CRITERIA
RMA/Resource Renewability
MAN-MM/Energy Usage
MAN-MM/Airborne Emissions
MAN-MM/Waterborne Effluents
MAN-PF/VOC Emissions
MAN-PF/Ozone Depleting Potential
MAN-PF/Global Warming Potential
MAN-PF/Inhalation Toxicity
• COLUMN TOTALS
CRITERIA SCORES
FIRST
SOLVENT
3
3
5
5
5
7
7
5
40
SECOND
SOLVENT
3
5
7
7
5
7
9
5
48
THIRD
SOLVENT
7
3
7
7
7
9
9
7
54
RMA; = Raw Materials Acquisition, MAN = Manufacturing Stage,
MM = Materials Manufacture, PF = Product Fabrication
Based on the sum of the individual scores, the first solvent switch
would be rated 1.20 (48/40) while the second solvent switch would be rated
1.35 (54/40). These ratings can then be compared with other P2 activities for
the lithographic printing industry. The higher the score, the lower the
environmental impacts.
CONCLUSIONS
While the results of this application of the P2/LCA framework to the
lithographic printing industry are only preliminary and involve hypothetical
data, the case study indicates good potential for evaluting P2 options based
on life-cycle considerations. The solvent changes made by this particular
printer for the blanket wash were selected based on a single criterion, i.e.,
low VOC emissions, however, the additional, applicable life-cycle information
for the new solvents indicates that these changes have effected overall
environmental improvements as well.
REFERENCES
1. '• Fava, J.A., F. Conso'li, R. Denison, K. Dickson, T. Mohin, and B.W.
Vigon. A Conceptual Framework for Life-Cycle Impact Assessment. Workshop
Proceedings. Society of Environmental Toxicology and Chemistry,
Sandestin, Florida, -1992.~
-------
THE UNIVERSITY OF TENNESSEE
KNOXVltLE
Center for Clean Products and Clean Technologies
327 South Stadium Hall
Knoxville, Tennessee'3 7996-0710
U;S:A.
(615) 974-8979
FAX (615) 974-1838
Chemical Ranking for Potential Health and Environmental Effects
Between 60,000 and 100,000 of the over 8,000,000 chemicals listed by the Chemical Abstracts
Services Registry are-commercially produced. Each of these has the potential of becoming
an environmental pollutant, but the time and resources required to perform thorough testing
of the potential human health and environmental effects of chemical exposure can be
prohibitive. Chemical ranking and scoring systems that evaluate the toxic effects of
chemicals and have an exposure component' are science-based screening tools that can
rapidly assess relative chemical hazards. Several chemical ranking arid scoring systems have
been or are being developed for the purposes of regulatory action, priority setting or impact,
evaluation, but there currently is no scientific, consensus oh chemical ranking. Furthermore,
many of the existing methods focus on a single environmental media, or have been
demonstrated on only a limited number of chemicals.
The University of Tennessee Center for Clean Products and Clean Technologies developed
a'chemical ranking and scoring method under EPA Cooperative'Agreement CR 816735, The,
Product Side of Hazardous Waste Reduction: Evaluating the Potential for Safe Substitutes.
The method was designed as'a priority-setting tool, to recommend a set of priority chemicals
for substitute evaluation. The method provides an approximate ranking 'of chemical hazards
based on their relative toxicity and potential for exposure. The method was demonstrated
using the chemicals for which toxic chemical release reporting is made in the Toxic Release
Inventory (TRI) as required under Section 313 of Title III of the Superfund Amendments
'and Reauthorization Act (SARA) of 1986 and high-volume pesticides selected from annual
pesticide usage data.
In the development of the chemical ranking method, three areas were examined: 1) the
availability of experimental data and what to do when experimental data are absent; 2) the
formulation of scoring criteria which could be used to estimate the toxic effects of chemicals
and the potential for exposure; and 3) the development of an algorithm, that combines and
weights these criteria in order to rank chemicals according to their potential human health
and environmental .effects:
(337)
-------
NMP-BASED COATINGS REMOVER
AT TOOELE ARMY DEPOT
Johnny Springer, Jr.
Waste Minimization, Destruction
and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
(513) 569-7542
Bruce Sass
Battelle
505 King Avenue
Columbus, Ohio 43201-2693
(614) 424-6315
INTRODUCTION
The goal of this study is to evaluate a replacement solvent
paint remover for methylene chloride and other hazardous
compounds that can be used to remove organic coatings, such as
enamels, lacquers, and varnishes from metal surfaces. Methylene
chloride is a primary component of many cold paint removers and
is one of the substances targeted by the 33/50 Program for use
reduction. U.S. EPA considers methylene chloride to be a
hazardous air pollutant because of its low exposure limit and
high volatility. The paint remover to be evaluated is based on
n-methy1-2-pyrrolidone (NMP) and also contains monoethanolamine
(MEA) . NMP is a highly versatile solvent that has been used for
more than 15 years in the chemical and petrochemical industries.
MEA is used as a co-solvent that helps accelerate removal of
paint and other organic contaminants.
Tooele Army Depot (TEAD) provided the site for this
technology demonstration. TEAD is a government-owned,
government-operated (GOGO) installation, located in Tooele, Utah
sinc
-------
The focus of this study is on the Parts Chemical Cleaning
System (PCCS), which is designed for depainting, cleaning and
applying conversion coatings to ferrous and nonferrous engine
parts and powertrain subassemblies. Only the nonferrous cleaning
line was the subject of this study. Application of conversion
coatings is a surface preparation method to provide corrosion
protection and increase adhesion of the paint coating.
PCCS is designed such that an automated overhead monorail
transports baskets of parts through tanks of paint remover and
various rinses, prior to application of conversion coatings. A
preprogrammed system controls the process by controlling the
material handling equipment that immerses the baskets into the
tanks for predetermined dwell times, drains the baskets, and
moves the baskets to succeeding tanks. The rinsewater from the
process contains paint remover, dissolved paint resins, pigments,
and other paint additives. This rinsewater is piped into a
holding tank and then is later treated by a fixed-film biological
reactor prior to merging it with a stream that feeds to the
Industrial Waste Treatment Facility. The system employs
automatic controls to regulate tank solution levels,
temperatures, agitation tank ventilation, tank heating and
solution filtration.
This study evaluated three objectives. First, the study
evaluated the ability of the replacement paint remover to remove
paints and compared these results with those using the old
technology (methylene chloride) . The pollution prevention
potential of the new paint remover and rinsewater purification
system were evaluated. Finally, the economic potential of the
new paint removal process was compared to the cost of using the
methylene chloride paint removal system.
METHODOLOGY
To evaluate product quality, test coupons were made and
processed through the paint remover system along with actual
parts. An equal number of coupons were coated with heat
resistant or chemical agent resistant coatings. The degree of
paint removal from the coupons was qualitatively evaluated. The
baskets of actual parts were evaluated on a pass/fail basis. To
evaluate the pollution prevention potential of the new paint
remover solvent system, three process streams will be evaluated.
First, the paint removal solvent was evaluated for dissolved
metals (Cd, Cr, Cu, Pb, Mn, Ni, and Zn) and concentrations of NMP
and MEA. The rinsewater entering the biological reactor was
analyzed for concentrations of metals, NMP, MEA, pH, total
suspended solids,-total organic carbon, and chemical oxygen
demand. The rinsewater was also analyzed after treatment by the
biological reactor for the same analytes.
8
-------
RESULTS
t
i Data was still being compiled at publication of this
extended abstract. Qualitative analysis of the test coupons and
parts batches indicates that the NMP-based solvent removed the
paint as well as the methylene chloride paint removal system.
Results of the study will be presented at the Symposium
-------
SELECTED TECHNOLOGY ASSESSMENTS AND EVALUATIONS UNDER THE WREAFS PROGRAM
Kenneth R. Stone
Pollution Prevention Research Branch
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati OH 45268
513/569-7474
INTRODUCTION
As Federal environmental objectives are identified for the 1990's, EPA programs are being
restructured to meet new requirements. Chief among these is the increased importance of
pollution prevention in all Federal activities. Further, all Federal departments and services have
been directed to develop pollution prevention opportunities in order to decrease the environmental
burden resulting from Government activities, and to provide opportunities in both the public and
private sector to reduce environmental risks.
In keeping with the Agency's responsibility to advise and cooperate with other Federal
departments on environmental risk reduction, the Pollution Prevention Research Branch (PPRB)
has managed a technical support effort knownas the Waste Reduction Evaluations At Federal Sites
(WREAFS) Program.
METHODOLOGY
The WREAFS Program was established to conduct research, develop and demonstrate
opportunities to reduce the generation of waste from Federal activities. Since 1988, WREAFS has
funded work on other Federal sites and it has supported RD&D with other Federal departments
through Interagency Agreements (IAG). WREAFS has sponsored pollution prevention opportunity
assessments, base-wide assessments, technology and product demonstrations, technology
evaluations, technology and methodology development, technical assistance and technology
transfer across the Federal community.
WREAFS has conducted cooperative RD&D activities with the following Federal departments
and services:
National Aeronautics and Space Administration,
Department of Defense,
Department of Treasury,
Department of Transportation,
Department of Energy,
Department of Interior,
Department of Agriculture,
Department of Veterans Affairs,
U.S. Postal Service.
WREAFS continues to provide integrated environmental support for: (1) primary research in
pollution prevention technology; (2) expanding cooperative research, development and
demonstration with other Federal departments; (3) developing technology transfer opportunities for
both public and private sector benefit. This presentation will discuss recently completed WREAFS
projects, highlighting the following two technology evaluations:
10
-------
NASA-LaRC Dry Powder Towpreg Technology Evaluation
One WREAFS project focused on the production of non-refractory composite materials and
aircraft structures made from those materials. Basic composites are graphite or carbon fibers
impregnated with some type of polymer - a process called, "prepregging." in solution prepregging,
polymer resin is placed in a solvent carrier and applied to the fiber. The solvent volatilizes off
as toxic air emissions. Polymer resins normally have a self life and must be refrigerated until the
solvent carrier has been extracted.
NASA-LaRC's Polymeric laboratory has developed a process called, "Dry Powder Towpreg", in
which the fiber stands are separated in a small air chamber and a finely-powdered polymer resin is
"dusted" onto them. The polymer dust fully impregnates the fibers just before being passed
through a furnace. The resulting composite "ribbon" is allowed to cool. Later, it can be formed
into a laminate and used in composite manufacturing without requiring changes in capital
equipment for application. This process has the potential to eliminate the use of solvents, and
reduce energy consumption because in dry powder form polymer resins do not require
refrigeration.
For this technology evaluation, EPA compared the solution and dry powder processes, using
methyl ethyl ketone to demonstrate solvent reduction for thermosets and n-methyl-pyrrolidone for
thermoplastics. The test was designed to evaluate the performance of the dry powder process
and to determine the environmental and energy impacts of each process.
MEK Substitute in Aircraft Radome Depainting
For this project, WREAFS has researched and evaluated substitutes for methyl ethyl ketone
(MEK) as cleaners and solvents in aircraft maintenance operations at Tinker Air Logistics Center
(ALC) in Oklahoma. Tinker ALC performs maintenance, including structural repair and
re-fabrication of USAF aircraft, notably the B-1B and the B-52. Tinker ALC reported using MEK at
an annual rate of 5,385 gallons to wipe-down aircraft, and 8,250 gallons to depaint aircraft
radomes.
From its research and through a CRADA with Texaco, RREL identified solvent formulations
based on propylene carbonate and n-methyl pyrrolidone as possible alternatives for MEK. Three
formulations were tested as proof-of-concept, using test coupons from condemned radomes. The
test focused on the ability of the chemical mixture to accomplish the job required, and meet the
same MILSPEC standard as MEK without modifying the operational procedures.
This study determined that a blend of 25% propylene carbonate, 50% n-methyl pyrrolidone
and 25% Dibasic ester (PC blend) removed paint in comparable time to the MEK, although a little
more scrapping was required for final removal. Hardness tests showed that the PC blend did not
embrittle the fiberglass/epoxy substrate of the radome, nor did it effect flexural properties.
Scanning with electron microscope indicated no significant damage to the fibers or fiber matrix
interface. Test samples were successfully re-painted and demonstrated complete paint adhesion.
For solvent properties, the PC blend compares favorably with MEK. The PC blend has a flash
point of 210OF (against MEK's 20OF), low toxicity, and lower evaporation rate.
CONCLUSION
Many of the WREAFS projects are selected on the basis of a pollution prevention technology
or technique's ability to prevent oir reduce compliance problems at Federal facilities. It is
11
-------
anticipated that the continuing compliance responsibilities of Federal facility operators will merge
with cleaner operating practices. That is, facility managers are looking for ways to avoid
producing wastes that require controlled disposal, and for clean substitutes for toxic chemicals,
some of which are being banned. In response, WREAFS is developing initiatives to integrate
pollution prevention with environmental compliance. In this environment, WREAFS will develop
opportunities for cutting edge RD&D and innovative technologies, driven by the compliance
incentive.
12
-------
RECYCLING NICKEL ELECTROPLATING RINSE WATERS
BY LOW TEMPERATURE EVAPORATION AND REVERSE OSMOSIS
Paul M. Randall
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Pollution Prevention Research Branch
28 West Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7673
and
Timothy C. LJndsey
Hazardous Waste Research and Information Center
One East Hazelwood Drive
Champaign, Illinois 61820
(217) 333-8940
INTRODUCTION
Electroplating operations generate rinse water wastes that are classified as hazardous due to
the presence of heavy:-metals. Typical metals and alloys used for plating include cadmium, copper, iron,
lead, nickel, gold, silver, platinum, brass and bronze. Rinse water wastes can be treated by either end-
of-pipe or in-plant recovery techniques. End-of-pipe treatments rely on chemical reactions such as pH
adjustment to precipitate metal and other plating chemicals. These methods provide an effective
means to remove the metal and cyanide species from the rinse water thus enabling reuse or discharge
of the water. However, in most plant treatment systems, waste streams from the various plating lines
are combined prior to treatment. Consequently, the resulting sludge must either be disposed of or
treated with a high temperature metals recovery system to recover the metals. These options tend to
be wasteful and expensive to implement.
Low-temperature evaporation and reverse osmosis techniques were each evaluated ( on a pilot
scale) on their ability to process rinse water collected from a nickel electroplating process. This project
was done in cooperation with Graham plating in Chicago, Illinois. The testing was conducted at
HWRIC's pilot scale laboratory.
METHODOLOGY
Low temperature evaporators(LTE) manufactured at Licon, Inc., in Pensacola, FL heat water
under a vacuum to produce steam at relatively low temperatures (150 to 160° F). The steam rises into
a condenser where distilled water results. The plating bath chemicals do not rise with the steam and
become a concentrated sjurty or solution of chemicals. The LTE unit is a model C-3, single effect, pilot
scale evaporator especially designed for conducting pilot scale tests on a variety of feed solutions.
Reverse osmosis (RO) is a pressure-driven membrane separation process in which a feed
stream under pressure (200 to 800 psi) is separated into a purified "permeate" stream and a
"concentrate" stream by selective permeation of solution through a semi-permeable membrane. The
13
-------
pressure required to force the permeate through the membrane is dictated by the osmotic pressure of
the feed stream. Membranes are constructed of a variety of materials such as aromatic polyamide,
cellulose acetate, and polyether/amide. The RO unit used in this project was an Osmonics model
PES/OSMO-19T-80SSXXC reverse osmosis machine for process evaluation. It was equipped with one
Osmonics model number 192T-MS05 thin-film, composite, spiral-wound membrane cartridge. The
solution was prefiltered through a 5 n cartridge filter for the RO testing.
Four, 55-gal drums of nickel electroplating rinse water were collected from the Graham Plating
facility and processed through the LTE(Drums A and B) and the RO (Drums C and D) systems. The
RO tests were conducted at two different operating pressures: Drum C at pressures of 250 to 300 psi
and Drum D at 350 to 380 psi. Samples of the concentrated feed solution as well as the distillate (LTE
system) and permeate (RO system) were collected at regular intervals throughout the tests as the rinse
water was processed. Nickel analyses were done to determine how efficiently the systems removed
nickel from the rinse water and concentrated it for potential recycling. Analyses for total organic carbon
(TOG) were done to indicate the fate of organic constituents (e.g., brighteners) in the rinse water.
Immediately after samples were collected, electrical conductivity measurements were made to indicate
the soluble salts present in the samples.
RESULTS
The LTE system concentrated the rinse water, which had exhibited initial nickel concentrations
of 2,540 to 4,140 mg/L to nickel levels as high as 13% to 18%. These levels are well above the 8%
required for placement into the plating bath. The concentrate, permeate, and distillate nickel
concentrations exhibited in samples collected throughout the tests have been summarized (Table 1).
Nickel concentrations increased at a steady rate until concentrations of approximately 25,000 to 30,000
mg/L were reached. Rgure 1 shows how nickel levels changed in the feed solution during the course of
the low temperature evaporation tests. This level corresponds to a point where approximately 80% to
85% of the rinse water volume had been processed. Beyond this point, nickel concentrations increased
dramatically until the final concentrations of 13% and 18% were achieved. The rinse water feed solution
volume was reduced by over 98% as a result of this process.
Table 1. Comparison of Nickel Concentrations in Concentrate, Distillate, and Permeata
Product
Low Tamp. Evap.
Drum A Drum B
Reverse Osmosis
Drum C Drum D
Concentrations at beginning of test (mg/L):
Concentrate 4.140 2.540
Distillate 2.5 2.2
Permeate — —
Concentrations at end of test:
Concentrate 179.000 128,000
Distillate 1 0.3
Permeate — —
Rallo of distillate permeate to concentrate:
Distillate - 0.02% 0.01%
Permeate — —
2,580
44.5
12,560
210
1.49%
1,425
14.5
18,200
790
1.54%
14
-------
200000 1
(00000 -
a Drum A
» Drum 8
20 40 60 80 100
% of Drum Volume Processed
Figure 1. Concentrate nickel concentration versus percent of
; drum volume processed; LIE tests.
The LIE system concentrated the organic constituents of the rinse water from initial TOG levels
of 550 to 990 mg/L to final levels of 25,000 to 26,000 mg/L TOC levels in the Graham Plating nickle
baths are normally maintained at approximately 14,000 mg/L The concentration rate of the organic
components paralleled the nickel concentration rate suggesting that little of the organic material was
lost to volatizat'on. As shown in Table 2, distillate produced by the LTE system was very low in nickel
concentration (average 0.37 to 0.71 mg/L). Additionally, TOC concentrations in the distillate were very
low (average 3.04 to 3.50 mg/L). Disadvantages of the LTE system include its relatively high ($140,000)
capital cost and high energy requirements ($20/ 1,000 gal processed). The implied rate of return of
10.6% and payback period of 6.9 yr determined in the economic assessment for this system suggest
that it is a marginal investment opportunity by today's standards. These estimates do not, however,
consider the reduced future liabilities brought about by drastically decreasing the hazardous waste
discharges from the facility.
The feed solution processed through the RO system contained initial nickel concentrations of
1,425 tp 2,580 mg/L (Table 1). Figure 2 depicts how nickel concentrations in the feed solution changed
as the solutions were processed. Nickel concentrations increased steadily until about 60% of the rinse
water volume was processed. At this point, nickel concentrations were about 4,000 to 5,000 mg/L in the
two drums. Beyond this point, nickel concentrations increased more rapidly until final concentrations of
12,560 mg/L (Drum C) and 17,900 mg/L (Drum D) were reached. The RO system exhibited superior
productivity at the beginning of the tests and productivity dropped off dramatically after about 60 % of
the feed solution had been processed. Beyond these levels, the productivity of the RO equipment
decreased dramatically as solids began to precipitate and foul the membrane. The final concentrations
achieved with the RO process were 12,560 to 18,200 mg/L (1.256% to 1.82%) and are well below the
8% nickel concentration required for the plating bath. Some of this solution could be used to replace
water losses in the electroplating process. The RO system, however, would probably produce excess
volumes of concentrated rinse, water composed of 1.2% to 1.8% nickel. This material would have to be
further processed with the use of an alternative technology such as LTE or be shipped to a facility that
could extract the nickel for use in other industrial processes.
The RO system concentrated the organic constituents present in the rinse water feed solution
from initial TOC levels of 340 to 540 mg/L to levels of 2,800 to 3,500 mg/L These concentrations
suggest that the organic bath constituents are concentrated by the reverse osmosis equipment at rates
15
-------
that parallel the nickel concentration rates. The quality of the cleaned rinse water permeate produced
by the RO equipment was directly related to the quality of the feed solution pumped into the unit.
Permeate, produced by the reverse osmosis system averaged 89 to 134 mg/L nickel (Table 2). These
levels are about 98.5% lower than the nickel concentrations present in the concentrated solution. This
solution would not, however, be acceptable for discharge to publicly owned treatment works.
7"»W« 2. Average Nickel Concentrations in Distillate^and Permeate
Distillate Ni Concentration
Standard
Mean Deviation
(mg/L) (mg/L)
Permeate Ni Concentration
Standard
Mean Deviation
/mg/L) (mg/L)
A (n-13)
B(n-l6)
C fn-22)
D (n-17)
0.71
037
0.63
0.52
89.55
134.38
49.22
202.19
20000 n
I
2
I
|
10000
D Drum C
* Drum D
20
40 60
% of Drum Volume Processed
100
Figure 2. Concentrate nickel concentration versus percent of drum
. volume processed; RO tests.
16
-------
Jhe nickel levels present in this solution could be further reduced by passjng this solution
through the reverse osmosis equipment again. TOC concentrations averaged 19.46 to 21.98 mg/L in
the permeate solution which suggests that some of the organic compounds were able to permeate the
membrane. The reverse osmosis equipment condensed the feed solution to final volumes that were
88% to 94% less than the original volumes of the two tested drums. Differences between the two tests
can be attributed to the difference in operating pressures used during the tests. Advantages of the RO
system include its relatively high production rates with respect to low concentration feed solutions.
Additionally, it would require lower capital investment (about $50,000) than a comparably sized LTE
system. Energy costs required to operate a RO system would be only about $2.50/1,000 gal processed
Disadvantages associated with a reverse osmosis system include its inability to concentrate the feed
solution to levels beyond the 12,560 to 18,200 mg/L levels revealed in this study. This factor alone
would prevent use of a stand alone reverse osmosis system at the Graham Plating facility because of
the economic impracticalities associated with the concentrate produced by the system. Another
disadvantage associated with the RO system is the lower quality permeate produced by the system.
This .solution would probably have to be reused within the plant or further processed through the
reverse osmosis system before discharge to the POTW.
Both the LTE and RO systems appear to offer advantages under specific operating conditions.
Based on this factor, the potential for utilizing these technologies in tandem was examined. The RO
system is: best adapted to conditions where the feed solution has a relatively low nickel concentration. It
can process the low concentration feed solution with relatively high efficiency to a level of 4,000 to
5,000 mg/L. At this point, the solution could be transferred to the LTE for further concentration. The
LTE system appears to be best adapted to processing solutions with relatively high nickel
concentrations. It can process these solutions so,that a concentrate solution composed of 8% or more
nickel is produced along with a very high quality distillate solution. Using the equipment within its
optimum Operating ranges would augment the ability of the systems to process the rinse water with
maximum efficiency while supplying the electroplating operation with high quality concentrate, distillate
and permeate solutions for reuse. Since the equipment would always be functioning within optimum
concentration ranges, smaller reverse osmosis and low temperature evaporation units could be
implemented than if the individual units were used alone. If this type of combined system were installed
at the Graham Plating facility, it would require a capital investment of $115,000 which would be paid
back in 2.8 yr through a 27.6% implied rate of return.
Electrical conductivity measurements taken during operation of both the low temperature
evaporation and reverse osmosis systems could be of great value during actual plant operating
conditions. The electrical conductivity data obtained in this project were well correlated with nickel
concentration, TOC concentration, and membrane flux characteristics.
CONCLUSIONS
This research evaluated LTE and RO systems on a pilot scale basis to tests their respective
ability to process rinse water collected from a nickel electroplating operation. Each system offered
advantages under specific operating conditions. The LTE system was best suited to processing
solutions with relatively high (greater than 4,000 to 5,000 mg/L) nickel concentrations. The RO system
was best adapted to conditions where the feed solution had a relatively low (less than 4,000 to 5,000
mg/L) nickel concentration. In electroplating operations where relatively dilute rinse water solution's must
be concentrated to levels acceptable for replacement in the plating bath, a combination of the two
technologies might provide the best process alternative. Initially, the RO system could be used to
concentrate the feed solution-. This could be followed by LTE processing to concentrate the solution to
levels acceptable for replacement in the plating bath.
For more information on this study or other pollution prevention research studies contact the
U.S.EPA Pollution Prevention Research Branch (513-569-7215) or Paul Randall (513-569-7673).
17
-------
SOURCE REDUCTION IN THE PULP AND PAPER INDUSTRY
Thomas J. Holdsworth
United States Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7675
SOURCE REDUCTION REVIEW PROJECT
The Source Reduction Review Project (SRRP) is an initiative of the U.S. Environmental
Protection Agency to evaluate pollution prevention alternatives during the regulatory
development process. The goal of the SRRP is to foster the use of source reduction measures
as the primary means of achieving compliance. At present the SRRP is focusing its efforts on
regulatory development within the Agency's Office of Water, Office of Air and Office of Solid
Waste. The Office of Research and Development (ORD) is working with these program offices,
as well as the Office of P9llution Prevention and Toxics and the Pollution Prevention Policy
Staff, Office of the Administrator.
The Risk Reduction Engineering Laboratory (RREL), ORD, is working on issues primarily
related to the Office of Water and the Office of Solid Waste. The primary focus of the SRRP-
related Industrial Wastewater Program is on industrial sectors that have been identified by
the Office of Water. These include the pulp and paper industry, the pesticide formulating,
packaging and repacking industry, the pharmaceutical industry and the metals and metal
products industry. At present, there are active research projects ongoing in all but the
metals and metal products industry.
PULP AND PAPER INDUSTRY
As outlined by the Environmental Defense Fund and the National Wildlife Foundation
settlement agreement, the Agency is investigating process changes and pollution prevention
strategies for the pulp and paper industry. A sampling effort has been completed at a
pulping plant in Alabama. The Agency sampled this plant's pulping, bleaching and wastewater
treatment technologies. Included in this evaluation was the Kamyr, or extended cooking
process, which is a pollution prevention technology because it focuses on removing lignin
from pulp prior to bleaching. This increases the quantity of organic material recycled and
burned in the recovery boiler for energy recovery rather than being removed in the bleach
plant as in the conventional process. The extended cooking process also decreases the amount
of bleaching chemicals required, thereby reducing the likelihood of forming and discharging
chlorinated organics. The key analytes of concern are volatile organics, chlorinated dioxins
and furans, chlorinated phenolics, adsorbable organic halogens (AOX), organic halides (OX),
COO BOD. TSS and color. Based on the samples collected, the dioxin and furan results
indicate very low "hits" (all liquid samples were 11 parts per quadrillion or less; the
sludge samples were detected at 1.0 parts per trillion or less). There were sporadic "hits"
of chlorinated phenolics in the pulp line and final effluent, but a review of plant operating
data indicates a discrepancy with these values.
A brief sampling effort at a pulp and paper plant in Virginia to characterize this
plant's pulping, bleaching and wastewater treatment technologies has been C9mpleted. This
plant has incorporated an ozone system for its bleaching process as a substitute for
chlorine. By using an oxidant such as ozone rather than chlorine, the likelihood of forming
and discharging chlorinated organics is eliminated. This study was conducted jointly with
the Office of Air to evaluate the effects of ozone bleaching in both the aqueous and air
media. However, the air analysis portion of this program was completed by making an estimate
of air emissions based upon contaminant concentrations determined in water and slurry
samples.
Over the past three years, ORD and the Office of Water have jointly investigated novel
pollution prevention processes within the pulp and paper industry. These processes have
included oxygen delignification, extended cooking, and ozone bleaching. The investigation of
18
-------
Totally Chlorine Free pulping processes is another classic pollution prevention technology
that;will be targeted for investigation by ORD. This technology is based on using oxygen
peroxide and other reagents to achieve a pulp with comparable strength and brightness as'
chlorine-bleached pulp. The European and Nordic nations currently lead-the-way in
investigating these technologies along their bleach lines
19
-------
PESTICIDE TREATABILITY DATA BASE. VERSION 2.0
T. David Ferguson
U.S. Environmental Protection Agency
Risk,Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 42568
(513) 569-7518
and
Brian K. Hoppy
SAIC
11251 Roger Bacon Drive
Reston, Virginia 22090
PESTICIDE TREATABILITY DATA BASE. VERSION 2.0
As a result of amendments to the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA '88),
the U.S. Environmental Protection Agency (EPA) no longer accepts stockpiles of pesticides that have
been suspended or canceled, but reviews instead the pesticide manufacturer's proposed disposal
methods In ongoing studies of various problems associated with pesticide disposal, the EPA has
identified and categorized treatability data and disposal methods for various pesticides. The
Pesticide Treatability Data Base was created for use by the Office of Prevention, Pesticide and Toxic
Substances (OPPTS) and for others to use to disseminate this important information.
The Pesticide Treatability Data Base lists more than 1500 pesticides currently in use in the
United States or removed from the market in the last 20 years. Physical and chemical properties of
the compounds treatability data, Freundlich isotherm data, and other names by which the pesticides
are known are listed, when available, as well as complete references for the sources from which the
data were derived.
A thorough search was performed of all available reference material and scientific and technical
journal articles concerning the general characteristics of pesticides and the treatability/destruction
of pesticides. All reference material retrieved was reviewed and all chemical and physical data
concerning the pesticides were transferred to a standardized form. These standardized forms were used
by a technician to transfer the information from the reference material to the data base.
All treatability/destruction selections were reviewed to ascertain if the appropriate technical
information was included. The technical information required included:
• Experimental procedures
• Influent levels
• Effluent levels or percentage of chemical removed
Once the appropriate information was retrieved from each article, the data were placed on
standardized forms and entered by a technician. A stringent Quality Assurance/Quality Check system
was devised and strictly adhered to in order to maintain accurate transfer of data.
The Pesticide Treatability Data Base allows access to information on the user's own PC. The
data base program can run on any version of DOS and is assembled in such a "user-friendly" way that
anyone witn even the most basic knowledge of computers can access the information successfully, saving
valuable time and resources.
Upon successful completion of the installation of the program and data base files, the data base
can be accessed. Due to the fact that each active ingredient can have a variety of names, the data
have been made accessible to-searches using any of the following information:
• Compound name
• Chemical Abstract Service (CAS) number
• Trade names and synonyms
20
-------
Choosing the "display""option allows the user to view"the first'sfTppfTrfahTp iT
provides the chemical/physical data for the active Ingredient h
This screen
Version 2.0 Pesticide Treatability Data Base
Search on: MALATHION
DIMETHYL PHOSPHORODITHIOATE OF DIETHYL MERCAPTOSUCCINATE, 0,0-
09/29/92
CAS Number • 121-75-5
Compound Class : I
Compound Formulation..: D, EC OS P ULV C
Molecu!ar Formula : C10-H19-06-P-S2
Compound Type..... : OP
CHEMICAL AND PHYSICAL PROPERTIES
Physical Form • |_
Molecular Weight(g/mole)..'.'-. 330 38
Melting Point(C).... • 2 9'
Boiling Point(C) : 156 TO 157
Vapor Pressure OTCO.TORR..: 4E-05 @ 30
Solubility in Water @T(C),mg/L : 145
Solubility in other solvents
Stability :
Log Octanol/Water Partiti6n'c6efficient:
Henrys Law constant,atm x m3mole-l
Pesticide Compound Code..: 05770
WP
REFERENCE
0008
0514
0509
0526
0526
0509
MOST ORG 0008
0002
512
NA
NA
F2-Reference Search
F8-Print search
F4-Abbreviations
F9-Exit search
TABLE 1. FIRST SCREEN
PgDn-Next screen
Version 2.0 Pesticide Treatability Data Base : : ~
Search on: MALATHION
Treatment Technology Data
Technoloav
Fil
HT
GAC
GAC
GAC
F2-Ref
F5-SIC
—
i Source
, Matrix
RCRA
WOS
C
C
C
Search
Code List
11 "
Seal
F
B
P
P
P
Influent"
I Level Units
' 31.7 U^TL
1.00 g/L
60 mg/L
100 mg/L
20 mg/L
NOTE: C
- calculated
Effluent
Level Units
0 mg/L
OC g/L
0 • mg/L
0.9 mg/L
1 mg/1
value
F3-Treatment Technology
FlO-to continue
F4-Othe
09/29/92
Percent
Improv.
100C
100
100C
99.1C
95C
Ref
0206~
0325
0206
0206
0206
21
-------
The user can hit F3 to define the "Treatment Technology" abbreviations used, for ^Pj^AC" is
-
down option is available to view the Isotherm Screen (Table 3).
Version 2.0 Pesticide Treatability Data Base
DIMETHYLnpHOSPHORODITHIOATE OF DIETHYL MERCAPTOSUCCINATE, 0,0-
FREUNDLICH ISOTHERM DATA
Absorbent Matrix K
MLSS CLW 403
F2-Reference Search
F5-SIC Code List
1/N
0.6
C
Units
ug/L
X/M
Units
ug/gm
F4-Other Abbrev.
F8-Print search
Reference
0129
PgUp-Previous screen
F9-Exit search
At this noint the user will have seen all the data and may wish to find out where the data came
fron By oLK F2 and tyoTng the reference number of the data one is interested in, a reference
screen will appear (Table 4).
WEBER
Data also available
Reference number: 0129
and Jones B.E., "Toxic Substance Removal in Activated
EPA Reoort No • EPA 600/S2-86/045, EPA Water
ilUUye dllU TMV^ 11 council), oj-.^v....^ . crM i\c|jui i. iiu. . "-'" Q^,
Engineering Research Laboratory.Cincinnati, OH (June 1986).
All data on table for activated sludge were from systems operated at:
SRT = 6 days
HRT =5.5 hours
MLSS = 3500 mg/L
Additional data available in reference at other SRTs and MLSSs.
on partitioning of pollutants to air and sludge.
Data in table for PACT were from systems operated at:
SRT = 6 days
HRT = 5.5 hours
MLSS = 3900 mg/L (excluding PAC)
PAC = 50 mg/L of Hydrodarco C
with two exceptions:
Lindane
SRT = 3 days
MLSS = 2100 mg/L
To!uene
PAC = 200 mg/L
Additional data availabfe in reference at other PAC dosages.
*END OF DATA* —^===
TABLE 4. SAMPLE REFERENCE SCREEN
22
-------
-
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, OH 45268
23
-------
THE ECO LOGIC GAS-PHASE CHEMICAL REDUCTION PROCESS
Gerard W. Sudell
Foster Wheeler Enviresponse, Inc.
Raritan Plaza I - Second Floor
Edison, NJ 08837-3616
908-417-2269
Gordon M. Evans
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 West Martin Luther King Drive (MS215)
Cincinnati, OH 45219-0963
513-569-7684
INTRODUCTION
The Eco Logic International (ELI) Gas-Phase Chemical Reduction Process treats organic hazardous
waste in a hydrogen-rich reducing atmosphere at approximately 900°C and ambient pressure. The reaction
products include hydrogen chloride (HCI) from the reduction of chlorinated organics such as polychlorinated
btohenyls (PCBs), and lighter hydrocarbons such as methane and ethylene from stra.ght-cham and aromatic
hydrocarbons. Water acts as a hydrogen donor to enhance the reaction. The absence of ree oxygen in
the reactor inhibits the formation of dioxin and furan compounds. The first f,ve reasons, through
hydrogenation, remove chlorine from PCBs and reduce the higher molecular W6I9^^V^0^™* <°
simpler more saturated compounds. The final reaction regenerates hydrogen. When treating h gh
crcentr^ion organic wastes, the ELI Process produces excess reformed gas It can compress the
reformed gas and store it for later use, either as fuel to fire the boiler, or as a fuel product for resale. A
portion of the reformed gas is burned in a propane-fired boiler before release to the atmosphere.
Fiaure 1 shows a process schematic diagram of the field demonstration unit. The demonstration-scale
reactor is 2 m (6 ft) in diameter and 3 m (9 ft) tall, mounted on a 15 m (45 ft) drop-deck trailer EU
desiqned the unft to treat 4 tons/day of waste oil, 10 tons/day of wastewater, and 25 tons/day of gas/water
vapor depending on the nature of the contaminants, their degree of chlorination, and the.r water content.
A thermal desorption unit orTDU, designed to remove most volatile, most semivolat e, and some metalHc
conteminants-treats the soil. A screw feeder delivers the contaminated soil to the TDU, which ,s swept wrth
hydrogen to chemically reduce, or prevent the oxidation of volatile and semivolatile spec.es created during
soil heating. The contaminant-laden hydrogen stream flows to the reactor vessel. Decontaminated soil exits
the desorber to a quench tank for proper disposal.
The reactor utilizes mixing, temperature, and residence time to convert the contaminants into a raw fuel
oas Nozzles spray steam-atomized waste liquid with associated suspended solids and PCB-nch oils into
the'reactor. For the Demonstration, a heat exchanger converted contaminated aqueous feedstock into
atomteed steam and heated liquor. The tangential entries swirl the reactants to provide effective mixing
ThTswirHng mixture traveis in the annulus formed by the reactor wall and *™«*™a™*Ł* PŁ
electrical heaters. By the time the mixture leaves the reactor, it has been heated to 900 C (1,650 F). The
reduction reactions occur asrthe gases travel from the reactor inlet to the scrubber inlet.
After auenchinq the gases flow through a countercurrent packed scrubber where contact with scrubber
water removes hydrogen chloride and fine particulates. A large water-sealed vent, acting as an emergency
24
-------
O)
«J
=0
I
T3
CO
DC
0)
O)
25
-------
oressure relief duct, passes scrubber water to a tank below. A heat exchanger, fed by cooling water from
an evaporative cooler, lowers the scrubber water temperature to 35°C (95°F). Caustic and makeup water,
added to the scrubber liquor, maintain HCI removal efficiency. The scrubber produces two effluent streams:
sludge and decant water.
Gases exiting the scrubber contain excess hydrogen, lighter hydrocarbon reduction products such as
methane and ethylene, and a small amount of water vapor. Approximately 95 percent of this gas, reheated
to 500°C (930°F) recirculates back into the reactor; about 5 percent of the hydrocarbon-rich gas serves as
supplementary fuel for a propane-fired boiler. The boiler produces steam used in the heat exchanger, and
polishes the reformed gas prior to emitting it; it produces the only air emissions.
METHODOLOGY
In preparation for the Demonstration, ELI first processed clean sand to adjust the system to peak
performance; followed by feed containing a tracer material to adjust sampling equipment and sampling
Sis Two test runs (Conditions 1 and 3) followed over the next 17 days. Condition 1 reated 2^3 tons of
wastewater contaminated with 4,600 ppm of PCBs and 3,665 ppm of perchloroethylene (PCEs). Condrt.on
3 treated 0.8 tons of waste oil contaminated with 24.5 percent of PCBs and 5,000 ppm of PCE Condition
2 which EPA regarded as a proof-of-concept test, processed PCB-contaminated soil through the TDU;
desorbed gases travelled to the reactor for further treatment. The two Condition 2 runs extended over n.ne
days, treating 1.1 tons of soil contaminated with 650 ppm of PCBs and 15,240 ppm of hexachlorobenzene
(HCB).
The SITE Demonstration Program objectives were as follows:
Demonstrate at least 99.9999% destruction and removal efficiency (ORE) for PCBs.
Demonstrate at least 99.99% destruction efficiency (DE) for PCE in the liquid feedstock.
Ensure that dioxin and furans were not formed.
Characterize products of incomplete reaction (PIR) emissions.
Characterize HCi emissions. n.^,~ .
Document compliance with Michigan Department of Natural Resources (MDNR) air permit.
Characterize criteria air pollutant emissions.
Document compliance with Toxic Substances Control Act (TSCA) permit requirements.
Validate cost assumptions used in economic analyses.
Characterize effluents and residual streams relative to disposal requirements.
Determine the suitability of the reformed gases for reuse/resale.
Demonstrate system reliability.
Develop a system mass balance, including metals.
Characterize critical process scale-up parameters.
Validate the ELI Chemical lonization Mass Spectrometer (CIMS).
Document system operation during test runs.
Samplinq and analysis of the feedstock, intermediate products, and residual streams followed standard
EPA procedures outlined in the Demonstration Plan. EPA subjected the entire sampling and analyses
program to a rigorous Category II quality assurance (QA) procedure designed to generate reliable test data.
The Demonstration Plan details the QA procedures.
EPA sampled three matrices-gas, solid, and liquid-^and analyzed all key input, intermediate, and output
streams for physical properties (flow rate, density, moisture, etc.), PCBs, PCDDs/PCDFs, PAHs,
26
-------
selected compounds.
RESULTS
' HCI' Q, C0" C0-
, THC, and other
The principal results of the reactor Demonstration are the following:
°r flreater' "* DEs for PCE'
• Formed no dioxins or furans.
• Exceeded MDNR permit limitations only on benzene emissions (73 to 106 ^g/dscm).
• Produced minimal HCI and criteria air pollutant emissions.
• Indicated system capacity factor in range varying from 20 percent to 55 percent.
CONCLUSIONS
sssissr*
FOR MORE INFORMATION:
Mr. Gordon Evans
U.S. EPA
SITE Project Manager
26 West Martin Luther King Drive
Cincinnati, OH 45268
513-569-7684
27
-------
VITRIFICATION of SUPERFUND SITE SOILS and SLUDGES
Emilio D. Spinosa
Ferro Corporate Research
7500 Pleasant Valley Road
Independence, OH 44131
(216) 641-8585 ext. 6657
Vitrification of nonnuclear, hazardous waste can offer the
advantage of chemically altering the inorganic, hazardous species
through their incorporation into the structure of stable glass. This paper
will review the development of glass compositions suitable for treating
soils contaminated with inorganic waste. Efforts directed toward
demonstration of the feasibility of continuously processing a waste
stream of mixed inorganic and organic waste into one of those stable
compositions will also be discussed. Initial cost comparison of such a
process and cementitious immobilization indicate that vitrification is cost
competitive when the cost of landfilling the reduced volume of glass is
included in the analysis.
TCLP results for 10 independent replicates.
No.
1
2
3
4
5
6
7
8
9
10
MEAN
ppm in extract liquid
As
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
Cd
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
Cr
0.02
0.02
0.01
0.02
0.03
0.02
0.01
0.02
0.02
0.02
0.019
Cu
0.20
0.22
0.79
0.13
0.38
0.76
0.18
0.20
0.23
0.46
0.355
Pb
0.10
0.10
0.10
0.10
0.20
0.10
0.10
0.10
0.10
0.20
0.130
Nl
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
Zn
0.29
0.21
0.71
0.26
0.34
0.25
0.20
0.24
0.21
0.22
0.293
A student's-t test confirms that all of the regulatory specifications for
toxic release are met at the 95% confidence level. In fact, a similar
analysis performed at 1/10 the regulatory limit is also passed at the 95%
confidence level.
28
-------
PHOTOTHERMAL DESTRUCTION OF OFF-GAS FROM SVF AMD
THERMAL. DESORPTION TREATMENT PROCESSES
Chien T. Chen
U.S. EPA, Releases Control Branch
2890 Woodbridge Avenue, Edison, NJ 08837-3679
(908) 906-6985
John L. Graham and Barry Dellinger
University of Dayton Research Institute
800 College Park, Dayton, OH 45489-0132
INTRODUCTION
Soil vapor extraction (SVE) has been proven to be very effective to remove volatile organic
compounds (VOCs) and some semi-volatile organic compounds (SVOCs) from contaminated subsurface
soil at many sites. Thermal Desorption (TD) is the most adapted technique for the separation of VOCs
and SVOCs from the excavated soil(1). The off-gas from these two processes are usually adsorbed on
activated carbon or condensed to liquid state. The disposal or treatment of these waste residuals is a
tedious and expensive task. The treatments of waste water contaminated with organic compounds
using the combination of ultraviolet (UV) irradiation and oxidants such as hydrogen peroxide and/or
ozone are used commercially®. However, the powerful photolysis of organic vapor is still not used for
the treatment of VOC air emission. Recently, the U.S. Environmental Protection Agency (EPA) has
conducted a study on the photolysis of VOCs from SVE at ambient temperature(3). The results showed
that the reactions were slow and failed to mineralize the VOCs. In order to solve the aforementioned
problem, we have developed a laboratory scale photothermal destruction unit to conduct the tests on
VOCs. Preliminary results have shown good results(4). The increase of light absorption and rate of
mineralization of some organic compounds with .increasing temperature will be shown and discussed in
this paper. An effort will be made to develop a detailed design for a prototype photothermal destruction
unit for use in processing the effluent from SVE and TD operations.
METHODOLOGY
Absorption spectra of materials of interest were studied at various temperature using a custom built
High Temperature Absorption Spectrophotometer (HTAS) with a flow cell as shown in Figure 1. An inert
carrier gas (nitrogen) is used for sample transport to avoid the decomposition or oxidation of the testing
material, The length of exposure of the sample to elevated temperature is kept short (typically 1 second)
to limit destruction of the sample.
Thermal and photothermal reactions were conducted in a Laboratory Scale-Photothermal
Detoxification Unit (LS-PDU) as shown in Figure 2. It is equipped with an irradiation source, a cold trap
to collect reaction products and unreacted material, and an analytical system which included a gas -
chromatbgraph and two detectors, mass spectrometer and flame ionization detector (GC/MS/FID).
Carbon monoxide (CO) and carbon dioxide (COJ were collected in a 1 liter Tedlar bag for analysis. The
initial concentration and irradiation source of compounds tested in the experiment are listed in
Table 1.
29
-------
• OMA Workstation
USD Workstation
X.
HF'OWorksta'ton
X-nonArc
Heactor ,
MSD Gas Chromalograptt
/ OoM Trap y
SampK Intall
Figure 1.
High Temperature Absorption
Spectropnotometer (HTAS)
Figure 2.
Laboratory Scale Photothermal
Detoxification Unit (LS-PDU)
TABLE 1. INITIAL CONDITIONS FOR COMPOUNDS EXPOSED
FOR 10 SECONDS IN DRY AIR
Compound
Initial Concentration
Irradiation Source
Trichloroethytene (TCE)
Dtchtorobenzene (DCB)
Monochtorobenzerte (MCB)
2.52X1Q-6 mol/L
39.4X1016 mol/L
25.9X1Q-6 mol/L
18.1 W/cm2 xenon
arc (A.>230 nm)
18.1 W/cm2 xenon
arc (X>230 nm)
0.883 W/cm2 pulsed
laser (X=280 nm)
RESULTS
Absorption of UV Radiation
Table 2 and Table 3 show the absorption strengths and the onset absorption wavelength of the
materials at various temperatures with an 18.1 W/cm2 xenon arc (a 883 mW, 280 nm laser was used for
MCB experiments). The results show that for each compound, the absorption intensity increased, the
onset absorption wavelength shifted to longer wavelength, and the overlap (data not shown) with the
radiation Increased as the temperature increased.
TABLE 2. INTEGRATED ABSORBILITY (X>230 nm) RELATIVE TO BENZENE AT 100°C
Temp °C
100
200
300
400
500
600
B
1.00
1.28
1.46
1.73
1.94
2.26
T
2.33
2.59
2.81
3.33
3.77
E
2.29
2.57
2.97
3.43
5.25
X
3.13
3.62
4.21
5.46
7.01
TCE
6.16
8.47
10.88
12.59
15.50
PCE
18.95
24.12
26.34
30.33
32.04
DCB
5.33
7.26
9.52
12.88
16.30
2.33
2.89
3.32
4.89
6.47
.0713
.139
.246
.400
.661
.975
3.67
4.35
5.05
5.54
7.29
30
-------
TABLE 3. WAVELENGTH (nm) FOR THE ONSET OF ABSORPTION
Temp °C
100
200
600
. B
267
288
T
278
286
E
281
290
X
285
303
TCE
262
275
PCE
270
288
DCS
290
310
MCB
285
303
CCI4
249
288
GAS
293
298
B:Benzene T:Toluene E:Ethyl Benzene X:m-Xylene TCErTrichloroethylene PCE:Tetrachloroethylene
DCB:1,2-Dichlorobenzene MCB:Monochlorobenzene CCI4:Carbon Tetrachloride GAS:89 Octane Gasoline
Destruction Efficiencies
Tables 4 - 6 summarize the results of thermal and photothermal destructions of TCE, DCS and MCB
at various temperatures. Photothermally a significant portion of each compound was destroyed at
300°C, the lowest temperature for which data is typically generated, the extent of destruction steadily
increases with increasing temperatures. Furthermore, these compounds may be photothermally
destroyed at temperature far lower than required for thermal destruction. All three compounds had
significant destruction at the starting temperature (300°C) photothermally. TCE did not start to react until
the temperature reached at 500°C while DCB and MCB did not have significant destruction below 600°C
thermally.
TABLE 4. DESTRUCTION OF TCE
Temperature(°C)
Thermal %
Photothermal %
*320°C ,
Temperature(°C)
Thermal %
Photothermal %
300
0
12*
400 500
0 0
30 57
550 600
21 64
82 93
650
99
100
700
100
100
TABLES. DESTRUCTION OF DCB
300
0
29
400 500
0 0
30 34
600 650
0. 32
53 73
700
100
100
TABLE 6. DESTRUCTION OF MCB
Temperature(°C)
Thermal %
Photothermal %
300
0
30
400 500
3 5
52 62
600 650
7 11
75 92
700
82
100
31
-------
Products of Incomplete Conversion (PICs)
Most of the three compounds studied were converted to CO and CO2- Small amounts of various
PICs are listed in Table 7-9. Except for carbon tetrachioride (CCIJ from TCE, the amounts of all other
PICs produced from photothermal process were relatively small at 500°C or above. The thermal
destruction of MCB produced many hydrocarbons at or above 600°C, while the photothermal process^
produced only small amounts of oxidized hydrocarbons which almost completely disappeared by 500°C.
The figures on the photothermal destruction of CCI4 and tetrachloroethylene (PCE) at high temperatures
are anomalous. At present, no explanation can be offered. Further studies on this are planned.
TABLE 7. PRODUCTS OF INCOMPLETE CONVERSION FROM TCE
TemperaturefC)
% of Fed TCE
320 400 500 550 600 650
CC14
Thermal
Photothermal
PCE
Thermal
Photothermal
0.0
0.0
0.0
0.018
0.0
1.7
0.0
0.095
0.0
9.8
0.0
0.220
3.0
17.8
0.163
0.358
14.4
18.2
0.308
0.323
10.7
7.8
0.403
0.188
Temp(°C)
TABLE 8. PRODUCTS OF INCOMPLETE CONVERSION FROM DCB
~~ % of Fed DCB
300 400 500 600 650 675
MCB
Thermal
Photothermal
2-CP*
Thermal
Photothermal
0.00 0.00 0.03 0.13 0.74 1.06
0.00 0.17 0.48 0.92 1.01 0.29
0.00 0.00 0.06 0.17 0.48 0.26
0.51 0.66 1.05 1.26 0.46 0.05
*2-Chk>rophenol
TABLE 9. PRODUCTS OF INCOMPLETE CONVERSION FROM MCB
TempfC)
% of Fed MCB
300 400 500 600 650 675 690 700
Benzene
Thermal
Photothermal
Phenol
Thermal
Photothermal
0.239 0.735 0.689 0.248
0.312 0.612 0.500 0.564
0.085 1.54 2.48 1.55 0.172
8.60 6.28 2.68 0.869
32
-------
TAB1.E 9. PRODUCTS OF INCOMPLETE CONVERSION FROM MCB (Continued)
0.165 1.33 1.90 1.36 0.327
2-Chlorophenol
Thermal
Photothermal
Malete Anhydride
Thermal
Photothermal
2(3H)Furanone
Thermal
Photothermal
3-Chlorophenol
Thermal — —
Photothermal — —
1-Butene-3-yne
Thermal — —
Photothermal — —
p-Ethynyftoluene
Thermal — -—
Photothermal — ——
2-Methylnaphthalene
Thermal —• —
Photothermal — —
1 -Methy (naphthalene
Thermal — —
Photothermal — —
0.685 0.354 0.102
1.22 2.31 1.71 0.125
0.447 0.647 0.207
0.106 1.49 1.84 1.33 0.284
0.680 1.64 1.43 1.40 0.586
0.102 4.67 0.093 0.081
0.146 2.45 0.136
1.65 0.130
—: not detectable
Performance Prediction
X mathematical model was
used to determine the conditions
required to achieve a specific
percentage of destruction given the
fundamental parameters as measured in
the laboratory(4). The results of TCE
are shown in Figure 3. The results
obtained using this model with 18.1
W/cm2 radiant energy were very close
to our experimental data. The model
predicts that 99% of TCE may be
destroyed by 500°C using a xenon arc
illuminating system delivering 100
W/cm2 of radiant energy.
Manured LS-PDU Performance On Trlchloroethylene
Using 0 And 18.1 W/cm2 Xenon Arc Radiation And
Predicted With 100 W/cm2 Xenon Arc Radiation
a
ai
E
Ł
o
1 0 0~*
8 0-
6 0-
4 0-
2 0-
0-
1 •-•Ł.- ;-•*-..• ' '
""a.
XQ
*» \
'. \
"a x\
*. *
b
— • — 0 W/cm2
--S--18.1 W/cm2 :&
— -*— 100 W/cm2
0 100 200 300 400 500 600
Temperature, °C
Figure 3.
Destruction of TCE with Xenon Arc Radiation
33
-------
CONCLUSIONS
The results obtained thus far have demonstrated that various VOCs, including hydrocarbons and
chlorinated hydrocarbons, absorb UV radiation at longer wavelength and with increasing efficiency.
These effects were greater than were previously anticipated. The photothermal process can destroy
VOCs at much lower temperatures than thermal process and can completely mineralize organic vapors
that conventional photochemical treatments cannot. After getting more information, a detailed process
computer model will be constructed to predict system performance with various reactor vessel
configurations, lamp types, and exposure conditions. This will result in the detailed design for a
prototype photothermal destruction unit to treat the off-gas from SVE and TD operations.
REFERENCES
1. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Technology
innovation Office: Innovative Treatment Technologies: Semi-Annual Status Report (Third Edition).
EPA/540/2-91/001, U.S. Environmental Protection Agency, Washington, DC, 1992. 71pp.
2. Chen, C.T. Assessment of the Applicability of Chemical Oxidation Technologies for the Treatment of
Contaminants at Leaking Underground Storage Tank (LUST) Sites. In: W.W. Eckenfelder, A.R.
Bowers and JA Roth (ed), Chemical Oxidation, Technologies for the Nineties. Vol. 3. Technomic
Publishing Company, Inc., Lancaster, PA, 1993. pp 225-248.
3. BIyston, P. Destruction of Organic Contaminants in Air Using Advanced Utraviolet Flashlamps.
EPA/540/F-93/501, U.S. Environmental Protection Agency, Cincinnati, Ohio, 1993. 2pp.
4. Chen, C.T., Graham, J.L and Dellinger, B. Photothermal Detoxification of Air Toxics. Paper
presented at 1993 Fall International Symposium of American Flame Research Committee on Impact
of the Clean Air Act, Tulsa, Oklahoma. October 18-20,1993.
34
-------
POTENTIAL APPLICATIONS AMD LIMITATIONS OFTiCX PHOTOCATALYT1C OXIDATION FOR THF
INACTIVAT1ON OF MICROORGANISMS ' — "
J.H. Owens, J.C. Ireland, E.W. Rice, R.M. Clark
Drinking Water Research Division, Risk Reduction Engineering Laboratory, U.S. Environmental
Protection Agency, 26 W. Martin Luther King Drive, Cincinnati, Ohio 45268
(513) 569-7235
lOTRODUCTION
The incorporation of titanium dioxide (TiOj) as a photocatalyst for the formation of hydroxyi
radicals (HO-) in aqueous solutions has prompted the investigation of HO- as a primary oxidant for
environmental remediation (1). Minimum HO- concentrations (-10* molar (M)) generated by irradi-
ating fixed TiO2 particles (anatase) with ultja violet (UV) light (X < 400 nanometers (nm)) are much
greater than HO- concentrations (<10'12 M) generated by other water treatment processes, such as
ozohation, direct photolysis of hydrogen peroxide, and radiolysis (2). The use of TiOa photocatalytic
oxidation for the degradation of anthropogenic organic compounds has been well documented, but
there is a dearth of information regarding the practical application, of this technology for microbial
inactivation. Preliminary investigations support tine hypothesis that significant HO- concentrations
generated by TiO2 photocatalysis would serve as an effective biocide because of its high oxidation
potential (1.6X and 2.0X that of chlorine and ozone, respectively) and non-selective reactivity (3).
METHODS
A photoreactor was used to treat several 12 liter (L) pure culture and natural water samples.
The reactor recirculated samples at 2 L/minute (min) over UV bulbs (X = 365 nm) bound by a fiberglass
mesh impregnated with TiOj, and sheathed by a stainless steel column (Figure 1). Escherichia coli (E.
coll) was suspended in dechlorinated tap water (2.0 x 107 CFU/milliliter (mL)) and recirculated through
the reactor.
PhokiMctor
800-400 nm UV temp
wflh
TOt cocted ffcar^
Samplng port-
^ «
Rgure 1. Schematic diagram of the experimental T!O2 photoreactor showing the flow configuration.
The evaluation of TiO2 photocatalytic oxidation for microbial inactivation, with regard to potential
practical applications, required the conversion of the existing recirculating reactor into a continuous flow
(single pass) system. A-smaller variable speed centrifugal pump was used to inject sample volumes
through the reactor with multiple flow rates. Effluent samples were collected in sterile 50 mL tubes. The
new design made it possible to ascertain rates of inactivation with more control and less subjectivity.
35
-------
RESULTS
Nine minutes of recirculation reduced the viable fraction of a 12 L pure culture E. coli sample
>7 orders of magnitude (Table 1). The recirculating reactor performed similarly on the heterotrophic
population of a natural surface water (Table 2) when hydrogen peroxide (H2O.,) was added (6.5 mM at
two time points) as an irreversible electron acceptor, used to eliminate HO- scavengers, i.e. detritus.
Replicate studies indicated equivalent H2O2 concentrations, in the absence of TiO2 photocatalysis, do
not have a significant effect on heterotroph viability.
TABLE 1. INACTIVATION OF E. coli OVER UV IRRADIATED TiO2 IN A RECIRCULATING REACTOR
Sample Description*
E. coli Concentration
(CFU/mL)
Dechlorinated Tap Water (DTW)
Spiked DTW
6-min DTW Recirculation (lamps off)
3-min Exposure
6-min Exposure
9-min Exposure
12-min Exposure
30-min Exposure
60-min Exposure
2.0 x107
TNTCt
TNTC*
2.6 X 102
adapted from Ireland, et al., 1993
'Exposure times are cumulative
1TNTC, too numerous to count
TABLE 2 INACTIVATION OF HETEROTROPHIC BACTERIA OVER UV IRRADIATED TiO2 IN A
RECIRCULATING REACTOR WITH H2O2 ADDITION
Sample Description"
Heterotroph Concentration
(CFU/mL)
Untreated pond water control
Untreated pond water control
plus 6.5 mM H2O2
3-min exposuret
6-min exposure
9-min exposure with first
H2O2 addition
12-min exposure
15-min exposure
18-min exposure with second
H2O2 addition
28-min exposure
38-min exposure
4.8 x103
2.1 x103
3.6x10*
23
9
14
4
2
6
2
adapted from Ireland, et. al., 1993
'Exposure times are cumulative
photoactivated
36
-------
Data generated from the reconfigured reactor design was less impressive than that of the original
(Table 3). The continuous flow system required much more contact time/unit volume than the recir-
culating reactor. Inactivation of similar E. coli concentrations required a reduction of sample volume
from 12 L to 200 mL Data generated from the different reactors do not necessarily conflict They mav
imply reactor performance is a function of dynamics unique to each system.
TABLE 3. INACTIVATION OF E. coli OVER UV IRRADIATED TiO2 CONTINUOUS FLOW REACTOR
> Flow Rate
Influent CFU/mL
Effluent CFU/mL
390 mL/min (no UV)
390 mL/min
155 mL/min
40 mL/min
*TNTC, too numerous to count
CONCLUSION
1.13x107
1.13x107
1.13 x 107
1.13 x107
1.02x 107
TNTC*
TNTC*
<1
The relative poor performance obtained with the current continuous flow system has prompted the
evaluation of new continuous flow reactor designs. Increasing the ratio of UV-irradiated TiO particle
surface area per unit sample volume has been the crux of design modifications, since HO- does not
occur randomly in solution; proximal distance of an HO- to an irradiated TiO, particle is infinitesimal
Subsequent studies will focus on the inactivation of samples similar to the aforementioned. As reactor
performance improves more resistant organisms, i.e. Cryptosporidium parvum, will be employed.
/,«= SUpp0rte^ '"P*? by. an appointment to the Research Participation Program at the Risk Reduction Engineering
Laboratory/US. Environmental Protects Agancy administered by the Oak Ridge Institute for Science and Engineering through an
.nteragency agreement between the U.S. Department of Energy and Environmental Protection Agency. glneenn9 throu9n •"
REFERENCES
1. Haag, W.R. and Yao, C.C.D. Rate constants for reaction of hydroxyl radicals with several drinkina
water contaminants. Environ. Sci. Tech. 26: 1005-1013, 1992.
2. Ireland, J.C. and Valinieks, J. Rapid measurement of aqueous hydroxyl radical concentrations in
steady state HO- flux systems. Chemosphere. 25: 383-396, 1992.
3. Ireland, J.C., Klosterman, P., Rice, E.W., and Clark, R.M. Inactivation of Escherichia coli by Titanium
Dioxide Photocatalytic Oxidation. Appl. Environ. Microbiol. 59(5): 1668-1670, 1993.
37
-------
SITE DEMONSTRATION OF THE
COLLOID POLISHING FILTER METHOD fCPFM)
Annette M. Gatchett
U.S. Environmental Protection Agency
26 West Martin Luther King Dr.
Cincinnati, Ohio 45268
513\569-7697
The
INTRODUCTION
A Superfund Innovative Technology Evaluation (SITE) demonstration and
evaluation was conducted for the CPFM process developed by Filter Flow
Technology Incorporated during the last three weeks of September, 1993. The
U.S. Department of Energy (DOE) and the EPA joined together in a cooperative
effort to test this technology at the Rocky Flats Plant (RFP) located in
Golden, Colorado.
The CPFM technology treated contaminated groundwater collected in an
interceptor trench constructed around solar evaporation ponds within RFP.
groundwater contained low levels of radioactivity with a concentration of
about 100 picoCuries per liter (pCi/L).
There were three primary objectives in the evaluation of the CPFM
technology; 1) to assess the technology's effectiveness in the removal of low
level radionuclides from contaminated water, 2) to determine reproducibillty
of the test system, 3) to determine the costs associated with the operation of
the system.
METHODOLOGY
The CPFM technology (figure 1) uses specially designed colloid filter
packs in a filter press unit to treat water contaminated with low
concentrations of radionuclides or heavy metals.
During the demonstration, radionuclide- or heavy metal contaminated water
was pumped from an interceptor trench pumphouse to open-top 500,000 gallon
storage tanks, which stored influent for the CPFM system. Treated effluent
was routed to a second 500,000 gallon tank. Influent from the storage tanks
was pumped into two 200-gallon mixing tanks for pH adjustment and chemical
pretreatment. Pretreatment was necessary to adjust water chemistry to the
optimum range for contaminant recovery by the filter packs. After
pretreatment, the water was gravity fed to a lifter station where it was
pumped to a mini-clarifier which removes suspended solids. Settled solids
from the bottom of the clarifier were dewatered in a small filter press
attached to the clarifier. The solids were then collected and stored in a
solids disposal container.
38
-------
Mlni-Clarifier
Stirring
Motor
Control
Panel
Stirring
Motor
Mixing
Tank
Colloid Filter-i
Press
Final pH .
Adjustment
-Tank
Colloid Filter
Packs
- Bag Filters
• Acid Tank
Lift Station
Filter Press
Lift Station
Contaminated
Water
COLLOID POLISHING
FILTER METHOD
Figure 1. CPFM Schematic.
was
microns in size. The system contains three bag filters narllleT so ha
water may be rerouted if the filters become clogged Effluent from the baa
rf T r°Ute2 t0 th! C0ll°id fi1ter Press »"1ts- Each colloid filter^
oid fme'r Sacks"' jl*^-* 2?if°ur fl'lter plates "ntSinlnS three
? !T .^p ?s *. The Co11oid filter packs contain an inorqanic
The pretreated water was evenly dispersed throughout the filter oacks
where physical and chemical mechanisms remove contaminants Chemical
a«?Mj^^
filter cake that contains about 60 percent solids. ™rnnng a
adius?mpnf
«Her packs was then pumped to a final PH
tSf sr-K p "
and
39
-------
The demonstration consisted of three tests. The first test consisted of
three runs of 4 hours each, treating about 5 gallons per minute. For the
second test, also run for 4 hours at 5 gallons per minute, the influent water
was treated with sodium sulfide in the pretreatment tanks to change the
oxidation state of the radioactive metals in the water. The third test was a
15-hour run designed to determine the amount of concentration each filter pacK
is capable of treating.
During the demonstration, samples of untreated influent, pretreated water
after passing through the mini-clarifier and bag filters, and treated water
that has passed through the filter packs, were collected. Samples were
analyzed to evaluate the technology's effectiveness. Adjustment of the
influent pH was not required at RFP because the influent water was within the
optimum pH range (7.5 to 9.0) for the technology. The pH of the effluent
water was monitored in the final pH adjustment tank and was treated to reduce
the pH to between pH 9.0 and 7.5 as required by RFP.
PRELIMINARY RESULTS
Bench scale laboratory studies were performed at RFP during the fall of
1991 using feed water spiked with uranium, radium, plutonium, and americium.
Results from this study showed promising removal for uranium-234, uranium-238,
plutonium and americium with no pretreatment and a pH of 7.6 (Table 1).
TABLE 1. CPFM LABORATORY RESULTS FROM SPIKED FEED WATER
Radiochemistry
=
Gross Alpha
Gross Beta
Radium-226
Uranium-234
Uranium-238
Plutonium
Influent (pCi/L)
========
166+15
124+8
13+7
56+10
35+6
7+1
22+4
Effluent (pCi/L)
23+6
57+7
7.4+7
.03+. 03
.01+. 03
.01+. 02
.01+. 01
Similar results were observed when sodium sulfide and sodium bisulfite
were added at pH 9.0. Sodium sulfide acts as a reducing agent which decreases
the solubility of metals in solution. Sodium bisulfide was added to alter
oxidation states from +6 to +4. No differences were observed in metals
removal when changing the valence state.
CONCLUSIONS
Key findings from the full scale field demonstration, including complete
analytical results and economic analysis, will be presented.
40
-------
CONTROLLING CQPPF.R. LEAD AND IRON IN SMALL SYSTEMS
Darren A. Lytle
Environmental Engineer
U.S. Environmental Protection Agency
26 W. Martin Luther King Dr., Cincinnati, OH 45268, 513-569-7432
Jonathan Clement
Process Engineer
Black and Veatch, Cambridge, MA 02140
INTRODUCTION
!The U.S. Environmental Protection Agency (USEPA) promulgated National Primary Drinking Water
Regulations for lead and copper on June 7, 1991.' The Lead and Copper Rule (LCR), as it is referred to
establishes "action levels" for lead and copper of 0.015 mg/L and 1.3 mg/L, respectively, at the consumer's
tap. Water suppliers that fail to meet either action level are required to employ corrosion control measures to
reduce metal levels. Corrosion control treatment strategies include: pH/alkalinity adjustment, silicate addition
orthophosphate addition, and blended (orth- and poly-phosphate combination) phosphate addition.
Determination of an optimum corrosion control strategy can be made through detailed pilot- and bench-
scale studies such as pipe loop studies or desk-top evaluations which incorporate the source water quality and
current knowledge of lead and copper solubility. Large (>50,000 persons served) and some medium (3,301-
50,000) sized water utilities will likely be required to conduct pilot- or bench-scale studies. However, the
majority of medium and small (<3,301) sized systems will simply implement some form of treatment due to
lack of financial resources and technical knowledge.
Selecting the appropriate corrosion control treatment technique is dependent upon a variety of water
quality parameters such as pH, dissolved inorganic carbonate (DIG), and alkalinity, as well as the source(s) of
lead and copper (e.g. lead service lines, lead-based solder, brass, or copper pipe). Secondary impacts of
corrosion control treatment such as disinfection effectiveness, turbidity, "red water" complaints, and conflicts
with other drinking water regulations must also be considered when choosing a strategy. Further, chemical
availability, safety issues, personnel, space, and cost are issues to consider when choosing a control strategy.
A concern faced by many utilities is the issue of simultaneous control of lead and/or copper corrosion
and "red" and "black" water complaints resulting from iron and manganese in the source water or cast iron lines
in the distribution system. Iron and manganese can easily be removed from groundwater by oxidation and
filtration processes. However, many of the small and medium sized systems with iron and manganese problems
may find construction, material, and sludge disposal costs associated with these treatments costly and may lack
the staff to operate and maintain such a facility. These utilities will be faced with finding a simple, economical
treatment alternative. Many utilities have alleviated the aesthetic problems by adding either polyphosphates or
silicates, to sequester iron and manganese.
A major conflict may arise at utilities that sequester iron and manganese with respect to meeting the
LCR. The sequestering ability of these chemicals generally performs best at pH values < 72-3-4 which are
considered to increase lead and copper solubility5-6. In addition, the metal-complexing abilities of these
chemicals may result in elevated lead and copper levels. Water utilities unable to construct iron and manganese
removal facilities must struggle with finding the appropriate set of water quality conditions and chemical dosages
to enable them to continue using the sequestering agents while meeting the requirements of LCR.
The first objective of this paper is to provide a fundamental understanding of the chemistry and
chemical conditions which polyphosphate- and silicate-based chemicals may be used to simultaneously address
41
-------
sequestration of iron and manganese, and reduce lead and copper solubility. The paper discusses the rationale
for choosing a treatment, silicate addition, for a medium sized utility in Massachusetts attempting to
simultaneously control lead and copper and sequestration of iron. This paper briefly describes the system and
baseline monitoring data thus far collected. Unfortunately, at the time of the manuscript preparation, treatment
results were not yet available. The basis for this study is an ongoing USEPA cooperative agreement through the
New England Water Works Association.
BACKGROUND
The dominant use of sodium silicates in water treatment is to sequester iron and manganese, to reduce
consumer complaints on these metals. Sodium silicate solution, is a proportioned mixture of silica and sodium
oxide. Sodium oxide causes a specific pH increase for a target dosage of silica. When sodium silicate is
introduced into solution and dissolved, silica will initially be in a polymeric form. After a short period of time
(15-30 min), the polymeric species likely reverts to a monomeric or polymeric form with shorter chain length.
The reversion most likely depends on temperature, pH and ionic strength.
Unlike polyphosphates that react with the non-oxidized form of iron and manganese, silicates react with
the oxidized form of these metals through a colloidal dispersion mechanism. For the sequestering process to be
effective, chlorine must be injected a few seconds ahead of the silicate2. The chlorine promotes rapid oxidation
of the iron which forms an iron colloid. Dissolved silica species prevent the growth of the iron colloids. PH
values of 7.8 - 8.0 and silica levels of 10 - 12 mg SiO2/L provided effective sequestering of 2 mg/L of iron.
When the calcium content is negligible (<5 mg Ca/L), higher PH values tend to promote more effective
sequestering. However, as the calcium concentration increases the optimum pH for sequestering decreases.
Higher silica dosages can be used to overcome the negative effects of calcium.
Manganese sequestering with silicate is more sensitive to pH shifts than iron. As pH increases over the
range of 7 to 8, the effectiveness is significantly reduced. Overall, sodium silicate is less effective in
sequestering manganese than it is for iron.
Historically, silicates have been used to control corrosion of hot water systems7-8. The effectiveness of
silicate for this purpose depends on pH and the ionic character of the water (e.g. calcium, magnesium, sulfate
and chlorides). Outside of this use, very little data exists on how silicates perform in reducing lead and copper
levels in distribution systems.
Unfortunately, there has been a lack of research on the ability of silicate to sequester iron and
manganese above pH 8.0. At the present time, the ability of silica to control lead and copper based corrosion is
unknown. As a consequence, utilities using or considering use of sodium silicate should rely on the better
documented benefits of the pH increase that results from the application of sodium silicate.
Polyphosphates can be used to sequester iron and manganese under proper water conditions.2-9
Polyphosphates revert (breakdown) over time to orthophosphate, which may under the appropriate water quality
conditions reduce lead corrosion. Polyphosphates are simple to use, requiring as little as a chemical storage
tank and chemical feed pump. They may be added to existing systems with little effort and change.
Polyphosphate products are commercially available in a wide variety of formulations. Specific
examples include sodium tripolyphosphate (Na5P3OIO) and sodium pyrophosphate (Na4P;jO7). More complicated
are the glassy polyphosphates which are also commonly referred to as "metaphosphates" or
"hexametaphosphates". The composition of glassy phosphates is not clear and their structure is not clearly
ordered. Glassy phosphates are produced by heating soda ash or caustic and phosphoric acid which then is
allowed to cool and crystalize into a "glassy" plate which is crushed. Amorphous mixtures of long chain
polyphosphates can be formed by differing the heating and crystallization conditions.10 Bimetallic
polyphosphates include zinc in the chemical formulation, variety of bimetallic polyphosphates are available with
42
-------
varying zinc to phosphate ratios, solubilities and inhibitor and sequestering capabilities. Zinc tends to slow the
retention rate and limit the dissolution of the polyphosphate, and is thought by some to play a role in corrosion
inhibition.
When added to water, polyphosphate revert, breakdown, or hydrolyze with time into mixtures of
shorter chain polyphosphates and the orthophosphate ion (PO/-). The rate of reversion varies with
polyphosphate chemical, water temperature, PH, and other ions such as calcium. Lowering the pH increases
the rate of reversion while increasing the effectiveness to sequester iron and manganese. Higher temperature
and calcium concentration also increase the rate of reversion. A consideration, particularly to utilities with long
retention times in the distribution system, is the issue of reversion and the loss of sequestering effectiveness at
distant sites from the treatment facility. The original chemical may revert to a form that is less likely to
sequester a desired metal at a distant site, resulting in aesthetic complaints from consumers Further the
experience of one researcher suggests utilities should consult with the polyphosphate manufacturer concerning
storage time and chemical deterioration.2
Robinson et al.3 pointed out that the degree of sequestration of iron and manganese depends on
chemical, pH, and calcium concentration. These authors found that turbidity may increase when phosphates are
used in waters, containing calcium as a result of the precipitation of calcium orthophosphate. A trade-off
between turbidity and color formation existed, and that by fine tuning polyphosphate dosage optimal conditions
could be achieved. Best iron sequestration was found at lower pH's (i.e., 6.0-8.0), however, as the pH
increased the solubility of calcium orthophosphate increased. It was also suggested that polyphosphate
hydrolysis and colloidal particle charge change with pH.
ch^Utf f P°1yPhosPhates to se "^ silicates, sequester by preventing iron and
manganese to be oxuhzed.* Strong oxidizing agents in the water system (i.e. laundry bleach) could overpower
the sequestenng capabilities of polyphosphates and oxidize bound iron or manganese If this is a
polyphosphate dosage adjustment may be required. . » «
< i. Hi °f "^ polyPhosPhates for lead ^ ^pper control is a controversial one. Holm and
Schock ! point out concerns and risks for the potential of increase lead solubility in drinking water systems.
Other researchers have observed similar findings under field conditions. "'**». Further, utilities treating for
iron and manganese are typically operating at low pH conditions as suggested for optimization, but the low pH
can adversely affect lead and copper solubility in the system. The reversion of polyphosphates to
orthophosphate may be beneficial towards lowering lead solubility. Orthophosphate used under the proper pH
conditions has been demonstrated to inhibit lead corrosion. . "Blended" phosphates are chemical manufacmrers
Send co ""I f § T f TT^ C°mplaintS «* C°ntr0llmg lead C0rrosion ** *•'«• cheS
Blends cons st of a imxture of polyphosphate (for Fe and Mn control) and orthophosphate for lead control
^ds and describe
43
-------
and effects on lead and copper solubility are described in the peer reviewed literature. This of course, requires
utilities to rely on manufacturers for technical support and guidance. As has been pointed out, the use of
feŁŁ* effeTvely is a function of many variables and is site specific. Issues to be consumed include:
quality (pH, temperature, calcium), iron and manganese, chlorination, reversion rate, copper and lead
Interpretation of previous experiences with polyphosphates is difficult do to the diversity of the
issues mentioned.
METHODOLOGY
A medium sized (serving 7600 people) utility in Massachusetts receives it's drinking water from five
gravel wells, each with a pump station. Two of the five wells contain elevated levels of iron (as high as 2
S and manganese (0.0? to 0.23 mg/L). A polyphosphate chemical has been added to those weUs to prevent
red and black water complaints. The remaining three wells are untreated. All sources have a pH of 5.9-6.2
wd^DIC of approximately 13 rng C/L. The community has exceeded both the lead and copper action levels
(90* percentile values of 0.077 and 5.87 mg/L of lead and copper, respectively).
The city has identified 14 sites to perform lead and copper monitoring over the duration of the study.
Three rounds of sampling for lead and copper (and a variety of other important water quality parameters) along
with 2 rounds of sampling for the LCR have been used to establish a consistent baseline. Currently, the system
is completing installation of chemical feed units at the well sites.
Early on, thought was given to using a "blended" phosphate for simultaneous control of lead, copper,
iron and manganese. Theoretically, a proper blend could sequester iron and manganese via the polyphosphate
portion and reduce lead via phosphate presence. The pH would need to be raised for optimal lead reduction to
approximately 7.3-7.6, which would likely benefit copper reduction, but, reduce sequestering capability.
Several problems exist, however. First, the blend used would have to be well-defined rn that polyphosphate
type, reversion rates, ratio of poly- to orthophosphate could be identified. Due to the proprietary nature of
these chemicals, no manufacturer would provide this information. Specific compositions of blended phosphate
compounds vary quite dramatically from manufacturer to manufacturer. Therefore, data generated from this
studywould be specific to the particular chemical used and could not be extrapolated to other polyphosphate
compounds. Secondly, we would be operating at a threshold between minimum iron and manganese
sequestration and optimal lead control, which results were somewhat uncertain. And, finally, the utility would
need to add two chemicals; a blended phosphate and a pH adjustment chemical.
Because of the above reasons, sodium silicate was chosen to be used at this community. As mentioned,
sodium silicate contains silica for control of iron and manganese and sodium oxide for pH adjustment. Sodium
silicate requires only one feed system and different ratios of SiO2 to Nap solutions are available. Lower ratios
result in products that are more basic, resulting in greater pH increase benefit. The PH range that silicates
control red water, extends to at least pH 8.0. Therefore, higher pH water values could be achieved for lead and
copper control without jeopardizing iron sequestration. Currently a cost/benefit analysis ,s being conducted to
identify which chemical ratio will be used. The system will aim to add 20 mg SiO2/L and a pH of 7.5.
CONCLUSION
A medium sized utility in Massachusetts is combatting the issue of simultaneous control of lead and
copper and iron precipitation complaints by customers. The system is limited from the standpoint of monetary
resources, personnel, and technical knowledge which are common deficiencies among many small and medium
sized utilities in the United States. It has been decided that sodium silicate addition would provide the utility
with the best treatment approach. Sodium silicate treatment would be simple, require little attention, and would
be relatively inexpensive. Baseline sampling has been concluded at 14 sites in the distribution system and
current progress is being made towards installing chemical feed pumps at the wells. Unfortunately at the time
of preparation of this manuscript data is not available on the treatment results.
44
-------
REFERENCES
1. Lead and Copper. Final Rule. Red. Reg., 56:26460 (June 7, 1991).
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Robinson, R. B., ET AL. Sequestering Methods of Iron and Manganese Treatment AWWA Res
Fdn., Denver, CO. (1990).
Robinson, R.B., Minear, R. A., and Holden, J. M. Effect of Several Ions on Iron Treatment by
Sodium Silicate and Hypochlorite. Journal AWWA, 79:7:116 (July 1987).
Robinson, R.B. and Ronk, S.K. The Treatability of Manganese by Sodium Silicate and Chlorine
Journal AWWA, 79: 11:64 (Nov. 1987).
Schock, M.R. Understanding Corrosion Control Strategies for Lead. Journal AWWA, 81(7):88
(1980).
Schock, M.R. and Gardels, M.C. Plumbosolvency Reduction by High pH and Low Carbonate-
Solubility Relationships, Journal AWWA, 75(2): 87 (1983).
Falcone, J.S., ed. Soluble Silicates, ACS Symposium Series 194, American Chemical Society
Washington, D.C., 1982. J>
Lane, et al. The Effect of pH on Silicate Treatment of Hot Water in Galvanized Piping. Journal
AWWA (August 1977).
Klueh, K. and Robinson, R. B. Sequestration of Iron in Groundwater by Polyphosphates. Journal of
the Environmental Engineer Div.,ASCE, 114:5:1192 (Oct., 1988).
AWWA Research Foundation/D VGW Forschungsstele. Internal Corrosion of Water Distribution
Systems. AWWARF Coop. Res. Rept. AWWARSF, Denver, CO. (1985).
Holm T.R. & Schock, M.R. Potential Effects of Polyphospfaate Products on Lead Solubility in
Plumbing Systems. Journal AWWA, 7:76:82 (July, 1991).
Berner, R.A. Rate of Concretion Growth. Geochim. Cosmochim. Acta, 32:477 (1968).
Berner, R.A. and, Morse, J. W. Dissolution Kinetics of Calcium Carbonate in Sea Water IV
Theory of Calcite Dissolution. American Journal of Science, 274-108 (1974).
337?Ł 1 ?79 '(J™Sl)°f Lead COrr°Si0n Wlth S°diUm Hexametopfaosphate. Journal AWWA,
Baily T.L. Corrosion Control Experiences at Durham, N.C. Proc. 1982 WQTC, Nashville, Tenn.
~
Sheiham, I. & Jackson, P.J. The Scientific Basis for Control of Lead in Drinking Water by Water
Treatment. Journal Inst. Water Engineers and Scientists, 35:6:491 (Nov. 1981).
Neff C.H. BTAL. Relationship Between Water Quality and Corrosion of Plumbing Materials in
Buildings. Final Report CR8566-02. EPA/600/S2-87/036, Cincinnati, OH (1987).
45
-------
ULTRAFILTRATinN MFMBRANE RESEARCH
FOR SMALL DRTNKING WATER SYSTEMS
James A. Goodrich
Benjamin W. Lykins, Jr.
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Drinking Water Research Division
Systems and Field Evaluation Branch
Cincinnati, Ohio 45268
(513/569-7605, 569-7460)
Both small community and non-community drinking water systems (serving
less than 3,300 people) have and will continue to have difficulty complying
with the ever increasing number of regulated contaminants. Currently, it is
estimated there will be nearly 100,000 violations of the Safe Drinking Water-
Act annually. Approximately half of these are for Maximum Contaminant Level
(MCL) violations. Of these, the majority are microbiological violations by
the small systems. The Surface Water Treatment Rule along with the future
Ground Water Disinfection Rule and Disinfection/Disinfection By-Products Rule
will especially impact small systems.
The Safe Drinking Water Act (SDWA) and it Amendments mandate the regula-
tions of all public water systems that have 15 or more service connections or
serve at least 25 people for at least 60 days each year. These public water
supplies include community and non-community systems such as schools,
factories, and hospitals that have their own water supplies but usually
exclude noncommunity systems that cater to transitory customers in non-
residential areas such as campgrounds, motels, and gas stations.
Of the approximately 200,000 public water systems in the United States,
about 30 percent are community water systems which serve primarily residential
areas and 90 percent of the population. Of the 59,266 community water
systems, which serve about 232 million people, 51,682 were classified as
"small" or "very small." These systems served populations of fewer than 3,300
people, with a total population served of about 25 million people (Table 1).
Small systems are the most frequent violators of federal regulations and
accounted for almost 88 percent of the violations posted in 1991.
Microbiological violations accounted for the vast majority of the cases with
failure to monitor and report (M/R) exceeding violations of the SDWA Maximum
Contaminant Levels (MCLs).
Membrane technology is being considered for designation as BAT for the
oendina regulations and in order to develop cost and performance information
the U.S. Environmental Protection Agency Drinking Water Research Division in
Cincinnati, Ohio has-developed a Small Systems Research Program. The research
incorporates both in-house and field investigations.
46
-------
r i
O
p
I
^H
Q
S3
§
a
H
1
*•* c
C OT5
O +3 C)
0,2 >
0) 3 0)
Q. ttCO
O
Q.
C
o ^
55-0 c
JH Q5
- ^^^ tfl% B^^*
S^BH ^^» WH
Q. ^^
C«J
(n
>.Q.w;
C/5^-
OJ
is.
l/>
CO
CD
CO
V
O5
00
a
C\J
CD
^
O
o
C3
co"
•75 '
E5
COLO
0
CO
CO
CM
LO
CO
00
00
CO
CvJ
CO
o^
co"
O
o
0
•>»
o
A
_
o
, c
o
r5
-z.
w
E
to
0
s
'E
=3
E
o
O
«
41
-------
Two field projects that have installed UF membrane package plants in
diverse communities are presented in this report. The first being in the
hills of West Virginia treating water from a collapsed mine shaft that serves
approximately 100 people. The second community is located on the Chemehuevi
Indian Lands on the western shore of Lake Havasu in the California desert.
This community serves approximately 500 people utilizing the impounded
Colorado River as the source of supply. During the hot summer months there is
a tremendous amount of recreational activity on the lake and there have been
reported cases of Giardiasis in neighboring communities.
UF ADVANTAGES FOR SMALL SYSTEMS
Ultrafiltration (UF) is an effective means of removing particulates and
dissolved polymers from water sources as diverse as sea, river and ground-
water. UF membranes fall in between the range of reverse osmosis (RO) and
raicrofiltration (MF). Since UF cartridges operate in a cross-flow mode,there
are similarities between UF and RO systems.
A major reason for small system use is that UF systems operate at pres-
sures up to 120 psi, offering much lower pressures than RO systems which
operate at hundreds of psi. Second. UF cartridges permeate at rates appro-
aching 30 6SFD (gallons per square foot of membrane per day), in contrast to
RO membranes which flux in the order of 15 GSFD-rejecting salts and other low
molecular weight components that need not be removed from municipal water.
Th rd UF systems produce much more water at lower operating pressures than RO
systems, and therefore, both the capital and operating costs for UF systems
are significantly less.
Other benefits of using ultrafiltration include:
Ultrafiltration eliminates the need for chemical coagulants or
filter aids.
Ultrafiltration produces higher quality filtrate than using granular
media filtration alone. In general, the silt density index of
ultrafiltered water approaches zero.
UF systems can operate at much higher recoveries than RO systems.
Ultrafiltration does not exhibit solids breakthrough associated with
granular media filtration.
UF cartridges, designed with proper cross-flow velocities and with
routine cleanings, can eliminate the bacteria typically found in
conventional flow-through filters.
-UF cartridges can be fed high turbidity water without frequent
rep!acements.
UF membranes can be sanitized with up to 200 PPM of NaOCl.
48
-------
BUCHANAN, WVA
The Buchanan,, WVA system is one of eleven operated by the McDowell Countv
Public Serv ce District (PSD). The Buchanan system serves approximate y 100
people requiring about 10,000 gallons over a 24 hour period. The water source
is an abandoned coal mine that has collapsed and been sealed off to keep
?HiS^?tOUt; The raw, water quality is very good with low levels of bacteria,
turbidity, trace metals, and inorganic contaminants. Previous treatment
included super chlorination of sand filtered water. Both pieces of equipment
were near collapse and although currently in compliance, produced unpalatable
*
HAVASU LANDING, CALIFORNIA
The Chemehuevi Indian Lands is located on the western shore of Lake
Havasu in the California desert. Approximately 87 homes and 400 people are
served in one of the two systems being operated by the tribal government
Water usage peaks at three times the average demand during the summer months
because of the use of evaporative cooling (swamp coolers) for the homes PeV
capita consumption can average 300 gallons per day for several days
continuously. Past water treatment consisted of a pressure sand filter
followed by chlorination. Because of capacity limitations and equipment
design, bacterial violations had been occurring and the system was under a
inn™fandT?LJ?JThVH hater qua"tyT. Raw water <*ua1ity experienced spikes
in TDS and Giardiasis had been reported in nearby communities. Algal blooms
also occur in the late summer. y Olooms
A package plant similar to the one in Buchanan, WVA was installed with a
40 gpm capacity. Pre-treatment consisted of the rehabilitated sand fi Her
(without chemical additions) and a cartridge filter prior to the UF bag filter
to prolong the UF membrane operation by removing the potentially high TDS
«ch»Tc1». to
RESULTS
BUCHANAN, WVA
Over the first 15 months of system operation at Buchanan, WV some
informative results have been documented. The UF plant cycles on and off in
operation based_on sensors in the 3,000 gallon storage tank. Daily system
operation ranged from about 9.5 to 17 hours/day. In order to maintain a
consistent permeate flow of approximately 10 gpm during operation, the
transmembrane pressure (TMP) was periodically increased in response to
membrane fouling. The system was operated with adjustments to TMP until the
inlet high pressure gauge indicated a membrane cleaning was needed. Also
prefilter inlet pressure was tracked to provide an indication of when bag
filter replacement was necessary. Generally, bag filters were replaced when
membrane cleaning was performed, and at some additional non-cleaning times
Membrane flux was consistently maintained at 20 gsfd or above
49
-------
Initial data indicates that the UF system is operating according to
soecifi cat ions Preliminary water quality data for the UF system show good
redSctions iS parti ces and bacteria. Little TOC reduction ^observed as
expected for a 10,000 MWCO membrane. During this sampling period, 1. 40 PPM
sodtSm hypoch?orite was injected into the reclrculatlon loop prior to the
membrane vessel . There are small increases in TOX, THMs, HAAs, and HANs.
Preliminary samples indicate some aldehydes in the raw water with Possible
biodegradlti on and/or chemical reaction with chlorine occurring An increase
in AOCis observed with a possible link to the aldehydes and chlorine
Edition The inorganic chemical parameters are not removed by the UF system
as expected Some iron and manganese precipitation onto the membranes
occurrtS requiring an acid wash" Hence, chlorine addition was moved from the
recirculation loop (pre-membrane) to post-membranes after about 1520
production hours.
HAVASU LANDING, CA
Since December 1992 when the ultrafiltration membrane package plant
became operational there have been a variety of modifications to the
operation treatment train, and frequency of maintenance Because of these
SaŁ chXieslX institutional problems, there has been little J opportunity to
fully evaluate and challenge the package plant. Initial emphasis was on
oroductivity and becoming familiar with the operation of the system. During
the first three weeks, the system operated well with only one pref liter change
beini Sees*??. However, the system was not cleaned throughout the first
month and productivity declined considerably. Membrane cleaning occurred at
i??eaular intervals over the next few months with productivity only partially
returning after cleaning. By April it appeared that the membrane was severely
but SI irreversibly fouled. It was decided to clean the membrane every two
weeks to try to return the membrane's productivity and move the point of
chlorination to the sand filter to prevent algal growth in the system It was
also recommended that the sand be replaced in the roughing filter since it was
very old It was also recommended that a larger water heater be installed in
order that warm water be available for the entire cleaning. Throughout the
summer mo^hs?cleaning intervals remained erratic although the sand was
eventually replaced in the roughing filter.
Monthly water sampling began in February 1993 for a wide variety of
oarameters The parameter! of most interest included total coliforms, fecal
?oliforms!*turbidity, TOC, and disinfection by-products. In most cases, it is
evident that the fin shed water was not in compliance given the few total
colifoVms found in the package plant permeate. The membrane nominal pore size
is 005 micros and should rlmove bacteria. Whether these occurrences are
blcause of lack of membrane integrity, sampling error, or biological 9™*-
thrS of the membrane is difficult to determine. Routine sanitizing of the
short runs of PVC pipe at the sampling port is being implemented to address
one coSiSation possibility. Given the low intermittent total conform
IPVPI? S-chlori nation of the permeate water will provide another safety
barr er and produce finished water in compliance with the Total Conform Rule.
50
-------
Finished water turbidity values were with only one exception verv aoori
Even values as low as 0.2 NTU required by the State of California Ire
routinely attainable by the package plant. TTHM 1 Lei s were very ?ow « 10
ug/L) even during periods of pre-chlorination. Other by-products were
negligible. As expected, the package plant had little effect on thftrace
metals or inorganic parameters. Raw water levels for these parameters were
not a concern nor did they pose any serious constraints to the oStion of
n'd^h'^^n^'h^3" th? aWareneSS °f some f
and the algal blooms already mentioned.
COST
«f TJe.B;!chanan» WVA Plant cost $25,000 to purchase. O&M costs have been
S J?te2iJ° be aPPr?ximate1y 10 dollayts P^ household per month Nearly half
of this O&M expense is for an operator to visit the plant daily for
inspection Retrofitting the system for remote monitoring w n reduce O&M
costs considerably. The Havasu Landing plant cost approximately $65 000
' °f Pa 5°0-
51
-------
POINT OF USE TREATMENT FOR ARSENIC REMOVAL
Kim R. Fox
513-569-7820
Drinking Water Research Division
Risk Reduction Engineering Laboratory
26 W. MLK Dr.
Cincinnati, Ohio 45268
could
This study was conducted to determine whether point-of-use (POU) reverse osmosis (RO) units
on in lieu of central treatment. The POU units were to remove arsenic and
ng water supply at San Ysidro, NM. POU treatment was evaluated for remova,
efficiency, cost, and management effectiveness.
Seventy-eight under-the-sink model RO units were installed in private homes. Of the 78 units
installed? 72 untts were monitored for about 18 months to evaluate operational and maintenance data
for POU treatment.
INTRODUCTION
The Village of San Ysidro is a small rural community of approximately 200 people located in the
north cental \S* the State of New Mexico approximately 45 miles (72 km) north of Albuquerque.
The VHteoe > washaving problems meeting water demands and was also out of compliance with the
SnkCwalTRegTtions for arsenic and fluoride. Feasibi.rty studies were performed tc .del ermine
wheS economical improvements to the system could be recommended (improvements that would
solve both San Ysidro's water quantity and quality inadequacies).
The Village has had a long history/of water supply problems including low water pressure and
at times no water at all. They also had water quality problems including taste, color, danty, and odor in
addSon to aSc and fluoride contamination and sporadic coliform violations. The water supply
source is an infiltration gallery that produces an average of 27,000 gpd in the winter and 36,000 gpd in
^ romi^lfo^^r ^e Village uses an average of 30,000 gpd, which equates to
gpd p^r perUn. This consumption rate pushed the production lunita of the gallery.
Central treatment of the entire water supply was not considered feasible for many reasons.
First a disposal problem exists with both the arsenic contaminated wastes from activated alumina
genera ton and the reject brine from an RO system. Second, the capital costs and the operabonand
rrSintenance oosts of central treatment were determined to be higher than POU treatment And, astly
^SSSS^ considered too complicated to be efficiently operated by a cornmuruty the size of
San Ysidro. The solution in this case was determined to be POU treatment with RO units.
METHODOLOGY
A notification letter was sent to each water customer, and a public hearing was held on
Sss Sheeting, the cooperative agreement between the Village and the EPA was
be thfvillagers to explain, the water qualfty problem and to discuss the procedures needed
to have fee RO POU devices installed, maintained, and tested during the study period. The Village
an ordinance requiring the installation of an RO system in each water customer's home if the
The ordinance was deemed necessary because POU treatment cou.d not
52
-------
be considered a viable alternative to central treatment for a public water system unless the utility
furnished safe drinking water to each water customer. Each water customer also had to siqn a
permission form to allow the Village to install the unit in their home and to allow access to the unit for
testing and maintenance. The permission form was necessary because an ordinance could not qive
the Village the authority to enter a person's house (only an individual can grant permission to the
Village to enter his home).
A Request for Proposal was prepared in which contractors were asked to prepare competitive
bid proposals for furnishing approximately 80 RO units. The bid proposals was to include installation
and 14 months of unit maintenance.
„ .. units.were re?uired to be "nder-the-sink models capable of producing a minimum of 5
gallons of drinking water per day with a storage capacity of 3 gallons. The system pressure range was
given as 40 to 60 psi maximum, with a minimum pressure of 20 psi.
A contract was given to both install and maintain the RO unit. Within the first 4 months of the
project, 73 RO units were installed, and 5 more were added by the end of the project period Of the 78
units installed, however, only 72 units were actually available for testing on a regular basis At three
homes, totalizing meters were installed on the inlet line to the RO units to measure the amount of water
used by the system.
RESULTS
The RO units were operated and monitored for an 18 month period. Water samples were
scheduled to be collected every other month for arsenic and fluoride analyses. In addition the units
were to be samples every 4 to 6 month for chloride, iron.and manganese. Because of various
restrictions, only 40 units were analyzed for total coliforms. Each month, an average of 31 units were
sampled for arsenic and fluoride; of these, 15 were also sampled for chloride, iron.and manganese and
i o Tor total coliforms.
uar- H f a U^f, by the R° SyStems recorded at the three nomes with totalizing meters
varied from 8.5 to 17.0 gpd. Water use varied because of the size of the families and their use of the
RO-treated water. Product water production (recovery) by the RO system depends on inlet water
pressure and TDS, but ranges from 20% to 30% of the total flow in the unit at 50 psi and less than
1 ,500 ppm TDS.
The RO units effectively removed arsenic and fluoride from the water. The RO units also
effectively removed chloride, iron, manganese, and TDS. Because of the high costs of arsenic and
fluoride analyses, conductivity measurements were evaluated as a substitute for arsenic and fluoride
tests. An analysis of the arsenic, fluoride and conductivity data showed a rule of thumb could be
established whereby a conductivity measurement of less than 600 micromohs/cm would maintain less
than 0.03 mg/L of arsenic and less than 1.0 mg/L of fluoride.
CONCLUSIONS
The following conclusions were drawn as a result of the San Ysidro study.
POU treatment of drinking water is an effective, economical, reliable, and viable alternative to
central treatment in a small community like San Ysidro to remove arsenic as well as other
contaminants.
53
-------
Adopting a POU treatment system in a small community requires more care than does a central
treatment system relative to time-keeping to monitor the individual systems.
POU systems require special consideration from regulatory agencies to determine appropriate
methods for record keeping, monitoring, and testing frequencies that may differ from existing
regulations.
The RO units with polyamide membranes installed in San Ysidro resulted in the following
removal percentages, bringing all of the contaminant levels well below the MCL's: arsenic (total) - 86%;
fluoride - 87%; IDS - 88%; chloride - 84%; iron - 97%; and manganese - 87%.
The cost to the customer of POU treatment per month ($7) in San Ysidro is less than half of
the estimated cost of central treatment ($30 to $40 per month). The cost per gallon of treated water,
however, is more than three times that of central treatment, since central treatment treats the entire
water supply and the POU device treats a small fraction of the supply.
54
-------
THE EFFECTIVENESS OF SORBENTS FOR SOLID-BED METAI r.APTllRF JN AN
INCINERATOR: SCRFFNTNG TESTS AT THF TNCINERATTON RFSEARCH FAHTI TTV
V
Gregory J. Carroll
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
Shyam Venkatesh
Donald J. Fournier, Jr.
Acurex Environmental Corporation
555 Clyde Avenue
Mountain View, CA 94043
INTRODUCTION
*he.concfn over the emission of hazardous constituent trace
nntt- inclnerators>uthere ^ currently considerable interest in the
potential use of mineral -based sorbents for capturing and retaining those
metals m the incinerator "ash" discharges (fly ash and bottom Ssh).
Most of the research completed to date has focussed on quantify inq the
thfnp6neSS °t VarTS Pr^°sed sorbents for "Paring vaporized meU?s from
the flue fas. In such applications, it is theorized that vaporized metals
will react with the sorbent particles at elevated incinerator temperatures or
homogeneously condense onto the sorbents as the flue gas cools In the
absence of available condensation sites, vaporized metals will orimarilv
undergo homogeneous condensation, forming a fine fume. Thus the gol with
flue-gas sorbents is to make particles available with which the metals can
UP-?i Wt1Ch the 2fals can condense. Metals bound to larger sorbent
h!i lt^m°re 6fff JiV6ly Sheeted ^ air pollution control systems
^n th I f, a S ?resen^d as a fine fume- studies completed to date
suggest that chemical reaction between the metal and the sorbent dominates
over physical adsorption offering the additional advantage of redSced
potential for metal leaching from collected particulate.
™HH o r"earchers have studied the incorporation of sorbents into the
solid feed. This approach seeks to capture and bind the metals in the
incinerator bottom ash preventing them from exiting with the combustion
9?!":.,Res!arch ^eted to date suggests that for this approach to be
effective, the metal should become volatile in the incinerator environment
and chemically react with the sorbent material. "Iera™r environment
annw, subject *est P™gram was designed to further investigate this second
approach by screening several minerals for their suitability ai sorbent
materials for capturing metals in the solid bed and preventing their release
reta n tnem fn^the S^t103-t0 Ca5tur1nf uthe meta1'' an ideal sorEent wo ?d
retain them in the ash when disposed, so that extraction of the ash by the
toxicity characteristic leaching procedure (TCLP) would yield a leachate with
metals concentrations below respective regulatory levels. Accordingly the
objective of th^,s screening program was to evaluate several candidate sorblnts
with respect to: (1) the degree to which they facilitate retention of t?ace
55
-------
metals in the ash/solid bed discharged from an incinerator; and (2) the degree
to which they retain trace metals in the solid bed when subjected to TCLP
extraction.(l)
METHODOLOGY
Test Equipment
The screening tests comprising this program were conducted in the
bench-scale thermal treatability unit (TTU) at the U.S. EPA Incineration
Research Facility (IRF). The TTU is a small commercial pathological
incinerator that has been modified to allow for continuous test material
feed and treated material (e.g., ash) removal; for variable and controlled
treatment temperatures; and for expanded process operation monitoring
(Figure 1).
The combustor portion of the TTU consists of three chambers. The charge
chamber is designed to accept the solid material feed stream and corresponds
to the primary combustion chamber (or kiln portion) of a waste incinerator.
The retention chamber, which directly follows the charge chamber, is designed
to effect further organic constituent destruction and corresponds to the
secondary combustion chamber (or afterburner) of a waste incinerator. The
breaching chamber serves as a second-stage afterburner. Volumes of ihe charge
chamber, retention chamber, and breaching chamber are 0.82 m (29 ft ),
0.67 m3 (23.5 ft3), and 0.10 m3 (3.5 ft3), respectively. Each of the three
chambers is fired with natural-gas-fueled burners.
A feed system transports quartz trays containing the test material
through the charge chamber using a variable-speed chain-drive mechanism. The
system can accommodate a number of different tray sizes; those used for the
subject tests were 18.5 cm (7.3 in) long by 8.5 cm (3.3 in) wide by 4 cm (1.6
in) deep, and were capable of holding up to 1 kg (2.2 Ib) of test material.
Test Program
The test program consisted of 50 tests, including two duplicates of one
test condition. Test variables were sorbent material type, solid bed
temperature, feed chlorine content, and metal form in the feed.
Six sorbents were evaluated. Four of the six (diatomaceous earth,
kaolinite, bauxite, and alumina) were selected based on the most promising
results from other researchers. The attapulgite clay used in past IRF trace
metals studies was tested as the fifth sorbent. The sixth material, quartz,
was selected as a "neutral" material, or system blank.
The approximate mineral content of diatomaceous earth, kaolinite,
and bauxite are given in Table 1. Alumina is presumed to be pure A1203
and the attapulgite clay is a hydrated magnesium aluminum silicate
[(Mg,Al)cSioO,p(OH),-4H20] containing some dolomite [Ca,Mg(C03)2],
calcite ICaci], and silica [Si02]. Quartz is presumed to be pure silica.
56
-------
>
(^\—
'• i W^
1
1
! XJ^V
' 1 Frcl
| TO MOV-S | ' ~
1 ^ 1
: TEMP INDICATING
i CONTROLLER
1
! 1 TOMOV-1 1
i TEMP INDICATING 1
| CONTROLLER
1
1 PESO
| CONVEYOR
i ' '
TEi-2
)
I .
I
FIGURE 1. IRF THERMAL TREATABILITY UNIT (TTU)
TABLE 1. APPROXIMATE SORBENT MINERAL COMPOSITION
Diatomaceous
earth Kaolinite Bauxite
SiOU 90.4 52 11
A12<53 6.5 45 84
FeJDj 2.3 1 5
Ti02 2
CaO 0.2
MgO 0.3
Other oxides 0.3
57
-------
Three solid bed temperatures were tested: 540°C, 700°C, and 870°C
(1000°F, 1300°F, and 1600°F). Two feed chlorine contents were tested: 0% and
4% by weight. Polyvinyl chloride powder was added to the mixtures as the
chlorine source.
Sorbent impact on the retention of the following trace metals was
evaluated in the program: arsenic, cadmium, chromium, lead, and nickel.
Two forms of incorporating the metals into the feed mixtures were evaluated.
Past trace metals tests at the IRF have used aqueous metal spike solutions
containing soluble nitrate salts of the metals (with the exception of arsenic
which has been added as As203). For continuity, this form was one of the two
used in the tests. The second form of metal spiking was a metal compound
"dispersion". The dispersion consisted of metal compound powders suspended in
a liquid carrier analogous to pigments dispersed in paint or ink. Chromium,
cadmium and lead were present as metal oxides in the dispersion, while arsenic
and nickel were introduced as sulfides and carbonates, respectively.
For each test, weighed amounts of the appropriate mixtures of sorbent,
PVC (for tests with chlorine-containing feed), and metal spiking formulations
were added to a TTU quartz tray. Charge depth was held constant at nominally
2 cm, corresponding to a charge volume in the tray of approximately 300 cm .
Feed metal concentrations were also held constant as follows: arsenic (250
mg/kg); cadmium (50 mg/kg); chromium (150 mg/kg); lead (250 mg/kg); and nickel
(150 mg/kg). Charge mass ranged from 50 g to 320 g, depending on sorbent bulk
density.
Tray residence time in the TTU was approximately 20 minutes at the
target solid bed temperature. Based on the results of other research, no
further reaction between the sorbent and metals was expected after the first
10 minutes at target temperature.
Sampling and Analysis
For each test, TTU gas temperatures and solid bed temperatures were
measured.
Samples of unspiked sorbent, aqueous metal spiking solution, metal
dispersion, TTU feed, and TTU discharge were collected for metal analyses.
Additionally, one TTU feed sample from each of the sorbent/metal formulation
combinations and TTU discharge samples from every test were subjected to TCLP
extraction, and the resulting leachates analyzed for trace metals.(2) Quality
assurance samples were prepared as well.
Aqueous liquid samples were digested using EPA Method 3010.(3) Solid
samples and the metal dispersion samples were digested using a microwave-
assisted HNO,/HF procedure.(4) Analyses of each of the latter digestates for
arsenic were by Method 7060 [GFAA].(3) All other digestates analyses were by
Method 6010 [ICAP].
58
-------
RESULTS/CONCLUSIONS
Jests were conducted from November through December 1993. Analytical
results were not available as of the writing of this abstract. As available,
results and conclusions will be discussed during the Symposium presentation.
REFERENCES
1.
2.
3.
4.
Acurex Environmental Corp. Quality Assurance Project Plan for Evaluating
the Effectiveness of Additives as Sorbents for Metal Capture Using the
Thermal Treatability Unit. Contract No. 68-C9-0038, U.S. Environmental
Protection Agency, Cincinnati, Ohio, 1993.
Method 1311 Toxicity Characteristic Leaching Procedure (TCLP). 40 CFR
Part 261, Appendix II.
Test Methods for Evaluating Solid Waste: Physical/Chemical Methods
SW-846, 3rd Edition, Revision 1, U.S. Environmental Protection Aqencv,
1992. » j"
Methodology for the Determination of Metals Emissions in Exhaust Gases
from Hazardous Waste Incineration and Similar Combustion Processes
40 CFR Part 266, Appendix IX.
FOR MORE INFORMATION
Gregory J. Carroll
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory (ML 481)
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
513/569-7948
59
-------
ABSTRACT
"PILOT-SCALE INCINERATION TESTS OF UDHH AND NITROGEN TETROXIDE
FORMER SOVIET UNION LIQUID BALLISTIC MISSILE PROPELLANTS"
Donald A. Oberacker*
US EPA RREL/WMDDRD/TDB/TRS
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
513/569-7510
and
Larry R. Waterland
Acurex Environmental Corporation
555 Clyde Avenue, P.O. Box 7044
Mountain View, California 94039
415/964-5145
*(Speaker)
[Note: This work was sponsored by the Defense Nuclear Agency (DNA)
under DNA IACRO #93-691 and Work Unit 00005.]
This is a summary report of a particularly interesting and unusual
missile propellent incineration research project conducted at the U.S. EPA's
Incineration Research Facility (IRF) during the winter of 1993-94. Data
generated from these recent pilot-scale incineration tests were useful toward
the goal of international peace - toward the elimination of propellants from
some of the world's strategic offensive arms. The presentation will begin on
a brief historical note as follows.
On April 4, 1993, Presidents Clinton and Yeltsin agreed in principle to
a draft agreement between the U.S. Department of Defense (DOD) and the
Committee for Defense Industry of the Russian Federation concerning
cooperation in the elimination of Russian Strategic Offensive Arms. This
agreement was signed and became effective on August 26, 1993. The agreement
obligates the DOD to provide the Russian Federation with an extensive list of
equipment for the Federation's use in eliminating strategic offensive arms in
accordance with schedules negotiated in the Strategic Arms Reduction Treaty
(START). Included in this equipment list are mobile incinerators, each
capable of destroying at least 750 metric tons per year of ballistic missile
liquid propellent comprised of unsymmetrical dimethyl hydrazine (UDMH) fuel
and nitrogen tetroxide (N204) oxidizer.
60
-------
DNA supported a series of tests at the EPA's IRF to demonstrate that the
exact U.S. chemical equivalents of Russian Ballistic missile propellant fuel
and oxidizer can be incinerated effectively and safely and in compliance with
applicable U.S. and Russian environmental laws and regulations. The tests
successfully obtained certain technical information which is helping the U S
to commercially procure the incinerators.
DNA selected the U.S. Environmental Protection Agency (EPA) Incineration
Research Facility (IRF) at Jefferson, Arkansas to conduct the tests. The
presentation will describe the specific test program objectives and the test
conditions for the tests, etc. The sampling and analysis procedures used to
satisfy the test objectives will also be described.
;The test Pr°9ram was conducted in the rotary kiln incineration system
(RKS) at the IRF which consists of a primary combustion chamber, a transition
section, and a fired afterburner chamber. The primary combustion chamber, or
kiln, is nominally 1.04 m (3 ft-4.75 in) in diameter and 2.26 m (7 ft-5 in)
long. The afterburner is nominally 0.91 m (3 ft) in diameter and 3 05 m
(10 ft) long. A 4.43-m (14 ft-6.5 in) long, 0.61-m
(2-ft) diameter refractory-lined afterburner extension follows the afterburner
and allows isokinetic sampling of hot flue gas prior to quenching. Both the
kiln and afterburner are fitted with 590 kW (2 MMBtu/hr) auxiliary fuel
burners. Natural gas was the auxiliary fuel, although liquid waste or fuel
can also be fired as was the case for these tests. Typical firing rates are
290 to 440 kW (1.0 to 1.5 MMBtu/hr) to the kiln and 440 to 590 kW (1.5 to
2.0 MMBtu/hr) to the afterburner. U
After exiting the afterburner, flue gas flows through a quench section
followed by a primary air pollution control system (ARCS). The primary APCS
for the planned tests was a venturi- scrubber/packed-column scrubber
combination. This scrubber system removed most of the particulate and acid
gas, such as HC1, SO,, or NO,, in the flue gas. Downstream of the primary
APCS, a backup secondary APCS, comprised of an activated-carbon adsorber and a
high-efficiency particulate air (HEPA) filter was in place. This secondary
APCS ensured that RKS stack emissions were acceptable even under incineration
failure conditions.
The objective of the Phase I tests were to establish that UDMH and N,0,
can be destroyed in an incineration system in a manner that meets U.S. and
Russian environmental regulations. The applicable U.S. environmental
regulations are the hazardous waste incinerator performance standards
established under the Resource Conservation and Recovery Act (RCRA). These
standards require that the incinerator achieve:
At least 99.99 percent destruction and removal efficiency (ORE) of
the principal organic hazardous constituents (POHCs) in the waste
feed to the incinerator
HC1 emissions of no more than 1.8 kg/hr or 1 percent of the HC1
entering"the incinerator's APCS, whichever is greater
The promulgated regulations require that particulate emissions be no
greater than 180 mg/dscm (0.08 gr/dscf) corrected to 7 percent 0,. However
recent EPA guidance, planned for incorporation into hazardous waste
61
-------
incinerator operating permits as they are issued or renewed, states that
particulate emissions be limited to 34 mg/dscm (0.015 gr/dscf) corrected to
7 percent 02.
In addition, hazardous waste incinerator permits currently being
enforced in the U.S. require that CO emissions be no greater than a 1-hour
rolling average of 100 ppm, corrected to 7 percent 02, and limit hazardous
constituent trace metal feedrates to levels designed to prevent exceeding
risk-based ambient levels. The hazardous constituent trace metals are
antimony, arsenic, barium, beryllium, cadmium, chromium, lead, mercury,
silver, and thallium. Finally, current guidance states that emissions of
total tetra- through octa-chlorinated dibenzo-p-dioxins and dibenzofurans
(PCDDs/PCDFs) be limited to 30 ng/dscm corrected to 7 percent 02.
Discarded or off-specification UDMH to be destroyed or disposed of would
be the listed hazardous waste U098. The POHC for this waste for which an
incinerator would need to achieve 99.99 percent ORE or better is the UDMH
itself. Discarded or off-specification N,0A would be listed waste P078. P078
is listed as nitrogen dioxide (N02). Although N204/N02 is not an organic
constituent, requiring 99.99 percent N204/N02 ORE or better was thought to be
an appropriate incinerator requirement.
The Russian environmental regulations limit the emissions of UDMH as
well as several potential UDMH products of incomplete combustion (PICs) from
the incineration of UDMH. The limits are occupational exposure limits in
terms of maximum permissible concentrations in workplace air.
The test program consisted of a series of individual incineration
tests. Three tests were performed under the same incineration system
operating conditions feeding each component of the missile propellant.
Triplicate testing is a requirement for U.S. hazardous waste incinerator trial
burns. In addition, triplicate testing establishes a measure of confidence in
test results by allowing the precision of the test program sampling and
analysis procedures to be evaluated. Two sets of triplicate tests feeding
UDMH (six total) were required to complete all the flue gas sampling
procedures for the UDMH feed tests. Thus, nine tests in total, six feeding
UDMH and three feeding N204, were performed.
The tests showed that environmentally acceptable incineration could be
achieved under at least one set of incineration conditions, although perhaps
not the optimum set of conditions.
The UDMH destruction tests were performed at kiln exit gas temperature
of 980eC (1,800°F). Only UDMH was fed to the kiln along with the required
atomization media and combustion air. The RKS auxiliary fuel, natural gas,
was zero during actual testing, although natural gas was used for incinerator
heat up, and to maintain incinerator temperatures overnight between tests.
UDMH was fed to the kiln via the liquid waste/fuel nozzle of the kiln's dual
fuel burner. The UDMH was directly pumped and metered from its nitrogen-
blanketed storage container to the burner nozzle.
The N20A destruction tests were also performed at a kiln exit gas
temperature of- 980°C (1,800°F). Diesel fuel served as the material to be
62
-------
oxidized by N20, for its destruction. The diesel fuel was fed to the kiln via
the liquid nozzle of the kiln's dual fuel burner. The N20, oxidant was added
to the burner primary air supply. 2 4
The mode of operation will require an N,0, evaporator between the N,0,
storage container and the kiln burner, which delivered NP0, vapor to the
burner air supply line. The evaporator was fed from the N,0, storage
container by pressurization of the storage container nitrogen blanket
Evaporator discharge vapor temperature was maintained at 32'C (90'F)
comfortably above the N204 boiling point of 21'C (70'F). The N20, supply line
to the kiln burner air supply line was heat-traced to ensure tha! no N 0
a!r°"Spp?f !?ne°atCU^nJhbUrne?.Hn6 "" """"^ 1nt° th6 bU™r >Amr*
For all test feeds, the RKS afterburner was fired with natural gas to
maintain an afterburner exit gas temperature of 1,090'C (2,000'F) The IRF
Sf2?^??^*??*?!:??!!?96!!!611* permiJ specifies a minimum afterburner'temperature
of 1,017 C (1,863 F) whenever a hazardous waste is being fed to the svstem
The 1,090'C (2,000'F) set temperature was chosen as a typical afterburner
operating temperature, arid that level also allowed a margin of flexibility
above the permit-mandated temperature.
Heat balance calculations were performed to estimate firing rates
associated with the target incinerator operating conditions The heat of
combustion of UDMH is 33.0 MJ/kg (14,200 Btu/lb); that of typical diesel fuel
is 45.3 MJ/kg (19,500 Btu/lb). The heat of reaction of diesel fuel and Nn
is about 45.7 MJ/kg of diesel fuel (19,700 Btu/lb). 2 4
For all tests, the RKS scrubber system was operated at its nominal desiqn
conditions and at as close to total recirculation (zero to minimum blowdown
as possible. Flow rates required to achieve the target kiln and system
operating conditions were based on scoping tests. These scoping tests
explored the ability to safely achieve stable incinerator operation and will
be discussed in greater detail in the presentation
63
-------
EVALUATION OF A ROTARY KILN INCINERATOR AS A THERMAL DESORBER
Justice A. Manning
EPA, ORD, CERI
513/569-7349
Robert C. Thurnau
EPA, ORD, RREL, TDB, WMDDRD
26 W. M.L. King Dr.
Cincinnati, OH 45268
513/569-7504
INTRODUCTION
Incineration is a common means of treating soils at Superfund sites contaminated with
organic hazardous constituents. Low to moderate thermal desorption is alternative technology.
Operation of a rotary kiln incinerator in the temperature range of a thermal desorber is an attractive
application for the remediation of soils. EPA decided to investigate such a possibility at its
Jefferson, AR, rotary kiln incinerator site. Advantages of operation in the low to moderate
temperature range include less fuel consumption, avoidance of slag formation, and less likelihood of
volatilizing trace metals that may be in the soil.
The objectives of the test program designed by EPA and its contractor were to study the
effects of five parameters believed to be of primary importance in decontaminating soils containing
organics and trace metals. Test parameters included: soil moisture content, treatment temperature
range, treatment time, solids bed depth, and degree of solids agitation. This paper is limited to
reporting preliminary results on: moisture content, treatment temperature versus type compound
(i.e., volatile vs. semivolatile), and soil feed rate versus treatment time.
METHODOLOGY
A synthetic feed material was prepared using a mixture of local top soil and clay. The
organic contamination was achieved by mixing nine organics ranging from volatile (benzene) to
semivolatile (pyrene) compounds (see Table 1) and adding this mixture to the soil. Soil moisture
contents of 10 and 20 percent were selected for testing. Soil moisture content was held constant
while the other test parameters were varied by changing the operating conditions of the incinerator.
These conditions were changed from test to test by varying the contaminated soil feed rate, kiln
exit gas temperature, and kiln rotation rate. Twelve tests under 11 different combinations of test
variables were performed; one duplicate test was performed.
Test combination targets consisted of: three different kiln exit gas temperatures (320°C,
480°C, and 650°C), two soil feed rates (70 and 210 kg/hr), and three kiln rotation speeds (0.2,
0.5, and 1.5 rpm). The peak solids bed temperature was achieved by changing the kiln exit gas
temperature. Bed temperatures corresponding to the above exit gas temperatures were about
120°C, 270°C, and 430°C (see Table 2). Residence time was achieved by varying the kiln
rotation speed. Total kiln solids residence times corresponding to the above kiln rotation speeds
were 60, 40, and 30 minutes, respectively. Solids bed depth was affected by both feed rate and
kiln rotation speed.
To allow for a more thorough evaluation of treatment effectiveness at varying treatment
times for each test condition, samples of the solids bed material were taken at four axial locations
along the kiln, in addition to a solids discharge sample. These samples corresponded to four
64
-------
different treatment times for each test condition.
RESULTS/CONCLUSIONS
• "'A summary of achieved operating parameters and resulting test conditions is given in Table
3. As noted by comparing Tables 2 and 3, target test conditions were met in most cases,
especially for the exit gas temperatures. As noted in the Introduction, the results reported in this
paper are limited to the effectiveness of treatment by comparing bed temperature vs. kiln exit
temperature; treatment temperature vs. type of compound; soil feed rate vs. type of compound;
and the effect of moisture on degree of treatment.
The effect of moisture at the different kiln operating temperatures with the low feed rate is
depicted in Figures 1 a through 1 f. By examining Figure 1, one can see that moisture affected the
treatment of benzene very little at any of the kiln temperatures, while for pyrene the higher
moisture content shifted the time required for desorption of pyrene significantly to the right.
Moreover, a minimum of about 20 to 30 minutes is needed for the bed temperature to vaporize the
moisture before treatment of the volatile organic compounds can be effected.
At lower operating temperatures high moisture content has a noticeable effect on the
volatile compound, benzene, while little noticeable difference is indicated for the semivolatile
compound, pyrene. Thus, depending on operating temperatures, moisture definitely adversely
affects the reduction of organic compounds from soils. For the semivolatile compounds, only about
99.5 per cent reduction is achieved regardless of the kiln temperature.
Figure 2a-d is a depiction of the effect of feed rate on the effectiveness of treatment.
Obviously, the faster feed rate adversely affects the reduction of pyrene. This results from the
insufficient time for the bed temperature to reach a level necessary for the semivolatiles to be
volatilized. Even benzene is affected at the lower temperatures.
The results to date indicate effects that were expected. More analysis will be performed
prior to the presentation of the paper and the additional analysis will be available at the Symposium.
TABLE 1. ORGANIC CONSTITUENTS IN THE SYNTHETIC CONTAMINATED SOIL
Compound
n-Hexane
Benzene •
Toluene
Tetrachloroethene
n-Octane
Chlorobenzene
Naphthalene
Phenanthrene
Pyrene
Molecular
weight
86.2
78.1
92.7
165.9
114.2
112.6
128.2
178.2
^02.2
Specific Melting point,
gravity °C
0.66
0.88
0.87
1.62
0.70
1.11
1.16
1.18
1.27
-94
6
-95
-22
-57
50
80
100
156
Boiling point,
°C
69
80
111
121
126
132
218
340
404
Organic liquid
mixture
composition,
wt%
15
15
15
24
15
10
3
2
1
Concentration in soil
at an organic liquid
fraction of 2%,
mg/ikg
3,000
3,000
3,000
4,800
3,000
2,000
600 .
400
200
65
-------
TABLE 2. TARGET TEST CONDITIONS
Test
1
2
3
4
5
6
7
8
9
10
11
12
Kiln exit gas
temperature,
°C(°F)
320 (600)
480 (900)
650 (1,200)
320 '(600)
480 (900)
650 (1,200)
480 (900)
480 (900)
480 (900)
480 (900)
480 (900)
480 (900)
Expected peak
solids bed
temperature,
°C(°F)
120 (250)
270 (520)
430 (800)
120 (250)
270 (520)
430 (800)
270 (520)
270 (520)
270 (520)
270 (520)
270 (520)
270 (520)
Kiln rotation
rate, rpm
0.2
0.2
0.2
0.2
0.2
0.2
0.5
0.5
1.5
0.2
0.2
0.2
Soil feedrate,
kg/hr (Ib/hr)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
70 (150)
210(470)
210 (470)
70 (150)
Soil
moisture
content,
%
10
10
10
20
20
20
10
20
20
10
20
10
66
-------
cc
2
O
I-M
&
Q
2
O
U
O
5
H
W
eu
O
1
J
1
u
<
Ed
09
Ł
g|
1 ^
'.S
0^
S;«J
5.
N
>
^
to
« ^
**
r*)
"1
«S
5.
V
1
#
0?
r^ S
NO O
,?
VH ^f
^^
^^
SSx
^
58 1
<-N
S§
0?
t~ S
p
1-H "*
PI
g|
P
j-N
S O
f
=9
g
al^^
gJ S3
T S
ss
sf
S
3
-3
5 *
^1 00 ^
„
•
N?
-N
rn
so
a
»
S
oj t-i o? r*
^ ^4 V) •
^-< «s rl
-
B « S R
T-< i-H ^H f\J
3T ?S *** **^
C. C' ^ 12
S i § Ł
|8 S1 g
JS 0*
*N «-•
*-" en
o?
ON 0 !Q
!•* f*J CM
T o
s fO ^
a a 5, 1
a .s a B
\o 55 o* ^*
SM^ >w^ VMS N—'
g s » a
,
U-j
NO
67
-------
BZ
Effect of Moisture, 12DOT, 150 Ib/hr
PVR
Effect of Moisture, 12001s, 1SOlb/hf
4T
TreatmantTImo (mfn)
BZ
Effect of Moisture, 9001=, 150 Ib/hr
i* s
TS
ft.
21
i I i I i r
OH—r
0 S3 40- 60 80
Troaimont Tfmo.(m'n)
PYR
Effect of Moisture, SOOT, 160 Ib/hr
0 20 40 60 60
Treatment Time (rrtn)
Figure la-d. Effect of Moisture vs. Kiln Exit Temperature (Low Feed Rate)
68
-------
BZ
Effect of Moisture, eOOT, 160 Ih/hf
0 20 40 60 80
Treatment TImt (mto)
PYR
Effect of Moteiuro, 60QT, 160 \bfrr
C*
0
T—r
20
I ' ?
40
60 60
Treatment Time (mtn)
Figure le-f. Effect of Moisture vs. Kiln Exit Temperature (Low Feed Rate)
69
-------
•fc
BZ
Sfoct of Food Rate
0 20 40 60
Treatment Time (mln)
PVR
Effect ef Food Rate
Treatment Time (mln)
w
0>
*
Ł
c
Ł
BZ
Effect of Fstd Rate
0 20 40 60
Treatment Tim© (min)
? 3
O
&
c
o
1 «
c
e
PYR
Effect of Food Rate
0 20 40 60
Treatment TImo (mln)
60
Figure 2a-d. Effect of Waste Feed Rate on Organic Treatment
70
-------
RESULTS OF THE MITE PROGRAM'S MATERIAL RECOVERY FACILITY (MRR EVALUATIONS
Lynnann Hitchens
Risk Reduction Engineering Laboratory
US Environmental Protection Agency
5995 Center Hill Road
Cincinnati, Ohio 45224
(513) 569-7672
INTRODUCTION
: Increasing environmental concern over the disposal of municipal solid waste (MSW) in the
United States has prompted municipalities to increase recycling activities in order to conserve natural
resources and minimize the amount of solid waste requiring disposal. Material Recovery Facilities or
"MRFs" have become a major component in solid waste recycling systems. A MRF is a facility that
accepts commingled recyclables (such as plastic, glass, paper and aluminum) and processes the
material for market. The incoming materials are collected at the curbside, usually from a bin or bag, and
delivered directly to the MRF for processing. There are over 200 MRF's in the United States, and
facilities are continually being sited and built.1 MRF designs vary greatly depending on the requirements
of the individual recycling program, economic constraints of the municipality, and the solid waste
generation rate of the area.
This evaluation is being conducted under the auspices of the Municipal Solid Waste Innovative
Technology Evaluation (MITE) Program. The purpose of the MITE program is to provide objective, third
party evaluations of solid waste management and recycling technologies and transfer this information to
municipalities and the public sector. The purpose of the MRF evaluation is to gather data and
information that will answer the many questions that still exist regarding MRFs. Though MRFs are
continually being sited and built, there has been no in-depth evaluation of existing facilities, examining
potential environmental problems and quantifying their impact on the existing solid waste management
system.
For the purposes of this evaluation, it was decided that a cross-section of facilities was
necessary to address the full range of operating characteristics. The MITE evaluation includes six
facilities, differing in size, ownership (public or private), geographic area, and separation scheme. The
six facilities that volunteered to participate in this study include:
Islip, New York
Montgomery County, Maryland
i - Albuquerque, New Mexico
Hartford, Connecticut
; - Rice County, Minnesota
Orange County, Florida
In order to adequately address the pertinent environmental and economic issues, each
evaluation consists of the following four components:
• Sampling of the air quality within and surrounding the facility.
Materials Recovery and Recycling Yearbook: Directory & Guide 1992-1993
71
-------
• A material mass balance and calculation of recycling rate.
• Calculation of the energy requirements, including the collection, separation, and
transportation of the recyclable material.
• Documentation of the capital and operational costs, within the full integrated solid waste
management system.
The results of this study will be used to provide municipalities and others with information on
successful programs, and to assist in economical and environmentally safe facility design and operation.
METHODOLOGY
A program plan was prepared considering the four components of the study. Each MRF
evaluation consisted of two distinct parts: 1) a five day on-site evaluation that included interviews with
facility managers and environmental sampling, and 2) a questionnaire, which gathered the other needed
Information on capital and operating costs, energy consumption, and recycling and recovery rates. The
same evaluation plan was used for each MRF selected.
On-S'rte Evaluation
The facilities were approached and selected by the Solid Waste Association of North America
(SWANA), and a planning visit was conducted by the sampling team prior'to any project scheduling.
Each facility operated under a specific sampling plan, tailored to the design and operating practices of
that specific facility. The sampling plans underwent quality assurance review as well as review by the
facility operators. The MRF operator review information was made available to help ensure that all
parties were informed about what measurements would be taken.
The facility sampling consisted of three types of environmental measurements: ambient outdoor
sampling, ambient indoor sampling, and personnel sampling. The ambient outdoor measurements were
taken at the facility fenceline to ensure the inclusion of emissions from truck traffic with three outdoor
sampling stations established, one upwind of the facility and two downwind. Inside the MRF both grab
samples and personnel samples were taken. The personnel sampling was performed using monitors
worn by selected personnel who were chosen as to represent different job tasks inside of the MRF. In
addition to the measurement of particulate and gaseous contaminants, noise measurements were taken,
both outside and insides the facility.
Since there was a lack of literature data on any of these contaminants in relation to MRFs, it was
decided that a large list of measurements would be used for the first two facilities. Based on the results
shown from the first two facilities, the list would be reduced by eliminating the measurements in which
non-detectable levels of contaminants were found. The services of an Industrial Hygienist were used to
determine which measurements were most appropriate. The reduced sampling plan was established
and used at the remaining four facilities. The field sampling team retained the flexibility of adding any of
the eliminated measurements, if there was the potential for detectable levels of that particular compound
to be present. Table 1 outlines the full list of contaminants, where the measurements were taken and
what measurements were retained in the Phase II testing.
EPA sampling and analytical methods were used as appropriate and all indoor contaminants
were measured using the appropriate National Institute of Occupational Safety and Health (NIOSH) test
methods.
72
-------
TABLE 1. ENVIRONMENTAL MEASUREMENTS
Contaminant
Total Suspended Particulates
PM10
Lead (Pb)
Carbon Monoxide (CO)
Mercury Vapor (Hg)
Acetone
Benzene
MEK
Methyl Chloride
Toluene
TCA
Xylenes
Noise
Total/Respirable Dust
Pesticides
Metals (Al, As.Cr, Ni)
Crystalline Silica
Microbiological
Measurement Location
Exterior
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Interior
X
X
X
X
X
X
X
X
X
Status for
Phase II
X
X
X
X
X
VOCs reduced to one
day only
X
X
Reduced level of
sampling
Eliminated due to non
detectable levels
X
X
Economic Assessment
A two part questionnaire was developed by project personnel and the contractor, Roy F.
Weston, Inc. and sent to each facility operator. The purpose of Part 1 of the questionnaire was to
document the cost elements of the integrated solid waste management system, including capital and
operating expenses, as well as education and outreach. The purpose of Part 2 was to gather data on
energy requirements, by collecting information on the electricity required for facility operation and fuel
consumption for transportation. This energy consumption/generation information is to be collected for
all components of.the solid waste management system, including disposal components such as landfills
or waste-to-energy systems? The questionnaires are being compiled and interviews with solid waste
department will be conducted to fill in any data gaps that exist.
73
-------
RESULTS AND CONCLUSIONS
The evaluation was set up to complete sampling at two facilities using the full range of tests and
measurements outlined previously. After receiving the results of these tests, areas in which no
detectable levels were found were eliminated from further testing. This allowed project costs to be
reduced without jeopardizing the integrity, of the sampling plan. It should be noted that elevated levels of
crystalline silica and aluminum have not been found, and these measurements were retained in the
sampling plan, since glass and aluminum containers are present in every feedstock.
A conclusion that has already been reached, prior to final data analysis, is that no two facilities
are the same, even if the design characteristics are identical. Each system must be designed and
adapted within the framework of the integrated solid waste management system that already exists.
Flexibility has also arisen as a major issue. The more successful MRFs are those that are designed to
be flexible in response to changing recycling markets and market conditions. This allows for changes in
processing requirements and making the facility more able to meet changes in markets.
One of the more difficult tasks of this project is documenting the costs associated with the
individual systems. Each facility has different approaches to accounting, financing, facility ownership
and operation. Different contractual arrangements are negotiated based on the needs of the individual
municipality or solid waste management district. Placing these costs on a level playing field in order to
document the true cost of recyclable collection and processing has proven to be a difficult task.
A report will be published based on the results of this evaluation. The report will outline the
major findings and conclusions of the project. The purpose of the report is to assist solid waste officials
in their decision making process, and expose them to technologies and approaches that they are
currently unfamiliar with. This evaluation can be used as a tool to educate the solid waste decision
official on the ranges of technology that are available.
REFERENCES
1.
2.
Berenyi, E. Gould, R. Materials Recovery And Recycling Yearbook: Directory & Guide 1992-
1993. Government Advisory Associates. 1992.
Peer Consultants, P.C. Material Recovery Facilities for Municipal Solid Waste.
EPA/625/6-91/031, U.S. Environmental Protection Agency, Cincinnati, Ohio, 1991.
FOR MORE INFORMATION: Contact Lynnann Hitchens, Risk Reduction Engineering Laboratory,
5995 Center Hill Road, Cincinnati, Ohio 45224, (513) 569-7672.
74
-------
FULL-SCALE LEACHATE-RECIRCULAT1NG MSW LANDFILL BIOREACTOR ASSESSMENTS
David A. Carson
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory (ML-CHL)
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268-3001 USA
INTRODUCTION
The integrated waste management hierarchy philosophy continues to develop as a useful tool
to solve solid waste issues in an environmentally responsible manner. Recent statistics indicate that
approximately two thirds of municipal solid waste in the United States is disposed in landfills. EPA
research has continued to develop and refine the variety of technologies, materials, and operational
techniques to further reduce risk from this most often used form of waste disposal.
Current landfill operational technique involves the preparation of a waste containment facility,
the filling of the waste unit, installation of the final cover, and the maintenance of the unit. The goals
being to isolate the waste from people, and to minimize infiltration of water, thus minimizing releases of
moisture and decompositional gases into the environment. This method of operation has proven to be
reasonably effective in waste disposal, effectively minimizing risk by collecting the liquid that percolates
through the waste, called leachates, at the bottom of the landfill, and controlling landfill gas with
collection systems. Effective gas collection often results in utilization, of the gas for other purposes.
. Concerns over the longevity of containment systems components present questions that
cannot be answered without substantial performance data. Landfills, as currently operated, serve to
entomb dry waste. Therefore, the facility must be maintained in perpetuity, consuming funds and
ultimately driving up waste collection costs. Further, there is a concern about rare but possible
environmental impacts from leakage through lining or cover components and possible catastrophic
failures of the landfill system.
• This presentation will describe a new form of solid waste landfill operation, it is a technique
that involves controlled natural processes to break down landfilled waste, and further minimize risk to
human health and the environment.
A landfill operated in an active manner will encourage and control natural decomposition of
landfilled waste. This can be accomplished by collecting leachate, and reinjecting it into the landfilled
waste mass. Keeping the waste mass moist will lead to a largely anaerobic system with the capacity
to rapidly stabilize the landfilled waste mass via physical, chemical and biological methods. The
system has proven the ability to breakdown portions of the waste mass, and to degrade toxic materials
at the laboratory scale.
METHODOLOGY
Experiments are designed to compare similar landfill cells as each is operated in a different
manner, either wet or dry. The experiments are time consuming and require long-term commitment to
research. The projects are also very costly because of their magnitude. To gather data and analyze
results, U.S. EPA embarked on a research program to study this operational technique for MSW
landfills in the early 1980s. Bench-scale laboratory research on the anaerobic bioreactor in MSW
landfills was conducted with the assistance of The Georgia Institute of Technology (1). Researchers
found that by simply reinjecting common MSW landfill leachate under controlled sequences that the
75
-------
following benefits could potentially be realized at full scale:
• expansion of landfill capacity through volume reductions induced by biological decomposition of
MSW in the landfill, reducing the actual number of landfills that must be sited
• improved quality and quantity of recoverable landfill gases (methane and carbon dioxide)
through controlled biological reactions
toxicity reduction of the MSW mass through biological decomposition and immobilization in the
waste mass resulting in lower pollutant concentration in leachate
• reduced post-closure monitoring time due to toxicity reduction
Building upon this fundamental research, EPA sponsored further research to take this
operational technique to the field in the form of pilot-scale landfill test cells (2). Anticipating the
opportunity to build on this research, full-scale landfills were sought and selected to prove the
technology at full-scale. One landfill was selected near Gainesville, Florida, and another landfill project
is in the construction phase near Rochester, New York.
Table 1. U.S. EPA Landfill Bioreactor Projects in 1994
Location
Georgia
Delaware
Florida
New York
Ohio
Scale
Lab
Pilot
Full
Full
Full
Description
Model Landfill Lysimeters
Test Ceils 2 each, 1 acre each, wet vs. dry
Active Fill - 6 acres
Active Fill - 10 acres
Remediation of Existing Landfill
Status
Completed '93
Covered
Filling
Filling
Planning
Ref
1
2
3
4
5
In addition to these activities, EPA is tracking related projects in the United States, and around
the world. EPA is also pursuing the application of this technology in a remedial manner at an unlined,
uncontrolled landfill.
Leachate-Recirculation Process
Prototype methods which have been utilized to date include surface spraying, surface ponds,
vertical injection wells with and without wicks, and horizontal surface infiltration devices. Generally,
additional costs for leachate recirculation materials are relatively low. For proper execution of a
leachate recirculation system at a modern MSW landfill, the following components are generally
considered to be necessary:
• a composite lining system comprised of a single or double composite compacted soil and a
geomembrane, with leachate handling system.
MSW placed at a density determined to accommodate leachate recirculation
daily cover that does not significantly affect the continuous passage of moisture through the
landfill from top to bottom (this rules out many traditional MSW landfill daily cover practices)
76
-------
a leachate reintroduction system that is concealed within the landfill enclosure, that only uses
native landfill leachate
an active gas collection system as part of a comprehensive landfill cap design
a landfill cover capable of providing modern landfill cap functionality, that can maintain integrity
as the leachate recirculation process causes landfill volumes to decrease,
a trained landfill operator who understands the daily operational requirements of a leachate-
recirculating landfill
Many of these issues remain as engineering challenges that need to be resolved in full-scale
implementation.
The leachate-recirculation process is fundamentally simple, requiring relatively minor changes
over the current designs. However, the number of variables involved to demonstrate control over the
reactions in these studies are numerous, and every attempt is made during experimental design to
control the number of variables. Operator training is critical, and outputs from these projects are
aimed at operator control. As a matter of practicality, experiments to evaluate materials that may be
added to the landfill during filling to enhance degradation have been relegated to future
experimentation. Possible additions include waste pre-processing, microbiological additives, waste
sludge, gases, or the control of temperature of the waste mass.
Concurrent with EPA research, other research projects are underway in the USA and around
the world. There are new projects being initiated in California, Florida and New Hampshire and
underway around the world in Canada, Germany, Denmark, Italy, Sweden, Japan, and United
Kingdom. The collective database formed in the study of these landfills will supply enough information
to assess the performance of this operational technique in the field.
Landfilling in the Future
As landfills continue to evolve as sophisticated waste disposal and decomposition facilities
there are new goals anticipated for the future. Foreseen is a waste management park that will involve
centrally located management of a variety of waste streams, of which a landfill will provide a necessary
service. Waste water, sludges, composted waste and green waste, MSW materials recovery facilities
incinerators, and recycling facilities require inputs and outputs that can be integrated resulting in more
efficient processing of MSW.
The landfill could produce an output of low to medium grade compost that could be used for
soil amendments in roadways and earth works. Gas from the landfill could be used to generate
electricity or power collection vehicles. The landfills could be arranged in a turntable-type arrangement
so that cell #1 of the landfill can be constructed, filled, covered, and placed under leachate
recirculation, then on to cell #2, where the process is repeated, and so on. When all cells are
complete, the operator returns to cell #1, where the contents could be mined (7), with recyclables and
compost removed, the bottom lining inspected and replaced when necessary, and the entire process
again. While this technique will likely not eliminate the need to site new landfills, it will significantly
increase the useful life of landfills when constructed to operate as bioreactors.
RESULTS
The status of active EPA sponsored projects are described here. The effort has benefitted
from the active participation of major waste management companies in the USA, and from the
77
-------
collaborative research efforts of EPA's colleagues, and their participation is gratefully acknowledged.
Results to date are derived from the following projects:
Laboratory Lysimeter Studies
Completed in 1993, the research performed at the Georgia Institute of Technology and
University of Pittsburgh showed leachate recirculation offers rapid and complete stabilization of the
landfilled waste masses. The system proved resiliency to toxic loadings that caused retardation but
not defeat of the stabilization process. The study showed that landfills are capable of biological and
physicochemical reactions to attenuate waste constituents via reduction, precipitation, and matrix
capture for heavy metals, and biotic and abiotic transformation, and sorption for organic materials.
The study also showed that the performance of the bioreactor landfill can be monitored with leachate
and gas parameters during significant phases of the reaction. The study concluded that the bioreactor
landfill design is viable, and offers a significant improvement over traditional landfill operation (1).
Pilot-Scale Landfill Test Cells - Delaware
Two 1-acre pilot-scale landfill test cells are currently operated by the Delaware Solid Waste
Authority in Sandtown, Delaware. Two identical ceils were constructed, one operated traditionally (dry)
and the other employing leachate-recirculation. The cells have been operating for approximately 1.5
years, and although the data is being analyzed, preliminary results have shown that the trends
described in the early laboratory research are repeated in the field (2).
Full-Scale Landfill Project - Alachua County, Florida
The first of two full-scale landfill projects was selected near Gainesville, Florida. The landfill is
operated by the City with assistance from The University of Florida, engineering firms, and others.
The landfill is approximately half full, and leachate recirculation systems are being installed as filling
progresses. A nearby landfill at that site has performed this technique in a less sophisticated manner,
utilizing surface infiltration ponds in previous years with success (3).
Full-Scale Landfill Project - Monroe County, New York
The second full-scale landfill project is in the construction phase near Rochester, New York.
EPA is assisting the New York State Energy Research and Development Authority (NYSERDA) in their
efforts to study how this landfill operational technique may achieve complementary goals of
environmental protection, and the potential for enhanced energy production through gas recovery (4).
Remediation Project - Ohio
An unlined landfill is under study in Cincinnati, Ohio for application of leachate recirculation as
an alternative to pumping and treating groundwater at an uncontrolled pre-regulatory landfill. The
project is significantly more complicated due to the absence of a bottom liner, but researchers are
convinced that groundwater can be controlled and barriers can be constructed to accommodate
leachate recirculation at the site (5). A feasibility study is in final draft status, and the project will soon
enter its second phase to complete pre-construction details.
Comprehensive Study - University of Central Florida
A comprehensive evaluation of active landfill bioreactor projects is underway at the University
of Central Florida (6). Researchers there are compiling data from these projects, other American
projects, and data from projects from around the world by site visitation and interactions with specialty
78
-------
groups like the International Energy Agency's Landfill Gas Expert Working Group Specific projects
aim to assess moisture distribution of the leachate into the waste mass through computer models
This will assist designers in the configuration of the injection system, and may lead to guidance on
placed waste densities that will accommodate leachate recirculation. A conference is planned for the
autumn of 1994.
CONCLUSIONS
Landfills are currently designed and operated to serve as containment systems While
originally designed to entomb waste, the modern landfill has evolved into a more technically advanced
containment system, with sophisticated controls and operational techniques. The landfill of the future
will be protective of human health and the environment, will degrade the waste mass in the landfill and
will be reusable, allowing the waste to be excavated for the cell to be inspected and refilled as Dart of
an integrated waste management park.
REFERENCES
1.
2.
3.
4.
5.
6.
United States Environmental Protection Agency. 1993. Behavior and Assimilation of Organic
and Inorganic Priority Pollutants Codisposed with Municipal Refuse Volumes 1 and 2
EPA/600/R-93/137a and b (NTIS PB93-222198 and 93-222206). Risk Reduction Engineering
Laboratory, Cincinnati, Ohio. . .
Vasuki, N.C. 1993. "Practical Experiences with Landfill Leachate Recirculation in Pilot and
Field Scale Units, Proceedings: 31st Annual International Solid Waste Exposition Solid Waste
Association of North America, San Jose, California, August 3.
Townsend, T, and W. L. Miller. 1993. Alachua County Southwest Landfill Bioreactor Project
Proceedings: Madison Waste Conference, Madison, Wisconsin, September.
Babcock, J., and Reis, J. 1993. "Monroe County's Leachate Recirculation Project" Modern
Double Lined Landfill Management Seminar, New York State Department of Environmental
Conservation, Saratoga Springs, NY, March 2-4.
United States Environmental Protection Agency. 1993. Leachate Recirculation for Remedial
Action at MSW Landfills, DRAFT FINAL REPORT. Risk Reduction Engineering Laboratory
Cincinnati, Ohio. .
Reinhart, D. R., and Carson, D. A. 1993. "Full-Scale Application of Landfill Bioreactor
Technology," Proceedings: 31st Annual Solid Waste Exposition of the Solid Waste
Association of North America, San Jose, CA (August 2-5, 1993).
7. United States Environmental Protection Agency. 1993. Evaluation of the Collier County
Florida Landfill Mining Demonstration. EPA/600/R-93/163. Risk Reduction Engineering'
Laboratory, Cincinnati, Ohio.
FOR MORE INFORMATION
Please contact David Carson, U.S. Environmental Protection Agency, Risk Reduction
Engineering Laboratory (Mb-CHL), 26 W. Martin Luther King Drive, Cincinnati, Ohio, 45268-3001, USA,
3t (513) 569-7527.
79
-------
PROTOCOL - A COMPUTERIZED SOLID WASTE QUANTITY AND COMPOSITION
ESTIMATION SYSTEM
by
Albert J. Klee
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 W. Martin Luther King Dr.
Cincinnati, Ohio 45268
INTRODUCTION
Traditional sampling theory generally follows the following paradigm:
SAMPLE SELECTION - SAMPLE OBSERVATION - SAMPLE ESTIMATION.
Typically, the samples are chosen by an unbiased procedure, such as simple
random sampling or systematic sampling where it is assumed that the population
is already in random order. In the sample observation or data recording
stage, it is further assumed that observation of the elements of the sample is
an independent process, i.e., that there is no queue of sample elements build-
ing up waiting to be observed while an observation on one sample element is
being made. Unfortunately, these assumptions do not fit the circumstances when
the problem is to estimate the quantity of solid waste arriving at a given
site a specific point in an industrial or commercial process. For one thing,
the sample comes to the investigator, which is the reverse of the situation
commonly described in standard sampling textbooks. Since the investigator has
no control over the arrival of the sample elements, the sample observation
process often is far from independent.
METHODOLOGY
Consider a situation where it is desired to weigh a random or systematic
sample of 10% of the vehicles arriving at a landfill. Suppose that it is
feasible to weigh up to 10 vehicles per hour and that the average interarnval
time of vehicles is 3.2 minutes. On the average then, with either random or
systematic sampling, one vehicle would be weighed every 32 minutes. If it
takes an average of 10 minutes to weigh a vehicle, then this is well within
the capability of the sampling system. Unfortunately, vehicles do not have
uniformly distributed arrival times. There may be peak arrival periods when
the number of trucks arriving and to be sampled overwhelms the weighing cap-
ability and one is forced to "default" on weighing some of the vehicles se-
lected by the sampling plan. One result is that fewer vehicles are weighed
than the sampling plan calls for, thus reducing the precision of the estimate
of solid waste quantity. More important, however, is that if the load weights
of the defaulted samples differ appreciably from the nondefaulted samples,
bias will be introduced. For example, at many landfills vehicles arriving
toward the end of the day tend to have smaller load weights than those arriv-
ing at other times. Since fewer vehicles arrive towards the end of the day,
the tendency would be to oversample these lightly loaded vehicles and to
undersample the normally loaded vehicles arriving at peak hours, thus intro-
ducing a bias._
80
-------
When the rate of arrival of vehicles is not uniform throughout the day
random usamples will produce some intervals of little or no activity and others
of frenzied activity. Once scales are rented and labor hired to make the
observations, it makes little sense not to use the equipment and labor to the
fullest extent possible. If one samples to the fullest extent of the sampling
capability, not only can the bias can be removed mathematically by equations
developed in this study, but the resulting estimate has a smaller variance
than random or systematic sampling.
SEASONALITY
It is well-known that the quantity of solid waste generated varies sig-
nificantly from month to month. Municipal solid waste generation, for example
is typically low during the months of January, February, November, and Decem-'
ber, and peaks during June, July, and August, although there are some varia-
tions depending upon geographical location. Thus it is not sufficient to
sample one week out of the year to estimate generation for the entire year A
Monte Qarlo simulation was performed using actual data obtained from a Boston
solid waste site. Two sampling protocols were simulated: (1) random sampling
and (2) systematic sampling. Sampling frequencies of from two to ten times per
year were investigated. Systematic sampling was found to be superior to random
sampling for all sampling frequencies except for a sampling frequency of two
where random sampling is identical to systematic sampling. At a sampling
frequency of four times per year, for example, the coefficient of variation
(i.e., the standard deviation as a fraction of the mean) of systematic sam-
pling was only 56% that of random sampling.
SAMPLE WEIGHTS FOR COMPONENT ESTIMATION
Traditional sampling theory assumes that there are sampling elements,
i.e., discrete entities comprising the population about which inferences are
to be drawn. When it comes to sampling solid waste for composition, however
there are no such discrete entities. There is, for example, no such thing as
a basic unit of paper or of textiles. Thus, sampling procedures based upon
discrete distributions (such as the multinomial or binomial) ,are not valid
Nonetheless, some basic unit weight of sample must be defined. In traditional
cluster sampling theory, a balance is achieved between the within-cluster and
between-cluster components of the total variability of an estimate If the
cluster (i.e., in this context, a sample of given weight) is too small, then
the between-cluster variability will be greater than the within-cluster vari-
ability^and will result in a large sample variability. If the cluster weight
is too large, however, the greater will be the time and expense of sampling
Since this is not a linear relationship (i.e., doubling the cluster size will
not necessarily double the precision of the estimate), the optimal procedure
is to select that cluster size where the precision of the sampling estimate
does not significantly improve with cluster size. For raw municipal solid
waste, this sample we-ight has been found to be between 200 and 300 Ibs For
processed waste streams, on the other hand, it has been found that particle
size distributions are adequately described by functions based upon the expo-
nential distribution, such as the Rosin-Rammler equation. Such equations are
also fully described in this study.
81
-------
COMPONENT ESTIMATION
When it is desired to place confidence intervals about the estimates made
in the sample estimation process, traditionally the distribution of either the
population or of the population parameter estimated is assumed to follow a
specific classical probability distribution; typically, the normal distribu-
tion is assumed. In the sample selection process, similar assumptions are made
when determining the number of samples to be taken. The method is applied in
an iterative, trial-and-error procedure.
Although such assumptions are justifiable in the estimation of solid
waste quantity, such is not the case in estimating solid waste composition.
For one thing, component fractions are bounded, i.e., there are no components
in solid waste that are present in fractions less than zero or greater than
one. These boundaries are generally located close to the means of their
distributions. Thus, solid waste component distributions are, at the very
least, positively-skewed (i.e., skewed to the right) and, at worst, are 0-
shaped. Nor does reliance on the Central Limit Theorem of statistics help
much, since even averages of component fractions do not approach normality
quickly, at least not within an economically feasible number of samples.
Distributions of component averages still tend to be positively-skewed. This
characteristic precludes the rote application of the traditional statistical
formulas for either the estimation of sample size or the construction of
confidence intervals. Although transformations can be used to construct the
proper asymmetric confidence intervals after the sample data is obtained,
these are of little help for estimating sample size before the sample is
taken. A knowledge of the effect of positive skewness on the actual level of
significance of a confidence interval, however, can be of help in determining
the number of samples to take.
Since the rationale behind using the Central Limit Theorem is to permit
the use of t-statistics to construct confidence intervals about the estimation
of the percentage of any component in the waste stream, an appropriate measure
of the ability to meet the normality requirements is the fraction of con-
fidence intervals that actually contain the true mean at a given level of
significance. For example, Monte Carlo simulations have shown that, given an
average of size 10, if a confidence interval at a nominal significance level
of alpha-nominal = .05 were constructed about the mean, the actual signifi-
cance level would be alpha-actual = .104. In other words, instead of a 95%
confidence interval, we would actually be constructing an 89.6% confidence
interval. As the size of the average gets larger, the discrepancy gets small-
er For example, at a nominal significance level of alpha-nominal = .05 and an
average of size 50, the actual significance level is alpha-actual = .138. For
moderately positively-skewed distributions, such as ferrous metals, these
discrepancies are very much smaller, even for very small sizes of the average.
From the data developed in this study, however, equations have been obtained
that relate alpha-nominal and alpha-actual.
Although the calculations dictated by these protocols - including those
for quantity, composition, and sampling size - are arithmetically tedious, a
computer program (designed to be run on personal computers with modest cap-
abilities) has been developed to carry out these tasks. The program, called
PROTOCOL, also contains routines that check input data for errors in coding,
and an editor for preparing and modifying input data files.
82
-------
CONCLUSIONS
Efficient and statistically sound sampling protocols for estimating the
quantity and composition of solid waste over a stated period of time in a
given location such as a landfill site or at a specific point in an industri-
al or commercial process, are essential to the design of resource recovery
systems and waste minimization programs, and to the estimation of the life of
landfills and the pollution burden on the land posed by the generation of
solid wastes. The theory developed in this study takes a significantly differ-
ent approach over the more traditional sampling plans, resulting in lower
costs and more accurate and precise estimates of these critical entities A
computer program, called PROTOCOL and designed to be run on personal computers
with modest capabilities, has also been developed to do the calculations
FOR MORE INFORMATION
Albert J. Klee
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 W. Martin Luther King Dr.
Cincinnati, Ohio 45268
(513) 569-7493
83
-------
PRICE INFORMATION AND THE HAZARDOUS WASTE
REMEDIATION INDUSTRY
Gordon M. Evans
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
(513)-569-7684
INTRODUCTION
The rapid growth of the hazardous waste remediation industry during the past 10 years has
resulted in the development of market imperfections which have inhibited it's efficient operation.
Through the use of the classical economic paradigm, this presentation will analyze one aspect of this
market, the role of price information. The prescriptions which are offered leave the hope that greater
efficiencies can be obtained through systematic efforts to gain consensus to the meaning of price
information.
METHODOLOGY
This economic inquiry starts with a problem that forms the foundation for all subsequent
analysis; the scarcity problem. One side of the basic economic problem is rooted in the unlimited
wants and needs of individuals. The other side reflects the limited availability of land, labor, capital,
and time. Together, the two problems create the condition of scarcity. Modern economic theory is
focused on examining the mechanisms employed by individuals and society to cope with the scarcity
problem.
Every society has faced the scarcity problem. Our society has relied on markets to help us
reach consensus on the allocation of society's scarce resources. Markets are dynamic mechanisms for
transmitting information on scarcity; this information comes in the form of prices. Trade occurs when
both buyer and seller benefit. In any market, price is a reflection of a mutual agreement between
demand and supply forces regarding scarcity. In efficient markets, the latest price information reaches
others almost instantaneously. The key idea is that markets produce prices which give information on
scarcity.
THE MARKET
It's natural for buyers and sellers to focus on that part of the scarcity problem which they are
most familiar with. Suppose a problem site contains soils and groundwater contaminated with PCBs.
Demand is likely to be driven by the public's understanding of the hazards involved and their outrage
over the problem. This manifests itself in a desire for a minimal impact, 0-residual, 0-emission,
destructive remediation. On the other side of the market are vendors prepared to provide various
technologies as a solution tojhe problem. Though both sides are reacting to the same reality, buyers
and sellers see the problem differently, and tend to act accordingly.
84
-------
It has become a common practice within the remediation industry to discuss prices in a form of
shorthand. This shorthand expression of price sets forth the terms of trade in dollars per quantity of
waste matrix treated (i.e., $/ton, $/gal, $/cu yd). Our problem begins here. The more distant one is
from negotiations for the procurement of remediation services (based on site specific conditions) the
more likety rt is that this shorthand price information will become part of the basis for decision making
While a shorthand method of conveying price information may be convenient, its use assumes that
buyers and sellers are in general agreement regarding the services rendered for the price quoted
Herein lies the source of a major market imperfection.
During the early stages of a site remediation (when various technology solution are under
consideration) buyers and sellers begin to discuss the price in shorthand terms. Unfortunately those
on the buying side of the market tend to infer that the terms of trade cover a greater range of activities
than sellers expect to deliver. At the present time, there is no common definition regarding the
activities which are reflected within the "dollar per unit treated" price quotes. The time has come to
adopt a more precise price definition.
THE SITE PROGRAM'S EXPERIENCE
A study of Applications Analysis Reports prepared under the auspices of the USEPA's
nrnPNomndf 'nT?^ Technol°9y Evaluation (SITE) Program provides some interesting insights into the
problem of ill-defined prices. Before examining that analysis, a brief review of the SITE Program and
it s efforts to report on technology costs is in order.
The SITE Program was established by Congress through the passage of the 1986 Superfund
Amendments and Reauthorization Act (SARA). The goal of the SITE Program is to assist Superfund
decision makers through the creation of reliable performance and cost data for unique and
commercially available hazardous waste destruction and treatment technologies. The special nature of
this public/private sector interaction suggested the need to establish a protocol which would quide
analysts as they try to project the costs of new hazardous waste treatment technologies Four
problems were identified which would limit the ability of a SITE demonstration to generate realistic cost
1)
2)
3)
4)
Each field demonstration is a mix of unique factors.
The research and development aspects of each demonstration will impact the observed costs
Each developer is a profit maximizer operating within a competitive marketplace
The SITE Program creates unique interactions between public and private sector forces
With the potential for dozens of demonstrations to be conducted over the life of the SITE
Program, the large number of independent variables involved with each demonstration made the
o?H6 n rT a, Smg'e; ri?id Pru tOC°' SUSpect lnstead' 9eneral guidelines were established that would
aid in the development of each cost projections. The idea was to create a high degree of uniformity
among all the SITE cost projections while allowing the unique conditions of each demonstration to
dictate how each projection would be buift.
The most important step in this process was to establish a set of cost categories which would
serve as a common framework for all base-case cost projections. These 12 categories, listed in Table
1, were an early attempt to provide some structure to the terms of trade with in this industry
85
-------
TABLE 1. COST CATEGORIES USED BY THE SITE PROGRAM
1. Site Preparation
2. Permitting & Regulatory Requirements
3. Capital Equipment
4. Start-Up
5. Labor
6. Consumables & Supplies
7. Utilities
8. Effluent Treatment & Disposal
9. Residuals/Waste Shipping & Handling
10. Analytical Services
11. Maintenance & Modifications
12. Demobilization
Under ideal conditions, a demonstration would generate enough information regarding each
cost category to build a complete "start to finish" cost projection. In reality, however, a lot of cost data
is unavailable. Thus, when assigning costs, the analyst is instructed to include only those activities for
which solid or defensible cost data is available. The analyst is instructed not to list costs for those
categories where there is limited or incomplete cost data. An empty category is not discarded, but
explicitly labeled as empty. This approach provides the reader with a clear and explicit definition of the
activities which are included for each cost projection.
In addition to the establishment of the 12 cost categories, four other rules were set forth to
guide the conduct of each SITE cost analysis. In summary, the five cost projection guidelines are as
follows:
1) Place each cost analysis within a common framework of 12 cost categories.
2) Present the cost projection as "Order-of-Magnitude° estimates (+50% and -30%).
3) Provide full disclosure of all assumptions and calculations used in the analysis.
4) Identify key operating parameters likely to have significant cost implications beyond
normal operations.
5) Offer developers with the opportunity to present their own cost analysis.
While no methodology can insure that projected costs are delivered with the same degree of
precision that engineering or chemical data is, these simple protocols go a long way to insuring a
consistent approach.
THE ANALYSIS
In an effort to characterize trie wide degree of variation found within the terms of trade, a
review of 23 published SITE Applications Analysis Reports was conducted. The economic analysis
found within each Applications Analysis Report was reviewed and their contents cataloged.
It was noted that costs projections were reduced to a simple dollar per unit treated cost quote
in 87% of the reports. This shorthand cost data was typically reported in both the Executive Summary
(where assumptions supporting the costs were usually ignored) and in the underlying cost analysis
(where the assumption were~explicitly stated).
The frequency with which each of the 12 cost categories were covered within the 23 reports
was computed and is presented in Table 2.
86
-------
TABLE 2. COST CATEGORIES, AND THE FREQUENCY OF USE
IN APPLICATION REPORTS
A. Utilities
B. Labor
C. Capital Equipment
D. Consumables & Supplies
E. Start-Up
F. Maintenance & Modifications
G: Site Preparation
H, Demobilization
I. Analytical Services
J. Permitting & Regulatory Req.
K. Residuals/Waste Shipping
L. Effluent Treatment
100%
100%
100%
96%
91%
78%
74%
74%
65%
48%
43%
39%
CONCLUSIONS
h** «. '° d!f 9;eemeunt between buyers and selle's with regard to the meaning of price information
has created a srtuation where the search for cost efficient remediation solutions is inhibted. InZ
short term, the practice of screening technologies on the basis of shorthand price data will likely lead to
shorfhanH ±rn rT** rd SJ9nifiCant C°St escala«<™ ^ur. In the long run, reliance on
IS?- ? ? f *? Pr°Vlde markets With imPerfect scarcity information, with the outcome being an
inefficient allocation of society's scarce resources.
87
-------
USING FOURIER TRANSFORM INFRA-RED SPECTROSCOPY (FT-IR)
TO MONITOR THE PROGRESS OF
PLANT BASED BIOREMEDIATION EFFORTS.
Lawrence C. Davis1, Charles T. Chaff in3, Larry E. Erickson2, William G. Fateley3,
Robert M. Hammaker3, Roseann M. Hoffman3, Narayanan Muralidharan2 and Vance P. Visser2;
Departments of Biochemistry1, Chemical Engineering2, and Chemistry3,
Kansas State University, Manhattan, KS, 66502
INTRODUCTION
Bioremediation usually provides a less costly means of correction for contaminant problems
than does a fully engineered approach (1). Frequently, intrinsic bioremediation is effective and
even less costly, but it may be more time-consuming. We have sought to develop a plant based
bioremediation in which root exudates provided by plants augment the intrinsic remediation ability
of a soil system by enhancing microbial populations. Plants, through their use of water may also
stimulate the remediation process by drawing contaminated water into the aerobic vadose zone
from the typically anaerobic ground water.
Monitoring is a major cost associated with any remediation effort. We have developed an
open path/ long path system of Fourier Transform Infrared Spectroscopy (FT-IR) to use in
monitoring gas phase concentrations of volatile compounds. All organic compounds and many
other volatiles of interest, including C02, S02, and CO, have useful infrared signatures. A
commercially available FTIR instrument has been used to monitor the gas phase above plants
growing in the presence of dissolved organic contaminants including toluene (2), The present work
describes extension of that study to trichloroethylene (TCE) and trichloroethane (TCA) in ground
water, gas phase and plant tissue.
METHODOLOGY
Design and construction of the plant treatment system was previously described (2).
Briefly, the plant growth chamber consists of two identical halves each having a channel 10 cm
wide, 35 cm deep and 1.8 m long folded in a U shape for convenience. Fluorescent lights 40 cm
above the ground surface provide illumination for growth of alfalfa planted at 10 cm intervals along
the channel. An aluminum and glass enclosure 26 cm high allows sealing of the gas phase above
the plants to monitor gas phase concentrations of volatiles. Ground water is maintained to a depth
of about 10 cm from the bottom, giving a 25 cm vadose zone. Compounds of interest are
Introduced into the water source at the desired concentration. In earlier experiments (2) toluene
and liquified phenol were added at levels of 500 pLI L water (2). For these experiments TCE was
added at 200 //L/L and TCA was added at 50 /A./L. Four sampling wells, spaced at nearly equal
intervals (33 and 66 cm from the ends) in each channel, allow monitoring of the concentration of
substances in the ground water.
Use of the MIDAC FT-IR to monitor gas phase concentrations of volatile organics was
previously described (2). Light from a glowing source enters from one side of the upper gas phase
enclosure through a KBr window. It is then reflected from a mirror set at c^ 45° to the end of the
chamber, from which it is reflected back to the 45° mirror and out through another KBr window
close to the entrance. The internal pathlength is 2.44 m while the external path is about 0.6 m.
Changes in gas phase concentrations of materials in the external path between collected spectra
affect the observed spectra but have not presented difficulties except in precision determination of
CO2 concentration.
88
-------
Concentrations of volatile organic compounds such as CH4, TCE and TCA in the ground
water have been monitored at frequent intervals by head space gas analysis using a gas
chromatograph. Peak positions relative to known standards were used for identification
Standards used are indicated below. They have also been measured with good sensitivity by use
of the FT-1R system. A 0.5 m pathlength cell closed with KBr windows and having an internal
volume of 697 ml has been used for calibration and headspace gas analysis. The sample to be
analyzed (3 mL of ground water) is collected with a syringe and nylon tubing as for gas
chromatography. It is placed into a vial (c, 10 mL), sealed with a reacti-vial cap (Pierce Chemical
Co.) and then shaken to equilibrate the head space. The gas ceil is evacuated. Then the head
space is transferred to the gas cell by opening a stopcock which has an attached needle inserted
through the closure of the reacti-vial. About 99% of the gas phase should be transferred in this
way. In this instrumental configuration the open path length is about 0.3 m while the internal path
length is 0.5 m. Only water, ammonia and C02 concentrations in the external path have affected
calibration measurements. None of these are compounds of interest for study of ground water in
these experiments.
To measure concentration of volatile materials in aboveground plant tissues the plants
were cut at 5 cm above the soil surface and quickly transferred to a glass jar which was then
sealed. The lid was modified by addition of a fitting with a rubber gas chromatograph septum to
aI°Wotra[1Sfer °f the 93S phase from the bottle to the mR celL TVPical Plant samples contained
1 0-1 2 mL of water.
Limits of detection from compounds of interest were determined by calibrating the 0 5 m
cell with known amounts of individual compounds. In each case four manometrically measured
partial pressures of the pure compound were diluted in N, gas. In each case Beer's law was obeyed
over the concentration range examined - up to 2.4 x 10'^M (mol substance/ L gas). The limit of
.
,~.~ estimated as 3 times tne Peak to peak noise in the spectral region of interest was below
10' M. Compounds calibrated included 1,1,1 -trichloroethane, 1,1,2-trichloroethane, 1 1-
dichloroethane, 1 ,2-dichloroethane, trichloroethene, 1 ,2-dichloroethene, c/s-1,2-dich'loroethene
fra/7s-1,2-dichloroethene, methane and carbon dioxide. Values for vinyl chloride were obtained from
earlier studies with a 0.05 m cell.
RESULTS
The initial planting of alfalfa was done in June 1 992 and plants have been maintained since
that time with harvesting of most of the aboveground biomass at about monthly intervals. Plants
in one half of the chamber were exposed to c.. 500 ppm phenol from June 1 992 to May 1 993
Following a washout of phenol, TCE and TCA were introduced at the end of June 1 993 Most'
experiments described here were conducted in September and October of 1 993 by which time the
levels of TCE and TCA had reached a steady state of introduction and degradation. The other half
chamber was used for toluene feeding and then a conservative tracer, KBr, was introduced to
determine the dispersion characteristics of the soil system. Comparison of elution rates indicates
that there is significant adsorption to the soil, even though it is low in organic matter (c_ 1 %)
Significant degradation of phenol in the ground water was observed (2, unpubl obs) while toluene
concentration in ground water was unaltered but a mass balance required significant loss of toluene
from the whole system. Outflow of water was from 10-30% of input, while the remainder was
lost through evapotranspiration. Toluene was below the limit of detection in the gas phase above
the plants (2), indicating that 70-90% of the input must have been degraded.
Treatment of the soil with phenol may have enhanced the population of bacteria containing
specific hydroxylase enzyme able to metabolize TCE (1 ). Degradation of TCA is generally
considered to happen only anaerobically; it is a rather recalcitrant compound (1). Ground water
89
-------
head space gas analysis indicated that CH4 was present in the ground water at the 3rd and 4th
sampling wells. Chloride ion concentration was also increased in the water at the outflow. This
indicates that some TCE or possibly TCA, was being degraded anaerobically to some extent. A
further portion of the TCE may have been degraded in the vadose zone. To determine the extent
of degradation of TCE relative to TCA, the gas phase above the plants was monitored by FT-IR.
Concentrations of both were also measured in the ground water by head space gas analysis using
FT-IR. Some typical results for ground water are shown in Figure 1.
0)
o
c
o
.0
l_
o
w
-------
C
o
O
CO
.0
O)
>
_o
o:
Methane
1305.7 cm-1
(cm-1) 1300 . 1280
Figure 2
CONCLUSIONS
ethenes could not constitute more than a few percent of the
observed concentrations of TCE and TCA.
In the ground water both TCE and TCA
concentrations decreased by about 1 /3 from inlet to outlet.
Calculated concentrations based on input volumes should
have been 2.23 and 0.5 mM respectively. Measurable
levels of methane were found in the ground water only in the
latter half of the flow path of the chamber (Figure 2).
Methane levels in the gas phase from ground water samples
were calculated to be 2.75, 3.72 and 4.8 x 10'6 M for wells
3,4 and outlet respectively. This corresponds to 1.2 mM in
the ground water sample. Chloride ion analysis indicated only
about 0.7 mM chloride, suggesting that some methane may
have come from materials other than TCE and TCA.
Complete conversion of 1 /3 of both TCE and TCA to
methane should have yielded about 1.8 mM methane and
2.& mM chloride in the water, assuming that none of either
was transferred to the vadose zone prior to exiting the
chamber.
Studies of volatiles accumulating in the gas phase
above the well grown plants, or with the plants cut to 5 cm
showed significant accumulation of TCA but not of TCE
within less than one hour. The detection limits for both
compounds are quite similar so that failure to detect TCE
indicates that it must have been metabolized in the vadose
zone. In the ground water, concentrations of the two
compounds were proportional (c. 1:4) as expected from the
input concentration ratio.
A FT-IR system has been successfully applied to determination of contaminant levels in the
gas phase, ground water and plant tissues. It provides rapid analysis from small amounts of
material and is capable of detecting both the initial contaminant and its volatile products if they are
produced as a significant fraction of the input material. The system is mobile and may easily be
transported to locations where remediation efforts are underway.
REFERENCES
1. Thomas, J.M. and Ward, C.H. In situ biorestoration of organic contaminants in the subsurface
Environ. Sci. Technol. 23: 760-766, 1989
2. Davis, L.C., Chaffin, C., Muralidharan, N., Visser, V.P., Fateley, W.G., Erickson, L.E., and
Hammaker, R.M. Monitoring the beneficial effects of plants in bioremediation of volatile organic
compounds in: L.E. Erickson, D. Tillison, S.C. Grant and J.P. McDonald (eds.). Proceedings of the
8th Annual Conference on Hazardous Waste Research . Engineering Extension Service, Kansas
State University, Manhattan, KS. 1993. pp 236-249,
FOR MORE INFORMATION-:
Lawrence C. Davis, Dept of Biochemistry, KSU. Phone 913-532-6124 FAX 913-532-7278 Bitmail
LDAVIS@KSUVM
91
-------
FIELD SCALE EVALUATION OF GRASS-ENHANCED
CONTAMINATED SOILS
BIOREMEDIATION OF PAH
Darwin L. Sorensen, Ronald C. Sims
Department of Civil and Environmental Engineering
Utah State University
Logan, Utah 84322-4110
and
Xiujin Qiu
Central Research and Engineering Technology
Union Carbide Corporation
South Charleston, West Virginia 25303
INTRODUCTION
The toxicity and carcinogenicity of polynuclear aromatic hydrocarbons (PAH) gives rise to concern
over PAH contaminated soils. Two and three ring PAH compounds are biodegradable and can serve as
growth substrate for microorganisms. Higher molecular weight PAH compounds may be cometabolized
by soil microorganisms. The enhancement of biodegradation of pesticides, and other environmental
contaminants, in the rhizosphere suggests that similar acceleration of the biodegradation of PAHs may
occur in this environment.
In situ degradation processes in low permeability, clay soils are often slow. This may be due to the
slow flux of nutrients and electron acceptors through the soil resulting in limited microbial respiratory
activities. PAHs are generally resistant to biodegradation under methanogenic or sulfate reducing
conditions in contaminated subsurface environments. Maintaining a well aerated condition in soil with low
permeability is a challenge.
While PAH degradation and humification may take considerable time, turf grasses provide short term
benefits by reducing the risk posed to human health and the environment. Dense turf grasses serves as a
botanical cap to prevent exposure to soil born contaminants by direct ingestion, inhalation, and skin
contact of soil dust. Turf grasses also reduce potential contaminant migration through surface runoff and
infiltration to ground water resources.
Union Carbide and Utah State University jointly initiated research on grass enhanced bioremediation
of PAH contaminated soil in 1989. A preliminary laboratory study showed that high molecular weight
PAHs in soil can be degraded in the presence of prairie grasses and cow manure additions. The study
also indicated a significant reduction of water infiltration through the grass root zone. Subsequent
laboratory mass-balance experiments, using radio labeled compounds, indicated that mineralization rates
of phenanthrene and pentachlorophenol were significantly enhanced in the presence of prairie grasses.
Enhancement of the volatilization rate of naphthalene in the presence of prairie grasses suggested that
the roots improved soil aeration. A seed germination study indicated that prairie grasses were tolerant to
the PAH contamination and may survive in clay soil with significant levels of PAH, oil and grease. It was
hypothesized that a deep, fibrous root system of prairie grasses may improve aeration in soil and broaden
the spectrum of degradative capability in the rhizosphere and the organic exudates from roots may induce
microbial cometabolism of high molecular weight PAH. Based on these results, a field pilot scale study
was started in 1992 to test this hypothesis. The primary objective of the pilot study is to evaluate the
enhancement of the rate and extent of PAH biodegradation by the establishment of a grass root zone in
clay soil in the field environment. Treatment performance is being evaluated by assessing degradation
kinetics of PAH compounds.
METHODOLOGY
The field plots were constructed at Union Carbide's Seadrift Plant. Site selection and
characterization were based on available information regarding regional hydrogeology, contamination
history background soil sampling and analysis, as well as long term land use planning and safety
considerations. The Seadrift Plant is located in Calhoun County, Texas, near the Gulf of Mexico. The test
92
-------
site is a leveled surface at an elevation of approximately 28 feet (8.5 m) above MSL. The ground water
table is approximately 6 feet (1.8 m) below the surface. The top soil contains 50 to 60% clay. The top of
the uppermost sandy aquifer is approximately 13 feet (4 m) below the ground surface. The ground water
is high in salinity. Local weather records (1951-1980) showed that the mean annual precipitation was 39.5
inches (100.3 cm) and the mean annual lake evaporation was 59 inches (149.9 cm). The annual average
daily low and high temperatures were 63 - 78° F (17.2 - 25.6° C). The monthly extremes were 45° F (7.2° C)
in January and 91 °F (32.8° C) in July and August.
Exploratory soil samples were taken at 1 to 3 ft (30.5 - 91.4 cm) depths below the ground surface
from ten candidate locations at the Seadrift Plant. Composite soil samples for each of the locations were
analyzed for PAH constituents. The final site was selected due to its representative PAH constituents and
concentration levels.
Two plots, each 10 X 20 ft (3.05 X 6.1 m), were constructed for the study. One plot was sodded with
'Prairie' Buffalo grass and the other was kept free of vegetation. Table 1 shows the three-factor factorial
experimental design for the plots. They were constructed between the subgrade foundations of an old
olef ins unit building in Union Carbide's Seadrift Plant. The building had been demolished after the
operation was shut down. The soil underneath the building had been contaminated by the process
material; mainly pyrolysis gasoline. The contaminants present in the soil are benzene, toluene, and PAHs
including naphthalene, fluoranthene, etc. The plots were divided into 10 subplots, each 4X5 feet (1.22
X 1.52 m) and a sample core was taken from each of the subplots at each sampling event. The location of
the core within the subplot was determined using systematic random methods and a rectangular grid.
Table 1. Experimental design.
Factor
Replicate samples
Vegetation
Sample depth
Time
Total soil samples
Description
Number
Plot 1 - Unvegetated control
Plot 2 - Vegetated
Surficiai (0 -1 ft)
Subsurface (1 - 4 ft)
4 - 6 month sample intervals
7
< 280
Sampling was designed to assure that enough samples were taken to accurately estimate the
variation of the contaminants within each plot while minimizing alteration of the soil environment so that
plant grtiwth, soil moisture dynamics, gas exchange, etc., were not interfered with. A limit of less than 1%
of the plot surface disturbance by sampling was chosen to protect against these interference. The
number of sampling points on each plot was no more than ten in each of the seven sampling events.
Performance evaluation was done by measuring extractable PAH concentration reduction in soil over
time and with depth in the testing plots. Soxhlet extractions were done using 1:1 (v:v) methylene
chloride:acteone followed by concentration using Kaderna-Danish apparatus. Concentrations of PAH
compounds in the extracts were determined using gas chromatography/ion trap (mass spectrometry)
detection.
Data analysis indicated that the distribution of PAH concentrations within each of the plots was log-
normally distributed. Analysis of variance procedures were, therefore, performed on the log + 1
transformed data. All values reported below the limit of quantitation were assigned the value zero for data
analysis purposes.
93
-------
RESULTS
Concentrations of naphthalene, acenaphthylene, acenaphthene, phenanthrene, and anthracene
were routinely monitored in soil samples from the plots. All of the other 16 priority pollutant PAH
compounds were not found in environmentally significant concentrations in the initial samples. In general,
only naphthalene, acenaphthylene, and phenanthrene were found in quantifiable concentrations
frequently enough to allow reliable use of normal statistics in the data analysis. For this reason, only data
for these three compounds will be presented. Average phenanthrene concentration decreased
significantly in both the vegetated and unvegetated plots. Shallow sample concentrations in the
vegetated plot tended to be lower than in deep samples. Acenaphthylene also decreased in both plots
over time and was present in higher concentrations in unvegetated shallow soil than in vegetated shallow
soil. Naphthalene concentrations have changed little over the experiment to date. As with phenanthrene
and acenaphthylene, naphthalene concentrations tended to be higher in the unvegetated shallow soil
than in the vegetated shallow soil.
CONCLUSIONS
The experiment described is still in progress and the data presented here must be considered to be
preliminary and the interpretation of the data is subject to revision. To date, it appears that PAH
degradation in the Seadrift plots is slow. Surprisingly, naphthalene degradation kinetics appear to be
particularly slow. There is some evidence that Buffalo grass sod planting and establishment has
enhanced degradation in the near surface soil.
FOR MORE INFORMATION
Darwin L. Sorensen
Utah Water Research Laboratory
Utah State University
Logan, Utah 84322-8200
Telephone: (801) 750-3207
94
-------
ISOLATION OF POLLUTANTS USING A BIQBARRIER TECHNOLOGY
: J. William Costerton
Center for Biofilm Engineering
Montana State University
Bozeman, MT 59717
Telephone (406) 994-1960
FAX (406) 994-6098
INTRODUCTION
For several decades Microbiologists have known that bacteria can produce
several times their own weight in slimey exopolysaccharides that can readily plug
porous media such as sand filters. This characteristic of bacteria has militated
against their use as agents of bioremediation, because it limits their mobility through
porous media, but it has led to numerous suggestions that they could be used
selectively to plug the high permeability zones that mediate water breakthrough in
secondary oil recovery by water flooding. We have examined the passage of bacteria
through porous media (1.6 darcy: 27 p,m pore throats) and we have found very
shallow penetration and almost complete plugging by these organisms (1). Because
these bacterial plugs were too shallow (< 1 cm) to be useful, and because they
could not be positioned deep within a porous medium, they have not been used in
selective plugging.
A decade previously Merita's group had discovered (2) that marine bacteria
reacted to starvation by producing ultramicrobacteria (UMB). These UMB are very
small (±0.3 urn in diameter) when compared to the vegetative cells from which they
were derived (± 1.0 |j.m) and they are metabolically dormant and produce only very
small amounts of exopolysaccharide (EPS). We reasoned that UMB would probably
penetrate porous media more readily than vegetative cells and we know from the
literature (2) that they would return to their full vegetative size and produce large
amounts of EPS if they were provided with nutrients at some location within these
porous media. It is now abundantly clear that the physiological state of bacteria has
a profound effect on their passage through porous media and that the successful
exploitation of this scientific fact will lie at the heart of successful technologies in oil
recovery and in bioremediation.
METHODOLOGY
UMB are produced from natural strains of bacteria recovered from produced
groundwater by starvation for 60 days in a high salts medium deficient in organic
carbon nutrients (3). The starvation process is monitored by light microscopy and
the final suspension of UMB, which can readily be produced at concentrations of
5 x 10s cells/ml, is centrifuged at 100,000 xg for 15 minutes and washed several times
in PBS to produce a uniform suspension of small dormant cells. The nutrient solution
95
-------
used to resuscitate these UMB in situ in the porous media is a solution of sodium
citrate containing several salts.
The physical simulator used in these studies is described in Figure 1 and it
consists of a sand-filled drum, maintained under simulated overburden pressure, in
which the UMB and the nutrient solution can be injected sequentially at a single point
and detected at 8 different probe locations to detect their radial distribution. In this
experimental design the UMB were injected until they could be detected at all 8
sampling probes and then the nutrient solution was injected at intervals while we
monitored the injection pressure to detect plugging of the porous medium. At the
end of each experiment, typically 21-40 days, the physical simulator was dismantled
and samples of the sand were taken at carefully selected locations to be analyzed for
their content of bacteria and EPS by chemical and morphological methods (3).
RESULTS
When one pore volume (PV) of a 1 x 106 cells/ml suspension of UMB was
injected into the physical simulator sandpack, which had an effective permeability of
1.3 darcys, these small dormant cells had reached even the most distant sampling
probe within 105 minutes (Table 1) after injection. The number of these UMB at each
sampling location varied from 1.2 to 3.6 x 105 at these locations. These data attest to
the ability of UMB to penetrate this porous medium.
When the nutrient solution was injected to initiate the resuscitation of these
UMB in. situ in the physical simulator the number of resuscitated vegetative cells/ml of
pore fluid reached at least 1.2 x 107 cells/ml by day 5 and at least 2.3 x 108 cells/ml
by day 18. The size of these vegetative cells (> 1.0 p,m) was confirmed directly by
light microscopy. The injection pressure of the physical simulator gradually increased
from an initial level of 8 kPa to > 100 kPa as the UMB were resuscitated by nutrients
In situ. When the experiment was terminated sand samples were taken from various
locations, as the physical simulator was dismantled, and direct microbiological and
chemical measurement of the numbers of bacteria and the p.g of EPS per gram of
sand clearly indicated that the porous sand matrix of the physical simulator had been
uniformly and completely filled by bacterial cells and their EPS. Scanning Electron
Microscopy (SEM) of the clean sandpack and the sandpack at the end of the
nutrient-stimulated UMB plugging experiment (Figure 2) clearly shows the complete
occlusion of the pore spaces by vegetative bacterial cells and their slimey EPS.
These data attest to the ability of UMB to respond to nutrient stimulation by
replicating and by producing EPS to effect complete plugging of this porous medium.
CONCLUSIONS
We have previously presented data to establish that vegetative bacteria are
capable of plugging porous media but that the resultant plug is very shallow (1)
because of biofilm formation on the closest available surfaces (4). In this paper, and
96
-------
(b)
15-2 7-6 7-6 15-2
Figure 1. The design of the three-dimensional sandpack 45 cm in
diameter by 38 cm in length, showing the location of the eight
probes (1-8) within the sandpack viewed (a) from above and (b)
as a cross-section. All dimensions are given in cm.
97
-------
Table 1.
Penetration of umb through the sandpack following their
injection in one pore volume.
Viable cell numbers (c.f.u. ml-') after (min):
Probe
no.
1
2
3
4
5
6
7
8
5
NS
NS
NS
, NS
NS
2-8 x 10*
3-1 x 104
2-0 x 10*
15
NS
NS
NS
1-3 x 10*
NS
1-3 x 10s
•9-1 x 10*
8-1 x 10*
45
NS
2-5 x 10*
NS
1-5 x 10*
NS
3-2 x 10s
1-9 x 10s
1-9 x 10s
75
7-3 x 10*
6-3 x 10*
9-1 x 10*
5-6 x 10*
1-2 x 10*
2-2 x 10s
2-0 x 10s
2-5 x 10s
105*
1-5 x 10s
1-3 x 10s
1-2 x 10s
1-5 x 10s
1-5 x 10*
1-5 x 10s
2-1 x 10s
3-6 x 10s
165f
(1-5 PV)
4-5 x 105
5-0 x 10s
2-7 x 105
1-8 x 10s
6-5 x 10s
6-5 x 10s
7-0 x 10s
4-0 x 10s
NS, Not sampled.
•After injection of 1 PV, UMB were
Mean = 1-78 x 10s c.f.u. ml"1, SD = 0-78.
t After injection of 1-5 PV, UMB were evenly distributed
Mean = 4-75 x 10s c.f.u. ml-', SD= 1-88.
evenly distributed throughout the sandpack.
throughout the sandpack.
98
-------
Figure 2. Scanning electron microscopy showing empty pore spaces in a
clean sandpack (A) and pore spaces completely occluded by
vegetative bacteria and EPS in a sandpack plugged by injection
of umb followed by nutrient stimulation (B). The bar indicates
99
-------
in previous publications (5, 6 and 7), we have established that most bacteria can be
starved to produce UMB which penetrate readily through porous media > 110
millidarcies in permeability and can then be resuscitated to plug these porous media
by cell proliferation and by EPS production.
The UMB deep plugging technology, which was developed for the Alberta
Oilsands Technology and Research Authority and is protected by US patent
#4,800,959, was initially developed for the selective plugging of permeable
breakthrough zones in secondary oil recovery by waterflooding. In this area of
application this technology has been tested extensively in the laboratory and is
currently in the early stages of application in the fields in Alberta and Saskatchewan.
The UMB plugging technology has also been used, in the environmental area, to
produce a 4-6 inch deep zone of bacterial growth and EPS production to seal mine
tailings so that oxygen cannot penetrate and react with pyrite to initiate acid mine
drainage (8). However, the area of application of this technology that is of paramount
interest in environmental research is in the isolation of organic and metallic pollutants
in the subsurface environment.
It is clear, from these data, that the UMB technology can be used to position
deep and effective biobarriers for the containment of pollutants in the subsurface
environment. Specifically, bacterial biobarriers can be placed to prevent the
movement of pollutants from underground plumes to sensitive ecosystems such as
streams and lakes. These bacterial biobarriers are especially useful because they
block groundwater flow and contain elements (bacterial cells and EPS) that can
potentially degrade organic pollutants and bind metallic ions. Bacterial biobarriers
are superior to alternate technologies, such as grout curtains, in that they are very
economical and in that they bind intimately to underlying bedrock and do not allow
groundwater seepage at the sand/rock interface. The UMB biobarrier technology has
already been pilot tested at the mesoscale level at the National Hydrology Institute in
Canada and it has been shown to produce a barrier that is elastic, long lived, and
99% efficient in reducing permeability and controlling groundwater flow. We suggest
that this technology, which involves the drilling of a series of shallow wells from which
UMB and nutrients are sequentially injected, can be used to prevent the passage of
underground pollutants into sensitive ecosystems. We plant to inject UMB into these
shallow wells and to then inject sufficient amounts of nutrient to produce overlapping
columns of soil in which the pore spaces are virtually sealed by bacterial proliferation
and EPS production. We are confident that the resultant bacterial biobarriers will be
very effective in isolating a wide variety of pollutants in the underground environment.
Following their effective isolation by this UMB technology it is clear that organic
pollutants will be biodegraded by the injection of degradative bacteria that will be
manipulated in the opposite way in order to maximize penetration and minimize EPS
production.
The UMB biobarrier technology is currently being pilot tested in a macroscale
facility and we expect that it will be ready for field trials within the next 18 months.
100
-------
REFERENCES
1. iShaw, J.A., Bramhill, B., Wardlaw, N.C., and Costerton, J.W. Bacterial fouling in
a model rock core system. Applied and Environmental Microbiology 50: 693-
;701, 1985.
2. Novitsky, J.A. and Morita, R.Y. Morphological characterization of small cells
resulting from nutrient starvation of a psychrophilic marine vibrio. Applied and
Environmental Microbiology 32: 617-622, 1976.
3. Cusack, F., Singh, S., McCarthy, C., Grieco, J., de Rocco, M., Nguyen, D.,
Lappin-Scott, H., and Costerton, J.W. Enhanced oil recovery - three-
dimensional sandpack simulation of ultramicrobacteria resuscitation in reservoir
formation. Journal of General Microbiology 138: 647-644, 1992.
4. Costerton, J.W., Cheng, K.J., Geesey, G.G., Ladd, T.I., Nickel, J.C., Dasgupta,
M., and Marrie, T.J. Bacterial biofilms in nature and disease. Annual Review of
Microbiology 42: 435-464, 1987.
5. Lappin-Scott, H.M., Cusack, F.M., MacLeod, F.A., and Costerton, J.W. Nutrient
resuscitation and growth of starved cells in sandstone cores - a novel
approach to enhanced oil recovery. Applied and Environmental Microbiology
54: 1373-1382, 1988a.
c
6. Lappin-Scott, H.M., Cusack, F.M., MacLeod, F.A., and Costerton, J.W.
Starvation and resuscitation of Klebsiella pneumoniae isolated from oil well
waters. Journal of Applied Bacteriology 64: 541-549, 1988b.
7. MacLeod, F.A., Lappin-Scott, H.M., and Costerton, J.W. Plugging of a model
rock system by using starved bacteria. Applied and Environmental
Microbiology 54: 1364-1372. 1988.
8. Blenkinsopp, S.A., Herman, D.C., and Costerton, J.W. The Use of Biofilm
Bacteria to Exclude Oxygen from Acidogenic Mine Tailings, in: Proceedings of
the Second International Conference on the Abatement of Acidic Drainage,
Montreal, 1991. pp. 369-377.
101
-------
KINETICS OF BIODEGRADATION. SORPTION AND DESORPTION OF PHENOL. SUBSTITUTED
PHENOLS AND POLYCYCLIC AROMATIC HYDROCARBONS IN SOIL SLURRY SYSTEMS
Henry H. Tabak
U.S. Environmental Protection Agency
ORD, Risk Reduction Engineering Laboratory
Cincinnati, OH 45268, U.S.A.
(513) 569-7681
and
Chao Gao, Xuesheng Yan, Lei Lai, Steven Pfanstiel, Chunsheng Fu, Rakesh Govind
Department of Chemical Engineering
University of Cincinnati
Cincinnati, OH 45221, U.S.A.
(513) 556-2666
INTRODUCTION
Bioremediation of Superfund soil and sediment sites requires a fundamental understanding of biodegradation
kinetics and the physicochemical factors that control the rate of biodegradation. The quantification of biodegradation
kinetics provides useful insight into the favorable range of the environmental parameters for the improvement of the
microbiological activity, and enhancement of the biodegradation rates of the contaminants in soil, sediments and
aquifers and consequently for enhancing the bioremediation of these environments (Tabak et al. 1992; Govind et al.
1993)1»5. Biological treatment systems, especially when used in conjunction with other physical/chemical treatment
methods, hold considerable promise for efficient, safe, economical on-site and in-situ treatment of toxic wastes.
The main objectives of this research are to quantitate the bioavailability and biodegradation kinetics of organic
chemicals in surface and subsurface soil environments, examine the effects of soil matrices and soil conditioning
(drying, aging, compacting), and develop a predictive model for biodegradation kinetics applicable to soil systems.
The specific objectives of this study are as follows: (1) Evaluate the availability of representative organics in selected
soils for biodegradation using biokinetic data; (2) Examine the effects of soil matrices and soil conditioning on the
biokinetics; (3) Examine the threshold inhibition of contaminated soils on the microbiota in contaminated soils; and
(4) Develop a group contribution predictive biodegradation model for biokinetics in soil systems. This paper
highlights biodegradation studies on phenol, several alkyl phenols (p-cresol, 2,4-dimethyl phenol, catechol,
hydroquinone, and resorcinol) and selected polycyclic aromatic hydrocarbons (PAHs) (napthalene, phenanthrene,
accnapthene, and acenapthylene. The studies incorporate the use of soil microcosms for acclimation of soil
microbiota, measurement of respirometric oxygen uptake and carbon dioxide evolution in soil slurry reactors,
measurement of carbon dioxide generation rates in shaker-flask systems, and determination of adsorption/desorption
equilibria and kinetics. A mathematical model for soil slurry reactors is used to determine the biodegradation kinetics
for the selected compounds from the oxygen uptake data. A comprehensive description of experimental data and
methodology for the kinetics of biodegradation, adsorption and desorption of the phenolic and PAH compounds was
presented and published elsewhere^"".
METHODOLOGY
Studies on Soil Microcosm Reactors: The methodology for the establishment of the bench-scale microcosm
reactors, including the reactor configuration and supportive equipment description, undisturbed soil cube sampling for
102
-------
inclusion in the microcosms, procedure for contamination of soil bed with homologous series of organic
compounds, description of the application of nutrients and the operation of the microcosm reactors (CO2 generation
analysis and chemical analysis of soil samples and reactor leachates for the parent compound and metabolites) has
been fully described elsewhere (Tabak et al. 1992; Govind et al., 1993) ]>6. Each microcosm reactor represents a
controlled site, which eventually selects out the acclimated indigenous microbial population in the soil for the
contaminating organics. Samples of soil are then taken from the microcosm reactors and used as source of acclimated
microbial inoculum for measuring: (1) oxygen uptake respirometricaily; (2) carbon dioxide generation kinetics in
shaker flask reactors; and (3) for studies with other soil reactor systems. The microcosm reactor units are also being
used directly to evaluate the biodegradability of the pollutant organics and to measure their average biodegradation
rate in this intact, undisturbed soil bed. One microcosm reactor was spiked with a mixture of phenolic compounds
dissolved in deionized distilled water so that the total chemical oxygen demand per kg of soil in the microcosm
reactor was 300 mg. Equal concentrations of phenol, resorcinol, catechol, 2,4-dimethyl phenol, cresol and
hydroquinone were used in the mixture. The second microcosm reactor was contaminated with 25 ppm each of
several polycyclic aromatic hydrocarbons (PAHs) dissolved in a mixture of deionized-distilled water and 0.5%
solution of a surfactant, Triton X-100. Control microcosm reactors were uncontaminated soil which was sprayed
with an equal volume of deionized-distilled water and soil contaminated with 0.5% solution of surfactant, Triton X-
100. Uncontaminated forest soil was selected for all studies presented in this paper.
Respirometric Studies with Soil Slurries: Studies were conducted with soil slurry reactors, wherein the oxygen
uptake was monitored respirometricaily. The extent of biodegradation and the Monod kinetic parameters for the
organic pollutant compounds by soil microbiota were determined from oxygen uptake data. Various concentrations
of soil (2, 5, 10%) and compound (50, 100, and 150 mg/L) were mixed with a synthetic medium consisting of
inorganic salts, trace elements and either a vitamin solution or solution of yeast extract and stirred in the
respirometric reactor flasks. The flasks were connected to the oxygen generation flask and pressure indicator cells of
a 12 unit Voith Sapromat B-12, electrolytic respirometer, and the oxygen uptake (consumption) data were generated
as oxygen uptake velocity curves. In the case of phenol and alkyl phenols, the compounds were soluble in water.
However, for polycyclic aromatic hydrocarbons (PAHs), a large fraction of the compound was insoluble in water.
The appropriate PAH chemical was weighed and added as a solid powder to the reactor flask. A more comprehensive
description of the procedural steps for the respirometric tests is presented elsewhere (Tabak et al. 1992; Govind et al.
1993)1-6.
Kinetic Analysis of Soil Slurry Oxygen Uptake Data: The oxygen uptake data were analyzed by computer
simulation techniques and curve fitting methods, using the Monod type equation combined with a nonlinear
adsorption isotherm, to determine the soil adsorption parameter, and the biokinetic parameters. The experimental data
were analyzed using a mathematical model (equations given below) which include the effect of chemical adsorption /
desorption on the soil particles and the adsorption / desorption of the microorganisms from the soil.
xs = Kbxw
1/n
(1)
(2)
02 = (Sto - St) - (Xt - Xto) - (Sp - Spo) (3)
dXt/dt = (WXWSW)/(KW + Sw) + SWXS SS/(KSW + Ss) - (bKwXt)/(Kw + St) (4)
dSt/dt = -(l/y)(wXwSw)/(Kw + Sw)-(l/y)(swXsSs/(Ksw + Ss)(w/v) (5)
dSp/dt = -yp(dSt/dt) (6)
where subscripts t, s, w and p represent the total, soil, water and degradation products respectively. S is the
concentration of compound, X is the concentration of biomass, and Sp is the concentration of the degradation
products. w, Kw, y and yp are the Monod equation maximum specific growth rate parameter, Michaelis constant,
103
-------
biomass yield and product yield coefficient, respectively for the water phase. sw, Ksw are the Monod equation
maximum specific growth rate parameter and Michaelis constant, respectively for the soil phase, b is the biomass
decay coefficient, Kj is the soil adsorption isotherm parameter, (1/n) is the soil adsorption intensity coefficient, and
Kb is the biomass adsorption parameter. C>2 is the cumulative oxygen uptake, w is the weight of soil in the slurry
reactor and v is the total volume of water in the reactor.
Measurement of Carbon Dioxide Evolution Rates: Carbon dioxide generation rates were measured in shaker-
flask soil slurry systems and in the electrolytic respirometry soil-slurry reactors in order to assess the rate and extent
of biodegradation/mineralization. The CC«2 generation rate measurement serves as an additional tool for quantitating
the biofate of these organics in addition to the cumulative respirometric oxygen uptake data from which the
biokinetics were derived. Soda lime was used initially as the absorbent for CC>2 evolution studies in the shaker flask
and in respirometric reactors. Subsequently soda-lime was replaced by KOH as the absorbent in the CO2 evolution
measurement studies, because of the observed slower rate of release of CC>2 from soda-lime during the analysis of
CO2- Four methods were used for CO2 measurement in the shaker flask soil slurry systems and in respirometric
reactors: (1) use KOH to trap CC-2 released from soda lime, and then titrate KOH using standard HC1 before and after
CO2 absorption; (2) use Ba(OH)2 to trap CO2 released from soda lime, and then titrate Ba(OH>2 using standard HC1
before and after CO2 absorption; (3) use KOH to trap CO2 released from soda lime, precipitate the KHCO3 and
K2CO3 formed, using BaCl2 to get BaCO3 which is filtered, dried and weighed; (4) use KOH to trap CO2, measure
pH of KOH solution before and after CO2 absorption, calculate CO2 generated according to the changes of pH values
using the developed computer program. The pH method was shown to be most accurate and provided the most
reproducible data. Shaker flask slurry systems data on CO2 production were generated for the same alkyl phenols
used in the respirometric soil slurry reactor studies.
Studies on Adsorption/Desorption Equilibria and Kinetics: Soil adsorption kinetics and equilibria are
measured using batch well-stirred bottles. The soil was initially air dried and then sieved to pass a 2.00 mm sieve. A
10 g of soil sample was placed in each bottle and mixed with 100 mL of distilled deionized water containing various
concentrations of the compound and mercuric chloride to minimize biodegradation. The soil solution ratio was
expressed as the oven-dry equivalent mass of adsorbent in grams per volume of solution. The liquid was sampled
after 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, and 36 hours. Before the liquid sample was taken, after the predefined time
has elapsed, the bottle contents were centrifuged and the liquid sample was withdrawn using a syringe connected to a
0.45 m porous silver membrane filter. The concentration of the chemical compound in the liquid sample was
analyzed using three methods: (1) standard extraction (EPA method 604 and 610) with methylene chloride followed
by GC/MS analysis; (2) HPLC analysis; and (3) scintillation counting of the 14C using radiolabelled compound.
All three analytical methods were calibrated using standard solutions of each compound with concentrations levels of
1,5, 10, 50,100 and 150 mg/L. The calibration data is used to convert the peak areas (GC/MS or HPLC) or counts
of disintegrations per minute (DPM) for radiolabelled compounds to the actual liquid concentration. In the case of
polycyclic aromatic hydrocarbons (PAHs), each chemical was weighed equal to 80% of standard solubility in hexane.
Standard solutions containing 60%, 40%, 20% and 10% standard solubility levels were obtained by diluting with
hexane. A HPLC column (Supelco LC-PAH 15 cm x 4.6 cm) was used. A plot of peak areas versus the
concentration was made to obtain the standard calibration curve for each chemical. For chemicals with extremely
low solubility (< 1 ppm), radiolabelled compounds were used with scintillation counting. Standard solutions were
made in the same manner and calibrated with counts of disintegrations per minute (DPM). Stock solutions of each
PAH chemical were made by first weighing chemical equal to at least twice standard solubility level in water. The
chemical was thoroughly mixed with ultrapure water using teflon coated magnetic stir bar for 48 hours. The
solution was allowed to settle for 48 hours. The saturated solution was decanted and filtered using a 5 m millipore
filter to remove all suspended particles. The liquid concentration was analyzed using either HPLC or scintillation
counting. Other stock solutions were obtained by diluting with ultrapure water to obtain solutions with 80%, 60%,
40%, 20% and 10% standard selubility levels.
Desorption studies were conducted by first adsorbing the chemical on to the soil until equilibrium was achieved. Two
methods were used for desorption studies. In the first method, 100 ml of deionized distilled water is mixed with 20
104
-------
grams of soil and specified concentration of chemical for adsorption. After adsorption equilibrium is attained, the
whole solution is centrifuged and 90 ml of supernatant is taken out and replaced with an equal volume of deionized
distilled water and with 20 gm/1 of mercuric chloride to inhibit biodegradation. Desorption begins from this time. A
20 ml sample was withdrawn at 4, 8, 16, 24, 48, 72, 96 and 120 hours from separate bottles. Each sample was
filtered with 0.45 m silver membrane filter and extracted with methylene chloride and analyzed using GC/MS
technique to determine the concentration of solute. HPLC analysis and 14C scintillation counting analysis for
radiolabelled compounds were also used as additional methods for comparative purposes. In the second method, 100
ml of deionized distilled water was mixed with 20 gms of soil and specified concentration of chemical for adsorption.
After adsorption equilibrium was attained, the sample was diluted with an equal volume of deionized distilled water,
and 20 gm/1 of mercuric chloride was added to inhibit biodegradation. A 20 ml sample was withdrawn at 4, 8, 16,
24, 48, 72, 96 and 120 hours from separate adsorption bottles. The sample was analyzed by extraction with
methylene chloride followed by GC/MS analysis. HPLC analysis and 14C scintillation counting analysis for
radiolabelled compounds were also used as additional methods to compare the analysis results in desorption studies.
Results from all three analytical approaches agreed closely, and subsequently HPLC was used exclusively.
Radiochemicai Techniques for Determining Biodegradatioii of Organics in Soil: A radiorespirometric
biodegradation protocol was developed for quantitating the biodegradation/mineralization of the organic pollutant
compounds in soil slurry systems. The protocol was shown to be particularly applicable to the study of
biodegradation of alkyl phenol compounds in the respirometric soil slurry reactors in that it was possible to
corroborate the extent of biodegradation data on these compounds as measured by the respirometric oxygen uptake
and carbon dioxide evolution in shaker flask systems. Studies were conducted with phenol and each of the alkyl
phenols in concentrations of 50, 100 and 150 mg/L in the respirometric reactors. The phenols were added to the
OECD nutrient solution made in deionized distilled water and mixed with either 5, 10 or 15% soil. One ml of 14C
fully tagged radiolabelled phenol (concentration equals to 1 ci/ml) was added to each of the experimental and control
reactors. Five ml of 2N KOH solution is used to absorb the 14CC<2 released from the solution. After appropriate
acclimation time, sampling of KOH in the special holder in the respirometric reactors was initiated by taking out 5
ml KOH volumes with absorbed 14CO2 every two hours from each of the reactors. The sampled KOH solution was
mixed with a cocktail solution (Ultima Gold) in a 1:10 ratio and analyzed via liquid scintillation counting
technology (Packard TRI-CARB 2500 TR Liquid Scintillation Analyzer). Liquid scintillation efficiency tracing
technique was used to quantitate the 14CO2 concentration released from the solution in the reactor systems.
Determination of Bacterial / Soil Sorption Isotherm: The adsorption isotherm for the bacterial cells was
determined by incubating soil microbiota with radiolabelled phenol in a respirometric reactor until an oxygen uptake
plateau was obtained, indicating that all phenol has biodegraded either into 14CO2, which is adsorbed in the KOH
solution and into *4C biomass. The soil suspension was allowed to settle for about 30 minutes. One ml of
supernatant is sampled and the ^4C activity was measured by liquid scintillation counting. Equilibrium amounts of
the '4C biomass adsorbed to the soil was determined by subtracting the '4C present in the biomass in suspension
and the 14C present as carbon dioxide absorbed in the KOH solution from the total '4C added initially. The ratio
of the biomass adsorbed to the soil and the biomass present in the suspension gives the biomass / soil adsorption
isotherm parameter, Kb. -:,.-.-.
RESULTS AND DISCUSSION
Studies on Soil Microcosm Reactors: With the use of these specially designed microcosm reactors, it was
possible to acclimate the indigenous microbiota to mixture of phenol and substituted phenols in one set of reactors
and similar performance was obtained for other set of microcosm reactors spiked with a mixture of PAHs. Soil
samples from the microcosm reactors are used as a source of acclimated microbiota; for measuring oxygen uptake
respirometrically to determine biodegradation kinetics; to determine carbon dioxide generation kinetics; and for
radiochemical studies performed respirometrically to quantitate biodegradation data. Cumulative carbon dioxide
generation as a function of time for the microcosm reactors, before and after spiking the microcosms (experimental)
with solutions comprised of phenolic compounds and PAHs and the control microcosm with OECD nutrients only,
105
-------
provided evidence of mineralization of respective constituents of the mixtures. The cumulative carbon dioxide
production increased after spiking with nutrients and the phenolic and PAH solutions. The GC/MS and HPLC soil
data on the residual parent compounds and intermediate metabolites and end products corroborated the CC>2 evolution
data and demonstrated conclusive evidence of mineralization of the respective phenolic compounds and the PAH
compounds.
Analysis of Oxygen Uptake Data: Respirometric oxygen uptake data were generated for several alkyl phenols
using soil slurry systems. The oxygen uptake data for 5% soil concentration with 0, 50, 100, and ISO mg/L of
phenol concentrations, indicate that at 0% phenol concentration, the oxygen uptake is due to degradation of soil
organic matter, and that the oxygen uptake increases with phenol concentration. The oxygen uptake curves did not
change appreciably when the soil concentration was increased, except that the lag time decreased for all phenol and
alkyl phenol concentrations. The same trends were demonstrated for substituted phenolic compounds. The oxygen
uptake data were analyzed using the mathematical model and the random search technique and the best-fit Monod
parameters were generated for the phenolic compounds studied. Except for Ks, the other biokinetic parameters do not
vary significantly with changes in phenol or soil concentration. The experimental value of the biomass adsorption
parameter, Kfc, was 167.0 (mg/Kg)(L/mg). A high value of K^ indicates that a significant amount of active biomass
remains attached to the soil particles as compared to the biomass existing as suspended culture.
Oxygen uptake data was generated with 5% soil slurry and the polycyclic aromatic hydrocarbons. 100 mg/L
concentration of each PAH was used in the experiments. Initially the PAH compound was insoluble in the nutrient
mixture and remained suspended as the mixture was mixed with the soil. However, after some acclimation time, the
PAH compound was emulsified in the aqueous mixture. The oxygen uptake data were analyzed using the
mathematical model and the best-fit Monod parameters for the PAHs were generated. The Monod kinetic parameter,
w, represents aqueous degradation of the dissolved or emulsified compound by the microorganisms present in the
aqueous phase. The maximum specific growth rate parameter, w, ranges from 0.127 to 0.482 for the ibur PAH
compounds, which is typical for aerobic aqueous biodegradation of most organic compounds. The yield coefficient,
y, is approximately 0.5 which is also typical for aerobic biodegradation. The Monod kinetic parameter, sw, for the
microorganisms immobilized in the soil matrix, was significantly higher than that for the aqueous phase. The
approximate first-order kinetic parameter for the immobilized biomass (SW/KSW) is lowest for phenanthrene and
highest for naphthalene. This agrees with literature values and with the fact that phenanthrene with three fused
aromatic rings is more difficult to biodegrade than naphthalene with two fused aromatic rings.
Measurement of CC>2 Generation Rate: Comparisons were made between the cumulative CC>2 concentrations
experimentally measured for the different soil and compound concentrations in the shaker flask soil slurry systems
and respirometric reactor soil slurry systems measuring oxygen uptake data. Ratio values were developed for the
CO2 experimentally measured in respirometer and shaker flask to the CC>2 calculated from oxygen uptake data. The
compatibility between the data on the CC>2 production in the shaker flask and respirometric vessel tests has been
established and verified and the measurement of CC>2 evolution in shaker flask soil slurry systems was shown to be
dependable for quantitative CC>2 analysis. The results on the relationship of the two reactor systems provided data on
the correlation between the oxygen requirements and CC>2 generation for metabolic activity of microbiota on toxic
organics in the soil slurry systems. The total amount of carbon dioxide generated agrees closely with the appropriate
amount of oxygen uptake required to mineralize the compound. This demonstrates that complete mineralization of
the compound was achieved in the respirometric experiments.
Measurement of 14CC>2 Generation Rate by Radiorespirometry: Data have been generated on radiolabelled
degradation of phenolic and PAH compounds in soil slurry systems spiked with unlabelled compounds for 5 and 10%
soil concentration levels. The DPM data indicate that phenols and PAHs are preferred over the organic matter
normally present in soil, once the indigenous microbiota are acclimated to these compounds. In the case of phenolic
compounds, the data demonstrate that measurement of a carbon dioxide evolution can be used for determining the
kinetics of biodegradation for phenol and alkyl phenols in soil slurry systems. Furthermore, analysis of the liquid.
106
-------
phase showed that little carbon residual remained in the liquid at the end of the experiment suggesting complete
mineralization of phenol.
Analysis of Adsorption/Desorption Data: The liquid concentrations for all phenojic compounds and PAHs did
not vary more than 5% after 48 hrs, which indicated that equilibrium was achieved. In addition, it can be concluded
that since the concentration of the compound did not change after 48 hours even though the experiments were
conducted for 96 hours, there was no biological degradation of the compounds under the experimental conditions.
The adsorption isotherm for each compound was analyzed using the Freundlich isotherm equation:
X/M = kaC1/na
(7)
The desorption isotherm for each compound was also analyzed using the Freundlich equation:
X/M = kdC1/nd (g)
where (X/M) is the equilibrium concentration in the soil phase, C is the equilibrium concentration in the liquid
phase, ka, k
-------
CONCLUSIONS
Respirometric studies with soil slurry reactors provides valuable insight into the biodegradation kinetics of
compounds in the presence of soil. It has been shown that a Monod kinetic equation in conjunction with a linear
adsorption isotherm can provide reliable estimates of the Monod kinetic parameters and the adsorption coefficient.
Carbon dioxide generation in soil slurry systems provides unambiguous measurements of the rate of mineralization
of the compound in the presence of soil. A protocol developed for quantitative measurement of ^CO2 evolution rate
by radiorespirometry provides confirmation of mineralization kinetics from carbon dioxide evolution studies and
ensures that the net CO2 is generated from mineralization of the compound and not due to natural soil respiration.
Further studies are planned for measuring the biokinetics of compounds in compacted soil systems, as opposed
to our current measurements in soil slurry systems. This will allow one to determine the impact of soil compaction,
oxygen transfer rates, and moisture content on biodegradation. A attempt will also be made to develop a detailed
mathematical model implemented as a computer program, that will use our experimental data and model parameters
to actually quantify rates of bioremediation at contaminated sites.
REFERENCES
1. Tabak, H.H. and R. Govind. "Determining Biodegradation Kinetics with use of Respirometry for Development of
Predictive Structure-Biodegradation Relationship Models." Presented at the 4th International IGT Symposium, 1991
Colarado Springs, CO. Published in Gas, Oil and Environmental Biotechnology. Volume 4. C. Akin, R.
Markuszeski, J. Smith (eds.) 1992. p.129
2. Tabak, H.H., L. Lai, X. Van, C. Gao, S. Kim, S. Pfanstiel and R. Govind. "Methodology for Testing
Biodegradability and Determining Bioavailability and Biodegradation Kinetics of Toxic Organics in Soil."
Presented at the 5th International IGT Symposium on Gas, Oil and Environmental Biotechnology, Chicago, IL,
Sept. 21-23, 1992. To be published in Gas , Oil, Coal and Environmental Biotechnology V, Institute of Gas
Technology, Chicago, IL., 1993.
3. Tabak, H.H., C. Gao, L. Lai, X. Van, S. Pfanstiel, S. Kim, and R. Govind. "Determination of Biodegradability and
Biodegradation Kinetics of Phenol and Alkyl Phenols". Presented at the 1992 ACS Industrial and Engineering
Chemistry Division Symposium on "Emerging Technologies for Hazardous Waste Management". Special
Symposium on Bioremediation of Soils and Sediments. Atlanta, Georgia, September 21-23, 1992. Published in
1993 Emerging Technologies in Hazardous Waste Management, ACS Symposium Series, Volume IV, 1993.
4. Tabak, H.H., R. Govind, C. Gao, L. Lai, X. Van, S. Kim, 1992. "Determination of Biodegradability and
Biodegradation Kinetics of Organic Compounds in Soils". Presented at the 1992 ACS Industrial and Engineering
Chemistry Division Symposium on "Emerging Technologies for Hazardous Waste Management". Special
Symposium on Bioremediation of Soils and Sediments. Atlanta, Georgia, September 21-23, 1992. Published in
1993 Emerging Technologies in Hazardous Waste Management, ACS Symposium Series, Volume IV, 1993.
5. Govind, R., C. Gao, L. Lai, X. Van, S. Pfanstiel, I. S. Kim and Tabak, H.H. 1993. " Development of Methodology
for the Determination of Biodegradability and Biodegradation Kinetics of Toxic Organic Pollutant Compounds in
Soils". Paper presented at the In-Situ and On-Site Bioreclamation, 2nd International Symposium, San Diego, CA,
April 5-8, 1993. Published in the 1993 On-site Bioreclamation Processes for Xenobiotic and Hydrocarbon
Treatment, Battelle Memorial Institute, Columbus, Ohio.
6. Tabak, H.H., C. Gao, X. Van, L. Lai, S. Pfanstiel, C. Fu and R. Govind. "Kinetics of Biodegradation, Adsorption and
Desorption of Alkyl Phenols and Polycyclic Aromatic Hydrocarbons in Soil Slurry Systems" Presented at the 6th
International IGT Symposium on Gas, Oil and Environmental Biotechnology. The Broadmoor Hotel, Colorado
Springs, CO, November 29-December 1, 1993. To be published in Gas , Oil, Coal and Environmental
Biotechnology VI, Institute of Gas Technology, Chicago, IL., 1994
108
-------
MEMBRANE BIOFILTRATION IN ANAEROBIC/AEROBIC APPLICATIONS
Amit Pundit and Rakesh Govind
Department of Chemical Engineering
Mail Location 171
University of Cincinnati
Cincinnati, OH 4S221
(513)5562666
•and
DoIloffF. Bishop
ORB, Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
(513) 569 7629
INTRODUCTION
The 1989 SARA emission summary for petroleum and chemical manufacturing companies shows
that the largest releases are of volatile organic compounds (VOCs) in air. Current technologies for
treating VOCs are limited and include activated carbon adsorption, wet scrubbing and incineration. An
emerging technology for treatment of VOCs involves the use of biofilters.
Currently, most biofilter studies have involved packed beds of soil, peat or compost, pelletized
packed beds or straight passages ceramic monoliths1'2^. In all these studies, the contaminated air
stream is contacted with a biofilm immobilized on a support material. Due to the presence of oxygen in
the contaminated air stream and high oxygen transfer rates, aerobic degradation of the contaminants
occurs in the biofilms. However, certain organics, like trichloroethylene or parachloroethylene, either
require a cometabolite under aerobic degradation conditions or biotransform only under anaerobic
conditions. The challenge in biofiltration of such compounds is to create anoxic/anaerobic conditions
when the contaminants are present in an air stream.
In this study, a hollow fiber porous membrane module was used as a biofilter. Biomass is
immobilized on the outside of the hollow fibers and the shell side is filled with nutrient mixture, which
flows at a low flowrate. The contaminated gas stream, containing toluene is passed through the'hollow
fibers. The contaminant, in this case, toluene, dissolves in the liquid in the membrane pores, diffuses into
the biofilm on the shell side and biodegrades. Since oxygen transfer is now limited by the dissolution and
diffusion steps, anoxic or anaerobic conditions can be created on the shell side biofilm. This enables the
biotransformation of compounds that are otherwise recalcitrant under aerobic conditions. Figure 1
schematically shows the mechanisim of pollutant dissolution, diffusion in the liquid phase and ultimate
biotransformation in the biofilm. The shaded zones shows the oxygen transfer into the biofilm. As the
biofilm grows on the shell side, the biofilm outside the shaded zone becomes anoxic/anaerobic due to
oxygen consumption in_the aerobic zone. The pollutant degrades in the anoxic zone and the
biotranformation products degrade in the aerobic biofilm as they diffuse back into the hollowfiber side.
109
-------
NUTRIENT
SOLUTION
ON SHELL
SIDE OF
MEMBRANE
MODULE
ANOXIC/
ANAEROBIC
ZONE BIOFILM
AIR WITH CONTAMINANT
POLYMERIC MEMBRANE
HOLLOW FIBER
AEROBIC ZONE
BIOFILM
Figure 1. Schematic showing the aerobic and anoxic zones
in a hollow fiber membrane biofilter. The
contaminant biotransforms in the anoxic zone
and the degradation products mineralize in the
aerobic zone.
110
-------
EXPERIMENTAL STUDIES
Figure 2 shows a schematic of the experimental apparatus. Two hollow fiber modules were
connected in series and toluene contaminated air was passed through the hollow fibers Nutrient mixture
was recirculated through the shell side of each membrane module. Table 1 summarizes the design and
operating conditions of the experimental system.
RESULTS AND DISCUSSION
Preliminary studies were conducted with an inlet toluene concentration of 200 ppmv As shown
in Figure 3, the biofilter removal efficiency increased steadily due to biofilm development on the shell side
of the hollow fibers. When the removal efficiency had increased to about 70% the inlet toluene
concentration was increased to 400 ppmv. After an initial decrease in removal efficiency the removal
efficiency increased attaining about 72%. Further increases in inlet toluene concentration resulted in a
similar response with an initial decrease in removal efficiency followed by a gradual increase to its initial
VEiUC. .
Further experiments are planned with contaminants such as trichloroethylene (TCE) and
parachloroethylene (PCE).
REFERENCES
1. Utgikar, V., R. Govind, Y.G. Shan, R.C. Brenner, and S.I. Safferman, "Biodegradation of
Volatile Organic Compounds in a Biofilter," In "Emerging Technologies for Hazardous Waste
Management H," Tedder, D.W., and F.G. Pohland, Eds., ACS Symposium Series 468 ACS
Washington, DC (1991). ' ''
2. Govind, R. and D.F. Bishop, "Environmental Bioremediation Using Biofilters" Presented at
Frontiers in Bioprocessing HI, Boulder, CO, September 19-23, 1993. Paper to be published in
Conference Proceedings.
3. ; Govind, JL and DolloffF. Bishop, "Advances in Gas Phase Biofiltration Applications for Volatile
Organic Compounds (VOC) Control", Paper presented at the HazMat 93 North-Am Conference
October 12-14, 1993, Detroit, MI. Paper published in Conference Proceedings
111
-------
13
12
1
—c
13
I
Figure 2. Schematic of the membrane biofilter experimental apparatus,
11
1 membrane module
2 hollow fiber
membrane
3 sample collection
point
4 nutrient
5 pump
6 mass flow controller
1 compressure
8 injection point
9 syringe with
Harvard pump
10 pressure drop
measurement set up
11 hood chamber
12 flow meter
13 biomass present
outside the hollow
fiber nuddle
112
-------
Table 1. Summary of the design and operating conditions for the membrane biofilter system.
Air flow rate
Amicon Hollow Fiber Cartridge
(Model NO: H1P10-43)
Number of hollow fiber middles
present per Cartridge
Length of each hollow fiber
middle
Internal diameter of each fiber
Outside diameter of each fiber
Active membrane surface area
per Cartridge
Total internal volume of
hollow fiber nuddles
(for 2 cartridge)
Residence time
Toluene concentration
Nutrient flow rate
pH of nutrient solution
temp
Removal efficiency
(at present)
80ml/min
51
12.5 cm
O.llcm
0.19cm
278.7 sq cm
21.26ml
15 seconds (approx)
600 ppm
lOml/min
7.24
25 oC
80 (%)
113
-------
•a
I
s
M
.9
I
1
o
1
>
(%) Xou3ioijj3 {BACKUS-^
114
-------
RECENT ADVANCES IN BIOFILTRATIQN
Rakesh Govind
Department of Chemical Engineering
University of Cincinnati
Cincinnati, OH 45221 U.S.A.
(513) 556 2666
and
DoIIoffF. Bishop
U.S. Environmental Protection Agency
OKD, Risk Reduction Engineering Laboratory
Cincinnati, OH 45268, U.S.A.
(513) 569 7629
INTRODUCTION
Biofiltration involves contacting a contaminated gas stream with microorganisms in a contactor
and is emerging as an attractive technology for the removal of low concentrations (< 600 ppmv) of
volatile organic chemicals (VOCs) from air. Compared to other technologies, biofiltration is inexpensive,
reliable and requires no post-treatment. It can be applied to any biodegradable VOC emission problem -
from manufacturing and processing units, wastewater and landfill leachate treatment plants (Utgikar et al,
1991)1, and soil remediation operations, such as vacuum extraction.
; Recently, studies have been conducted on the use of pelletized and ceramic structured media in
biofiltration (Utgikar et al., 1991; .Govind et al., 1993a, ^Sb)1*2*3. Conventional applications of
biofilters use fine or irregular support media, such as soil, peat or compost. Biofilters with pellets or
structured media potentially have significant advantages over conventional biofilters, including better gas
distribution, improved pH control by using buffers in the liquid medium trickling through the bed, and the
capability for removal of excess biomass from the media.
These improved biofilters have been used at bench and pilot scale to treat VOCs air stripped from
landfill leachate streams using adsorbent and non-adsorbent media. Nearly complete removal of almost all
of the aerobically degradable compounds were achieved in these biofilters. Intrinsic microbial degradation
kinetics in immobilized biofilms were determined using novel experimental systems. Mathematical
models were developed to describe VOC biodegradation in these biofilters.
Ongoing research and development on improved biofiltration include evaluation of the
operational dynamics, media cleaning and scale-up requirements and use of alternate media capable of
sustaining different kinds of microorganisms.
LIMITATIONS OF CURRENT SYSTEMS
Soil biofilters are_relatively large compared to filters using other media, primarily due to the fact
that soil pores are smaller and compounds have much lower permeability in soil (Bonn, 1992)4. Soil
biofilters also have limited depths due to problems associated with maintaining humidity in soil and
minimizing pressure drop. Furthermore, soil sorption capacity is limited and residual contaminant is
vented immediately to the atmosphere (Bohn and Bohn, 1988)5.
115
-------
Peat/compost biofilters are suitable for treating large volumes of air containing easily
biodegradable VOCs at low concentrations. However, both soil and peat/compost biofilters are susceptible
to channeling and maldistribution of the air stream. This leads to uneven biogrowth as well as drying of
the bed. Both of these effects adversely impact biofilter performance. Soil and peat/compost biofilters
require humidification, and nutrients cannot be supplied continuously. End products of biomass decay
cannot be washed out of the filters and ultimately require media replacement. When VOCs contain
organic chlorine, sulfur or nitrogen, the support media has to neutralize degradation products which are
acidic.
Developmental needs for improving biofiltration include: (1) improved control of moisture and
nutrients in the biofilter bed; (2) improved control of acids (pH) and other inhibitory degradation
products; (3) methods for control of excess biomass to prevent clogging and media replacement; and (4)
improved approaches and media that use adapted or engineered organisms and novel reactor
configurations for degradation of recalcitrant VOCs.
CURRENT DEVELOPMENTS
To meet these needs, extensive work has been conducted by EPA's Risk Reduction Engineering
Laboratory and the University of Cincinnati on biofilters with novel media and operational approaches
(Govind et al., 1993a, 1993b, 1993c)2>3>6. The research featured studies to: (a) demonstrate system
feasibility for degrading representative VOCs; (b) evaluate intrinsic kinetics of VOC degradation; and (c)
develop mathematical models for improved biofilters.
Representative VOCs included compounds with a range of aqueous solubilities and octanol-water
partition coefficients. These compounds were iso-pentane, toluene, methylene chloride, trichloroethylene
(TCE), ethylbenzene and chlorobenzene. Support media used in these studies included activated carbon
pellets' ceramic celite pellets and corrugated plates (Manville Corporation, Denver, CO), and extruded
ceramic monoliths with high surface area to volume ratio (Corning Glass, Inc., Corning, NY).
Comparative studies were conducted with peat/compost biofilters. Control studies were also conducted to
investigate adsorption/desorption of contaminants on the media in the absence of bioactivity.
RESULTS AND DISCUSSION
Biofilter performances were expressed in terms of removal efficiencies of the compounds.
Removal efficiency of the biofilter for each compound was calculated from the amount of compound
removed per unit time in the biofilter, expressed as a percentage of the amount of that compound entering
the biofilter per unit time. Design and operating conditions for each biofilter are summarized in Table 1.
Packed bed peat and compost biofilters
Abiotic control tests with both peat and compost materials showed poor adsorption of iso-
pentane. For both materials, the exit gas composition of isopentane became equal to the inlet
concentration (350 ppm) in about 30 minutes. For both materials, the removal efficiency increases with
decreasing inlet gas concentration and increasing gas residence time. The compost material showed
higher removal efficiencies for isopentane at 360 ppmv inlet concentration than the peat material. At
higher inlet concentrations, the two materials showed similar results.
The effect of water content in the peat and compost material was studied experimentally.
Experiments were conducted with both increasing and decreasing water content in the peat/compost
material. For peat, when the water content decreased below a critical value of 0.48 gms water/gms of dry
peat, the biofilter did not recover its removal efficiency when the water content of the material was
increased. For compost, a similar effect was observed below water content of 0.58 gms water/gms of dry
compost.
116
-------
a S:
§ a
o\
i
pentane (3
so-pentane (3
w v-'
I §
I1 &
I I
Ł I
««Ł
g 'C .
3 O Ł>
e S §
H.U C
1
f
I
•
a o
"3 J3
H U
g
.
i« —
o >,
j- 5
U Ed
0
«
Sis
_
•3
a
a
W)
I
§.
O
•e
§
VO
Ł a
Ł ^
•a-
I-
a
S H
ts
e
I"
a
u
^1
III
^!< ^ f^
o
'§
S
03
W
c e iw M
i a g.a 1
c/> o, S a. S
117
-------
This irreversible damage to the peat/compost material was caused by substantial irreversible
shrinkage of the peat/compost material below the critical water content, resulting in cracks and voids in
the bed through which the gas could easily bypass the bed. Addition of water to the material did not cause
the material to expand and fill the cracks and voids caused by the shrinkage. This showed that
maintaining water content through proper humidification of peat/compost materials was essential for
continued efficient biofilter operation.
The isopentane removal efficiency increased with increasing water content until an optimal value
was attained (0.56 for peat and 0.65 for compost). When the water content was increased above this
optimal value, the removal efficiency decreased for both materials. This decrease in removal efficiency
probably resulted from the partial filling of bed void fraction with free water causing mass transfer
limitations for the degradation of isopentane. This again emphasizes the importance of maintaining water
content in peat/compost materials at or near the optimal value.
Removal efficiencies increased with temperature for both materials. Both materials showed
maximum removal efficiencies above bed temperature of 35°C. Below 25°C, the removal efficiency
decreased almost linearly with temperature. If the bed temperature is below 20°C, substantial
improvements in performance could be achieved by heating the inlet air to increase the bed temperature.
Activated carbon biofilter
Removal efficiency of the biofilter was studied as a function of time for toluene, methylene
chloride and trichloroethylene (TCE) at an inlet gas flow rate of 150 ml/min. This flow rate produced a
gas retention time of 2 minutes in the biofilter bed. Initially the biofilter exhibited high removal efficiency
for all three compounds due to adsorption of the compounds on the activated carbon. However, as the
carbon became saturated, the removal efficiencies decreased until the microorganisms started to grow. The
removal efficiencies increased, attaining nearly 100% in about 80 days after the initial start up of the
system. The carbon biofilter, as the first system studied, did not have optimum startup since insufficient
ammonia in the nutrients limited biomass growth. Subsequent increase in nutrient ammonia accelerated
biofilm development after about 40 days, producing 100% removal efficiencies for all three compounds.
The activated carbon biofilter after acclimation effectively treated (> 99% removals) toluene,
methylene chloride and trichloroethylene for about 4 months without loss of treatment efficiency. After 4
months of operation, the gas flow rate was reduced to 50 ml/min. since the biofilter was beginning to
flood at higher flow rate. Flooding resulted from plugging of the biofilter bed due to excessive biomass
growth. Biomass growth and plugging of the biofilter was reflected in greater pressure drop across the
bed.
Ceramic Celite packed bed Biofilter
The removal efficiency of the ceramic celite packed bed biofilter was studied for five compounds
(toluene, methylene chloride, TCE, ethylbenzene, chlorobenzene) as a function of time. Removal
efficiencies gradually increased to nearly 100% for all the compounds except trichloroethylene (TCE).
This rise indicated that biomass growth occurred during this period, ultimately building enough biomass
to degrade the compounds completely.
The celite biofilter was as successful in removing the compounds, except for trichloroethylene, as
the activated carbon biofilter. Folsom et al. (1990)7 have shown that trichloroethylene is degraded only as
a secondary substrate in the presence of a primary metabolite, such as toluene or phenol. The
biodegradation rate of toluene is much faster than that of TCE. As toluene entered the bottom of the
biofilter, it was rapidly removed in the bottom third of the biofilter. The removal was confirmed by
determining the concentration profile of the compounds in the celite biofilter through analyzing samples
withdrawn from side ports located along the height of the biofilter. As a result, toluene was not available
as a primary metabolite to trigger TCE degradation in the rest of the biofilter.
118
-------
With the activated carbon biofilter, toluene fed through the gas phase was also consumed rapidly
However, toluene adsorbed on the activated carbon during the initial period provided primary carbon
source for TCE degradation in the rest of the column. A toluene reservoir is not possible in case of celite
which is essentially a non-adsorbent support medium.
After 6 months of operation, the celite packed bed biofilter plugged due to biomass growth
Studies were conducted on cleaning the pellets using water at room temperature. At biomass loadings
above 5 weight percent of the clean celite pellets, biomass loss was high initially. At low biomass
loadings, loss of biomass from the pellets was small. Further, the amount of water required to clean the
pellets was significant. High ( > 5) water to pellet weight ratios may be a problem since significant
amounts of wastewater would be generated and adequete water handling systems would be required.
Ceramic Celite Plates Straight Passages Biofilter
The removal efficiency of the biofilter was studied as a function of time for all the five
compounds. The removal efficiency curves indicate that near complete removals of all the compounds
except TCE were obtained at the flow rate of 600 ml/min. Gas chromatographic analysis of the exit air
stream did not detect any degradation products. As in earlier studies, the carbon dioxide increase in gas
streams and the chloride ion increase in the nutrient stream were monitored. The results indicated that
removed compounds had undergone full biological mineralization.
Biomass loss in the nutrient solution was characterized by collecting samples of the nutrient
solutions and drying them at 100 °C. The effect of nutrient flow rate on biomass loss rate from the
biofilter was measured. The biomass loss from the biofilter is affected by nutrient flow rate and the total
biomass loading in the biofilter, which depends on the gas flow rate and its inlet composition.
To increase TCE removal efficiency and demonstrate that lower TCE removal efficiencies were
due to lack of toluene, another co-metabolite, phenol, was added to the nutrient media at a concentration
of 5 mg/L. Increased TCE removal efficiency was observed after phenol addition to the nutrient media
Complete removal of TCE was achieved when inlet TCE concentration was reduced from 25 pom to 10
ppm. **
Degradation rates achieved in. the experimental straight passages biofilter were mainly limited bv
the low surface area of the celite plates. With recent developments in extruded ceramic monoliths
significantly higher surface area per unit volume can be obtained commercially Using such high surface
area extruded monoliths, significantly higher biomas loadings and hence higher degradation rates per unit
biofilter volume can be realized.
CONCLUSIONS
Biofilters using porous pellet or structured (straight-passages) media and recycled nutrient and
buffer solutions have potential advantages over conventional soil or peat/compost biofilters for removal of
biodegradable VOCs from contaminated air. These include:
• improved distribution of air flow and moisture control in the media, increasing process
performance;
• improved neutralization of mineral acid degradation products (pH control);
• increased capacity for treating higher VOC loadings (300 -600 ppmv); and
• capability for removal of excess biomass from the biofilter, thus preventing clogging and
eventual media replacement.
119
-------
Biofilters with porous pellet media, however, require development of effective cleaning methods
for field-scale applications. Biofilters with structured media (straight passages) offer easier cleaning
options, including release of biomass (self cleaning) into the recycling nutrient and buffer solution.
Basic research areas currently being investigated to exploit the full potential of structured media
(straight passages) biofilters include the following:
• quantification of biomass release, which depends on the shear force exerted on the
biofilm by trickling nutrients, amount of biomass present in the biofilter and substrate
load;
• quantification of biofilm adhesion to the media surface, which controls the release
of the biomass. Biomass attachment also impacts biofilter removal efficiency during
startup and transient operation.
Other research needs in biofiltration include:
• evaluation of biofilter dynamic responses to variable flowrates and contaminant
concentration;
• proper design guidelines for selection of operational parameters in industrial biofilters.
These guidelines would enable proper selection of media and media dimensions,
flowrate and composition of nutrient solution, countercurrent or cocurrent mode of
operation, methods for initial seeding of biomass, biofilter media cleaning methods and
frequency of cleaning.
Acknowledgement
We gratefully acknowledge support from the U.S. Environmental Protection Agency
(Cooperative Agreement No. CR816700). Other participants in this project are: Yongui Shan (Visiting
Scholar), Vivek Utgikar (Graduate Student), and Zhao Wang (Graduate Student).
LITERATURE CITED
1. Utgikar, V., R. Govind, Y.G. Shan, R.C. Brenner, and S.I. Safferman, "Biodegradation of
Volatile Organic Compounds in a Biofilter," In "Emerging Technologies for Hazardous Waste
Management II," Tedder, D.W., and F.G. Pohland, Eds., ACS Symposium Series 468, ACS,
Washington, DC (1991).
2. Govind, R. and D. F. Bishop, "Environmental Bioremediation Using Biofilters", Presented at
Frontiers in Bioprocessing III, Boulder, CO, September 19-23, 1993. Paper to be published in
Conference Proceedings.
3. Govind, R. and Dolloff F. Bishop, "Advances in Gas Phase Biofiltration Applications for Volatile
Organic Compounds (VOC) Control", Paper presented at the HazMat 93 North-Am Conference,
October 12-14,1993, Detroit, MI. Paper published in Conference Proceedings.
4. Bonn, H.L., "Consider Biofiltration for Decontaminating Gases," Chem. Eng. Progr.. 88(4), 34
(1992).
5. Bohn, H., and R. Bonn, "Soil Beds Weed out Air Pollutants," Chem. Eng.. 95(6), 73 (1988).
120
-------
Govind, R. and Dolloff F. Bishop, "Development of Novel Biofilters for Treatment of Volatile
Organic Compounds (VOCs)H, Paper presented at the IGT Symposium on Gas, Oil and
Environmental Biotechnology, Colorado Springs, CO, November 29-December 3, 1993. Paper to
be published in Conference Proceedings.
Folsom, B.R., PJ. Chapman, and P.H. Pritchard, "Phenol and Trichloroethylene Degradation by
Pseudomonas cepacia G4: Kinetics and Interactions Between Substrates," Appl Environ
Microbiol.. 56, 1279 (1990).
121
-------
RADIATION OF CONTAMINANTS WITH BIOLOGICAL ACTIVATED CARBON SYSTEMS
Thomas C. Voice1 • Xianda Zhao1 • Jing Shi2- and Robert F. Hickey2
1 DeDartment of Civil and Environmental Engineering, Michigan State University, East Lansing, Ml 48824
phone: (517)353-9718
2. Michigan Biotechnology Institute, 3900 Collins Road, Lansing, Ml 48909
INTRODUCTION
The conventional approaches to above-ground treatment of groundwater contaminated with volatile
organic compounds (VOCs) involve either liquid-phase adsorption by granular activated carbon (GAG) or
air stripping. In these processes, the VOCs are simply transferred from one phase to another without
degradation. As a result, further treatment or disposal of the receiving phase is often required. As an
alternative biological treatment has the potential to completely destroy contaminant compounds.
Biological processes, however, are frequently perceived as being less stable than physical-chemical
approaches, and thus are not always considered for groundwater treatment.
Biological activated carbon (BAG) integrates biological removal and GAG adsorption into a single unit
process [i 2] In systems designed as adsorbers, it has been shown that biodegradation increases the
period between GAG regeneration cycles [3]. In biofilm systems, the use of GAG as a biomass carrier has
been shown to remove the compounds resistant to biodegradation and provide enhanced removal during
loading transients [4,5].
In this research, laboratory and pilot-scale fluidized bed systems (BAC-FBR) employing GAG as a
biomass carrier were used for the treatment of groundwater contaminated with the gasoline constituents,
benzene toluene and xylene (BTX). The systems were evaluated during start-up, under pseudo-steady-
state conditions and following a step increase in influent loading. The Investigation focused on isolating
the contributions of adsorption and biodegradation to system performance.
METHODOLOGY
FBRs of three different sizes were used in this work. The dimensions and operational parameters are
listed in table 1. The reactors were operated as one-pass systems without recycle (Figure 1). The
qroundwater was oxygenated to a concentration sufficient to maintain an effluent dissolved oxygen
concentration (DO) of more than 4 mg/L and supplemented with nitrogen and phosphorous at a weight
ratio of 100-5-1 COD:N:P. The target compounds were injected into feed water using a syringe pump and
dissolved using a mixer unit. This contaminated water was then pumped to the bottom of the reactors
using peristaltic pumps. The reactors were inoculated with a mixed culture obtained from a pilot-scale FBR
that was originally seeded with activated sludge and has subsequently been supplied with BTX as the sole
carbon source for more than two years.
This project was divided into three phases. 1) Two 1" laboratory-scale fluidized bed systems were
operated in parallel for the comparison of adsorptive and non-adsorptive biofilm carriers. One reactor
(FBR) was charged with baker product, a non-adsorptive carbon, and another (BAC-FBR) was charged
with GAG The systems were evaluated in a start-up, pseudo-steady-state and step-load increase
conditions During the step-load increase experiment, the concentration of BTX was increased in 4-fold
for 8 hours while keeping the flow constant. 2) A 2" laboratory-scale BAC-FBR charged with GAG was
used to treat toluene contaminated groundwater. The amount of adsorbed toluene on BAG particles was
measured at regular intervals during start-up, and before and after a 77-hour 4-fold step-load increase. 3)
Toluene contaminated groundwater was treated in a 3" pilot-scale BAC-FBR for a period of 1 year. During
this operational period, BAG particles were collected from the reactor and subjected to adsorption
isotherms for the measurement of adsorption capacity. Isotherms for toluene were performed using a
bottle-point method [5].
BTX concentrations were determined using a gas chromatograph equipped with a flame ionization
detector (FID), a helium carrier gas, and an automated headspace sampler. The accuracy for measurement
122
-------
of a sample containing a concentration of 1.5 tig/L was ± 0.1 ng/L Influent and effluent DO were analyzed
using a polarographic electrode coupled to a digital pH/millivolt meter.
TABLE 1. THE DIMENSIONS AND OPERATION PARAMETERS OF FBRS
[
Dimensions of FBRs
Diameter
Height :
Working Volume
Column Material
Carrier Media
Flow Rate
Hydraulic Flux Rate
Empty Bed
Retention Time
Target Compounds
Organic Loading
Rate
Temperature
1 " laboratory-scale
FBR
2.54 cm (1 in)
92/1 84 cm (3/6 ft)*
0.5/1 liter
glass
Baker Product}:
20/30 mesh
(0.75 mm)_
0.14/0.3 liter/min
0.29/0.61 m3/min-
m2
3.6/3.3 min
BTX
3.4 kg COD/m3-day
6.0 kg COD/m3-clay
15 °C
1" laboratory-scale
BAC-FBR
2.54 cm (1 in)
92/1 84 cm (3/6 ft)*
0.5/1 liter
glass
GAG F-400§
20/30 mesh
(0.75 mm)
0.14/0.3 liter/min
0.29/0.61 m3/min-
m2
3.6/3.3 min
BTX
3.4 kg COD/m3-day
6.0 kg COD/m3-day
15 °C
2" laboratory-scale
BAC-FBR
5 cm (2 in)
184 cm (6 ft)
4 liter
glass
GAC F-400§
20/30 mesh
(0.75 mm)
0.9 liter/min
0.44 m3/min-m2
4.4 min
Toluene
5.3 kg COD/m3-day
15 °C
3" pilot-scale
BAC-FBR
7.62 cm (3 in)
322 cm (10 1/2 ft)
14.68 liter
PVCt
GAC F-400§
1 8/20 mesh
(0.75 mm)
2.2 liter/min
0.48 m3/min-m2
6.7 min
Toluene
1.4kgCOD/m3-day
20°C
1" laboratory-scale fluidized bed systems were set in two configurations. The 0.5 liter setting was used
for low organic loading experiment and 1 liter setting was used for high organic loading experiment.
t polyvinyl acetate;4 Baker product is the non-adsorptive carbon; § Calgon Fitrasorb 400.
Treated
Water-
Fluidized
Bed
Reactor
Oxygen
Nutrient Tank
Nutrient Pump
Syringe
I Pump
Figure 1. Schematic of a fluidized bed system.
RESULTS
In phase one of this work, two FBRs were operated in parallel. The BAC-FBR utilized both adsorption
and biodegradation as removal mechanisms and the FBR operated with biodegradation as the only
significant removal mechanism. During the start-up period, the BAC-FBR performed as an adsorption
column. Effluent benzene concentration increased up to 50% of influent level at 140 operation hours.
After this period, the DO consumption and the carbon bed height started to increase and the
123
-------
concentration of benzene in the effluent decreased. This indicated the development of significant
biological activity and biofilm growth in the BAC-FBR. During the entire start-up period, effluent
concentrations of toluene and p-xylene were maintained at very low level of 15 \ig/L and 18 ng/L,
respectively In the FBR system, breakthrough of BTX occurred soon after introduction of the influent
BTX After approximately 400 operation hours, sufficient biomass had developed on the baker product
and the concentrations of BTX in the effluent started to decrease. The DO consumption and bed height
also increased. As a result, a large amount of BTX was not treated by the FBR: average removals were only
33% 40% and 40% for benzene, toluene and xylene, respectively. In contrast, the BAC-FBR was
achieved 91%, 98%, and 98% for benzene, toluene and xylene, respectively, during start-up.
After a mature biofilm developed on the carbon, effluent BTX and DO consumption levels stabilized
and the FBRs were assumed to be operating under pseudo-steady-state conditions. In the first test, the
average total COD in the influent of the BAC-FBR was 9.6 mg/L (an organic loading rate of 3.8 kg COD/m3-
dav) and removals were 99.7%, 99.6%, and 92% for benzene, toluene and xylene, respectively (Table
2) A 7 3 mg COD/L (an organic loading rate of 2.9 kg COD/m3-day) was applied on FBR and the removals
were 94 6% 94 7%, and 90.7% for benzene, toluene and xylene, respectively (Table 2.). In the second
test the organic loading rates were increased in both reactors to 6.0 kg COD/m3-day. The two FBRs
performed similarly and the removal averaged 94%, 90%, and 81% for benzene, toluene and xylene,
respectively (Table 2.).
TABLE 2 COMPARISON OF PSEUDO-STEADY-STATE BTX REMOVAL BETWEEN FBRS
Reactor
BAC-FBR
FBR
BAC-FBR
FBR
Compound
Benzene
Toluene
p-Xylene
Benzene
Toluene
p-Xylene
Benzene
Toluene
p-Xylene
Benzene
Toluene
p-Xylene
Influent
(W3/L)
1066+303
1 023+340
996+340
834±151
758+380
732+340
1571 ±303
1531+330
1382±302
1 628±1 42
1479±170
1267±152
(W3/L)
3±2
5±5
81+53
45±40
40+40
68+42
86±40
152±60
254±1 00
92±52
148±85
237+123
(%)
99.7
99.6
92.0
94.6
94.7
90.7
94.5
90.1
81.6
943"
90.0
81.3
Rate
(ka COD/m3-clay)
3.8
2.9
6.0
6.0
After the achievement of a pseudo-steady-state at an organic loading rate of 6.0 kg COD/m3-day, the
BAC-FBR and FBR were subjected to 4-fold step-load increases for a duration of 8 hours to determine
whether the use of GAG as a biofilm carrier contributed to system stability. During the increase, the
removals of BTX in the BAC-FBR were 82.3%, 75.3% and 65.2% for benzene, toluene and xylene
resoectivelv (Table 3). Under the same conditions, removals in the FBR were 64.1%, 54.7% and 46.5%
for benzene, toluene and xylene, respectively (Table 3). The BAC-FBR removed 57% more COD from the
influent during this period than the FBR did. The ratio of DO consumed to COD removed was lower in the
BAC-FBR than in the FBR, and lower than the pseudo-steady-state value of 0.68. This indicates that the
BAG carrier was responsible for the additional removal in the BAC-FBR system. Using DO consumption in
the BAC-FBR to estimate biological removal, it was calculated that approximately 23% of the COD was
removed COD by adsorption.
In phase two of this study, a 2" laboratory-scale BAC-FBR was operated to treat toluene
contaminated groundwater. The adsorbed toluene on the BAG was monitored by collecting a small carbon
sample from five different heights on the reactors. The system was fed 2.5 mg/L of toluene (an organic
loading rate of 5.3 kg COD/m3-day). The amount of adsorbed toluene on the carbon increased
dramatically during the first 5 days (Figure 2.) of operation. There was no appreciable DO consumption
during this period. The DO consumption increased sharply between day 4 and 11 to the maximum value
124
-------
of 8 mg/L and then slowly decreased to a pseudo-steady-state average value of 5.6 mg/L. The amount of
adsorbed toluene increased until day 10 and then began to decrease gradually as the degradation
capability of the biofilm developed. In an attempt to assess the influence of biofilm thickness on
adsorption and biological removal, a large amount of biomass was removed on day 39. Following removal,
the amount of adsorbed toluene increased until the full biodegradation capability was reestablished. After
the reactor resumed pseudo-steady-state operation, based upon constant DO consumption
commensurate with that required for degradation of the added toluene, the concentration of toluene in
the influent was increased to 11 mg/L (an organic loading of 22.6 kg COD/m3-day). After 77 hours, the
concentration was decreased to 2.3 mg/L. The amount of adsorbed toluene increased during the period
of increased loading, but then began to decrease when the influent was returned to a lower level (Figure
2.). A high removal rate was achieved in all periods of operation, even during the step-load increase
(99.3%, 99.4%, 95.1% and 98.5% for start-up, pseudo-steady-state, step-load and recovery,
respectively).
TABLE 3. SUMMARY OF FBRs RESPONSE TO STEP-LOAD INCREASE
Reactor
BAC-
FBR
FBR
Compound
Benzene
Toluene
p-Xylene
Benzene
Toluene
/>-Xylene
Influent (^g/L)
7863+550
7503±525
63131568
6918±600
62221610
4911+450
Effluent (ng/L)
1 403±280
1 8501336
21971340
24871390
28211410
26251400
DO Consumed (mg/L)
23
19
Average Removal (%)
82.3
75.3
65.2
64.1 .
54.7
46.5
10
8 -•
6 --
•§
1 4
31
2 - -
Step Recovery
Increase r
10
20
30
40 50
Time (days)
60
70
80
90
Figure 27 Total mass of adsorbed toluene in the BAC-FBR system
In the third phase, toluene contaminated groundwater (2.1 mg/L of toluene) was treated in a pilot-
scale BAC-FBR. The average removal rate was 97.2% during pseudo-steady-state operation. A sample of
the BAC carrier was periodically collected from the middle of the reactor and an adsorption isotherm and
125
-------
analysis for adsorbed toluene was performed. The isotherm parameters were estimated using an
extended Freundlich isotherm equation:
qe+q0=KfCen
Where: qe is additional amount adsorbed toluene at equilibrium (mg/g). qo is initial amount adsorbed
toluene (mg/g). Ce is equilibrium concentration (mg/L). Kf and n are constants. An isotherm was also
performed using clean GAG. It was found that the Freundlich K{, which indicates adsorption capacity,
decreased to approximately one-half the value found for clean carbon. This represents a substantial
decrease in the toluene adsorption capacity over this period of operation.
CONCLUSIONS
Biological activated carbon systems integrate biological degradation and GAC adsorption
mechanisms in a single unit process. The results from this study demonstrate the value of such an
integrated approach. The BAG system exhibited a shorter start-up period than a similar biological only
system and consistently maintained low concentrations of target compounds in the effluent. The
performance of the two systems was similar under pseudo-steady-state conditions, suggesting that
adsorption is relatively unimportant when influent concentrations are stable. Under influent loading
transients, however, the BAG system maintained significantly lower effluent concentrations than the non-
adsorbing system. This suggests that the activated carbon serves as a buffer to smooth out influent
fluctuations and to dampen the effect of changes in the biological capability of the system. The data show
that compounds adsorbed during the concentration increase were subsequently desprbed and
biodegraded when the bulk liquid concentration returned to lower levels. The adsorption capacity of BAG
particles decreased significantly over the course of one year of BAC-FBR reactor operation. Based upon
the results of this study, several general conclusions and recommendations can be drawn.
1. At an organic loading rate of 3.8 kg COD/m3-day and a concentration of 1.0 mg/L for each of BTX,
the BAC-FBR can achieve greater than 90% removal in 3.6 min of empty-bed retention time.
2. BAC-FBR systems have a relatively short start-up period. At an organic loading rate of 3.8 kg
COD/m3-day and a concentration of 1.0 mg/L for each of BTX, the reactor can have full biological
capability for the treatment of BTX contaminated groundwater within one week. The use of
activated carbon as the biofilm carrier insures a high effluent quality during start-up.
3. BAC-FBR systems should provide better effluent quality during loading transients.
4. Additional DO and supplemental nutrients should be provided during a increase of substrate
concentration in the influent and continued for at least 2 weeks after the increase.
5. Due to a decrease of adsorption capacity of the GAC in the BAC-FBR system, it may be desirable
to periodically replace the carrier material.
REFERENCES
1. Weber, W.J., L.D. Friedman, and R. Bloom, Biologically- extended physicochemical treatment, in
Advance in Water Pollution Research, S.H. Jenkin, Editor. 1972, Pergamon Press (Oxford): p.
641-56.
2. Digiano, F.A.;. Influence of biological activity on GAC performance, in Conference on Applications
of Adsorption in Wastewater Treatment. 1981. Enviro Press, Nashville, Tenn. p.285-305.
3. Speitel, G.E., Jr., et al., Biodegradation of trace concentrations of substituted phenols in granular
activated carbon columns. Environ. Sci. Technol., 1989. 23(1): p. 68-74.
4. Suidan, M.T., P. Fox, and J.T. Pfeffer, Anaerobic treatment of coal gasification wastewater.
Wat.Sci.Tech., 1987-. 19: p. 229-36.
5. Voice, T.C., et al., Biological activated carbon in fluidized bed reactors for the treatment of
groundwater contaminated with volatile aromatic hydrocarbons. Water Research, 1992. 26(10): p.
1389-401.
126
-------
SOIL SLURRY BIOREACTORS
BENCH SCALE STUDIES
Jennifer
John A. Glaser1, Majid A. Dosani2, Paul T. McCauley1,
ifer S. Platt2, Edward J. Opatken1, and E. Radha Krishnan2
United States Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Dr.
Cincinnati, phio 45268
(513) 569-7568
2IT Corporation
11499 Chester Rd.
Cincinnati, Ohio 45246-4098
INTRODUCTION
The effectiveness of slurry-phase reactors to biologically
degrade toxic organic compounds in contaminated solids (soils,
sediments, and sludges) has been the subject of extensive evalua-
tion studies . x An EPA Best Demonstrated Available Technology
(BOAT) study showed remarkable pollutant disappearance rates of
polynuclear aromatic hydrocarbons in creosote contaminating
soil. One field study has used slurried lagoons for the treat-
ment of sludge deposits. This information tends to strengthen the
claim that slurry bioremediaiton is a reliable, predictable, and
fieldable technology. In spite of great acclaim, there remain
some ; fundamental issues requiring resolution before slurry
treatment matches current claims.3
BACKGROUND
The use of soil slurry bioreactors to treat contaminated
solid materials has increased to the point that 180,000 gal.
127
-------
reactors are being used as part of a remedial action plan at a
waste site. In spite of the continued use of slurry bioreactors,
there is no generally recognized source of information to guide
the operation of these reactors. As part of a general program to
develop a knowledge base supporting engineering design and opti-
mization of slurry bioreactors, we have initiated bench scale
studies as a diagnostic step to aid process definition for larger
scale operations.
Our research program is divided into two discrete endeavors.
At a basic level of inquiry, we assess slurry reactor operation
with a desire to adequately model the slurry treatment process.
At a more applied level of investigation, we assess the
performance of a slurry bioreactor using the best current
information to select optimal operating conditions. Problems
encountered at the applied level are considered for further
investigation as part of the basic program.
METHODOLOGY
The slurry-phase bioreactor is a continuously stirred tank
reactor (CSTR) in which contaminated solids are suspended in
water. For aerobic treatment, the bioreactor is supplemented with
air or oxygen, nutrients, and, where necessary, microorganisms
capable of degrading the targeted pollutants. Size classification
of the raw contaminated soil precedes slurry formation. The soil
fraction most highly contaminated is selected for processing.
Sand fractions and other non-interacting fractions are generally
excluded, since they dilute the feed stream and do not contain
much of the targeted pollutants. Aggressive agitation and
aeration provide an environment where biodegradation of the
organic contaminants is considered optimal. Limiting nutrients
can be provided and-physical contact between contaminants and
microorganisms can be optimized in a slurry system to enhance the
rate of biological reactions through improved mass transfer.
The reactors used in this study are special designs made of
Pyrex glass having a total volume of 8 L and a working volume of
6 L. They are made in the general design of glass resin kettles
with a four-necked cover. The reactor body has three sampling
ports attached located at 5, 10 and 15 cm on the vertical
dimension from the reactor bottom. The four openings in the
reactor cover permit slurry loading, introduction of gas transfer
lines, temperature measuring probes, and the agitator shaft. The
reactors are housed in individual plexiglass chambers for spill
and emission control.
The agitator attached to each reactor is powered by a 1/10
Hp direct drive, 20-1800 rpm variable speed mixer. The agitator
shaft has two impellers attached to provide maximum suspension of
solids. Removable stainless steel baffles provide turbulence for
efficient mixing. Air or oxygen is introduced through a small
gauge stainless steel tube below the lower impeller to insure the
128
-------
best gas dispersion. Needle-valve flow meters permit manual
adjustment of gas flow rate.
A host of operating factors are measured routinely during
the operation of slurry-phase bioreactors. Our initial studies
have focussed on physical limitations to mixing and the
conditions of suspension required for optimal reactor operation.
There are several possible mixing criteria that may be useful for
slurry processing: percent suspension, suspension uniformity,
off-bottom suspension, suspension height, and particle size
concentration profile as a funcion of suspension height.
: The bioslurry reactor was fitted with a R100 paddle at the
end of the agitator shaft and an A310 axial flow impeller fitted
7.5 cm from the end of the shaft. The minimum rotational rate
for complete off-bottom suspension of soil was recorded for
conditions of 10, 20, 30, and 40 scfh gasflow rates and 10, 20,
30, and 40% soil solids compositions. The distance separating the
impeller blades was varied and conditions for off-bottom
suspension were recorded.
RESULTS
These studies showed that when the impellers are separated
by 7.5 cm, air flow raised the speed of the mixer required for
complete off-bottom suspension in the range of gas flow of 0-20
scfh. For flow rates of 20-40 scfh, mixing speeds required for
off-bottom suspension did not increase with increased gasflow
rates. These results were seen in 10, 20, 30 and 40% soil-water
suspensions.
When the distance between the impellers was increased to 10
cm, the mixing speeds required for off-bottom suspension
increased over the range of all gas flow rates (0-40 scfh). When
the distance between impellers was narrowed to 5 cm, the speed
required for off-bottom suspension were equal to or slightly less
than the results for the impeller separation of 7.5 cm.
CONCLUSIONS
These studies were designed to evaluate the physical
characteristics of the reactor operation leading to off-bottom
suspension of the the soil slurry for the 8 L bench scale
reactors. The results suggest that the conditions for off-bottom
suspension are sensitive to gasflow and solids concentration in
the slurry. This investigation provides a systematic evaluation
of slurry mixing characteristics which will be coupled with other
research studies to mo're clearly define the optimal operation of
soil slurry bioreactors for hazardous waste treatment. Continuing
studies evaluating additional controlling factors will serve as
the basis of modelling slurry reactor operation.
129
-------
REFERENCES
1. Ross, D. Slurry-Phase Bioremediation: Case Studies and Cost
Comparisons, Bioremediation 1, 61-74, (1990/1991) Winter.
2. EPA, Technology Evaluation Report: Pilot-Scale Demonstration
of a Slurry-Phase Biological Reactor for Creosote-Contaminated
Soil, in print.
3. Glaser, J.A. and McCauley, P.J. Soil Slurry Bioreactors: A
Perspective. Paper presented at In-Situ, On-Site Bioremediaiton:
The Second International Symposium, San Diego, CA. April 5-8,
1993.
4. Jerger, D.E., Cady, D.J., and Exner, J.H., Full-Scale Slurry-
Phase Biological Treatment of Wood Preserving Wastes. Paper
Presented at the IN-Situ and On-Site Bioremediaiton, The Second
International Symposium, San Diego, CA.,April 5-8, 2993.
FOR MORE INFORMATION:
John A. Glaser
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Dr.
Cincinnati, Ohio 45268
(513) 569-7568
130
-------
Use of Composting Techniques to Remediate Contaminated Soils and Sludges
James H. Johnson, Jr.
Lily W. Wan
Department of Civil Engineering
Howard University
2300 Sixth Street, N.W.
Washington, DC 20059
(202) 806-6570
INTRODUCTION
As the number of sites on the EPA's National Priority List (NPL) increases (currently in excess of
50,000 identified and several hundred thousand leaking underground storage tanks) solid phase
bioremediation is becoming more important for NPL site cleanup due to its cost-effectiveness. The U.S.
Army recently evaluated 57 technologies for removal of contaminants in soil. Composting technology
was the only innovative technology selected for development.
Traditionally, composting is the controlled, solid-phase aerobic thermophilic oxidation of organic
matter by a consortia of microorganisms to yield partially stabilized residual organic matter (compost).
As a result of the temperature rise, pathogenic organisms are killed.
An emerging use of composting is for the treatment of sludges and soils containing hazardous
materials. In this application, the objective is to create an environment which encourages
microorganisms to utilize the contaminant as its carbon source and/or electron donor. Work has recently
been completed or is currently undeiway at Howard University for using this technology to remediate
soils contaminated with pyrene, TNT and BTEX.
The pyrene and BTEX studies build upon our understanding of the catabolism of these com-
pounds and were aimed at quantifying the parameters needed for optimization of the engineering
process. The objective of the TNT studies was to develop a treatment scheme incorporating composting
technology with a chemical oxidation step. The chemical oxidation step was incorporated in an attempt
to overcome the lack of understanding of the catabolism of TNT. The results of these activities are
described below.
MATERIALS AND METHODS
Pyrene studies were conducted using sludge from the Blue Plains Wastewater Treatment Plant
(Washington, DC) and a sassafras sandy loam soil obtained from a well characterized U.S.D.A. site in
Beltsville, MD. The sludge and soil were spiked with 10-15 mg/kg (dry wt. basis) of unlabeled and 15 ju
Ci/kg of labeled pyrene. Previously composted sludge and wood chips were used as bulking agents.
Pyrene;sludge degradation studies were conducted in a laboratory scale composter developed by Sikora,
et al. (1) and a series of batch composters (125 ml amber colored Erienmeyer flask). The laboratory
scale composter permitted continuous quantification of CO2 and 14CO2 production while the batch
composters allowed the measurement of pyrene disappearance with time.
TNT studies were conducted using sludge from Blue Plains as well as aqueous suspensions in
order to determine and track metabolites. All TNT studies were conducted using batch composters
Use of compostingiechnology for remediation of a BTEX contaminated soil was site specific. A
gasoline contaminated sandy soil was obtained from Sleeping Bear Dunes National Park in Traverse
City, Mi. The BTEX concentration in the sample received at the laboratory was very low. The soil was
spiked to increase the BTEX concentration to 2000 mg/kg (dry wt. basis), which represented the higher
concentration range of samples at the site. Amber colored bottles (250 ml) covered with mininert tops
131
-------
were used for these studies. Because of the volatility of BTEX, other systems which did not provide leak
proof conditions resulted in high losses of BTEX in abiotic samples.
RESULTS
Pvrene: During composting of a raw and limed sludge with woodchips as a bulking agent,
pyrene mineralization was observed approximately eight days after the beginning of the composting
period. Seventeen to twenty percent of the initial pyrene activity was recovered as 14CO2 from a basic
and neutral sludge, respectively (see Figure 1). The highest pyrene reduction achieved was 75%. A
factorial design analysis of pH, temperature and moisture content based upon four experimental
conditions indicated that pH and temperature had moderate inverse effects, and moisture had a
moderate direct effect. The combined interaction of all the parameters produced the most significant
effect (see Table 1). The target impact for the analysis was pyrene mineralization.
Figure 1. Labeled Carbon Dioxide
Evolution Profiles During Pyrene
Composting
CCL gr/kg compmL d«y
tlnwfdayt)
Table 1. Factorial Design and Effect Calculation of Composting Parameters
Condition
No.
1
Factorial
Conditions
a (high pH, low
moisture and
low temp.)
c (low pH, high
moisture and
low temp.)
b (low pH, low
moisture and
high temp.)
abc (high pH,
high moisture
and high temp.)
Parameters
pH Moisture Temp
Pyrene
Reduction
12
12
70
80
70
80
60
60
80
80
17.81
20.23
18.15
18.14
Parameter Effect
Calculation
pH Effect = -0.53
Temp. Effect = -0.36
Moisture Effect =
+0.77
Overall Effect =
+18.74
In the soil with composted sludge as a bulking agent studies, as the amount of soil content
increased the percent reduction in pyrene concentration decreases. Up to a soil loading of 30%
(soil/composted sludge ratio) the percent pyrene reduction increases. Pyrene disappearance for
132
-------
soil/sludge ratios of 0.5:1 to 4:1 fitted a first order kinetic model with k values ranging from 0.019 to
0.034 per day.
TNT: The rationale for these bioremediation studies was the chemical oxidation of TNT to
trinitrobenzoic acid or picric acid and subsequent biodegradation was analogous to the transformation of
BTEX and lightly chlorinated RGBs to their corresponding benzoic acid and phenols. The disappearance
of TNBA to date has only been demonstrated in the batch composter with raw sludge.
Attempts to reproduce these results in an aqueous system to permit identification and quantifica-
tion of metabolites have been unsuccessful.
BTEX: Ninety-three and sixty-five percent BTEX disappearance based upon head space
analyses occurred in these experiments at composting temperatures of 25°C and 60°C, respectively.
Because this project was site specific, only temperature and bulking agent quantity were varied.
Factorial design analysis of these at low and high values, indicated temperature had a significant inverse
effect, bulking agent content was significant proportionally and the combined effect was very significant
proportionally. The target impact was BTEX disappearance. Figure 2 shows typical results.
Figure 2. BTEX Disappearance with Mininert at 2SC Based on Headspace
Results
•Abiotic
m Siotic (sludge to soil ratio 1:1)
• Biotic (sludge to soil ratio 2:1)
468
Time - Days
10
CONCLUSION
Bioremediation using composting technology is very feasible. In cases where the biochemistry
involved in the catabolism of the compounds are known, the development of engineering parameters for
optimization can be obtained using a factorial design approach. The factorial design approach will not
only provide information about the impact of individual parameters but will quantify the interactions of the
parameters.
In cases where direct and broad based biodegradation of compounds are not yet possible, e.g.
TNT. Composting technology appears to be appropriate as one of the treatment processes suitable for
integration in a treatment scheme.
REFERENCES
1. Sikora, L.J., Ramirez, M.A. and Troeschel, T.A. Laboratory Composter Simulation Studies. Journal
of Environmental Quality 14(3): 434-438,1983.
133
-------
FOR MORE INFORMATION
James H. Johnson, Jr., Department of Civil Engineering, Howard University, 2300 Sixth Street, N.W.,
Washington, DC 20059. (202) 806-6570; (202) 806-5271 (FAX).
ACKNOWLEDGMENT
Funding for the work was provided by the Office of Research and Development, U.S.
Environmental Protection Agency under grant R815750 to the Great Lakes and Mid-Atlantic Hazardous
Substance Research Center. Partial funding of the research activities of the Center is also provided by
the State of Michigan Department of Natural Resources. The content of this publication does not
necessarily represent the views of either agency.
134
-------
USE OF CHEMICAL DISPERSANTS FOR OIL SPILLS IN MARINE WATERS
Daniel Sullivan
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
2890 Woodbridge Avenue
Edison, New Jersey 08837
INTRODUCTION
Despite many precautions, oil spills into both marine and fresh waters continue to occur.
Ideally, none of the spilled oil reaches the shoreline or other environmentally sensitive areas,
because the slick is usually contained with booms and physically picked up with skimmers. The
recovered oil is then either recycled or thrown away.
jUnder the best circumstances, collection efforts account for only about 10 to 15 percent of
the spilled oil. Ultimately, if the oil remained in the open sea, then as much as 30 to 40 percent of
it could evaporate, 50 percent could be biologically metabolized, and about 10 to 20 percent of the
heaviest compounds could form tar balls or sink to mid-depths or the bottom. Within the first few
days of a spill, after much of the oil evaporates, the oil becomes a thick, sticky, difficult-to-handle
residue: As time goes on, it becomes even more difficult to handle. In rough seas, emulsification
will also occur, leaving a water-oil "mousse" (water-in-oil emulsion) which is extremely stiff and
sticky and has a volume up to five times greater than the oil alone; needless to say this mousse is
difficult to remove or otherwise effectively handled.
. Although they have not been used in the United States, chemical dispersants are one of the
tools available in oil spill responses. Dispersants do not help in removing the oil, but promote its
distribution into the water column. They can help to prevent a slick from reaching and adversely
impacting sensitive areas such as wetlands or aquatic breeding areas. At best, however, only
about 30 percent of the spilled oil can be "removed" by dispersants. Dispersants have been used
regularly in European spills, where rougher seas often preclude use of mechanical cleanup
equipment.
OIL SPILL RESPONSE CHEMICALS
As required by federal regulation, the National Contingency Plan (40 CFR 300), only
chemicals that are listed with the U.S. Environmental Protection Agency may be used in U.S.
waters. In 1993, forty-eight dispersant products were listed of which thirty seven were actually
dispersants; the others were surface washing agents.
Dispersants enhance the natural dispersion process by facilitating the formation of small
(less than 20 micron) oil droplets. When this small, the oil droplets rarely coalesce, and the
dispersed oil rapidly becomes diluted in the sea water. The chemical products are surfactant
formulations that have botfi water-compatible (hydrophilic) and oil-compatible (lipophilic) properties.
With a minimum of mixing energy, the surfactants will diminish the tension at the oil-water
interface and promote the breakup of the slick into small oil droplets.
135
-------
DISPERSANT EFFECTIVENESS
Laboratory test results of the effectiveness of the dispersant chemicals on the NCP Product
Schedule range from less than one percent to over 100 percent. Under new NCP regulations,
proposed on October 22, 1993, dispersant products would have to meet a fifty percent
effectiveness criterion before being listed. This is expected to reduce the number of allowable
products. Since very few dispersants are stockpiled and readily available for use, only one or two
dispersants will likely be used in a real spill.
Laboratory tests are routinely used to evaluate the relative effectiveness of chemical
dispersants. Over 25 tests have been developed, and about six are routinely used throughout the
world. Actual performance of a dispersant in a spill is not likely to be the same as that achieved in
the laboratory because of application difficulties, weather conditions, and other factors. Five
simplified field tests are also available, but these are not routinely used; visual observation is
usually the only way to determine effectiveness in the field.
The following greatly influence effectiveness: type of oil, including chemical composition
and viscosity; weathering of the oil; nature of the spill; sea state; sea characteristics, including
salinity; type of dispersant, and dispersant application method.
ENVIRONMENTAL FACTORS
The decision to use dispersants always involves an evaluation of relative risks. One must
weigh the tradeoffs between the potential impact to the intertidal and shoreline communities by the
untreated oils (and the subsequent cleanup operation) and the potential adverse impact to water
column and possibly benthic organisms by chemically dispersed oil. Although dispersants can
reduce the severity of impact to fragile habitats, they may impose temporary stresses in other
offshore areas because of higher short-term exposure to the toxic components of oil. Most
researchers have concluded that the toxicity of modern dispersant-oil mixtures is not any greater
than that of the oil alone.
Use of dispersants (commercial de-greaser) in 1967 on the Torrev Canyon spill off the
English coastline lead to extensive mortalities of animals and algae, with a severely slowed natural
recovery afterwards. Many concluded that dispersants should not be used under any
circumstances. Since then, a number of non-toxic formulations have been developed. Under NCP
regulations, products must be tested for toxicity (using the Revised Standard Dispersant Toxicity
Test) and that data be published for use by spill response personnel.
There are certain environmental habitats in which dispersants should probably not be used.
These include coral reefs and shallow (less than one meter depth) waters, lagoons and atolls, fish
breeding areas, bird habitats, and in sensitive intertidal habitats such as mangroves and salt
marshes.
136
-------
REGULATORY AND ADMINISTRATIVE ASPECTS
The federal law pertaining to oil spill response actions is the Clean Water Act (CWA) as
amended by the Oil Pollution Act (OPA). The National Oil and Hazardous Substances Pollution
Contingency Plan, or more simply, the National Contingency Plan (NCP), was promulgated pursuant
to Section 311 of the CWA to regulate all response activities.
Dispersant use requires preplanning before an incident occurs. Due to weathering,
dispersants will usually not be effective if applied more than 24-48 hours after the spill starts. The
preapproval plan allows for quick action in responding to specific spills. Some regional response
teams (RRTs) are taking the following approach in preapproval plans: designating zones where
dispersants may be used (usually the deeper offshore waters), designating zones where dispersant
use may be tried (applying limited quantities, usually a barrel or two, to see if they work), and
designating zones where dispersants may not be used. In almost all cases, dispersants may not be
used in inland freshwater spills.
Under the NCP, regional and area contingency plans have been developed and are being
updated as the plans for responding to oil spills. Under the proposed NCP regulations, RRTs and
area committees must address the use of dispersants as a response option.
APPLICATION OF DISPERSANTS
One of the most critical factors in determining if a dispersant will be effective in actual use
is whether it is properly applied. Two of the most significant hurdles to overcome are the lack of a
dispersant supply and insufficient availability of application equipment. These should be addressed
in response planning documents.
Ineffective application can result from poor mixing, use of inadequate application
techniques (such as poor targeting and distribution of aerial sprays), the possibility that the oils
were not dispersible, or poor formulation of the dispersant (National Research Council, 1989).
Also, a major controversy concerns how to determine the field effectiveness of the dispersant.
Generally, the following principles apply: apply dispersant without dilution; ensure contact
of the dispersant with the oil; apply dispersant to leading edge of the spill or at the thickest part;
apply dispersants as soon as practical after the spill occurs; apply dispersant as close to the
continuing spill source as possible; do not apply dispersant to oil sheen; observe the action of the
dispersant from both the air and the surface; adjust application rate to achieve optimum
effectiveness; and do not apply dispersants in the immediate vicinity of recovery operations.
Usually, the initial response should be based on an application of 1:10 to 1:20 of dispersant to oil;
this is about 50 to 100 liters per hectare (5 to 11 gallons per acre).
Aerial application is preferred to vessel application because large areas can be covered in a
short time.
137
-------
DECISION PROCESS FOR DISPERSANT USE
Numerous oil spill response decision processes have been described in the literature, often
illustrated in the form of a decision tree. A simplified decision tree has been prepared for use by
spill response personnel. The decision process should be included in the pre-spill planning
documents required by NCP regulations.
PRACTICAL CONSIDERATIONS FOR DISPERSANT USE
Under the federal government's spill response policy, the On-Scene Coordinator (OSC) has
limited authority to take action, unless an action involves human health and safety issues.
Otherwise, various federal and state agencies must reach agreement for a decision to use oil spill
dispersants. Therefore, the ground work needs to be laid in pre-spill planning documents.
Not all dispersant products on the NCP Product Schedule are equally effective. The data
reported are based on laboratory analyses with a single crude oil and the results are not directly
transferable to the field. Further, many dispersants show an extremely low effectiveness (less than
10 percent), even under ideal laboratory conditions.
Two considerations should guide the decision to actually use dispersants: whether there is
a reasonable probability of measurable success and whether there is consensus agreement between
potentially affected parties.
REFERENCE
National Research Council. 1989. Committee on Effectiveness of Oil Spill Dispersants. Using Oil
Spill Dispersants on the Sea. ISBN 0-309-03882-0. National Academy Press, Washington, D.C.
FOR MORE INFORMATION
Daniel Sullivan, U.S. EPA, Office of Research and Development, Risk Reduction Engineering
Laboratory, Edison, New Jersey, telephone (908) 321-6677, FAX (908) 906-6990.
138
-------
NUTRIENT APPLICATION STRATEGIES FOR OIL SPILL BIOREMEDIATION IN THE FIELD
Albert D. Venosa1, Makram T. Suidan2, Brian A. Wrenn2, John R. Haines1
Kevin Strohmeier2', Edith Holder2, arid Loye Eberhart2
; 1U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Drive
• Cincinnati, OH 45268
2University of Cincinnati
Department of Civil and Environmental Engineering (ML 71)
Cincinnati, OH 45221-0071
INTRODUCTION
During the summer of 1994, EPA, in cooperation with the Delaware Department of Natural
Resources and Environmental Control (DNREC), plans to conduct a small-scale field study on the
shoreline along Delaware Bay involving bioremediation of crude oil released in small quantities on 15
plots each measuring approximately 36 m2. The goals of this research project are to obtain sufficient
statistical evidence to determine if bioremediation with inorganic mineral nutrients and/or microbial
inoculation enhances the removal of crude oil contaminating mixed sand and gravel beaches to
compute the rate at which such enhancement takes place, and to establish engineering guidelines on
how to bioremediate an oil-contaminated shoreline. To be certain that the study is carried out property
an engineering study is necessary to characterize the hydraulic conductivity, hydrodynamics, and nutrient
flow on the selected beach. Information from such an investigation will be essential to determine the
proper frequency of application of fertilizer to the plots. This paper discusses the design and conduct of
such a study and presents some preliminary site characterization data. The project was conducted durina
the weeks of November 12 to December 3, 1993. y
Several nutrient application strategies have been put forward to enable proper contact of the
nutrients with the degrading microorganisms. According to hydrodynamic modeling studies conducted bv
researchers from Auburn University, EPA's Environmental Research Laboratory (Athens, GA) and
EPA s Risk Reduction Engineering Laboratory (Cincinnati, OH) [Wise et a/. (1993)], a strategy of nutrient
application was postulated in which nutrients applied in a trench situated at the high tide line to a depth
below the water table would be dissolved by the fresh groundwater and be made available to the
degrading populations by the underlying freshwater lens that ebbed and flowed with the saltwater tides
This method has not .been tested in the field yet. Another strategy is incorporation of a sprinkling system
to apply water soluble nutrients to the contaminated site at regular intervals. This was one method used
in the Alaska project, but it was not optimized from an engineering basis. Both of these methods need to
be tested systematically in the field so that design guidelines for proper application and maintenance of
the right amount of fertilizer to fulfill physiological needs can be developed. The method of testing in this
study consisted of applying a water soluble tracer, lithium nitrate, to both plots via the trench and
sprinkler methods and following the disappearance of tracer with time.
METHODOLOGY
Site Location. The location of the field study is on a sandy and slightly gravelly beach south of
Slaughter Beach, Delaware. The stretch of beach used for this study was located approximately 500 m
south of a locked access gate in Slaughter Beach. The surface morphology consists of a loose upper 25
mm thick layer of smooth gravel ranging in size from 4.75 mm to 19.1 mm lying atop coarse sand havina
a moderately homogenous particle size distribution (approximately 75% of the mass ranging in size frorn
0.25 to 0.425 mm in diameter). The stretch extends uniformly for 3 km and is quite homogeneous The
average distance from the low tide line to the high tide line is approximately 30 to 32 m.
Plot Layout. Two identical plots were set up as shown in Figure 1. Each plot measured 5 m by 10 m
Two types of wells were situated within and outside the vicinity of each plot: piezometers and sampling
wells. The piezometers consisted of black iron rods about 2.5 m in length and 3.2 cm inside diameter
139
-------
(ID) The bottoms of the piezometers were fitted with a specially made fritted brass tip that allowed water
to enter the well filtered of fine sand or peat particles characterizing the deeper zone of the beach. The
piezometers were equipped with pressure transducers connected to a data logger mounted to a wooden
post in back of and between the plots. The pressure transducers were used to measure water head
continuously to provide accurate readings of water levels during tidal cycles.
The sampling wells were constructed of stainless steel and were also about 2.5 m long. Openings of
approximately 3.2 mm ID were drilled into the sides of the wells starting at 15 cm from the bottom tip and
extending upward at intervals of 15 cm over a total length of 1.8 m. Stainless steel tubing of the same
diameter was welded to these openings. The tubing extended inside the wells from the openings to
above the tops of the wells, where plastic tygon tubing was attached for collection of water samples via
syringe The openings in the sides of the wells were covered with a fine-mesh stainless steel screen to
filter out particulate matter that might clog the tubing. Thus, water samples at each depth interval were
totally independent from other water samples, which enabled measurement of tracer concentrations at
one depth without influence from tracer concentrations at other depths.
Tracer Application Systems. A staging area was positioned between and above the two plots. It
consisted of a wooden platform supporting a gasoline-powered generator, an electric 4,200 Uh pump for
delivering tracer to the plots, an 850-L Nalgene plastic reservoir containing the tracer dissolved in 800 L
of fresh water, and the mounting for the data logger described above. The sprinkler system consisted of
50 mm PVC piping arranged in an H-pattern and connected to a feeder pipe leading to the pump.
Sprinkler heads each rated at 400 Uh were connected to the PVC piping at the four corners of the H. The
linear dimensions of the sprinkler system were 5 m x 5 m. The tracer was applied in two steps, once in
the top half of the plot followed immediately in the bottom half.
Approximately 2.5 m landward from the top of each plot, a 10-m wide trench was dug by a back hoe
to a depth of approximately 0.6 m. PVC piping about 75 mm ID was placed in the bottom of the trench.
Holes were drilled into the pipe every 30 cm to allow uniform application of LiNO3 to the bottom of the
•Benchmark
TrMKnee
-5m-
South
Plot
-10m-
o o
olo
6 ""o
North
Plot
5m
Spring low Tide Line
13.* m
Figure 1. Schematic layout of experimental plots.
140
-------
trench. After positioning the pipes in the trenches and connecting to feeder pipes coming out of the
trenches, the trenches were covered over by the sand.
Application of Tracer. The original plan was to conduct two replicate experiments, one involving
the trench and the sprinkler in the north and south plots, respectively, then repeating the identical
experiment with the trench in the south plot and the sprinkler in the north plot. Time and inclement
weather, however, precluded the second experiment. The experiment is planned to be repeated in the
spring.
For the sprinkler plot, 20 kg of LiNO3 was dissolved in 800 L of fresh water (the fresh water was
kindly provided by the volunteer fire department of Slaughter Beach, DE). Fresh water was used to allow
the density of the water after dissolution of the tracer to approximate the density of the natural salty
groundwater. For the trench application, 30 kg was dissolved in the 800 L, because the trench, being 5 m
wider than the plot width, required more tracer for an equivalent amount to reach the desired area of the
plot. The above amounts were estimated by calculating the volume of water occupied by a wedge of
beach 5 m x 10 m in area with a slope of 10% and using a sediment porosity of 30%. Computed volume
was 7.5 m3. The minimum concentration of LiNO3 needed for detection in 7.5 m3 (detection limit of about
0.1 to 0.5 mg/L) was increased by three orders of magnitude to allow for losses through dilution over
time.
The tracer was added twice via the sprinkler system, once in the upper half of the plot and once in
the lower half. The total time of application was approximately 22 minutes (11 minutes per plot half)
Application occurred at low tide on November 22, 1993. Approximately four hours later, two hours before
high tide, the tracer was added to the trench in one application. Total time of delivery was approximately
25 minutes.
Sampling. As mentioned above, small lengths oftygon tubing were attached to the stainless steel
tubing protruding from the tops of the sampling wells. The tubes were color-coded and numbered to
match the position of the holes along the length of the wells. Plastic, two-way, Luer-lock valves were
inserted into the ends of the tygon tubing to allow attachment of 60 mL plastic syringes for collecting
water samples. When samples were collected, the syringe was connected to the valve, approximately 30
to 50 mL of water was collected and wasted, then about 100 to 120 mL of water (2 syringefuls) was
collected and placed into 125 mL plastic screw-capped bottles that had previously been labeled The
sampling schedule consisted of four sampling events during the first two days after application of tracer
(corresponding to the two low tides and two high tides every day). Following that, sampling occurred
during one low tide and one high tide daily. Samples were collected from every well at each sampling
event. Some of the openings were clogged, and others at times contained no water because the water
level was below the location of the opening. Approximately 1,500 samples were collected over the entire
10-day project period.
In addition to the water samples, sediment samples were collected at low tide every day The
sediment samples were collected by inserting a bulb planter into the sand, removing the column of sand
from the subsurface, and placing the sand into a plastic, screw-capped, 500 mL bottle. Six sediment
samples from each plot were removed each time, corresponding to the approximate locations of the
sampling wells shown in Figure 1.
Analyses. Water samples were split as follows: approximately 10 mL of each water sample was
poured into previously acid-cleaned, plastic, screw-capped test tubes, the test tubes frozen under dry ice
and the frozen samples shipped to Dr. Ken Lee of the Canadian Department of Fisheries and Oceans for
nitrate determination. The method used was the cadmium reduction method (1) using an autoanalyzer
Another aliquot of approximately 50 rnL was removed for determination of nitrate in the field by the
hydrazine reduction method (2) The latter was done on only a select few samples to aid in determining
approximately how much tracer was remaining at any time. The remaining 50 to 60 mL were brought
back to RREL-Cincinnati for measurement of lithium by atomic absorption spectrophometry (3) The
sediment samples were also brought back to RREL-Cincinnati for later measurement of lithium in the
pore water remaining in the interstices of the sediment samples
141
-------
Beach Profiles. To aid in accurate measurement of beach geomorphology and its changes over
time, elevation readings were collected daily at each well location on both plots. In addition, the entire
beach profile was measured several times over the coarse of the 10 days of the experiment. These
measurements were taken in accordance with the method of Emery (4). Two wooden profile rods each
152 cm in length marked in cm increments were used to measure elevation. A steel fence post was
inserted by a sledge hammer into the ground on the vegetated dune landward of the beach. This served
as a benchmark to which all noteworthy elevations were referenced. Two persons made the elevation
readings. One remained at the benchmark holding a profile rod while the other stood seaward of the
benchmark with the other profile rod. The distance between the two experimenters was measured by a
metric tape held level between the two points. The experimenter standing at the benchmark would line
up the top of the second experimenter's profile rod with the horizon and read the elevation from his own
rod also lined up on the horizon. Subsequent readings were made as the second experimenter moved
more and more seaward. Permanent elevation readings to the tops of all wells were made once (these
readings were constant), while elevation readings at the base of the wells changed according to weather
conditions and tidal flows.
RESULTS
At the time of this writing, only a limited amount of data were available for presentation in this paper.
The lithium concentration measurements had only been completed for the first eight sampling events
(i.e., the first 47 hours of the experiment) from the sprinkler plot. These results are presented in Figures
2 and 3 in the form of contour plots in a transect along the right side of the south plot. The Y-axis data
are the absolute elevation readings (cm) of each opening in the wells. The X-axis data are the distances
(m) from the top edge of the south plot to the position of the bottom two sampling wells 2.5 m below
(seaward) of the bottom edge of the plot. The contours are isoconcentration lines computed
(interpolated) from the measured lithium levels from the wells at various depths.
1st Sampling Event (Time 0)
Low Tld«, Right Sld«
Figure 2. Isoconcentration profile of lithium in the south plot at time 0.
Figure 2 shows that most of the tracer immediately after application was centered in the top half of
the plot. The highest concentration was focused approximately 80 cm below the top edge of the plot and
15 cm below the mid-point. It must be emphasized that the concentration profile shown represents the
tracer concentrations measured in the water column only. Subsurface samples above the water column
shown in the figure were dry, so no water samples were able to be collected. The figure does not reflect
the level of tracer in the pore water within the interstices of the sand. These measurements were not
available at this time. Figure 3 shows the lithium contours after the eighth sampling event (47 hours or
eight full tidal cycles after application).
142
-------
8th Sampling Event (47 Hours)
Mid Tide, Right Side
0.0 2.5 5.0 7.S 10.8 12.5
Distance, m
Figure 3. Isoconcentration profile of lithium in the south plot at 47 hours after application.
Clearly, the concentrations of tracer not only have not changed much even after 8 full tidal cycles
have elapsed but also the tracer has spread slowly by diffusion throughout the water column to at least
2.5 m seaward from the plot. Other graphs for sampling events 2 through 7 showed approximately the
same picture, with an evolving lens of tracer slowly dispersing seaward and deeper into the beach. The
tracer lens tended to rise and fall with the tides. Only one sample from one of the outer wells (i.e.,
outside and lateral to the plot) from among all those collected during the eight sampling events was
positive for lithium.
CONCLUSIONS
It is difficult to reach firm conclusions at this early stage in the analysis of the results. Only about
half of the tracer results from the sprinkler plot were available at this time, none of the results were
available from the trench plot, no confirmatory nitrate data had been received, and none of the critical
sediment tracer analyses had been completed. Nevertheless, the little information that has been
presented suggests that nutrients will likely exist for relatively long periods of time (days to at least a
week or more) in a beach setting such as exists at Slaughter Beach. Tracer tends to remain in the
groundwater underneath the beach surface and moves up and down with the tides. It migrates mostly in
a vertical direction (landward to seaward) and does not spread laterally to a significant extent. These
results should prove favorable toward bioremediation.
REFERENCES
1. American Public Health Assn. Automated cadmium reduction method 4500 F. JnjStandard
Methods for the examination of water and wastewater, 17th Ed. APHA, Washington, D.C.
2. U.S. EPA. Methods for Chemical Analysis of Water and Wastes. EPA-600/4-79-020, 197.9.
3. American Public Health Assn. Direct air-acetylene flame method 3111B. ln:Standard Methods
for the examination of water and wastewater, 17th Ed. APHA, Washington, D.C.
4. Emery, K. O. "A simple method of measuring beach profiles. Limnol. Oceanog. 6:90-93, 1961.
143
-------
BIOVENTING OF A JET FUEL SPILL IN A COLD CLIMATE WITH SOIL WARMING: A FIELD
EVALUATION
Gregory D. Sayles and Richard C. Brenner
U.S. EPA Risk Reduction Engineering Laboratory, Cincinnati, OH 45268
Robert E. Hinchee and Andrea Leeson
Battelle Laboratories Columbus Division, Columbus, OH 43201
Catherine M. Vogel
U.S. Air Force Armstrong Laboratory, Tyndall Air Force Base, FL 32403
Ross N. Miller
U.S. Air Force Center for Environmental Excellence, Brooks Air Force Base, TX 78235
INTRODUCTION
Bioventing is the process of supplying oxygen in-situ to oxygen deprived soil microbes by
forcing air through unsaturated contaminated soil at low flowrates1. Unlike soil venting or soil
vacuum extraction technologies, bioventing attempts to stimulate biodegradative activity while
minimizing stripping of volatile organics, thereby destroying the toxic compounds in the ground.
Since the bioventing equipment (air injection/withdrawal wells, air blower, and soil gas monitoring
wells) is relatively non-invasive, bioventing technology is especially valuable for treating
contaminated soils in areas where structures and utilities cannot be disturbed, such as at the Hill
AFB site.
The U.S. EPA Risk Reduction Engineering Laboratory, with resources from the Office of Solid
Waste and Emergency Response's program, the U.S. EPA Bioremediation Field Initiative, began a 3-
year field study of in-situ bioventing in the summer of 1991 in collaboration with the U.S. Air Force
at Eielson Air Force Base (AFB) near Fairbanks, Alaska. The site has JP-4 jet fuel contaminated
unsaturated soil where a spill has occurred in association with a fuel distribution network. The
contractor operating the project is Battelle Laboratories, Columbus, OH. With the pilot-scale
experience gained in these studies and others, bioventing should be available in the near future as a
reliable, inexpensive, and unobtrusive means of treating large quantities of organically contaminated
soils. This report summarizes the first 2% years of operation.
METHODOLOGY
Site history, characterization, installation and monitoring were summarized previously.2'3 Fig. 1
shows a plan view of the project. Briefly, four 50 x 50 ft test plots have been established, all
receiving relatively uniform injection of air. The four test plots are being used to evaluate three soil
warming methods:
([} Passive warming: enhanced solar warming in late spring, summer, and early fall using
clear plastic covering over the plot, and passive heat retention the remainder of the year by
applying insulation on the surface of the plot.
(ii) Active warming: warming by applying heated water from soaker hoses 2 ft below the
surface. Water is applied at roughly 35°C and at an overall rate to the plot of roughly
144
-------
0) O
'
-
Ł g
.
w Ł 8
1
X
Ł
w
T O
co
? °1 ° o
COCO HI- t- ^
_^ H "1 •«°"D'''8
O to o Ł0 tocoSS Ko S
1 -™ OS-
0 , , , , • n
0 0 H /\,
• •• s IU
0 " I 1 H
* 1.- » Is- --U? W
1C*.- ^..^i^? \^
Ł ^ ^ A^^- —
bill
tj w (*
5
-------
1 gal/min. Five parallel hoses 10 ft apart deliver the warm water. The surface is covered
with insulation year-round.
(iii) Buried heat tape warming: warming by heat tape buried at a depth of 3 ft and distributed
throughout the plot 5 ft apart. The tape heats at a rate of 6 W/ft giving a total heat load onto
the plot of roughly 1 W/ft2.
(iv) Contaminated control: Contaminated soil vented with injected air with no artificial method
of heating.
The passively heated, actively heated, and control test plots were installed in summer 1991
and the heat tape plot was installed in September, 1992. Air Injection/withdrawal wells, soil gas
and temperature monitoring points are distributed throughout the site (see Fig. 1). Heating of the
actively heated plot was discontinued in July 1993 in order to compare heated and unheated
biodegradation rates at the same location.
Periodically, in-situ respirometry tests4 are conducted to measure the in-situ oxygen uptake
rates by the microorganisms. These tests allow estimation of the biodegradation rate as a function
of time, and therefore, as a function of ambient temperature and soil warming technique. The rate
of oxygen use can be converted into the rate of petroleum use by assuming a stoichiometry of
biodegradation.4 Quarterly comprehensive and monthly abbreviated in-situ respiration tests were
conducted.
Final soil hydrocarbon analyses will be conducted in late 1993 and compared with the initial
soil analysis to document actual hydrocarbon loss due to bioventing.
RESULTS
Evaluation of soil warming methods
Figure 2 displays the average temperature of each plot and at an uncontaminated background
location as a function of time during the study. By applying warm water to the plot, the
temperature of the actively heated plot was maintained in the range of 10-25°C compared to the
contaminated (unheated) control plot where the minimum winter temperature is roughly 0°C.
When heating of the actively heated plot was terminated in July 1993, it's temperature followed
the temperature of the unheated control plot very closely, as expected.
The ability to control temperature in the passively heated plot was not as successful. The
temperature of the passively heated plot roughly mimicked the contaminated control plot
temperature except during the summer 1992 when the passively heated plot was roughly 5°C
warmer than the control plot. The insulation applied during the winter has been marginally
successful at best, providing 1 -2°C temperature elevation in the passively heated plot relative to
the control plot.
Heating by buried heat tape in the surface heated plot has been successful at maintaining
between 10 and 22°C year round. The temperatures achieved in this plot in the summer were
much higher than those maintained in the winter because, although the heat input was constant,
the ambient temperature was much higher in the summer.
146
-------
Rate of Biodegradation
The rate of jet fuel biodegradation, as estimated by in-situ respirometry tests, as a function of
time for each plot is shown in Figure 3. The influence of temperature on the rate is clear: the
actively warmed and surface warmed plots maintained rates 2 to 3 times greater than the unheated
control plot year round. The small difference in temperature between the passively warmed and
the qontrol plots (see Fig. 2) is reflected in the small difference in the respective rates measured in
these plots.
Note that the rate js non-zero (roughly 0.5 mg/Kg/day) in the unheated control plot in the
middle of winter in Alaska when the average temperature of the plot is roughly 0°C (see Fig. 2). It
is commonly believed that bioremediation systems should be shut down for the winter in any cold
climate because it is thought the microbial activity is always zero at these low temperatures. In
this case, the activity is significant.
CONCLUSIONS
I
The application of warm water and heat generated by electrical resistance has been successful
at maintaining summer-like temperatures in the soil year round. The enhanced temperatures in the
plots provided elevated rates of biodegradation. The passively warmed plot has performed only
marginally better than no heating (the contaminated control) with respect to temperature and rate.
At the conclusion of this study, a cost-benefit analysis will be conducted to compare the
performance of the heating methods in terms of rate enhancement versus cost of heating.
REFERENCES
1.
Hoeppel, R.E., R. E. Hinchee and M. F. Arthur, Bioventing Soils Contaminated with Petroleum
Hydrocarbons, J. Indust. Microbiol., 8: 141 (1991).
2. Sayles, G.D., R.C. Brenner, R. E. Hinchee, C. M. Vogel and R. N. Miller, Optimizing Bioventing
in Shallow Vadose Zones and Cold Climates: Eielson AFB Bioremediation of a JP-4 Spill,
Symposium on Bioremediation of Hazardous Wastes, May 5-6, 1992, Chicago IL EPA/600/R-
92/126, August 1992.
3. Leesbn, A., R.E. Hinchee, J. Kittel, G. Sayles, C.M. Vogel and R.N. Miller (1993) Optimizing
Bioventing in Shallow Vadose Zones and Cold Climates, Hydrologiical Sci., 38(4), 283-295.
4. Ong, S.K., R. E. Hinchee, R. Hoeppel and R. Schultz, In-Situ Respirometry for Determining
Aerobic Degradation Rates, in In-Situ Bioreclamation; R. E. Hinchee and R. F. Olfenbuttel, Eds.
Butterworth-Heinemann, Boston, 1991, pp. 541-545.
147
-------
en
ON
ON
(N
ON
ON
ON
ON
(0
OT
(0
CD
I
I
la
m
LJL
I
V
m
c
o
•fi
o
JO
•o .
II
O) M
8
S-i
1 'g
li
o t:
o o
Sg
g'S
TJ
(O
s
a
o
Ł
1
QJ
a.
ai
S
0)
CM
0}
O)
148
-------
CO
2
a
OJ
•a
Ł
3
<0
0)
a
I
Q)
n '^
2*5
§
T3
s
o>
(D
•o
O
!a
H «N
U OS
O ! ^H
i
a>
CD
2
-------
SITE DEMONSTRATION OF PNEUMATIC FRACTURING AND HOT GAS INJECTION
Uwe Frank
US EPA
2890 Woodbridge Ave.
Edison, NJ 08837
(908) 321-6626
Herbert S. Skovronek
Science Applications International Corp.
Hackensack, NJ 07601
(201) 489-5200
John J. Liskowitz
Accutech Remedial Systems, Inc.
Cass Road at Route 35
Keyport, NJ 07735
(908) 739-6444
John R. Schuring
Hazardous Substance Management Research Center
New Jersey Institute of Technology
Newark, New Jersey 07102
(201) 596-5849
INTRODUCTION
Soil vapor extraction (SVE) is a method that has been found to be very effective for the
remediation of soil contaminated with volatile organic compounds (VOCs) and petroleum
hydrocarbons. It has gained ever increasing popularity because it can treat large amounts of soil at
relatively low cost, with some estimates as low as $10 per cubic yard. This compares favorably
with virtually every other remediation treatment technology, with cost estimates ranging from $80
for some forms of bioremediation to well over $1200 per cubic yard for hazardous waste
incineration. Critical to the application of SVE, however, is the ability to achieve adequate vapor
flow through the contaminated soil. SVE is only applicable to sites with soil types that permit the
flow of contaminant vapors through subsurface formations for extraction and eventual, remediation.
It is necessary that air can flow through all of the contaminated soil at a site. Such vapor flow in
the vadose zone depends in part upon soil characteristics such as air permeability, water content,
porosity and soil homogeneity. Relative to SVE, air permeability is the measure of a soil's ability to
transmit fluids based on laboratory or field airflow tests. The density and viscosity of vapors
combined with the permeability of soil significantly influence the ability of the vapor to flow
through subsurface strata. Permeability of soil is usually the single most important soil parameter
to be considered in the successful application of SVE. It is a key parameter not only in deciding if
SVE is a feasible remedial option, but also for establishing SVE system design criteria. SVE is
typically more applicable to soil types with permeability values greater than 10'7 cm/second. This
includes subsurface strata-of gravel, sand, silty sand and some limestone, basalt and metamorphic
rock formations. Sites consisting of igneous rock, shale, clay, dense silt, and glacial till usually are
not amenable to SVE. Consequently, a process called Pneumatic Fracturing Extraction (PFE)(SMI
150
-------
was developed by Accutech Remedial Systems (ARS), Inc. and was evaluated in a SITE Program
demonstration to improve bedrock permeability for SVE remediation. This project was carried out
under a New Jersey Environmental Cleanup Responsibility Act (ECRA) Cleanup Plan for an industrial
site in Somerville, NJ where the soil and groundwater were found to be contaminated with VOCs
primarily trichloroethene (TCE). The Plan called for decontamination of the vadose zone by vapor'
extraction, where the formation was shale with very limited air permeability.
METHODOLOGY
The PFE process may create new or enlarge existing fractures and involves injecting bursts (10
to 20 sec) of compressed air (up to 500 psig) into narrow (2 ft) intervals of one or more wellbores
Each interval is isolated by a proprietary injector unit equipped with packers during pneumatic
injection. The new fractures provide increased connections and an enlarged radius of influence
thus making vapor extraction from the vadose zone more efficient. Vapor extraction can be carried
out from all wells, or with some wells either left open to allow passive air introduction or capped to
direct the subsurface air flow. ^PM«U 10
The demonstration included a series of radially placed, 6-in. diameter monitoring wells and a
central 4-in. diameter fracture well that were drilled to a depth of about 20 ft. Each well was
cased to about 8 ft below land surface (bis) and left uncased for the remaining depth.
Pressures, extracted air flow rates, and TCE mass removal rates were compared before and after
fracturing. A second prefracture test was carried out after a 24-hr dormant period to document the
recharge effect commonly observed when vapor extraction is interrupted and then restarted
Samples of the extracted gas were collected using EPA Method 18 and analyzed on site by gas
chromatography for TCE concentrations. These values were converted to TCE mass removal rates
using the air flow rates. Surface heave during each fracture event was estimated with electronic
tiltmeters.
Pressures and air flow rates also were measured while extracting individually from each
monitoring well to determine whether fracturing had established connections between the fracture
well and the monitoring wells and to provide information on the radius of influence created by
fracturing. The effectiveness of passive air inlet was evaluated by uncapping from one to four
monitoring wells while extracting from the fracture well. Pressures, air flow rates, and TCE
removal rates were then determined.
A second test was also conducted because ARS is also developing a catalytic oxidation system
for the aboveground destruction of chlorinated hydrocarbons, which may by investigated in a Phase
II study. Accutech hypothesizes that further improvements in VOC removal rates can be achieved
by injecting the hot exhaust gas (600 to 1000 °F) into one or more wells. The catalytic unit is not
yet available, consequently a compressor was used in this Phase I study to produce hot air (200 to
250 °F) in order to develop and calibrate a model to evaluate the effects of injecting heated air.
Two experiments were carried out to evaluate the effects of hot gas injection. Heated air
(-200 to 250 °F) was injected into a central well while extracting from one or more monitoring
wells. Temperatures in selected monitoring wells were then measured. Also, pressures, air flow
rates, and TCE mass removal rates were determined for the extracted air.
151
-------
RESULTS
The results from the PFE tests confirmed the developer's claims. A comparison of the 4-hr
postfracture data with the data from the restart test demonstrated an air flow rate increase of
between 400% and 700%, averaging about 600%. Although TCE concentrations after fracturing
were only slightly higher than before fracturing (58 ppmv.vs. 50 ppmv, avg), when coupled with
increased air flow rates the mass removal rate was increased by about 675%.
It was also found that a more complex gas mixture was extracted after fracturing, with higher
concentrations of benzene, chloroform, and tetrachloroethene. Fracturing may have improved
connection with pockets of these compounds, making them more accessible for extraction.
Extraction at each peripheral monitoring well individually before and after fracturing confirmed that
connections were significantly improved even at wells 20 ft from the fracture well, as shown by
extracted air flow rates in Table 1. Attempts were made to determine whether vertical connections
existed or were created by fracturing between adjacent 2-ft intervals, but the data were
inconclusive, probably because of perched water in the vadose zone and the wellbore.
The results of the hot air injection tests were generally too ambiguous to confirm the
developer's claims. During the first hot gas injection test (90-hr), temperatures in the monitoring
wells remained essentially constant over the first 10 hours at approximately 58°F. At that time the
TABLE 1. MONITORING WELL EXTRACTION TESTS
Distance
from
fracture
well (ft.)
10s*
10o/s
10d
10s
20s
7.5 d
20 d
Well
No.
FMW 1
FMW2
FMW 3
FMW 4
FMW 5
FMW 6
FMW 7
pre-fracture
<.62#
<.62-.88
<.62
<.62
<.62
<.88
<.62
Air Flow
scfm
post-fracture
5.15-6.36
6.99-5.22
5.11-9.35
5.7-8.11
5.48-7.46
4.83-7.1
1.94-1.96
Increase
>7.3 -
>4.9-
>7.2 -
>8.2 -
>7.8 -
>4.5 -
>2.1 -
9.2
10.3
14.1
12.1
11.0
7.1
2.2
* s = strike; d = dip; o/s = off strike and dip
# Some prefracture air flows are based on Accutech data.
FMW = Fracture Monitoring Well
thermocouples were raised from below 14 ft bis to 8 ft bis. Elevated temperatures were
immediately observed and continued to increase over the next 10 hour before stabilizing as shown
in Table 2. It is unknown whether one or more thermocouples were immersed in water in the well
at the 14 feet depth.
In addition, only very low concentrations of TCE (~ 1 ppmv) were found in the extracted air.
152
-------
both before and during hot air injection. Even with the increased air flow rates during injection the
calculated TCE mass removal rate actually decreased during hot air injection (Table 3), possibly due
to changes in the subsurface air flow directions when the system configuration was changed from
extraction only to extraction and injection.
A second experiment, lasting 24 hours, was carried out in another area of the site where higher
TCE concentrations were anticipated. A new hot gas injection well and an additional extraction
well were installed. Air was extracted from two wells, each 10 feet from the injection well.
Although initial temperatures in the extraction wells were somewhat higher (-65 to 75 °F) than in
the first hot gas injection test, well temperatures did not increase further in this case Compared to
an extraction-only pretest, TCE mass removal rate did increase about 50%, reflecting both
increased air flow rates and increased TCE concentration in the extracted air as shown in Table 4
Perched water in the wells may explain some of the inconsistencies in the air flow rate and
temperature results from the two hot air injection experiments.
Well
TMW2
TMW4
TMW1
TMW3
FMW4
FMW2
TABLE 2
Distance
from Injection
Well, ft
5
5
7
, 7
10
10
. HOT AIR INJECTION WELL TEMPERATURES
Monitoring Well Tenrmeraturfi, °F
Initial
to 1 0 hr
@ 1 4 f t
59
58
57
58
57
56
After
11 hr,
@ 8ft
72
71
75
75
65
64
After
21 hr,
@ 8ft
77
74
75
76
70
68
After
89 hr,
@ 8ft
75
73
72
74
71
69
TMW = Thermal Monitoring Well
FMW = Fracture Monitoring Well
TABLE 3. HOT AIR INJECTION TEST RESULTS
Test
Pre-hot air
Hot air injection
(one well extraction)
Air Flow,
Inject.
69.3JJ5.4
scfm avg
Extract.
11.64.1-5
55.84.3.4
TCE Mass Removed,
lbx10-6/min
172^18
20.4jt.32.0
Hot air injection
(four wells extraction)
73.0 + 3.4
82.6 + 7.1
31.2JJ0.3
153
-------
TABLE 4. SECOND HOT AIR INJECTION TEST RESULTS
Test
Pre-hot air inject
Hot air Inject
(2 wells)
Air Flow,
Inject.
to 26.1*
scfm avg
Extract.
3.7jM.8
9.2±4.7
TCE Mass Removed,
Ib x 1 0-6/min
63±27
97±33
Increase, %
150
54
'Some data lost due to leak in manifold; measured values ranged from 10.9 to 26.1.
Costs for the PFE demonstration were based on one year of operation, during which 1209 kg
(2,600 Ib) of TCE would be removed. It is estimated at $371,364, equivalent to about $307/kg
($140/lb) of TCE removed. Table 5 provides a summary of the costs and the percent each
subcategory contributes to total cost.
No cost estimate was developed for the effect of hot gas injection; the planned catalytic
oxidation unit was not used as the source of hot gas.
TABLE 5. OPERATING COST OF FULL SCALE PNEUMATIC FRACTURING EXTRACTION
Cost Item
Site preparation
Permitting/regulatory
Capital equipment (1.5 yr)
Startup
Labor salary
Consumables/supplies
Utilities
Emission control
Residuals (water, etc.)
Analytical services
Repair, replacement
Demobilization
Total Cost,
$
42,000
1,750
82,074
8,200
107,640
4,000
17,000
70,000
37,200
N/A
N/A
1,500
Cost/lb TCE,
$/lb
15.79
0.66
30.85
3.08
40.47
1.50
6.39
26.32
13.98
—
—
0.56
% of Total
11.3
0.5
22.1
2.2
29.0
1.1
4.6
18.8
10.0
—
—
0.4
Total
$371,364
139.60
100.0
CONCLUSIONS
For properly selected formations, PFE can significantly improve vapor extraction effectiveness.
154
-------
The nature of the formation, air permeability, moisture content, uniformity, water table, and the
presence of obstacles or potential sources of short circuits must all be considered when evaluating
PFE as a remediation option.
In the demonstration, Accutech's claims were far exceeded: fracturing increased extracted air
flow rates by 400% to 700% and TCE mass removal rates by almost 700% when operating with a
single fracture/extraction well and no air inlet sources. With passive air inlets, the air flow rate and
the TCE mass removal rate after fracturing were increased by 19,000% and 23,000%,
respectively, when compared with the prefecture results. Also, the radius of influence can be
increased significantly by fracturing, with 1^100% to 1,400% increases in extracted air flow rates
in wells at distances of 10 feet and 200% to 1,100% even in wells at 20 feet from the extraction
wells.
The estimated cost for PFE remediation of a site such as that in Somerville, NJ, is $307/kg
($140/lb) of TCE removed. Labor, capital equipment, and emissions control were the three major
cost factors.
The effects of hot air injection were inconclusive. Increases in the temperature of the formation
may be produced if sufficient heat is introduced, but this does not necessarily increase the TCE
mass removal rate.
FOR MORE INFORMATION:
Uwe Frank, U.S. EPA, Office of Research and Development, Risk Reduction Engineering Laboratory
Edison, New Jersey, Telephone (908) 321-6626.
155
-------
CASE STUDY OF THE APPLICATION OF SOIL VAPOR EXTRACTION-
AIR SPARGING TECHNOLOGY TO LEAKING UST SITE
Chi-Yuan Fan and Anthony N. Tafuri
Releases Control Branch
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
2890 Woodbridge Avenue
"Edison, NJ 08837
(908) 321-6635
INTRODUCTION
Cleaning up releases of petroleum hydrocarbons or other organic chemicals from the leaking
underground storage tanks (UST) in the subsurface environment is an increasing problem with over
240,000 confirmed releases. Depending on the distribution and type of contamination, cleanup of the
saturated and unsaturated zones should not be considered as unrelated activities. To develop a
remediation plan for selecting effective in situ treatment system(s), one must properly identify the
locations and phases of contaminants.
When a nonaqueous phase liquid (MAPI) is released from a leaking underground storage tank, it
will be present in 13 different phases-locations in the subsurface environment, and the most likely
locations of the contaminants in the subsurface areas of a site are:
o NAPL and/or mixture of organics and water between soil particles in unsaturated and saturated
zones,
o Gas phase between soil particles in unsaturated zone,
Dissolved in soil pore water in unsaturated zone,
Dissolved In groundwater,
o
o
o Adsorbed on soil particles in unsaturated and saturated zones, and
o Diffused into soil particles in unsaturated and saturated zones.
Mobility, a critical factor in determining pollutants transport in subsurface," is to indicate how readily
Individual constituent of leaked NAPL moves into air and water in the subsurface environment. It is also
used for evaluating the effectiveness of in situ remediation systems in terms of abiotic processes. The
primary indicators in determining mobility are vapor pressure (how easily will the contaminant volatilize
into soil pore space) and solubility (what is the contaminant's affinity for water). For a recent continuous
release, most of the petroleum products are likely to be in the nonaqueous liquid phase, with somewhat
smaller portions existing in the vapor and dissolved phase but will increase over time. On the other
hand, a large portion of the weathered petroleum hydrocarbons have already evaporated in soil vapor
and dissolved in pore water and groundwater with residual fractions being sorbed on and diffused into
soil/rock particles. The latter two phases would be less mobile but may act as a continuing source over
a fixed time.
Soil vapor extraction-air sparging (SVE-AS) systems combine in situ aeration of groundwater with
vacuum enhanced recovery of hydrocarbon-laden vapors that are present in the unsaturated zone. The
injection of air into the saturated zone functions both to volatilize residuals of petroleum hydrocarbons
and to increase the dissolved oxygen concentration in groundwater. Enhancement of dissolved oxygen
may in turn promote biodegradation by increasing the population of indigenous microbes. Because of
the dual action of sparging and soil vapor extraction, under certain geological conditions and proper
engineering design, the system may provide effective means for immediate abatement of dissolved
contaminants in the saturated zone and contaminant vapors in the unsaturated zone.
156
-------
Figure 1 presents a schematic diagram of these processes. As shown in this figure, air is sparged
into the groundwater, promoting partitioning of volatile organic compounds into the soil pore spaces
within the sparged air streams. The air streams rise to the groundwater table, and the released
hydrocarbon-laden air is removed by the SVE system. If volatile hydrocarbons are present in the
unsaturated zone, the SVE system will also function to remove these compounds by volatilization and
advection. Many practitioners claim to effectively remediate hydrocarbons contaminated soils in both
saturated and unsaturated zones by using SVE-AS system, however, limited performance data support
their claims.
This paper presents an approach for evaluating the effectiveness of SVE-AS technology with an
actual leaking LIST case study that system evaluations are being conducted at the site. The site is locat-
ed in Cleveland, Ohio and evaluated by a joint effort by the BP Oil Company, the Ohio State Fire
Marshall Bureau of Underground Storage Tank Regulations, U.S. EPA Region V Office of Underground
Storage Tanks, and U.S. EPA Risk Reduction Engineering Laboratory. A site specific Quality Assurance
Project Plan (QAPP) was developed by the EPA to ensure the validity of data collected from an existing
SVE-AS system installed at the site.
METHODOLOGY
1. Case Study Site Description - An SVE-AS system has been installed and operated at an active
gasoline retail-service station as indicated in Figure 2. The surrounding area is largely residential, with
light commercial development. Geologically, the site is located upon the Eastern Lake and Till Plains of
the Central Lowland Province. The soil is composed of sandy loam, silt, and clay and is characteristic of
a remnant beach ridge. Depth to first groundwater table at the site is approximately 19 feet below
grade.
Groundwater flow direction is towards Lake Erie to the north. The groundwater gradient at the site
is approximately 0.04 foot/foot. In situ aquifer testing has not been performed at the site. Based upon
the lithology, the site hydraulic conductivity is likely to be highly variable.
The current system operation provides remediation across the east half of the site. The locations of
SVE-AS wells, monitoring wells, and soil borings are indicated in Figure 2
2. SVE-AS Processes Evaluation Experimental Design - For the purpose of the initial remediation
activities at the site, the radius of. influence for each vapor extraction well was projected to be 30 feet",
and the radius of influence for each air sparging well at 15 feet. The system was designed such that the
influence of the air sparging well would be encompassed completely by the influence of the extraction
well. The well configuration consists of two remediation cells. Each cell contains one central SVE well
surrounded by three air sparging wills.
The SVE portion of the remediation system consists of two SVE wells manifolded to a 5-horsepow-
er regenerative blower. The SVE wells are constructed of 10-foot sections of 4-inch PVC continuously
wound well screen. Twenty-two vapor monitoring points were installed on site to monitor the area of
influence of both the air sparging and soil vapor extraction systems.
The AS portion of the remediation system consists of six air sparging wells manifolded to a 15-
horsepower, oil-free air compressor. Air sparging monitoring wells provide data used to determine the
operating effectiveness of the AS system.
3. SVE-AS System Performance Monitoring - The purpose of monitoring is to determine the amount
and movement of pollutants in the subsurface environment before, during, and after remediation. An
effective monitoring program includes the design of a reliable well network to ensure a complete
determination of and assessment of the site conditions.
157
-------
Vapor monitoring - To evaluate the performance of the SVE-AS systems, periodic field sampling
and data collection activities are necessary. Vapor monitoring of the remediation system at the site is
conducted at the SVE wells, vapor monitoring points (VMPs), and the SVE stack emission.
Groundwater monitoring - Groundwater monitoring at the site is performed at the groundwater
monitoring wells (MWs) and the air sparging monitoring wells (ASMs).
Soil Sampling - Three soil sampling events are scheduled to be collected during the field demon-
stration activities. These samples are collected and analyzed for BTEX and TPH. Results are compared
with the samples collected prior to the system start-up and at the conclusion of the field evaluation to
assess the effectiveness of the technology for removing contaminants from the soil.
4. SVE-AS Systems Operation Monitoring - Several process control parameters are periodically
measured to determine the adequacy of system operation as follows:
Soil vapor flow and temperature,
Ambient conditions (air temperature and barometric pressure),
System vacuum/pressure,
Soil vapor Oxygen and Carbon Dioxide,
Soil vapor volatile organic compounds,
Injection air pressure and flow rate,
Groundwater table elevation,
Groundwater dissolved oxygen,
Groundwater temperature, conductivity, and pH.
RESULTS
1. Soil Vapor - After two years of operation, the vapor concentrations from all four extraction wells
have began to reach asymptotic mass removal rates. Since January 1993, the average constant levels
of total BTEX/TPH from VEW-1, VEW-2, VEW-3, and VEW-4 were 20/200, 10/50, 4/10, and 10/60,
respectively.
2. Groundwater - Results from laboratory analyses indicated TPH levels in the groundwater samples
were reduced from 40 mg/l (ppm) to below laboratory method detection limits during the first two
months of air sparging operation. Total BTEX concentrations reduced from as high as 650 ppb to less
than 3 ppb in January 1992, and the concentrations have been maintained at these levels with
intermittent operation of SVE-AS systems.
3. Soil - A preliminary assessment of the SVE-AS systems performance in terms of changes con-
taminant levels in the subsurface soils in silty clays in the unsaturated zone was made by comparing of
soli samples collected in three different stages (i.e. August 1991, October 1992, and August 1993).
Contaminants concentrations in soil samples varied considerably throughout the site. The average
values of total BTEX/TPH concentrations in the initial soil samples ranged from 15/280 to 420/140 ppm
with the highest reading of 1088/320 ppm from the VEW-1 vicinity. After 13 months of SVE and 12
months of AS system operation, soil borings were collected, and the measured BTEX/TPH levels were in
the range of 0.54/62 to 30/63 ppm. However, in the August 1993 soil sampling survey, the averaged
values of BTEX/TPH ranged 0.005/42 to 228/125 ppm in the clay-rich soils in the unsaturated zone.
CONCLUSIONS
Based on initial results, hydrocarbons levels in the subsurface vapor and groundwater have been
decreased to below detection limits. However, contaminants concentrations in soil samples have
decreased substantially from initial concentrations but have not reached the State cleanup levels. In
158
-------
addition, contaminants concentrations varied considerably due to the complexity of lithology of the site.
All the data obtained during the two-year monitoring program are being evaluated.
ACKNOWLEDGEMENT
The authors extend their sincere thanks to the following persons for their contributions of this field
evaluation project: Gerald Phillips and Gilberto Alvarez of U.S. EPA Region V; James R. Rocco and Peg
Chandler of BP Oil Company; Ray Banary, Lorie Beabes, and Greg Jones of Engineering Science; and
Roy Chaudet, John Morse, and Tom Clark of IT Corporation.
Figure 1 - Soil Vapor Extraction-Air Sparging System Schematic
159
-------
S
5
I1
03
I
ai
CO
6
CO
CM
160
-------
PARTICLE SEPARATION (SOL WASHING) PROCESS
FOR THE TREATMENT OF CONTAMINATED SOILS
Peter Wood, Mike Pearl, Steve Barber
Gary LeJeune, Ian Martin
Warren Spring Laboratory
Gunnels Wood Road
Stevenage
Hertfordshire
SG1 2BX
United Kingdom
011-44-438-741122
INTRODUCTION
Warren Spring Laboratory has developed a soil separation and washing process under the SITE
Emerging Technology Program. This paper describes pilot scale test work conducted on soil at a gas
works that was contaminated with PAHs, petroleum hydrocarbons and heavy metals. The soil from the
gas works was investigated in the laboratory to characterize the distribution of contamination on the
basis of soil particle size, density, magnetic susceptibility and hydrophobicity (1, 2, 3, 4). This
information was then used to design the pilot scale flow sheet.
The main contaminants in the soil were: arsenic (> 30 mg/kg), lead (> 800 mg/kg), PAH (<200
mg/kg) and petroleum hydrocarbons (> 600 mg/kg).
METHODOLOGY
The pilot scale circuit was operated for a period of some four weeks. The nominal throughput rate
was 1 ton per hour with about 40 tons being treated. The soil was treated as a water based slurry with
minimal reagents used as additives. The overall flowsheet that was tested is shown in Figure 1. This
includes all of the various options that were investigated. The major operations involved are size
separation, washing and attrition, separation based on combined size/density differences, flotation and
magnetic separation. The contaminated soil is initially screened at 50 mm under high-pressure water
jets. The jets help break up any agglomerated material and wash off adhering fine particles from
coarser particles. The > 50 mm material is anticipated to be a clean gravel product.
The < 50 mm material enters a washing mill containing granite boulders. Here further washing
takes place, and the boulders assist in the breaking up of any clay balls. The overflow from the
washing mill is screened at 1 mm again under a water jet to remove any adhering fine particles from
coarser particles. The < 50 - > 1 mm material is then further screened at 10 mm to produce two
products: < 50 - > 10 mm, which is expected to be a clean fine gravel; and a < 10 - > 1 mm
product containing some contaminated carbonaceous (primarily coal/coke) particles, which will be
stockpiled for further treatment by jigging. Jigging is a process by which particles of different specific
gravity undergo vertical stratification,
Tr^e < 1 mm material passing the 1 mm screen enters a primary hydrocyclone operating to
separate at approximately 10 |im. The overflow from this hydrocyclone will be a highly contaminated
fine (< 10 jim) product which will require further treatment (for example by bioslurry treatment) or
161
-------
product
> 50 mm product
product
10mm
treated product
Magnetics
concentrate
Organics
concentrate
Metals
concentrate
1 Alternative option is to use spiral classifier
2 Alternative option is to use multi-gravity separator
Figure 1. Pilot Test Flowsheet for Treatment of Gas Works Soil.
162
-------
disposal. The underflow from the hydrocyclone will be < 1 mm - > 10 jim and enters a spiral
classifier operating as a densifier. The densifier stage is necessary in order to increase the solids
concentration of the coarser material entering a later attrition stage. Any fines incorporated in the water
removed at the densifier stage enter a secondary hydrocycioning stage.
Within the attrition scrubber the attrition action will remove any surface contaminants from the
coarser-particles. Dispersants are added to maintain removed contamination in suspension and to
discourage the finer particles from re-adhering to particle surfaces. Following attrition, the material
undergoes a separation stage based on the combined parameters of size and density. Both a spiral
and a hydrosizer were investigated. With a hydrosizer, three product streams result: a stream with the
coarsest and heaviest particles; a stream with the finest and lightest particles; and a stream with particles
of intermediate density and size. Trie coarse and heavy particle stream from the hydrosizer will
essentially be coarse sand and will be stockpiled. If chemical analysis shows this material is not
sufficiently clean then more treatment by a further gravity separation process may be necessary.
The fine and light particle stream from the hydrosizer will consist primarily of fine particles of clay
and contaminant liberated by the attrition process and also lighter (but possibly larger) carbonaceous
particles that may have adsorbed organic contamination. However, as the separation principle of a
hydrocyclone does not result in products of a definite specification, then some of the particles in this
stream will be transitional in size/density with the particles in the intermediate stream. Consequently, as
it may be possible to clean these particles by flotation, the fine and light particle stream is taken to a
secondary hydrocyclone (where it is combined with the fines from the densifier) so that they can be
removed as underflow prior to disposal of the overflow as a contaminated fine product that will require
further treatment or disposal. It is not sensible for the fine material to undergo a floatation process,
because excessive fines require large quantities of reagents. The fines can also coat large particles,
thus inhibiting collector attachment to particle surfaces.
The stream from the hydrosizer with particles of intermediate density and size will include particles
of similar hydraulic density but varying physical properties. For example, some particles may be
relatively larger but lighter (e.g. coke or other carbonaceous material that may have high levels of
contamination) while other are relatively smaller but heavier. Consequently, as tests have shown that
material > 500 jim is difficult to float successfully, this stream is screened at 500 iim to remove any
remaining > 500 urn particles as a contaminant concentrate for further treatment or disposal.
The < 500 urn material from the screen is combined with the underflow from the secondary
hydrocyclone. More dispersant is added to minimise the coating of larger particles by smaller particles
and so avoid any adverse effects on the performance of the remaining processes. This material then
enters an optional stage of magnetic separation where a heavy metal contaminated product may be
removed that may be suitable for recycling. Or it may enter directly into a flotation stage designed to
remove organic contamination. At this flotation stage, a collector reagent is added to remove
selectively organic contamination as a froth concentrate for further treatment or disposal while the
remainder of the material proceeds to the final process.
The final process needs to treat any metals remaining in the material, and both flotation and a
gravity separation process (using a multi-gravity separator) were investigated. With flotation, another
collector reagent is added to remove selectively metal contaminants as a froth concentrate for further
treatment or disposal while the remaining material represents the treated product.
163
-------
RESULTS
Samples of process streams were collected around the circuit to make possible a mass balance of
both contaminants and soil. All products were included in this sampling regime. These products are:
* > 50 mm clean gravel product
* 10 - 50 mm clean fine gravel product
* 10 -1 mm product that may require further treatment by jigging
* < 10 jim contaminated product from the hydrocyclones that may be treatable by bioslurry process
* < 1 mm coarse sand from the hydrosizer that may require further treatment by gravity separation
* < 1 mm - > 500 urn contaminated light gravity concentrate from screen after hydrosizer
* froth concentrate with organic contamination
* froth concentrate with metal contamination
* magnetic concentrate with metal contamination (optional)
* < 1 mm treated material.
At the time of this writing, the results of chemical analysis are still coming in. Data reduction and
evaluation will provide a material balance around the complete flowsheet for soil, individual
contaminants, and process water. This will enable performance determination for the complete process
as well as for individual unit operations. Additionally, the percentage removal for each contaminant in
any clean products will be determined by comparison with the original soil.
REFERENCES
1. Wood, P. and Pearl, M. Characterisation and Treatment of Contaminated Soils using Mineral
Processing Techniques, |n; Nineteenth Annual RREL Hazardous Waste Research Symposium -
Abstract Proceedings. EPA/600/R-93/040, U.S. Environmental Protection Agency, Washington,
DC, 1993, 192-196.
2. Pearl, M. and Wood, P. Separation Processes for the Treatment of Contaminated Soil. ]n:
Contaminated Soil '93. Kluwer Academic Publishers, Dordrecht, The Netherlands, 1993,1295-
1304.
3. Pearl, M., Wood, P., Martin, I., Barber, S., LeJeune, G. and Bardos, R. Particle Separation
Techniques for the Treatment of Contaminated Soils. Third Annual Symposium on Groundwater
and Soil Remediation. BIOQUAL, Quebec, 1993, 269-284.
4. Pearl, M., Wood, P., Martin, I., Barber, S., LeJeune, G. and Bardos, R. Using Separation
Processes from the Mineral Processing Industry as an Enabling Technology for Soil Treatment -
Laboratory and Pilot Plant Study: Interim Report. Presented at: International Conference of the
NATO/CCMS Conference on 'Demonstration of Remedial Action Technologies for Contaminated
Land and Groundwater' Phase II, Quebec City, September 1993.
164
-------
SITE PROGRAM DEMONSTRATION OF IN SITU STEAM ENHANCED RECOVERY PROCESS AT THE RAINBOW DISPOSAL
SITE IN HUNTINGTON BEACH. CALIFORNIA'
Paul de Percin
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Office of Research and Development
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513) 569-7797
Kyle Cook, Ruth Alfasso
Science Applications International Corporation
10240 Sorrento Valley Road, Suite #303
San Diego. CA 92121
(619) 552-7805
TECHNOLOGY DESCRIPTION
Steam Enhanced Recovery Process (SERP) is designed to remove volatile compounds such
as halogenated solvents and petroleum hydrocarbons, and semi volatile compounds from contami-
nated soils in situ. The vapor pressures of most contaminants will increase by the addition
of steam, causing them to become more volatile and mobile. The technology operates through
wells drilled in the contaminated soil. Injection wells deliver high pressure steam (15
psig, 9Q% dryness) to the soil, while extraction wells draw a vacuum on the soil. The
pressure gradient drives the steam, water, and vaporized contaminants to the extraction wells
where they can be removed for disposal or recycling. Figure 1 illustrates the operation of
the process beneath the soil surface.
A site to be treated with in situ SERP must have predominantly medium to high
permeability soils. A geological confining layer below the treatment depth and a confining
layer above the treatment zone help to contain the flow of steam. Injection and extraction
wells are arranged 9n the site in a pattern designed to promote even distribution of the
steam. Site-specific factors determine the number of wells used, their arrangement on the
site, their construction, and the above-ground process equipment to be used.
Extraction of Vapors
And Liquids
(to treatment system)
Injection of High Quality Steam
(from boilers)
EXTRACTION
WELL
DEPTH,
INTERVAL OF
TREATMENT
nem s
SOIL SURFACE
Vapor, Water and Contaminants Steam Front (moving through soil)
(CONTAMINATED SOIL)
i
INJECTION
WELL
INJECTION
INTERVAL
CONFINING LAYER (e.g., clay layer)
Figure 1. Conceptual Operation of SERP Below the Soil Surface
165
-------
A full-scale SERP system with thirty-five injection wells and thirty-eight extraction
wells was used to treat an acre of soil up to 40 feet deep at the Rainbow Disposal site in
Huntington Beach. California. This site was contaminated with diesel fuel compounds. Figure
2 shows the above-ground process equipment used for the SERP system. Water for steam
generation was pumped from an on-site deep water well, treated by ion exchange and the
addition of chemicals, and heated in one of two natural gas-fired steam boilers. High
pressure steam was delivered to all the injection wells using a manifold system. Air-lift
pumps were used to remove accumulated oily water from the extraction wells, and a vacuum pump
maintained a negative pressure on the soil and removed the vapors from each extraction well.
Heated liquids (condensate) from the extraction wells were routed to a heat exchanger
used to pre-heat the boiler feedwater and then treated in a gravimetric oil/water separator.
The oily product from the separator was collected in a storage tank. The remaining water was
then treated further using filtration and activated carbon before being discharged directly
to an underground sewer. Liquids entrained in the extracted vapors were removed using a
knock-out drum. The vapor stream was then treated in a thermal oxidizer unit which used
electrical heating to oxidize the vapors before discharging them to the atmosphere through a
stack.
The full-scale SERP system used at the Rainbow Disposal site was operated for
approximately two years. The system was operated for 16 hours a day, 5 days a week for the
first year. Twenty-four hour operation, 6 days a week was used for the remaining year.
During most of the treatment, steam injection and vacuum extraction were used simultaneously.
Vacuum extraction alone was used when the boilers were inoperable, during interim soil
Natural Gas
_ From City
High Quality
Steam
Wi
Soft
\
^ ' ' "" ^"t Bki
^ •*
««r Boiler Feed Tank
snera Steam Boiler
Heat Recovery •
Heat Exchangers A
WeU Water Stack ,->-.
1 . . ... Vacuum
Blower
Thermal Oxidizer Unit II
V Knock-Out _^— _
ITlAIr Cooled Drum "^
-------
sampling activities, and at the end of the remediation. Vacuum extraction alone after steam
injection is expected to dry and cool the soil and remove contaminants from lower
permeability soils.
DEMONSTRATION METHODOLOGY
The U.S. EPA Superfund Innovative Technology Evaluation (SITE) Program became involved
with the in situ SERP technology developer after the system was installed on the Rainbow
Disposal site. Pre-treatment soil sampling and analyses occurred prior to full SITE Program
participation. Therefore, the focus of the Demonstration was on the condition of the soil
after treatment to determine if the technology met the site-specific cleanup criteria of
1,000 ppm of total petroleum hydrocarbons (TPH, diesel). An economic analysis of the system
was also a primary objective for the Demonstration.
Before treatment, one to four samples from each of twelve boreholes within the defined
perimeter of contamination were sampled by the technology developer and analyzed for TPH.
Some pf the soil samples were also analyzed for benzene, toluene, ethylbenzene, and xylenes
(BTEXO, and semivolatile organics. The pre-treatment results will be used by the SITE
Program in a non-critical evaluation of the removal efficiency of the treatment.
The SITE program (EPA) performed post-treatment soil sampling and analyses, including
sampling from boreholes adjacent to (within four feet) and at the same depths as those
sampled before treatment. Additional samples at other depths were collected in most of these
boreholes. Twelve additional boreholes were also sampled, including six outside the defined
perimeter of contamination for a total of 24 boreholes. Samples collected after treatment
were analyzed for TPH, BTEX, and total recoverable petroleum hydrocarbons (TRPH). Six sets
of triplicate samples were collected at randomly determined sampling locations to assess soil
inhomogeneity. The data collected from post-treatment sampling and analysis will be used in
a geostatistical model to determine the likely distribution of contanrination remaining in the
soil and the statistical significance of the results.
' A detailed economic analysis of this full-scale technology application will be
performed utilizing monitoring data (i.e., water, chemical, and gas usage; waste generation:
and maintenance needs). This data was collected by the developer with oversight by the SITE
Program during the course of operation. This analysis will focus on the actual costs of the
full-scale remediation as well as theoretical costs at another site.
Key findings from the demonstration, including complete analytical results and the
economic analysis, will be published in an Innovative Technology Evaluation Report. This
report will be used to evaluate the in situ SERP technology as an alternative for cleaning up
similar sites across the country. Results will also be presented in a SITE Technology
Capsule and a videotape. . :
FOR FURTHER INFORMATION
EPA Project Manager:
Paul de Percin
U.S. Environmental Protection Agency
Office of Research and Development
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513); 569-7797
167
-------
ELECTROKINETIC SOIL REMEDIATION- A PILOT SCALE STUDY
Yalcin B. Acar and Akram N. Alshawabkeh
Civil Engineering Department
Louisiana State University
Baton Rouge, LA 70803
Phone: (504) 388-8638
Robert J. Gale
Department of Chemistry
Louisiana State University
Baton Rouge, LA 70803
Robert E. Marks
ELECTROKINETICS Inc.
The Louisiana Business and Technology Center
Suite 102, LSU, South Stadium Drive
Baton Rouge, LA 70803
(504)388-3992
INTRODUCTION
The demand to develop innovative and cost-effective in-situ remediation technologies in waste
management stimulated the effort to employ conduction phenomena in soils under an electric field to
remove chemical species from soils (1 -15). This technique variably named as electrokinetic remediation,
electro-reclamation, electrokinetic soil processing, electro-chemical decontamination or electrochemical
soil processing uses low-level DC in the order of mA/cm2 of cross sectional area between the electrodes
or an electric potential difference in the order of a few volts per cm across electrodes placed in the ground
in an open flow arrangement. The groundwater in the boreholes or an externally supplied fluid
(processing fluid) is used as the conductive medium. Open flow arrangement at the electrodes allows
ingress and egress of the processing fluid or the pore fluid into or out of the porous medium. The low-
level DC results in physico-chemical and hydrological changes in the soil mass leading to species transport
by coupled and uncoupled conduction phenomena in the porous media. Electrolysis reactions prevail at
the electrodes. The species input into the system at the electrodes (either by the electrolysis reactions, or
through the cycling processing fluid) and the species in the pore fluid will be transported across the
porous media by conduction phenomena in soils under electric fields. This transport coupled with
sorption, precipitation and dissolution reactions comprise the fundamental mechanisms affecting the
electrokinetic remediation process. Extraction and removal are accomplished by electrodeposition,
precipitation or ion exchange either at the electrodes or in an external extraction system placed in a unit
cycling the processing fluid (16-18).
Electrokinetic remediation technology has recently taken significant strides. Geokinetics corporation of
Netherlands has reported some field studies (12,19) and Electrokinetics Inc. of Baton Rouge, USA has
completed large-scale pilot studies, is conducting a field study expected to lead to a demonstration
study under the SITE program. This paper presents a summary of the results of some of the pilot-scale
studies conducted by Electrokinetics Inc. in collaboration with Louisiana State University. The main
objectives of the pilot-scale studies were to assess the efficiency of the technique when bench-scale
testing conditions are up-scaled to size and geometry more representative of field conditions and to
evaluate the techgnique by conducting a well-documented large-scale study with proper quality
168
-------
assurance. A second objective was to use the results of these studies to compare with the results of the
numerical implementation of a theoretical model developed at Louisiana State University.
METHODOLOGY
Two pilot-scale tests are conducted using kaolinife spiked with lead at initial concentrations of 850
mg/kg, 1,500 mg/kg and another with using fine sand and kaolinite mixture spiked with lead at 5,000
mg/kg. Figure 1 presents a schematic diagram of the container used for the pilot-scale study . The
dimensions of the container are 91 cm by 182 cm in width and length and 91 cm in depth. These
dimensions are selected in an attempt to minimize boundary effects, simulate one-dimensional transport
conditions and represent an intermediate electrode spacing between bench-scale and full-scale field
implementation. Inert graphite rods are used both for the anode and the cathode. Constant current
conditions are employed in testing . Tensiometers are installed across the electrodes to measure
development of pore pressures. Voltage probes allowed development of electrical potential gradients.
Specimens are taken across the electrodes in time to record the changing chemistry.
The kaolinite used had lead adsorption capacity of about 1,100 mg/kg. Lead nitrate salt is used as the
source of lead. Tap water is used both as the catholyte and the anolyte. Positively charged species
prematurely precipitate close to the cathode when the chemistry of the electrolyte at the electrodes is not
altered or controlled (unenhanced electrokinetic remediation) . Acar and Alshawabkeh (2) discuss the
need for enhancement and propose different enhancement techniques which prevent the encountered
precipitation. An enhancement technique is not used in the pilot-scale studies reported herein since the
specific tests also served in evaluation of a theoretical model of electrokinetic remediation. Other details
of the testing are presented by Alshawabkeh (20). Enhancement techniques and enhanced pilot-scale
studies using a soil from a lead contaminated site are ongoing at Electrokinetics Inc.
PLAN SECTION
CATHODE
4RRAY ~
ANODE
ARRAY"
CATHODE
ARRAY —
-|-*"Ov o o c. o
4 at 12 in = 48 in
7 in
35 in
7 in!
35 in
Electrodes
C: Thermocouple
S : Sampling Location
T : Tensiometer
V : Voltage probe
Figure 1. A Schematic Diagram of the Pilot-scale Test Set-up and the Data Collection Locations
169
-------
RESULTS
Thedetailed results of this study are presented by Alshawabkeh (20). Some findings are herein
presented. Figure 2 presents a contour diagram of lead distribution across the electrodes after 2952
hours of testing at 130 |xA/cm2-. An apparent electrical gradient of 2 V/cm was attained. The energy
expenditure in this test was close to 700 kWh/ms of soil processed. The power expenditure was almost
constant at 270 W/m3 after an initially lower power expenditure period. The apparent electrical conductivity
was about 70 nS/cm. The precipitation close to the cathode is confined to the lastIO cm of length. The pH
outside the zone of precipitation increased from about 2 to 4. The pH within the zone of precipitation was
mostly 5 and the it increased to 10 to 11 in the catholyte. There was not any electroosmotic flow in this
test. Excluding the zone of precipitation, lead was remediated by 98 %. Remediation at discretepoints
outside the last 10 cm ranged between 90 % to 98 %. Final lead concentrations across the soil ranged
between 30 mg/kg to 70 mg/kg with some points as high as 150 mg/kg. 70 % of the lead was in the last 2
cm and 15 % precipitated on the filter at the cathode compartment.
The acid front generated at the anode flushes across the soil mass and meets with the base front
generated at the cathode at a zone close to the cathode compartment. The lead adsorbed on the soil and
in the pore fluid moves across the soil mass by ionic migration as there is no electroosmotic flow. At the
zone where the low pH front meets the high pH front, lead precipitates. Most of the electrical resistance
across the electrodes is due to the low conductivity developed within this zone due to the merging pH
fronts. Power expenditure is mostly due to resistance offered by this zone. Enhancement by
depolarization of the cathode reaction prevents development of this zone and hence leads to even lower
Cathode
Anode
Cathode
Concentration
Ratio C/Co
80-
°>
CO
w
70-
60-
50-
40-
30-
20-
10-
nitial Concentration
CO = 1533 Ing/kg
I I
20 - 40 60 80 100
Axis across the electrodes (cm)
i
120
140
Figure 2. Contour diagram of lead removal across the electrodes in unenhaced electrokinetic remediation
of lead spiked Georgia kaolinite
170
-------
power expenditures (4). However, the acid used in depolarization is then an additional cost to the
process.
CONCLUSIONS
These studies demonstrate that the efficiency achieved in bench-scale studies in electrokinetic
remediation is attainable at the pilot-scale. It is demonstrated that lead is efficiently removed by 90 % to 98
% across the electrodes even when there is no electroosmotic flow across the medium. In most
contaminated soils and site conditions, ionic migration will be the predominant transport mechanism in
electrokinetic remediation (2). As in the unenhanced bench-scale tests, precipitation occurred very
close to the cathode with the increase in pH. Most of the energy expenditure is due to formation of this
low resistivity zone.
These results together with ongoing pilot-scale studies and studies investigating different
enhancement techniques demonstrate that the efficiency and cost-effectiveness of the technique in
remediation depend upon the complexity of contamination encountered together with employment of
an effective enhancement technique. A comprehensive chemical speciation/characterization and site
specific bench-scale and pilot-scale treatability studies may be necessary in field implementation of the
technique at a specific site.
REFERENCES
1. Acar, Y B. Electrokinetic Soil Processing; A Review of the State of the Art . jn: Proceedings of the
ASCE Grouting Conference, ASCE Special Publication No. 30, v. 2, pp. 1420-1432, 1992.
2. Acar, Y.B., Alshawabkeh, A. Principles of Electrokinetic Remediation . Environmental Science and
Technology, vol. 27, n. 13, pp. 2638-December 1993.
3. Acar, Y. B., Alshawabkeh, A., Gale, R. J. Fundamentals of Extracting Species from Soils by
Electrokinetics. Waste Management, Pergamon Press, New York, 12 (3): 1410-1421 , 1993.
4. Acar, Y.B., Puppala, S., Marks, R., Gale R.J., Bricka, M. An Investigation of Selected Enhancement
Techniques in Electrokinetic Remediation, Report presented to US Army Waterways Experiment Station,
Electrokinetics Inc., Baton Rouge, Louisiana, 1993, 160 p.
5. Acar, Y.B., Gale, R.J., Ugaz, A., Marks, R., Feasibility of Removing Uranyl, Thorium and Radium from
Kaolinite. Im Proceedings of 19th Annual RREL Hazardous Waste Research Symposium, EPA/600/R-
93/040, 161.
6. Acar, Y. B., Li, H., Gale R. J. Phenol Removal from Kaolinite by Electrokinetics, Journal of Geotechnical
Engineering, v.118, No. 11, pp. 1837-1852, 1992.
7. Acar, Y. B., Gale, R. J., Ugaz, A., Puppala, SM Leonard, C. Feasibility of Removing Uranium, Thorium
and Radium from Kaolinite by Electrochemical Soil Processing. EK-BR-009-0292, Final Report-Phase I of
EK-EPA Cooperative Agreement CR816828-01-0, Electrokinetics Inc., Baton Rouge, Louisiana, 1992,
243 p,
8. Acar, Y. B., Gale R. J., Leonard, C. QAPP-Phase I; Project EKSITE. Quality Assurance Project Plan
171
-------
submitted to Office of Research and Development, RREL, Electrokinetics Inc., Baton Rouge, Louisiana,
EK-BR-002-0390, 1990, 126 p.
9. Acar, Y. B., Gale, R. J., and Putnam, G. Electrochemical Processing of Soils: Theory of pH Gradient
Development by Diffusion and Linear Convection. Journal of Environmental Science and Health. A25 (6):
687-714, 1990.
10. Alshawabkeh, A., Acar, Y.B. Removal of Contaminants from Soils by Electrokinetics: A Theoretical
Treatise. Journal of Environmental Science and Health v.27, 7, pp.1835-1861, 1992.
11. Hamed, J., Acar, Y. B., and Gale, R. J. Pb(ll) Removal from Kaolinite Using Electrokinetics, Journal of
Geotechnical Engineering. ASCE, Vol. 112, pp.241-271, 1991.
12. Lageman, R.; Wieberen, R; Seffinga, G. Electro-Reclamation:Theory and Practice. Chemistry and
Industry, Society of Chemical Industry, London, pp. 585-590, 1989.
13. Pamukcu, S.; Wittle, J. K. Electrokinetic Removal of Selected Heavy Metals from Soil. Environmental
Progress. 11 (3), pp. 241-250, 1992.
14. Probstein, R.F.; Hicks, R. E. Removal of Contaminants from Soils by Electric Fields. Science. 260, pp.
498-504, 1993.
15. Runnells, D.D.; Wahli, C. Insitu Electromigration as a Method for removing Sulfate, Metals and Other
Contaminants from Ground Water. Ground Water Monitoring Review and Remediation.. 11 (3), 121,
1993.
16. Wieberen, Pool. A Process for electroreclamation of soil material, an electric current system for
application of the process and an electrode housing for use in the electric current system. EEC Patent No:
EP 0 312 174 A1, April 19, 1989.
17. Probstein, R.F., Brookline, P.C.R., Shapiro, A.P. Electroosmosis Techniques for removing materials
from soils. US Patent No. 5,074,986, Dec. 24,1991.
18. Acar, Y. B., and Gale, R. J.. Electrochemical Decontamination of Soils and Slurries. US Patent No.
5,137,608, Commissioner of Patents and Trademarks, Washington, D.C., August 11, 1992.
19. Lageman, R. Electroreclamation-Applications in the Netherlands. Environmental Science and
Technology, v. 27, pp. 13, pp. 2648-2650.
20. Alshawabkeh, A.N. Theoretical and Experimental Modeling of Removing Contaminants from Soils by
Electrokinetics. PhD Dissertation. Louisiana State University (1994).
FOR MORE INFORMATION
Randy A. Parker, Office of Research and Development, Risk Reduction Engineering Laboratory, EPA, 26
West Martin Luther King Drive, Cincinnati, OH 45268. Phone: (513) 569-7620, Fax: (513) 569-7620.
172
-------
COGNIS TERRAMET" LEAD EXTRACTION PROCESS
William E. Fristad
COGNIS, Inc., 2330 Orcadian Way, Santa Rosa, CA 95407
Phone: (707) 576-6235 Fax: (707)575-7833
INTRODUCTION
The COGNIS TerraMet™ lead extraction process leaches and recovers heavy metals, and
specifically lead, from contaminated soil with one of several proprietary aqueous leaching solutions. The
leachant is specifically matched with the substrate and level and type of lead contaminants, in this way
the system can deal with most types of lead contamination including lead salts, oxides and metallic lead.
Complete leaching of lead Is difficult considering the nature of the problem: the variability,
inhomogeneity, the presence of lead in both metallic and ionic form, and the natural binding properties
of soils containing days, iron and manganese oxides and humus. The leaching conditions studied in
this project were chosen to be physically, economically, and ecologically acceptable, and the leachant
had to allow recovery of the leached metals so that the leachant could be reused directly without pH
adjustments. The metal recovery process had to deliver the metals in a compact form which would be
easily recycled and reused.
The bench-scale results from an EPA SITE Emerging Technology Project are described here. In
addition, a brief summary of results from a full-scale implementation of the TerraMet technology at the
Twin Cities Army Ammunition Plant (TCAAP), New Brighton MN will be shown.
METHODOLOGY
The leaching-metal recovery concept is illustrated in Figure 1. The contaminated soil is
contacted with a leaching agent which dissolves a portion of the lead. The lead is removed from the
lead-rich leachant in a metal recovery stage. The lead-depleted leachant then contacts the soil a second
time. The process is repeated until the lead concentration in the soil is acceptably low. This approach
completely recycles the leaching agent Thus, the metal bearing leachant never leaves the process, no
liquid waste streams are generated, and metal recovery can be tailored to the site.
TerraMet Soil Remediation Systems
Metal loaded
leachant
Recovered
Metals
regenerated
leachant
Figure 1. Lead Leaching - Recovery Concept
173
-------
The preferred recovery process for lead is direct reduction of the dissolved ionic lead to lead metal.
Other metal recovery options have been tested, and the most appropriate recovery process for a
specific site depends on the leachant required and the amount and type of metals present.
The goal of the SITE Emerging Technology project was to test the leaching-lead recovery
concept at the bench-scale, study the efficacy of various leachants, and verify that the lead recovery
stage worked. Figure 2 illustrates the steps involved. This bench-scale testing was done with 2 g of
soil. This allowed a single soil sample to be weighed, leached multiple times, and the entire leached soil
sample to be digested after leaching without having to transfer soil from the leaching vessel. The lead
contained in the original contaminated soil sample is then equal to the lead contained in the n-leachant
solutions plus the lead in the treated soil (measured by EPA acid digestion). After verification of the
concept on a particular soil, the leaching process was scaled up and converted to a continuous leaching
process on a 500 g soil scale. The continuous bench-scale process proved that lead could be removed
from the lead-rich leachant and that the lead-depleted leachant could be recycled for continued leaching
of the soil.
Bench-Scale Leaching Experiments (//Contacts)
Fresh Leachant
Untreated Soy.
return for
leaching (n-1) times
Treated Soy
Loaded
-»- Leachant
(/> samples)
Figure 2. Process Flow Diagram for Bench-Scale Testing
RESULTS
Table 1 demonstrates bench-scale leaching of soil contaminated with lead as a result of a leaking
underground storage tank. The soil was excavated from around the leaking fuel tank and therefore
contains mineralized inorganic lead residues from the decomposition of tetraethyl lead in the released
gasoline. The initial contamination concentration is relatively mild (250 - 300 ppm), but high enough to
fail the TCLP test. Consecutive leaching contacts resulted in residual lead concentrations in the soil of
20 - 25 ppm.
Table 1.
LEACHING OF SOIL FROM LEAKING UNDERGROUND STORAGE TANK
Cumulative % Pb Leached _„,..„,,
Leaching Contact #
1
63
50
2
80
66
3
85
89
4
87
91
5 '
89
92
6
90
93
initial
JPfaf
(ppm)
256.5
311.5
FJnai
CPfa]
(ppm)
25.5
22.5
174
-------
fable 2 contains results from leaching of sandblasting waste containing leaded paint chips and
dust. This material was contaminated in the range of 4,000 - 4,500 ppm lead initially, and with six
consecutive leaching contacts it proved possible to remove all but approximately 200 - 250 ppm lead.
This residual concentration was well below the regulatory goal of < 500 ppm lead.
Table 2.
LEACHING OF LEADED PAINT SANDBLASTING WASTE
Cumulative % Pb Leached
Leaching Contact #
1
73
75
2
87
88
3
91
92
4
93
94
5
94
95
6
95
96
Initial
IPbJ
(ppm)
4329
4596
Final
HPb]
(ppm)
237
206
Soil from an ammunition test burning area (TCAAP) was also studied. At this site both metallic
lead fragments as well as ionic lead was found. Because of the high density of lead relative to most soil
constituents, the coupling of soil washing with density separation was a logical pretreatment to leaching.
Soil was first size classified, then density separated using standard mineral processing equipment to
remove heavy lead fragments. Tables 3 and 4 show the results on leaching of the fines and density
pretreated sand fraction. Leaching the fines was very effective and gave residual lead concentrations of
<20 ppm with leachants #2 or #3. The results on the sand were equally satisfactory. Residual lead
concentration of < 70 ppm was also achieved with leachants #2 or #3.
Tables.
LEACHINQ OF TCAAP -200 MESH FINES
i WITH VARIOUS LEACHANTS
TabJo 4.
LEACHINQ OF TCAAP DENSITY PRETREATED 8 to 200 MESH
SAND WITH VARIOUS LEACHANTS
Laacnant
1
1
2
2
3
3
Cumulative % Pb Leached
teaching Contact*
1
53
50
59
64
80
81
2
71
68
90
92
94
96
3
79
76
96
97
97
98
4
83
81
97
98
98
99
5
86
84
97;
98
98
99
Initial
i«g
leached
teaching Contact #
1
66
69
49
75
71
80
2
80
82
71
87
82
92
3
84
86
81
90
86
95
4
87
89
85
92
89
96
S
88
90
88
93
90
97
initial
tPOl
(ppm)-
255
205
585
190
221
223
flnat
t«>]
(ppro)
31
21
69
14
21
<7
After the small bench-scale experiments proved the success of the multiple leaching concept,
additional larger scale continuous leaching experiments verified the leaching results obtained earlier.
The continuous-scale apparatus more closely approximates full-scale treatment. It employs an agitated
leaching vessel from which a soil slurry is pumped into a clarifier. The clarifief separates the slurry into a
clarified feed at the overflow and a thickened slurry at the underflow. The undertow is continuously
returned to the leaching vessel. The overflow is pumped into the metal recovery unit where the lead is
175
-------
removed from the leachant and the lead recovered as solid lead powder. The lead-depleted leachant is
then returned to the leaching vessel for continued leaching. After leaching is complete, the soil-leachant
slurry is dewatered and neutralized. Thus, the entire leaching, clarification, and metal recovery process
operates continuously on the batch of soil in the leaching vessel. Table 5 illustrates typical data on
ammunition plant soil. Routinely < 100 ppm residual lead and TCLP passage was observed. The lead
concentrations shown under the influent and effluent columns are the concentrations of lead in the
leachant entering and exiting the metal recovery unit.
Table 5.
CONTINUOUS-SCALE LEACHING EXPERIMENT TCAAP SOIL (DENSITY PRETREATED)
Matrix
Soil (Avg)
Replicate 1
Replicate 2
Replicate 3
Leachate
Sample 1
Sample 2
Sample 3
Sample 4
Sample 5
Sample 6
Sample 7
Lead Concentration (f/g/g)
Pre-Leaching
250 - 350
Influent Oug/mL)
15.4
10.6
4.6
2.5
1.4
0.9
0.5
Leached
31.1
32.6
28.2
33.2
Effluent (pg/mL)
2.5
2.1
1.3
0.7
<0.5
<0.5
<0.5
CONCLUSIONS
The COGNIS TerraMet™ lead extraction process has been successfully proven at the bench-
scale on several sites: a leaking underground leaded gasoline tank, leaded paint-sandblast waste, and
on the Twin Cities Army Ammunition Plant (TCAAP) soil.
The TCAAP results led to the construction of a full-scale lead remediation plant. The plant
utilizes continuous, multiple leaching contacts and the continuous metal recovery concept proven in the
SITE Emerging Technology project The treatment train at TCAAP also uses a Bescorp soil washing
plant which physically pretreats the soil to remove the dense paniculate lead fragments and separates
the soil into three fractions: a clean gravel fraction and contaminated sand and fines fractions. The
sand and fines are leached in two parallel streams in the TerraMet plant. The combined plant utilizes full
recycle of all aqueous solutions and 1,600 tons of lead-contaminated soil were processed through the
full-scale system at TCAAP in mid-September through October, 1993.
176
-------
For More Information:
Mr. Michael D. Royer,
EPA SITE Emerging Technology Project Manager
Technical Support Branch
U.S. EPA
Bldg. #10 (MS-104)
2890 Woodbridge Ave.
Edison, NJ 08837-3679
Phone: (908) 321-6633
Fax: (908) 321-6640
177
-------
ACID EXTRACTION TREATMENT SYSTEM (AETS)
FOR TREATMENT OF METAL CONTAMINATED SOILS
Stephen W. Paff
Brian E. Bosilovich
Center for Hazardous Materials Research
320 William Pitt Way
Pittsburgh, PA 15238
- (412) 826-5320
Fax: (412) 826-5552
INTRODUCTION
Through a Cooperative Agreement with the U.S. Environmental Protection Agency's (EPA)
Risk Reduction Engineering Laboratory, the Center for Hazardous Materials Research (CHMR)
developed the Acid Extraction Treatment System. The project was conducted with support from
Interbeton bv and The Netherlands Organization for Applied Scientific Research (TNO). AETS is
intended to reduce the concentrations and/or leachability of heavy metals in contaminated soils to
render the soils suitable to be returned to the site from which they originated. Additional
applications may include treatment of contaminated sediments, sludge, and other heavy metal-
containing solids.
The objective of the project was to determine the effectiveness and commercial yiability of
the AETS process in reducing the concentrations and leachability of heavy metals in soils to
acceptable levels. This abstract briefly describes the activities conducted during the project, the
experiments performed, and the results.
A pilot scale system was designed, constructed, and used to test different soils. Five soils
were tested, including EPA Synthetic Soil Matrix (SSM), and soils from four Superfund sites (NL
Industries in Pedricktown, NJ; King of Prussia site in Winslow Township, NJ; Silver Bow Creek site
in Butte, MT; and Palmerton Zinc site in Palmerton, PA). These soils contained elevated
concentrations of arsenic, cadmium, chromium, copper, lead, nickel, and zinc.
Many experiments were conducted with this pilot-scale system, and the data obtained from
these experiments were used to produce several cost estimates for processing the soils through a
full-scale plant. These cost estimates are based on several different parameters, such as size of
site, levels of contaminants, and residence time.
METHODOLOGY
Process Description
A simplified block flow diagram of the AETS process is shown in Figure 1. Full-scale units
are anticipated to be able to process between 10 and 30 tons per hour. The first step in the full-
scale AETS process is screening to remove coarse solids. These solids, typically greater than 4
millimeter (mm) in size, are anticipated to be relatively clean, requiring at most a simple rinse with
water or detergent to remove smaller attached particles.
After coarse particle removal, the remaining soil is scrubbed in an attrition scrubber to
physically remove the metals and break up agglomerations. Then it is contacted with hydrochloric
acid (HCI) in the extraction unit. The residence time in the unit may vary depending on the soil
type, contaminants, and contaminant concentrations, but is anticipated to range between 10 and
40 minutes. The soil and extractant mixture is continuously pumped out of the mixing tank, and
the soil and extractant are separated using hydrocyclones. The solids are piped to the rinse
system, while the cyclone overflow (extractant) is treated using a proprietary technology which
removes the metals from the liquid and regenerates the acid.
i
The soils are rinsed with water to remove entrained acid and metals. The metals are
removed from the rinsate using the same technology that regenerates the acid. After rinsing, the
soil is dewatered using hydrocyclones and (if required) dewatering screens. In the final step, the
soils are mixed with- lime and fertilizer to neutralize any residual acid and return the soil to natural
conditions.
178
-------
CONTAMINATED
SOIL
MAKE-UP
ACID
RINSE
WATER
CLASSIFICATION
(SCREENING)
±
.^COARSE SOIL
PARTICLES
EXTRACTION
UNIT
REGENERATED ACID
EXTRACTANT
RINSE / DEWATER
1
RINSATE
ENTRAINED
son s
ACID
REGENERATION
NEUTRALIZATION
& STABILIZATION
f
HEAW METALS
TREATED SOIL
FIGURE 1. AETS BLOCK FLOW DIAGRAM.
Experimental Procedure
The pilot-scale AETS unit was constructed and tested in CHMR's Technology Development
Laboratory. This system is capable of processing between 20 and 100 kilograms (kg) of soil per
hour. The soils were initially characterized for total metals and Toxicity Characteristic Leaching
Procedure (TCLP) metals content. The soils were screened to remove the + 8 mesh fraction on a
mechanical shaker prior to being placed in the lab-scale attrition scrubber, where the soil was
slurried with water (or regenerated hydrochloric acid from previous experiments) prior to entering
the extraction vessel.
Next, the soil was contacted with the hydrochloric acid for residence times between 10 and
40 minutes in the extraction tank. The pH of the mixture was maintained between 1.8 and 2.2.
During extraction, the solids were separated using two hydrocyclones, and returned to the
extraction tank. The extractant was pumped to the acid regeneration system, and then returned to
the extraction tank.
At the end of the experiment, the soil was dewatered using two cyclones and a mechanical
shaker with a 200 mesh screen to separate the solids and the extractant. The extractant was then
regenerated to be used as the acid in the next experiment, and the solids were prepared for the
rinsing step.
The solids were rinsed in water to remove any residual acid. The metals were removed from
the rinse water using a separate regeneration system than the one used during the extraction. The
clean solids and all liquids were then analyzed for total and TCLP metals to form a material balance.
The extractant and rinse water were ready for use in the next experiment, so no waste streams
were generated during the experiments.
Experimental Soils .
This section gives a brief discussion of the soils used during the laboratory- and pilot-scale
investigations. These soils included the EPA's SSM, and soils from: NL Industries Site in
Pedricktown, NJ, Silver Bow Creek Site in Butte, MT, King of Prussia Site in Winslow, NJ, and the
Palmertoh Zinc Site in Palmerton, PA.
179
-------
The SSM is produced by the EPA specifically for use in research and development of
emerging or innovative technologies. The soil is a mixture of clay, silt, sand, gravel, and topsoil
that is blended together to form the soil matrix. Organic and inorganic contaminants are added
based on typical hazardous materials at Superfund sites. This soil contained approximately 10,000
milligrams per kilogram (mg/kg) total lead and 27 milligrams per liter (mg/L) of TCLP lead.
The NL Industries site, located in Pedricktown, NJ, was an integrated battery breaking and
lead smelting facility. The soil is contaminated with copper, lead, and zinc, but was chosen for this
project due to the high levels of lead. The total and TCLP lead values were 29,000 mg/kg and 520
mg/L, respectively.
The King of Prussia site was used to neutralize acid streams from an adjacent site. The soil
is contaminated with chromium, copper, and nickel, and it is not hazardous by RCRA standards.
The site was placed on the National Priorities List (NPL) because of high levels of chromium. The
soil contains approximately 1,200 mg/kg of total chromium.
The Silver Bow Creek Site contains a very sandy soil, with very little clay. The soil is
contaminated with copper and zinc, with the total metals of 120 mg/kg and 1,250 mg/kg,
respectively. The TCLP values were 1.5 mg/L for copper, and 5.0 mg/L for zinc. The Butte soil is
not considered hazardous soil.
The Palmerton site is an old zinc smelting facility. Only one experimental extraction was
conducted on this material, due to a lack of soil. This soil was chosen because of its high levels of
zinc, and because it contained lead, cadmium, and copper. The Palmerton soil contained 1,000
mg/kg total lead, and 10,000 mg/kg total zinc.
RESULTS
Experimental Results
Table 4 qualitatively summarizes the soil treatability for the soils and metals tested. For soils
with a high amount of clay, that clay fraction was treated and analyzed separately with the AETS
system. The table reports the results as if all fractions were remixed at the conclusion of the
experiment. The results show that AETS treated virtually all the soils tested to reduce both the
total and TCLP metals concentrations to below currently regulated concentrations. The only
exceptions were cadmium, which consistently failed the TCLP for SSM soil, and lead, which failed
both the TCLP and total metals requirements for SSM soils.
TABLE 1. QUALITATIVE RESULTS OF EXTRACTIONS
Metal
SSM
Butte
King of Prussia Pedricktown Palmerton
As
Cd
Cr
Cu
Ni
Pb
Zn
*, T, L *, T, L
*,T
*, T, L *, T, L'
*, T, L *, T, L *, T, L
*. T, L *, T, L
* *, T, L
*, T, L '. T, L
*, T, L
*, T, L
*, T, L
*, T, L *. T, L
*, T, L *, T, L
* - Metal is present in the soil
T - Successful treatment for total metals
L - Reduction in leachability to below standards
'Bold and large fonts indicate high initial metals content (at least double regulatory standards)
Table 2 shows typical results obtained from the lead contaminated soil from the NL Industries
Superfund site in Pedricktown, NJ. The table shows over 90% reductions in total metals
concentrations, and a 99% reduction in TCLP. Further work indicated that the TCLP and total lead
180
-------
in the soil could be reduced to below 5 mg/L and 1000 mg/kg, respectively. These are the EPA
recommended levels for remediation of lead in soil. The experimental work was completed during
January 1993, and the final report has recently been issued.
TABLE 2. NL INDUSTRIES SOIL
Metal Initial
Pb 29,200
Total Metals (mg/kg)
Final % Removal
1,310 95.5
Initial
520
TCLP Metals (mg/L)
Final % Removal
5.1 99.0
Economic Analysis
CHMR constructed an economic model of a full-scale AETS system based on the results of
the pilot-scale experiments. This economic model used such parameters such as: typical removal
efficiencies, metal concentrations in the soil, extraction residence time, site size, etc. Some cost
estimates for various situations are presented in Table 3 below. CHMR has estimated that a full-
scale AETS processing unit could be built to process soil at between $80 and 240 per cubic yard
TABLE 3. AETS COST SUMMARIES UNDER VARIOUS CONDITIONS
Capital and
Operating Costs
($/yd3) .
77
96
138
118
122
193
168
241
Feed Rate
(yd3/hr)
22.9
15.3
15.3
15.3
11.5
11.5
11.5
7.6
Extraction
Residence
Time (min)
24
24
36
24
24
36
36
36
% Fines
«50//m)
15
15
30
15
1.5
30
15
30
Metals
Cone.
(mg/kg)
5,000
1 5,000
15,000
15,000
5,000
15,000
5,000
15,000
Site Size
(1000yd3)
114.7
76.5
45.9
61.2
45.9
22.9
22.9
15.3
The following notes apply to this table:
1. The plant is operating for only 1 eight hour shift per day.
2. No metal recovery value is assumed. All metal sludge is disposed.
This table includes costs for mobilization, pilot plants, excavation, replacing soil, and reseeding the
ground as well as the actual soil treatment. Thus, the costs represent total costs of treatment
using the AETS. Note that the table conservatively assumes that the capital costs of the AETS are
amortized over only 1 site, and that the plant operates only one eight hour shift per day. Finally
the economic model assumes that the metal sludge is stabilized and disposed, and not reclaimed
The metals at many sites may be reclaimable. Relaxing all of these conservative assumptions will
reduce the estimated treatment costs by 20 to 30%.
CONCLUSIONS
The conclusions derived fram the study are summarized below:
• AETS is capable of treating a wide range of soils, containing a wide range of heavy metals to
reduce the TCLP below the RCRA limit and to reduce the total metals concentrations below
the California-mandated total metals limitations.
181
-------
• In most cases, AETS is capable of treating the entire soil, with no separate stabilization and
disposal for fines or clay particles. For most of the soils tested, AETS can do this to the
required TCLP and total limits. The only exception to this among the soils tested was with
the SSM, which may require separate stabilization and disposal of 20% of the soil because of
lead. This soil was successfully treated for other metals, including arsenic, chromium,
copper, nickel and zinc.
• Costs for treatment, under conservative process conditions, range between $80 and 240 per
cubic yard of soil, depending on the site size, soil types and contaminant concentrations, as
well as several other parameters.
ACKNOWLEDGEMENTS
The Center for Hazardous Materials Research gratefully acknowledges the financial and
technical assistance received from the United States Environmental Protection Agency, Interbeton
bv, and The Netherlands Organization for Applied Scientific Research. This work has been funded
wholly or in part under Cooperative Agreement CR-815792-01 -0 to the Center for Hazardous
Materials Research, under the Emerging Technologies Program.
FOR MORE INFORMATION
Naomi Barkley
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513) 569-7854
Stephen W. Paff .
Center for Hazardous Materials Research
320 William Pitt Way
Pittsburgh, PA 15238
(412) 826-5320
182
-------
TREATMENT OF ORGANIC WASTES IN AQUEOUS MATRICES BY X-RAY
Esperanza Piano Renard1, Vemon Bailey2, Randy Gurry3,
- " Heinz Lackner2, Gilbert Young4, and Norma Lewis5
1 U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory,
2890 Woodbridge Ave., Edison, N.J. 08837, (908) 321-4355
2 Titan/PSI, 600 Me Cormick Street, San Leandro, CA 94577 (510) 632-5100
3 Tetra Corporation, 3701 Hawkins St., Albuquerque, NM 87109 (505) 345-8623
4 Consultant-Department of Chemistry, University of S. California, CA 90089 (213) 740-7020
U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory,
26 West Martin Luther King Drive, Cincinnati, OH, 45268 (513) 569-7665
ABSTRACT
Titan/Pulse Science Inc. (PSI) is presently funded jointly by the U.S. Environmental Protection
Agency (US EPA) and the Department of Energy (DOE)under the SITE Emerging Technology program
to develop and demonstrate the x-ray (Bremsstrahlung radiation) processing concept for the treatment
of solid and liquid wastes contaminated with volatile (VOCs) and semi-volatile (SVOCs) organic
compounds.
X-ray treatment technology is based on the in-depth deposition of ionizing radiation. Collisions of
energetic photons (x-rays) within the contaminated waste generate a shower of lower energy secondary
electrons which ionize and excite the atomic electrons, break up the complex contaminant molecules,
and form radicals that react with the VOC or SVOC contaminant to form compounds such as water
carbon dioxide and oxygen.
A 1.2 MeV, 800 A, 55 nsec electron beam from a linear induction accelerator (LIA) impacts a high-Z
tantalum target which converts the electron pulse into x-rays. The LIA is pulsed at the rate of one pulse
per second giving a dose rate at the sample of * 5 rads per second.
Preliminary results of the compounds in aqueous matrices evaluated included contaminants in actual
Superfund well water samples as well as neat samples of trichloroethylene (TCE), tetrachloroethylene
(PCE), benzene, toluene, and carbon tetrachloride in deionized water. The concentration ranqed from
100 to 64,000 ppb.
Irradiated contaminants with concentrations ranging from 100-4000 ppb were decomposed at x-ray
doses of less than 200 kilorads (krads). Most of the other irradiated contaminants decomposed at dose
levels less than 200 krads with no residual by-products detected. However, a limited number of
experiments conducted for TCE at higher concentrations (64,000 ppb) showed the presence of
hazardous by-products which mineralized as the x-ray dose was increased, however, it is not clear
whether toxic residuals are formed and/or completely mineralized and, if so how much? This is still
under investigation. This paper will present only preliminary results obtained from the study.
INTRODUCTION
Electron beam processing has been established as highly effective in the destruction of organic
compounds in the removal of chemicals SOX and NOX the chemicals that cause acid rain, from power
plant and incinerator stack gases; (Ref. 1) disinfection'of sewage and plastics processing waste. (Ref. 1 -
10) High energy photons, gamma-rays are routinely used to sterilize medical supplies. (Ref. 12) It has
been observed that all forms of ionizing radiation (electron beam, gamma-rays and x-rays) have
183
-------
approximately the same chemical effects for a given dose. Electron beams with energies in the range of
1-5 MeV have been generated using conventional accelerators. For those energies the range of electron
penetration is limited to 0.3-2 cm for a material with a density of 1 g/cm3. The use electron beams
would require that the material will in the form of a thin stream or slab which may require substantial
material preprocessing and handling.
Gamma-rays, from radioactive cobalt offers greater penetration depths, however, the inability to turn
off the radiation from the source presents an acceptable health hazard. X-rays produced when high
energy electrons are stopped offer the same advantages as gamma-rays, penetrating more than five
times further than electrons for the same electron beam source while providing the safety of an electron
beam source which can be shut off when not in use. It is expected that a large number of organic
wastes can be decomposed at an x-ray dose of ( 1 megarad (Mrad) or 10 joules/gram.
A high throughput x-ray treatment system requires the availability of a high average power electron
accelerator capable of generating beams with energies of 5-7 mev. The upper limit on energy is a
compromise between increased x-ray conversion efficiency at higher energies and the requirement that
the energy be chosen small enough to avoid activation. (Ref. 1)
The U. S. EPA and the DOE under the Superfund Emerging Technology Program have entered into a
cooperative agreement with Titan/Pulse Science Inc. (PSI) to evaluate the x-ray treatment process for
the remediation of solid and liquid wastes contaminated with volatile (VOCs) and semi-volatile (SVOCs)
organic compounds. The overall objective of the project is to demonstrate and develop the X-ray
treatment concept for the destruction of organic contaminants in soil and water. The specific objectives
are:
1) To demonstrate in pilot scale experiments the efficacy of x-ray treatment for solid and liquid
waste contaminated with volatile organic compounds (VOCs) and semi-volatile compounds
(SVOCs).
2) To determine the x-ray doses required to reduce organic contamination to acceptable levels,
and
3) Develop the conceptual design for a high-throughput x-ray treatment system.
The linear induction accelerator (LIA) was first modified to include a high-Z x-ray target to convert
the electron beam pulses into pulses of x-rays. Experiments to determine the x-ray doses needed for
complete mineralization were begun during the later part of the first year of the program. During the last
portion of the program, much higher concentrations of VOC/SVOC contaminated aqueous solutions will
be irradiated with x-rays to investigate hazardous by-product formation and subsequent concentration
levels of the by-products as the dose is increased.
METHODOLOGY
The proposed technology is based on the fact that as x-rays, gamma rays, and energetic electrons
penetrate gases, llqukfe and solids they deposit energy primarily through ionizing collisions. These
collisions generate energetic secondary electrons within the material which are effective in breaking up
complex molecules and in forming chemically active radicals which will react with the contaminant
materials.
The physical mechanism by which VOC and SVOC contaminants are removed is primarily dependent
on the substrate. For example, in oxygenated water the primary reactant is the hydroxyl radical. This
kinetic mechanism is also expected to play an important role in nonaqueous matrices, because of the
184
-------
presence of moisture in contaminated soil, sludge and sediments.
In the presence of water the following reaction is known to occur:
H20 -* OH, e".,, H3CT, H, H2, H202
x-ray
High throughput waste treatment applications for Superfund sites is made feasible by the use of high
average power linear induction accelerator (LIA) for generating the 5-7 MeV electron beams required for
maximizing the x-ray conversion efficiency while operating at an energy low enough to avoid producing
radioactive by-products. (Ref. 1)
The operating principle of the LIA is analogous to that of an electrical transformer. (Ref. 1) The LIA
at PSI has been modified to include an x-ray target which is used to convert electron beam pulses to
pulses of x-rays. The target consists of a water cooled 2-mil titanium window, 5 mil tantalum foil, and a
0.2 inch thick carbon which acts as an electron stopper and debris shield. The 5-mil tantalum foil has
been tested at up to 1000 A beam current. The accelerator is operating at a beam current of about 800
A at an electron beam energy of 1.2 MeV. The x-ray dose at the sample is measured using a
radiochromic film. Scattering of electrons and x-rays in the converter serves to produce a broad x-ray
beam which allows samples of 40 cm3 in volume to be irradiated simultaneously.
Samples sealed in 40 ml vials with teflon lined septums were placed in a rotating sample holder 10
cm from the tantalum foil and irradiated within a nitrogen filled irradiation chamber. The LIA is pulsed at
the rate of one pulse per second to accumulate x-ray dose on the sample which is measured with
radiochromic film. Since the average dose rate is approximately 5 rads per pulse or second, 200 pulses
are required to deposit 1 Krad of energy in the sample. After irradiation the samples are analyzed by
Gas chromatograph.
RESULTS AND CONCLUSION
Results of the 20 aqueous matrices are shown in Table 1. The first 15 were prepared using neat
solutions of SVOC and VOCs in deionized water. The additional five matrices shown utilized well water
drawn from either contaminated Superfund sites or working wells. The contamination level for the neat
solutions varied from 13 ppb for cis-1,2-dichloroethene to 64,000 ppb for trichloroethylene indicate that
the x-ray dosage needed to reduce the contaminants to <0.5 to 54 ppb is between 4 to 224 krad. As
shown, compounds such as carbon tetrachloride require a much higher x-ray dose to destroy them than
VOCs which react with the hydroxyl radical. In the presence of OH scavengers, such as carbonate ions,
the x-ray dose required to destroy the VOC's increase somewhat. The spiked well water and Superfund
water showed slightly higher x-ray doses when OH scavengers are present in the water. Both the
spiked water well water and the two Superfund samples had high concentrations of carbonate and
bicarbonate tons. This phenomena has been shown by Cooper et. al. (Ref 13 ) The carbonate and
bicarbonate ions appear to act as OH scavengers. Benzene concentrations of 240 ppb and 262 ppb
were irradiated in deionized water and well water. In deionized water 8.8 krads mineralized the benzene
while 30.9 Krads was required for the mineralization of benzene in the well water. In addition the
reaction rate depended on the diluting solution. The ethyl benzene and xylene samples of spiked well
water also required slightly more x-ray dose to mineralize the contaminant.
A set of preliminary equations have been developed which allow the x-ray dose levels required to
destroy VOCs in deionized water to be scaled to multiple VOC matrices and matrices containing OH
scavengers, such as bicarbonates and carbonate ions. The scaling equations are based on the G-values
calculated from the concentration-dose curve generated experimentally. The G-value is defined as the
number of molecules of the contaminant which is formed or destroyed per 100 electron volts deposited
185
-------
Table 1. Summary of Radiolysis Experiments: X-rays Treatment of VOC/SVOCs In Aqueous Matrices
Compound
TCE
TCE
TCE
TCE
PCE
Chloroform
Methytene Chloride
Trans 1,2
Dichloroethene
1,1,1 Trtehloroethane
Carbon tetrachloride
Benzene
Toluene
Ethylbenzene
Xylene
CIs 1,2 Dichloroethene
Benzene/carbon
tetrachloride
Ethylbenzene/carbon
tetrachloride
D-xylene/carbon
tetrachloride
Matrix
Deionized Water
FIU well water
FIU well water
FIU well water
Initial
Concentration
(ppb)
64,000
2100
1000
490
230
2000
270
260
590
180
240
150
890
240
13
262/400
1000/430
221/430
Final
Concentration
(ppb)
<.5
<.5
<.5
<.5
3.6
4.4
3.1
.78
54
14
<.5
<.5
3.6
1.2
<.5
<.5/196
<. 5/70.9
<.5/85
X-ray
Dose
(kR)
180
20
10
10
4.2
178
145.9
10.6
207.1
224
8.8
4.83
20.4
5.6
10.6
39.9/93.8
33.2/185
20.5/171
Test
Method
8010
8010
8010
8010
8010
8010
8010
8010
8010
8010
8020
8020
8020
8020
8010
FIU
FIU
FIU
Superfund Well Water Sample #1:
TCE
PCE
1,1-dichloroethane
1,1,1 -trichloroethane
CIs 1,2 dichloroethane
SFWW1
SFWW1
SFWW1
SFWW1
SFWW1
3400
500
21
13
47
<.5
<.5
1
2.0
<.5
99.0
99.0
145.4
145.4
49.9
8240
8240
8240
8240
8240
Superfund Well Water Sample #2:
TCE
PCE
Chloroform
Carbon tetrachloride
1,2 Dichloroethane
1,1 Dichloroethane
Freon
SFWW2
SFWW2
SFWW2
SFWW2
SFWW2
SFWW2
SFWW2
5000
490
250
14
38
11
71
<1.0
1.6
81
4
17
6.8
32
291
291
291
291
291
291
291
8240
8240
8240
8240
2840
8240
8240
SFWW1: Superfund well water #1 was a superfund sample which was comprised of the six
contaminants shown.
SFWW2: Superfund well water #2 was a superfund sample comprised of the four contaminants shown.
186
-------
energy. (Ref 14) The G-value is used for scaling the chemical reaction. Typical G-values reported range
from 7-12. (Ref 14) Table 2 shows G-values derived from the experiments. The G- value has been
found to be a measure of the efficiency of a reaction, the higher the G-value the more efficient is the
reaction.
It was initially thought that any by-product formed during irradiation were also destroyed by the
irradiation. The x-ray irradiation data for initial concentration levels of up to 2000 ppb showed no
unknown compounds using GC or GC/MS test methods, although a few traces of C6 and C12 atoms did
appear on one or two reports. One possible explanation was that the concentrations of the by-products
were below the detection limit To confirm this, a 64,000 ppb concentration of TCE was prepared and
irradiated. At least four by-products were found to be formed and then at least partially destroyed as
the radiation dose level was increased. Chloroform and 1,1,2 trichloroethane were formed and
destroyed at a dose of 275 Krads. Detectable levels of chloromethane and Methylene chloride remained
at a dose of 275 krads. It appears that the higher contaminant concentration showed presence of
residuals. Experiments now underway, using much larger contaminant concentrations are attempting to
determine if the organic compounds are completely destroyed, forming carbon dioxide, water, and
carbon or if additional by-products are formed. Preliminary results of the investigation will be included in
the paper.
REFERENCES
1. J. R. Bayless, C. P. Burkhart and R. J. Adler, "Linear Induction Accelerators for Industrial
Applications," Proceedings 10th international Conference on the Application of Accelerators in
Research and Industry, Nov. 7-9,1988, Denton, Texas.
2. T. D. !Waite, et al., "Disinfection of Wastewater Effluents with Electron Radiation," Proc. Natl. Conf.
Environ. Engin., Austin, TX, July 10-12, 1989.
3. W. J.1 Cooper, et al., "Treatment of Complex Mixtures of Toxic Chemicals Using an Innovative
Treatment Process: High Energy Electron Irradiation," Proc. 63rd Annual Conf. Water Pollution
Control Federation, Washington, D.C., October 7-11, 1990.
4. W. J. Cooper, et. al., "High Energy Electron Irradiation: An Innovative Treatment Process for the
Treatment of Aqueous Based Organic Hazardous Wastes," Proc. 5th Annual Aerospace Hazardous
Waste Minimization Conf., Costa Mesa, CA, May 22-24, 1990.
5. A. Singh, et al., "Radiolytic Dechlorination of Polychlorinated Biphenyls," Rad. Phys. Chem., Vol. 25,
pp. 11-19,1985.
6. P. Gehringer, et al., "Removal of Chlorinated Ethylenes from Drinking Water," Rad. Phys. Chem.,
I; Vol. 15, pp. 456-460, 1990.
7. N. Getoff and W. Lutz, "Radiation Induced Decomposition of Hydrocarbons in Water Resources,"
Rad. Phys. Chem., Vol. 25, Nos. 1-3, pp. 21-26, 1985.
8. M. G. Nickelsen, et al., "High Energy Electron Irradiation of Oxygenated Secondary Waste Water
Effluent for the Removal of Benzene and Substituted Benzene Compounds," Proc. 198th Amer.
Chem. Soc. Natl. Mtg., Miami, FL, Sept. 10-15, 1989.
9. Dr. William J. Cooper, Drinking Water Research Center, Florida International University, private
communication, 11-90.
187
-------
Table 2.1 Preliminary Peak G-values calculated from experimental data obtained for the deionized water
experiments.
Compound
TCE
TCE
TCE
TCE
PCE
Chloroform
Methytene chloride
Carbon tetrachloride
Benzene
Toluene
Ethylbenzene
Xylene
Matrix
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Deionized Water
Initial Concentration
(ppb)
64,000
2100
1000
490
260
2000
450
180
240
140
890
240
peak
8.81
8.81
5.4
3.52
.61
6.1
.054
.040
2.21
1.46
1.5
1.35
Table 2.2 Spiked well water/ethylbenzene. Preliminary
experimental data
and G,- constants calculated from the
Ethylbenzene
Carbonate (CO32~)
Bicarbonate (HCCy)
(molX)
2.57E-06
2.35E-06
2.26E-04
(L/mol-s)
3.00E+09
3.90E+08
8.50E+06
oeak
1.5
Gi
Calculated
1,096
QI
Experimental
0.77
188
-------
Table 2.3 Spiked well water/xylene. Preliminary G k and G, constants calculated from the
; experimental data.
Xylene
Carbonate (CO32~)
Bicarbonate (HCCy)
(mol/L)
2.00E-06
2.35E-06
2.26E-04
ki
(L/mol-s)
3.00E+09
3.90E+08
8.50E+06
Gpeak
1.35
Gi
Calculated
0.917
Gi
Experimental
1.13
Table 2.4 Superfund well water sample #1 (SFWW1). Preliminary G k and Gf constants calculated
from the experimental data.
Trichlproethylene (TCE)
Perchloroethylene (PCE)
1,1-Dichloroethylene
Carbonate (CO32")
Bicarbonate (HCO3")
(mol/L.)
2.59E-05
3.02E;-06
4.54E-07
3.00E-04
2.20E;-03
ki
(L/mol-s)
4.20E+09
2.60E+09
6.80E+09
3.90E+08
8.50E+06
Gpeak
8.81
8.8
8.8
Gi
Calculated
3.8
0.27
0.11
Gi
Experimental
4.0
0.38
0.087
189
-------
10. C. N. Kurucz, et al., "Full Scale Electron Beam Treatment of Hazardous Wastes - Effectiveness and
Costs," Proc. 45th Ann. Purdue Univ. Ind. Waste Conf., W. Lafayette, IN, 5/8-10/90.
11. N. Frank, K. Kawamura and G. Miller, "Electron Beam Treatment of Stack Gases," Rad. Phys.
Chem., Vol. 25, p. 35.
12. Edward H. Bryan, "Disinfection of Wastewater and Residual Sludges," Proc. Conf. on Disinfection of
Wastewater, Effluents and Sludge, University of Miami, Miami, FL, May 7-9, 1984.
13. Nicesen, W. Cooper, C. N. Kurucz, T. Waite, "Removal of Benzene and Selected Alkyl Substituted
Benzenes from Aqueous Solution Utilizing Continuous High energy Electron Irradiation",
Environmental Science and Technology, Vol. 26, 1992.
14. Buxton, C. Creenstock, W. Helman, A. Ross "Critical Review of Rate Constants for Reactions of
Hydrated Electrons, Hydrogen Atoms and Hydroxyl radicals in Aqueous Solution, Journal of
Physical and Chemical reference Data, Vol. 17, No. 2, 513-886, 1988.
190
-------
RECOVERY OF LEAD BY A CHELATION/ELECTROMEMBRANE PROCESS
Ronald J -Turner
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
and
Mary Beth Foerst, Radha Krishnan
IT Corporation
11499 Chester Road
Cincinnati, Ohio 45246
INTRODUCTION
This paper summarizes the results of a bench-scale treatability test to
investigate key process parameters influencing an innovative chelation-
electirodeposition process in which lead is recovered while simultaneously
regenerating the chelating agent. Lead-contaminated soils from battery
reclamation sites typically contain 1 to 5 percent lead by weight. The forms of
lead found at these sites are lead dioxide, basic lead carbonate, elemental lead,
and lead sulfate.
The design of the bench-scale electromembrane reactor experiments was
derived from previous soil washing studies using chelating agents and
electroplating for lead recovery. (1) The goal of the recent study was to recover
the lead on the cathode, at the same time regenerating the chelating agent in its
sodium salt form in the cathode chamber. The chelating agents for this study
were di-sodium and tetra-sodium ethylene diamine tetraacetic acid (EDTA), and
diethylenetriamine pentaacetic acid (DPTA). A synthetic lead-chelate solution
was tested rather than lead extracted from a contaminated soil, as soil chelation
has been previously studied.(2) The lead solutions used in the tests represent
the range of lead concentrations that can typically be extracted by a soil
washing process employing chelating agents.
EXPERIMENTAL APPARATUS
The reactor was constructed from a 10-gallon aquarium with 1/4-inch
plexiglass walls. It was divided into two chambers by the cation-exchange
membrane which is almost impermeable to anions and prevents oxidation of the
chelating agent at the anode. Each chamber was 2 inches long, 12 inches high,
and 10 inches deep. The membrane was 7-by-7 inches and was mounted using a frame
and gasketing material.
Two types of membranes were used in the study: Ionics 61AZL386 and DuPont
Nafion. The Ionics membrane is a modacrylic fiber-backed cation-exchange
membrane with a specific weight of 14 mg/cm2, a thickness of 0.6 mm, a burst
strength of 8 kg/cm2, and a 2.7 meg/dry gram resin capacity. The Nafion membrane
is a perfluorosulfonic acid cation-exchange membrane, Teflon-reinforced, with a
weight of 6.3 gm/cm2, and thickness of 0.43 mm. Both membranes have low
electrical resistance, high permselectivity, high burst strength, long-term
resistance to aqueous acid, alkaline, and mild oxidizing solutions.
The lead electrodes were approximately 7 by 10 inches and were suspended
one-inch from the membrane surface using a wooden dowel rod. In a second set of
tests, cadmium electrodes were used. The electrodes were connected to a DC power
supply which controlled amperage and measured both current and voltage.
191
-------
EXPERIMENTAL DESIGN
The type of chelating agent, type of membrane, current density, Lead
concentration, and reaction time were varied to examine the effects each
parameter has on lead recovery. Table 1 presents the experimental matrix for the
bench-scale membrane reactor study. A total of 24 tests were performed.
Both di-and tetra-sodium EDTA and DPTA were selected for study because of
their relatively high lead stability constants (18-19 @ pH=9) and prior use in
soil washing studies. Based on jar tests conducted with various chelating agents
concentrations, solutions of 1:1 di-sodium EDTA-to-lead molar ratio, a 1.5:1
tetra-sodium EDTA ratio, and a 2:1 DPTA-to-lead molar ratio were determined to
be adequate for complete lead chelation. The forms of lead initially added in
the chelant solution consisted of lead sulfate, basic lead carbonate, elemental
lead, and lead dioxide. None of the lead dioxide or elemental lead was
successfully chelated. Only lead sulfate and lead carbonate were used to make
the synthetic lead-chelate solutions of 0.8 percent and 4 percent for the
electromembrane tests.
The current densities of 15 and 25 milliamps (ma/cm2) were selected for the
study. A current density above 30 ma/cm2 was previously determined to be
limiting in industrial electrodialysis systems.(2) Current density is an
important factor which influences the design of electromembrane reactors and
power supply.
EXPERIMENTAL PROCEDURE
The cathode chamber was filled with 4 liters of lead-chelate solution
adjusted to the experimental pH with sodium hydroxide or sulfuric acid. The pH
values were determined for the anode and cathode chamber solutions during each
experiment. Twice the stoichiometric quantity of 5 percent sodium carbonate
solution (2 moles sodium per mole of lead plated is required to regenerate the
sodium salt form of the chelating agent) was placed in the anode chamber to
prevent depletion of sodium ions.
The electrodes were weighed and placed into the chambers approximately one-
inch from the membrane. Current densities were adjusted on the power supply
unit, and the reaction initiated. Samples of the anode and cathode chamber
solutions were taken at 30-minute intervals for the first three hours to
determine the lead plated and the depletion of sodium ions. Experiments were
conducted for a total period of 3 hours to 5 hours. After the third hour,
samples were taken at 1-hour intervals to obtain the optimum time required for
plating out the lead. The anode chamber samples were analyzed for sodium content
and the cathode chamber samples were analyzed for sodium and lead by atomic
absorption.
192
-------
TABLE 1. ELECTROMEMBRANE/CHELATION STUDY EXPERIMENTAL MATRIX
Chelating agent
Tetra-sodium EDTA
Di -sodium EDTA
Tetra-sodium EDTA regenerated
solution8
DTPA (diethylenetriamine
pentaacetic acid)
DTPA^
DTPA regenerated solution15
DTPA
DTPA
DTPA
DTPA regenerated solution13
DTPA
Tetra-sodiura EDTA
Tetra-sodium EDTA"
Di- sodium EDTA
Di -sodium EDTA regenerated
solution0
DTPA (Cadmium electrodes)
DTPA (Cadmium electrodes)
DTPA (Cadmium electrodes)
Tetra-sodium EDTA (1.5% iron)
Tetra-sodium EDTA (1.5% iron)
Tetra-sodium EDTA (1.5% iron)
DTPA (Ionics membrane)
DTPA (Ionics membrane)
DTPA (Ionics membrane)
* Experiment was performed using
Pvn at* inian^ i*ia c- nav*<Łs\v«mjis4 no 4 MA
Run No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Current
density,
ma/cm
25
25
25
.15
25
15
25
25
15
25
15
25
25
25
25
25
15
25
25
25
25
25
15
15
the Tetra-sodium EDTA
Lead
cone. ,
0.8
0.8
0.8
0.8
0.8
0.8
0.8
4
4
4
4
4
4
0.8
0.8
4
0.8
4
4
4
4
4
0.8
0.8
solution
Reaction
% Membrane time, hr
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
DuPont Nafion
Ionics
Ionics
Ionics
from runs 2 and 12.
(? .._ _ * _ 1 A
3
3
5
3
3
3
5
3
5
4
5
3
3
3
3
3
3
3
3
3
3
3
3
3
PH
9
9
9
9
9
9
9
9
9
9
9
9
9
5
5
9
9
9
7
9
11:5
9
9
9
_„„,_„, ,,,™.. „ ,.„„ frf^i iVI in^u MWIII^J 1^1 iv- L/i*f-\ jv i u \f i un i cycuc i QUCU i i uiti i Uilo *r dflU y. r col
Experiment was performed using the Di-sodium EDTA solution regenerated from run 14.
193
-------
LEAD RECOVERY
The amount of lead recovered on the cathode is critical to making the
electromembrane process cost-effective. Representative lead recoveries achieved
by the various chelating agents (Na2 EDTA, Na2 EDTA/regenerated, Na4 EDTA, DPTA)
for 0.8 percent lead and 25 ma/cm2 current density are presented in Figure 1.
The highest lead removal (99.3 %) occurred when using di-sodium EDTA. The
regenerated di-sodium EDTA achieved a 91 percent lead removal rate. Similar
results were obtained for the tetra-sodium EDTA (not shown on Figure 1)
A comparison of the data obtained in the tests performed using initial
target lead concentrations of 0.8 and 4 percent showed that a higher percentage
of lead was recovered from the 0.8 percent lead solutions, but the mass of lead
recovered was greater in the 4 percent lead tests. One possible explanation the
lead removal rates were lower in the 4 percent solution tests is that the surface
area of the cathode may have been "saturated", and unable to plate additional
material. Additional data will be provided during the presentation.
REFERENCES
1. PEI Associates, Inc. Innovative Electromembrane Process for Recovery of Lead
from Contaminated Soils. National Science Foundation Grant No. ISI-8560730.
July, 1986
2. Assessment of Current Treatment Techniques at Superfund Battery Sites.
Bureau of Mines. W. Schmidt. 1990
194
-------
03
O
Ł
03 O)
0) <
-
o .Ł•
•+-«
C 03
Q. CD
Ł CD
CD
D
O)
LL
1C
C
3
T3 CM
« Ł
CO
CJ ~*r "O CM
Ł « E
^ o> 5
^o CO
E
Z! co co
: Q
LLI
DC
CM
CO CO
SIS
^> O CO
^^
1 O CM
O
o>
o
Tf ^ 73 CM
CCM ^ E
QCZ°'cM
4
195
-------
ULTRAVIOLET LIGHT DEGRADATION OF POLYCHLORINATED BIPHENYLS
Marilyn Barger, Department of Engineering, 133 Hofstra University, Hempstead, NY 11550-1090
(516-463-6019)
Teresa Cain, Stephen Parus, Department of Chemistry; and
Elizabeth Carraway, and Walter J. Weber, Jr., Department of Environmental Engineering and Water
Resources Engineering, University of Michigan, Ann Arbor, Michigan 48109 (313-763-2274)
INTRODUCTION
Polychlorinated biphenyls, PCBs, and other chlorinated aromatic compounds have had
widespread industrial usage during the past years due to their chemical and thermal stability and
excellent dielectric properties (1,2,3,4,5,6). Currently, they are illegal for most applications and are
being phased out of others. Highly chlorinated PCB congeners have been implicated as carcinogens
and mutagens, and all congeners are thought to alter liver functions and negatively impact
longevity (1,7). Ironically, the desirable properties of PCBs, which made them so desirable for
industrial and commercial applications are the properties now responsible for their long term
persistence in the environment (1,8,9).
There have been many approaches to treating PCB contaminated waters using ultraviolet
light. Degradation of organics can be induced by exposure of the target compound to
electromagnetic radiation of a wavelength which the compound absorbs. With sufficient overlap,
light may be used to effect the photolytic reaction directly. For poor spectral overlap, the
degradation can be induced, for example, using hydrogen peroxide as a radical initiator in the
presence of light (10). Additionally, the light can be used indirectly with a variety of chemical
sensitizers or surfaces that enhance the rate of degradation by offering a more efficient and/or less
energetic pathway (11,12,13).
The photolysis of PCBs has been studied for over twenty years. Many of these studies
were carried out using natural solar radiation (A> 290 nm) or artificial light sources that emit
continuous radiation in the same region of the electromagnetic spectrum (4). The environmental
fate of PCBs contaminating the environment was the focus of these early research activities
utilizing sunlight. The results of these environmental fate studies show that limited amounts of
dechlorination occurred, and that the more highly chlorinated congeners are the most susceptible to
the dechlorination. Additionally, for long exposure times in natural systems, undesirable
condensation products were identified. The results led to the conclusion that minimal degradation
of PCBs contaminating air, water, and soil environments could be expected from the natural
interaction with sunlight. Unfortunately, the lowest wavelength emitted by the light sources used
in the fate experiments, including the sun, overlap only the low energy tail of the primary
absorption band of the substituted biphenyls. This poor overlap of the light source and the
absorption bands of the target compounds results in very inefficient and very slow photolysis.
Specific results of early photolysis studies of PCBs include some product identification and
rate information. Degradation experiments performed in methanol produced methoxylated products
via nucleophilic substitution (5,13). Results of sensitized photodegradation produced similar
products (14,15). Photolysis in water yielded hydroxylated biphenyls and perhaps some ethers,
including the more toxic dibenzofurans and dioxins (6,16). The low solubility of PCBs in water
often necessitates the addition of an organic co-solvent and acetonitrile has been used most often
(3). Condensation products have also been identified as products in some of the longer exposure
time experiments (6). The rates of the degradation in these experiments were on the order of
hours to days.
196
-------
The objectives of this research are to investigate the feasibility of using direct photolysis for
the degradation of PCBs at discrete wavelengths between 185 and 266 nm. At these
wavelengths, direct photolysis should be enhanced due to the increased overlap of the source
radiation and the absorption band of the target chemical. Most PCB's have ultraviolet absorption
bands near those of unsubstituted biphenyl, which absorbs strongly at 202 and 248 nm. The main
thrust of this work is to understand the influence of the operating parameters (wavelength,
exposure time, and power level) in aqueous solutions, organic solvents, and in pure compound
phases, on degradation rates, efficiencies, and final photoproducts. Photolysis kinetics and
mechanisms were determined as necessary in support of developing optimal operating conditions.
METHODOLOGY
All PCB congeners were obtained in the purest forms commercially available from Ultra
Scientific Co. Stock solutions (0.00026 M) of the PCBs (2-chlorobiphenyl, 3,4-chlorobiphenyl,
2,2'-dichlorobiphenyl, 2,2',6,6'-tetrachlorobiphenyl, and 2,2',4,4',6,6'-hexachlorobiphenyl) were
prepared in spectral grade methanol and acetonitrile purchased from Mallinkcrodt. Aqueous
solutions were prepared as needed in distilled deionized water (Millipore-Waters mill-Q) water at
concentrations near saturation of the individual congeners.
The output power of the mercury lamp was measured with a Newport Corp. Model 815
power meter with ultraviolet sensitive photodiode. A bandpass filter passed only the 254 nm line.
Some mercury lamp experiments used a colored glass filter to pass the 254 nm line and exclude
the 185 nm line. Such experiments are termed filtered lamp experiments. The mercury lamp
power was also determined by ferrioxalate actinometry (17). Although low pressure mercury lamps
are characterized predominately by emission at 254 nm, they are also known to emit small
amounts (< 1 %) of 185 nm light (18).
' The photolysis experiments of 3 ml samples of solution phase PCBs were conducted as
continuously stirred, batch experiments in a one centimeter square suprasil quartz cuvette, (Spectra
Cell). Mercury lamp photolysis was performed at a distance of one centimeter from the lamp and
laser experiments at convenient distance inside the focal point of the focused laser beam. The
latter distance was adjusted so that the laser beam profile would be the same size as the cuvette
containing the PCB solution. The filtered mercury lamp experiments were run with the 185 nm
absorbing filter placed directly between the sample cuvette and the mercury lamp. During the
photolysis experiments, 200-microliter samples were taken at timed intervals for analysis.
Samples were analyzed as soon as possible after photolysis for the starting PCB congener,
unsubstituted biphenyl, as well as substitution and rearrangement products using a Hewlett
Packard, HP 5890 Series II gas chromatograph, with a 30 meter x 0.53 mm (id) DB-1 column
(J&W Scientific Co.) column fitted with an HMD photoionization (PID), and/or a 30 meter x 0.53
mm (id) DB-5 column (J&W Scientific) with an electron capture detector (ECD). Ultraviolet spectra
and absorption measurements were made on a Varian DMS 200 Ultra Violet Spectrophotometer.
Mass spectra were obtained using a Finnigan MAT INCOS 50 Mass Spectrophotometer interfaced
to an HP 5890 gas chromatograph with a 30.0 meter x 0.24 mm (id) DB-5 column.
Unfiltered mercury lamp (185 and 254 nm) experiments were performed on the stock
solutions of individual congeners: 2-chlorobiphenyl, 3-chlorobiphenyl, 4-chlorobiphenyl,
2,4-dichlorobiphenyl, 2,4'-dichlorobiphenyl, 2,2'-dichlorobiphenyl, 2,2',6,6'-tetrachlorobiphenyl
(tetPCB), and 2,2',4,4',6,6-'-hexachlorobiphenyl (HexPCB) in either acetonitrile or methanol.
Additionally, 2-chlorobiphenyl was photolyzed in water. Small amounts of acetonitrile or methanol
were added to aqueous samples to increase the solubility.
197
-------
Aqueous degradation experiments on 2-chlorobiphenyl were monitored by observing
changes in the uv spectrum during the reaction. Loss of the biphenyl ring structure's absorbance
band near 250 nm as well as the growth and decay of the smaller secondary absorbance band near
270 nm of the hydroxylated product were followed. The degradation appears to be complete in
less than 10 minutes, and the rate constant is approximately 0.398 min-1. Photolysis of the
2-chlorobiphenyl in water was also monitored using a HPLC with UV and fluorescence detectors.
The degradation rate of the congener was 0.635 min-1, slightly faster than the degradation in pure
methanol.
RESULTS
This presentation will summarize and discuss the rate data for the photolysis of chlorinated
biphenyl congeners in water, methanol, and acetonitrile using the 254 nm and 185 nm light of the
mercury lamp. The first order rate constants varied with solvent and wavelengths from a maximum
of 0.7074 to 0.0094 minutes"1. Experiments using the laser as the light source with wavelengths
of 193, 248, and 266 nm produced first order rate constants ranging from 0.101 to 3.30 joules"1.
Results of several mechanistic experiments also confirmed the reaction pathways to be homolytic
or heteolytical cleavage depending upon the solvent system.
CONCLUSIONS
Direct photolysis of methanolic solutions of low chlorinated PCB congeners using the low
pressure mercury lamp with a primary emission at 254 nm occurs through a highly efficient
heterolytic cleavage of the carbon-chlorine bond. Alternatively, direct photolytic degradations in
acetonitrile occur via a homolytic cleavage at a considerably slower rate. The fastest degradation
occurred in water, however, this occurred only slightly faster than that in pure methanol, and small
amounts of methanol were added to increase the solubility of 2-chlorobiphenyl in the water. The
primary photoproducts are the same that have been identified previously by others for direct
photolysis of these same congeners under exposure to longer wavelength (>290 nm) radiation.
Low wavelength photolysis occurred much faster than previously reported rates at these longer
wavelengths.
Comparison of mercury lamp filtered and unfiltered experiments reveal some decline in the
photodegradation rates for experiments on the same congener. Rate constants were lowered an
average of 35%, which includes approximately 10% attenuation by the optical filter itself.
Photodegradation experiments with a monochromatic source with an output only near 185 nm and
having approximately the same power, should help elucidate the effects, if any, of this emission
from the low pressure mercury lamp. Lack of significant degradation of 2-chlorobiphenyl and other
congeners using the ArF excimer laser (193 nm) suggest that the methanol and acetonitrile used as
solvents were absorbing much of this low wavelength light, regardless of its intensity. Using the
KrF excimer laser with emission at 248 nm or low pressure mercury lamp with emissions at 254
nm and 185 nm, very efficient degradation was achieved in all solvents. Finally, destruction of the
biphenyl ring structure has been observed during the short wavelength photolysis, indicating a more
extensive breakdown of the parent chlorinated biphenyl.
REFERENCES
1. Hutzinger, 0., S. Safe, and V. Zitko, 1974, The Chemistry of PCBs, CRC Press,
Cleveland, Ohio.
2. Safe, S., 0. Hutzinger, 1971, "Polychlorinated Biphenyls: Photolysis of 2,2',4,4',6,6'-
198
-------
3. Bunce, N. J., 1982, "Photodechlorination of PCBs: Current Status,"Chemosphere, 11:8,
707-714.
4. Bunce, N.J., and Y. Kumar, 1978, "An Assessment of the Impact of Solar Degradation of
Polychlorinated Biphenyls in the Aquatic Environment," Chemosphere, 11:2, 155-164.
5. Ackerman, D. G., 1983, "Destruction and Disposal of PCBs by Thermal and Non-Thermal
Methods," in Pollution Technology Review, No. 97, Noyes Data Corporation.
6. Epling, G. A., E. M. Florio, A. J. Bourque, X. Qian and J. D. Stuart, 1988, "Borohydride,
Micellar, and Exciplex-Enhanced Dechlorination of Chlorobiphenyls," Environ, Sci and
Technol., 22:952-956.
7. Brown, J. F., 1992, "The Human Health Effects of PCBs: Update'91", Proceedings: 1991
EPRI PCB Seminar, Baltimore, MD.
8. Hallet, D., 1988, "Ecosystem Management of Persistent Toxic Contaminants and
Ecosystem Health: Back to Basics," in Toxic Contaminants and Ecosystems Health: A
Great Lakes Focus, John Wiley and Sons, New York, New York.
9. Simmons, M., 1984, "PCB Contamination in the Great Lakes," in Toxic Contamination in
the Great Lakes, John Wiley and Sons, New York, New York.
10. Bull, R. A. and J. D. Zeff, 1991, "Hydrogen Peroxide in Advanced Oxidation Processes for
Treatment of Industrial Process and Contaminated Groundwater," in Proceedings: First
International Symposium on Chemical Oxidation: Technologies for the Nineties, Technomic
Publishing Co., Nashville, Tennessee.
11. Ollis, D. F. and C. Turchi, 1990, "Heterogeneous Photocatalysis for Water Purification:
Contaminant Mineralization Kinetics and Elementary Reactor Analysis," Environmental
Progress, 9:4, 229-234.
12. Ollis, D. F., E. Pelizzetti and N. Serpone, 1989, "Heterogeneous Photocatalysis in the
Environment: Application to Water Purification," in Photocatalysis: Fundamentals and
Applications, John Wiley and Sons, New York, New York.
13. Safe, S., N. J. Bunce, B. Chittim, 0. Hutzinger, and L. 0. Ruzo, 1976, "Identification and
Analysis of Organic Pollutants in Water," in Photochemistry of Bioactive Compounds.
Photodecomposition of Halogenated Aromatic Compounds, Ann Arbor Press, Ann Arbor,
Michigan.
14. Bunce, N. J., 1982, "Photolysis of Aryl Chlorides with Aliphatic Amines", J. Organ. Chem.,
47, 1948-1955.
15. Epling, G. A., Q. Wang and Q. Qiu, 1991, "Efficient Utilization of Visible Light in the
Photoreduction of Chloroaromatic Compounds," Chemosphere, 22:9-10, 959-962.
16. Crosby, D. G. and K. W. Moilanen, 1973, "Photodecomposition of Chlorinated Biphenyls
and Dibenzofurans^ Bull. Environ. Cont. and Toxicol., 10:6, 372-377.
17. Hatcher, C. G., and C. A. Parker, 1956, Proceedings of the Royal Society of London Series
; A, 235, 518-536.
199
-------
18. Rabek, J. F., 1982, Experimental Methods in Photochemistry and Photophysics, Part I,
John Wiley and Sons, New York, New York.
For More Information: Marilyn Barger, Department of Engineering, 133 Hofstra University,
Hempstead, NY 11550-1090, 516-463-6019.
200
-------
REMEDIATION OF METAL-CONTAMINATED SOIL BY ELECTRIC FIELDS
Ronald F. Probstein .
Department of Mechanical Engineering
Massachusetts Institute of Technology
77 Massachusetts Avenue
Cambridge, MA 02139
(617) 253-2240
INTRODUCTION
In electric-field restoration, electrodes are placed in the contaminated soil and a direct-current (dc)
potential is applied across them (1). The electric field drives the contaminants to the electrode wells
from where they can be brought to the surface. The process is essentially one of soil flushing, but has
several advantages over the usual pressure-driven pumping process. For example, the transport rate
induced by an electric field is not reduced by decreasing soil permeability, and the path followed by the
contaminants is confined by the electric field to the region between the electrodes. Electric-field
restoration can consequently be used in soils of low or variable permeability and in situations where
precise control over the movement of the contaminants is required.
The electric field drives the contaminants toward the electrodes by two mechanisms. One is
electroosmosis, an electrokinetic phenomenon, in which the saturating liquid and dissolved substances
flow toward an electrode. The electroosmotic flow rate is proportional to the applied electric field
strength and the charge difference or "zeta potential" across the soil-liquid interface. The value of the
zeta potential is very sensitive to soil properties and the composition of the saturating liquid. Water in
silt and clay formations typically has a zeta potential of -10 to -100 mV, and on application of a
100 V/m electric field will flow toward the cathode at a velocity of the order of 10 cm/day.
If the dissolved contaminants are ionic and carry a charge, then they will experience a force in the
electric field that will result in an additional velocity being imposed upon them. This electromigration
velocity is proportional to the product of the applied electric field and the charge on the ions. Because
metal containing wastes typically have high ionic strengths which suppress the zeta potential,
electromigration may be the dominant or only transport mechanism when treating heavy metal wastes.
Removal of heavy metals by electromigration might be expected to be a relatively straightforward
process. The velocity of the ions is generally from one to several orders of magnitude higher than the
electroosmotic flow, and is not affected by variations in the zeta potential and soil type. However,
laboratory tests have indicated that whereas removal of dissolved organic (uncharged or only weakly
dissociated) molecules by electroosmosis is invariably successful, poor removal efficiencies are
sometimes obtained with metal wastes.
In this paper, the results of an experimental investigation in which zinc is removed from clay are
presented to highlight the anomalous results that may be obtained with metals. The underlying physics
leading to the poor removal is then discussed, and means of attaining high removal efficiency
described. A mathematical model is then developed to simulate the metal removal process, and is
solved numerically.
Much of this work has appeared or will appear in other publications (1-3).
201
-------
METHODOLOGY
Experiments were conducted to evaluate the effect of the changes in soil pH that occur during the
electrorestoration process on the efficiency of metal removal. Georgian kaolin clay was saturated with
a 500 mg/L zinc solution, placed in a 20 cm long Plexiglas tube, and held in place by stainless steel
screens covered with filter paper. The tube was then connected on each end to an electrode well
containing carbon electrodes across which a d.c. potential was applied. The stainless steel screens
served as reference electrodes for measuring the potential drop across the soil. Plastic tubing
connected the electrode wells to reservoirs which contained the purge solutions and rinses. Pressure
induced flow through the soil was prevented by equalizing the static heads in the connecting tubing.
Ports in the electrode wells vented the oxygen and hydrogen produced by the electrolysis of water at
the carbon electrodes.
The pH of the saturating solution was adjusted to 6 with sodium hydroxide solution. As a result, the
solution had a relatively high concentration of sodium ions (-0.1 M). In some experiments, pH-
indicators were added to visualize pH changes in the clay. The saturated clay was allowed to
equilibrate for one day before the prepared sample was placed in the Plexiglas cylinder using a
vibrating table. The system stood for another day before applying the electric field. The experiments
were run with a constant voltage of 20 V across the clay as measured at the reference electrodes.
Most experiments were run for about 9 days. During the experiment, the volume of fluid collected in
the cathode reservoir was monitored as well as the current through the cell and the voltages across the
reference and working electrodes. Samples were withdrawn periodically from the electrode wells for
determination of pH and dissolved zinc concentration. Zinc concentrations were determined using a
high pressure liquid chromatograph or a spectrophotometer.
At the end of each run the anode and cathode solutions were analyzed for dissolved and total zinc. In
addition, the clay was divided into ten sections and each section analyzed for its pH, average zinc
concentration and liquid fraction. The pH was estimated by placing a pH-paper in contact with the clay.
To determine the average dissolved zinc concentration, the pore solution was separated from a portion
of a clay section by centrifugation. Another portion of the clay sample was acidified with nitric acid
before centrifugation for determining the total zinc concentration. The water content was determined by
comparing the weight of the remaining clay section before and after drying in an oven.
To model the process, the convective diffusion equation was solved numerically. For 1-D flows, the
overall transport rate for species i is:
8Cf
~dt
dx
t dx
dx
(1)
where C is the concentration, D the diffusion coefficient, T the tortuosity, z the charge number, v the
mobility, F the Faraday constant, the voltage, uc the convection velocity, and Rj and R," are the molar
rates of production due to chemical reactions and sorption, respectively. The first term inside the
brackets is the molar flux due to electromigration and the second term is the molar flux due to
convection. This convective transport originates from the electroosmotic flow. The rate of transport
due to diffusion is normally negligible relative to migration and osmosis, except in locations where very
steep concentration profiles develop.
There are usually a number T3f chemical species that have to be considered in electric field restoration.
These include the contaminants and co-contaminants that have percolated from the surface, as well as
naturally occurring substances such as soil electrolytes, carbonates, and humic materials. These
substances affect the transport process by modifying the electrical properties of the media, and most of
202
-------
them participate in chemical reactions. Reactions include dissociation, for example of water or a weak
organic acid, into ions, sorption onto the soil, and complexation or precipitation reactions.
The sorption rate, R*, is essentially a constraint characterized by a particular isotherm, and it relates
the concentration in the adsorbed phase to that in the bujk solution. The reaction rate, R,, is eliminated
from the equation by using the equilibrium constant for the reaction and the conservation of mass of the
elements to determine the equilibrium concentration of the species. The equilibrium assumption is
justified in that the time constants of the reactions and sorption processes are normally much smaller
than the transport time constants. This may not be valid for dissolution of precipitates.
After satisfying the electroneutrality condition, SZ|C, = 0, there are two unknowns remaining in Eq.(1),
namely the electric field and the convection velocity. The electric field is obtained from the equation for
the current density, i
(2)
The current remains constant along the length of the cell, so the electric field varies with position to
compensate for variations in concentration. Because the total applied voltage is known, Eq.(2) provides
a relationship between 3<|>/3x and Civ .
At the electrodes, species are oxidized or reduced with release or capture of electrons from the external
circuit. The rate of the electrode reactions is fixed since the current in the electrolyte (flow of ions)
must match the current in the external circuit (flow of electrons). In the usual case of inert electrodes in
which the electrode reaction is the electrolysis of water to produce hydrogen (anode) or hydroxyl
(cathode) ions, the reaction rate and the current need not be known. Instead, the concentration of
hydrogen and hydroxyl ions is obtained from the electroneutrality condition.
The second unknown, uc, is constant in space and is given by
«c =
(3)
where e is the permittivity of the fluid, u its viscosity, C the zeta potential, and < > denotes a volume
average. In the zinc experiments, the convection velocity was small and does not significantly affect
the overall metal transport.
RESULTS
Electrolysis of water occurred at the electrodes, with hydrogen ions being produced at the anode and
hydroxyl ions at the cathode. The pH was found to reach its final value of around 1 at the anode, and
about 12 to 13 at the cathode, within the first 15 minutes of a run, and then to remain essentially
constant. The hydrogen and hydroxyl ions produced by electrolysis migrated into the soil as waves of
low and high pH. The velocity of both front? remained fairly constant, with that of the acid front
measured at 0.56 m/day, which is 1.9 times the 0.28 m/day measured for the alkaline front.
The measured distribution of pH and total (dissolved plus precipitated) zinc in the clay at the end of a
9-day run is shown in Figure 1. The steepness of the pH jump and its influence on the zinc distribution
can be clearly seen. Near the anode the zinc is at a relatively low concentration and is in the dissolved
state. Near the cathode the pH is high so the zinc is present as the hydroxide precipitate jn amounts
203
-------
Q.
CL
14
12
10
8
6
4
2
0
pH
0.0
0.5
1.0 0.0
1 1 1 1 1 1 I 1 1
Concentration
Initial
concentration
It
12
10
8
6
4
2
n
i i i i i i i i i
- -
- -
_ . Initial pH
-/ t—
fr*
..•---•--•-•••/ ~
• i i i l 1 r i i i
Initial
concentration
• • • Experiment
Model
1.0
0.5
0.0
0.5 1.0 0.0 0.5
Normalized distance from anode
0.0
1.0
C
o
o
c
8
o
C
N
Figure 1. Distribution of pH (left) and total zinc concentration (right) after a 9-day test.
Top: Without electrode rinse. Bottom: Cathode rinsed with tap water.
approximately equal to the initial concentration. Some 40 to 60 percent of the original zinc is found
precipitated in a narrow band at the position of the pH jump, where its concentration reaches some 10
times the original concentration. The measured removal efficiencies were poor at from 2 to 10 percent.
To eliminate this isoelectric focusing effect (1), the hydroxyl ions must be prevented from entering the
soil. This can be achieved by rinsing them away or neutralizing them as they are formed, or by
promoting conditions for an alternative electrode reaction that does not produce hydroxyl ions. In our
experiments, the hydroxyl ions were simply rinsed away with tap water. This kept the pH at the
cathode neutral with only slightjncreases at the beginning of the experiment due to the initially high
currents. This process of rinsing the electrodes enhanced the zinc removal efficiency to 98%. The final
pH and zinc distributions in the clay for the case of the cathode rinse are also shown in Figure 1. The
pH is reduced to below 2 throughout most of the soil, and only approaches its maximum value of 4 at
the cathode. The measured zinc concentration is everywhere less than one-quarter its initial value.
204
-------
The numerical solutions to Eq.(1) indicate that very sharp changes develop in the pH and species
concentration profiles. These kinematic wave fronts are formed because of the very limited role of
molecular diffusion relative to electromigration and convection. Due to its small length scale, the
diffusion term is important only in the vicinity of the discontinuities. To simulate this effect properly, the
length scale of the computational elements had to be very small. In addition, the time increment also
has to be reduced considerably to avoid numerical instabilities.
The numerical solutions are included in Figure 1. To simulate the electrode rinse, the pH at the
cathode was fixed at pH 7. As can be seen, the mathematical model shows good agreement with the
experimental data for both the rinsed and unrinsed cases.
CONCLUSIONS
When the cathode was not rinsed, the zinc was not removed from the soil but concentrated at an
intermediate point between the electrodes by a process known as isoelectric focusing. The focal point
coincides with the pH of minimum solubility of the metal, and occurs at the position where the pH
increases sharply.
Application of the model identified the physical phenomena that importantly effect metal removal
efficiencies. The value of the pH on either side of the jump is dependent on the concentration of
soluble ions present in the saturating solution. These ions migrate and accumulate near the electrodes.
To satisfy electroneutrality, hydrogen ions accumulate with the anions, and hydroxyl ions accumulate
with the cations. In the presence of a metal contaminant, the increase in pH in the cathode region is
buffered to some extent by consumption of the hydroxyl ions to form the hydroxide precipitate. As the
pH increases further, the precipitate may redissolve to produce, typically, negatively charged hydroxo
complexes and hydrogen ions, further buffering the pH rise. The metal cations and negatively charged
complexes migrate toward each other and precipitate at the position of the pH jump.
The conductivity is also dependent on the concentration of ions in the saturating solution. It is
relatively high everywhere in the clay excepting at the position of the pH jump where it is very low. At
this position, the hydrogen and hydroxyl ions are at their minimum values, the metal ions have
precipitated, and other soluble ions have migrated toward the electrodes. Consequently, the electric
field is very high in the region of the pH jump. The growth of the electric field is relatively unstable
because any ions in this region migrate away rapidly under the action of the increasing field strength.
Both the model and experiment show that problems related to isoelectric focusing can be prevented by
simply rinsing away the hydroxyl ions generated at the cathode, resulting in better than 98 percent zinc
removal. To simulate specific site conditions, the model must be extended to include the chemistry of
soil electrolytes, organic matter, and co-disposed substances, and should also include thermal effects.
REFERENCES
1. Probstein, Ronald F. and Hicks, R. Edwin. Removal of contaminants from soils by electric
fields. Science. 260:498-503,1993.
2. Jacobs, Richard A., Sengun, Mehmet Z., Hicks, R. Edwin and Probstein, Ronald F. Model and
experiments on soil remediation by electric fields. Paper presented at Emerging Technologies in
Hazardous Waste Management V, American Chemical Society, Atlanta, Georgia. September
27, 1993.
3. Hicks, R. Edwin and Tondorf, Sebastian. Electrorestoration of metal contaminated soils. To be
submitted to Environ. Sci. Technol. 1994.
205
-------
FATE OF TERRENES IN THE ACTIVATED SLUDGE PROCESS
Franklin R. Alvarez
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 West Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7631
Daniel L. Perrin
IT Corporation
11499 Chester Road
Cincinnati. Ohio 45246
(513) 782-4700
INTRODUCTION
Chlorinated solvents are used extensively throughout industry for metal and printed circuit board
cleaning applications. The U.S. Environmental Protection Agency's (EPA) Office of Pollution
Prevention and Toxics, previously known as the Office of Toxic Substances, has been investigating
alternative materials for metal and printed circuit board cleaning applications because of the
potential adverse environmental and occupational health effects associated with chlorinated solvents.
Terpenes and certain aqueous cleaners have been found to have potential for replacing chlorinated
solvent usage in many cleaning applications. Recently, the industrial demand for these alternative
cleaning materials has risen sharply.
The use of terpenes is anticipated to have many environmental and occupational health advantages
over chlorinated solvents; however, the fate of these compounds in wastewater treatment processes is
largely unknown. Concerns regarding the potential aquatic toxicity of these compounds and their
anticipated increase in usage as solvent substitutes prompted EPA to sponsor this study of terpene
fate in the activated sludge process.
METHODOLOGY
Blank Activated Sludge Test with d-limonene
During this test, terpene laboratory analytical-methods were tested and verified through a "blank
activated sludge test", wherein the pilot-scale activated sludge unit does not contain biomass but is
fed with tap water spiked with dilute concentration of a terpene-based cleaner. DUSQUEEZE, a citrus-
based industrial cleaner containing d-limonene, was used for the tests. Operation in a "blank
activated sludge test" mode eliminated biodegradation as a removal mechanism. Therefore, the validity
of the liquid and air measurement techniques were verified by monitoring the pilot systems inlet and
outlet streams and performing a material balance by comparing the sum of the outlet streams with the
quantity measured in the inlet. Specifically, the analytical technique's were successful in providing
mass balance closure within 10% around the aeration basin of the pilot scale activated sludge system.
Three pilot-scale activated sludge units were operated using 0.5 L/min of municipal wastewater
from the City of Cincinnati spiked with terpene-based industrial cleaners.
Fate of d-Limonene in Wastewater
Operation of three pilot-scale activated sludge systems was initiated in late October 1992, and
has continued to the present. The units were allowed to acclimate to the d-limonene for approximately
two months. Performance of the activated sludge unit was monitored through analyses of effluent total
suspended solids (TSS) levels.and total chemical oxygen demand (COD) removal efficiencies. Monitoring
of d-L1monene fate was officially initiated in January 1993 and lasted one month. All d-limonene
analyses were performed utilizing a gas chromatograph at the T&E Facility.
206
-------
CONCLUSIONS
The following conclusions were drawn from the analysis of the data collected from the three
activated sludge units in January 1993:
• Greater than 95% of the d-limonene was removed from.the typical activated sludge process at
wastewater influent concentrations less than 10 mg/L. Removals greater than 99Ł were
achievable. D-limonene discharge concentrations below 10 /jg/L. were typical.
• The primary method of d-limonene removal from a typical activated sludge process was
biodegradation. Between 47 and 81% of the d-limonene entering the primary clarifier could be
biodegraded. Between 55 and 95% of the d-limonene entering the aeration basin could be
biodegraded.
207
-------
EXPERIMENTAL AND DETAILED MODELING STUDIES OF PYROLYTIC AND
OXIDATIVE PROCESSING OP CHLOROCARBON / HYDROCARBON SYSTEMS
Joseph W. Bozzelli and Robert B. Barat
Dept. of Chem. Engineering, Chemistry, and Environmental Science
New Jersey Institute of Technology
Newark, NJ 07102
(201) 596-3459 (201) 596-5605
INTRODUCTION
Incineration is a viable, but often ill perceived hazardous waste disposal
option. In order to optimize incinerator performance and allay public fears,
a fundamental understanding of the chemical and physical mechanisms of waste
combustion is needed. This is especially true for chlorinated hydrocarbon
(C1HC) waste incineration, where concerns include flame stability, formation
of products of incomplete combustion (PIC's), and economy of operation.
The ClHC-induced combustion instability has at least two consequences. The
first is the formation of PIC's. During periods of intermittancy, high levels
of unburned hydrocarbon (HC) and ClHC are emitted. The second is the supple-
mental amounts of expensive HC fuel required to maintain flame stability and
temperature in the presence of ClHC.
There is a need for improved incinerator design, especially regarding after-
burners. Conditions in the upstream rotary kiln incinerator are often diffi-
cult to control, so the downstream afterburner is critical for prevention of
PIC's. There is a need to determine the mechanism of a PIC emission; i.e.,
either from a chemical constraint or a mixing constraint. Our study will
contribute to this understanding. In addition, there is a need to influence
Federal regulations regarding afterburner design and regulation to reflect the
recent findings regarding issues such as inhibition.
OBJECTIVES OF THIS RESEARCH
Our objectives are: a) to perform well characterized experiments on the oxida-
tive and pyrolytic reactions of ClHC characteristic of those found in inciner-
ator environments, b) to perform well .characterized experiments on the ef-
fects of trace quantities of ClHC and their oxidation products (primarily HC1)
in typical post-combustion conditions, c) to develop detailed comprehensive
reaction mechanisms for the oxidation and pyrolysis of ClHC, d) to determine
the fundamental pathways of ClHC-induced inhibition of HC/air combustion, e)
to utilize knowledge obtained in the above tasks to construct applied models
for designing and optimizing incineration processes, and f) to apply these
mechanisms and applied models to characterize a large scale afterburner.
APPROACH
Experiments at NJIT-have been performed on two devices: 1) Externally heated
tubular flow reactors with stable species sampling and analyses by gas chroma-
tography (GC), and 2) a Laminar flat flame with a laser induced fluorescence
diagnostic for free radical measurements, plus stable species sampling and
analyses by GC. Detailed reactions mechanisms have been constructed which are
208
-------
based on fundamental thermochemical kinetic principles. The mechanisms have
been used in plug flow reactor and premixed flat flame computer simulations to
model:the observed species profiles obtained in our experiments. Further
modeling has identified the key reaction pathways for chlorocarbon inhibition.
A complex model (based upon groupings of idealized reactors) has been con-
structed for used with NJIT mechanisms to characterize a large scale incinera-
tor afterburner at the EPA Air and Energy Engineering Research Laboratory in
Research Triangle Park.
TUBULAR FLOW REACTOR STUDIES
Experimental research with the isothermal NJIT tubular flow reactors has been
completed for reactions in the following ClHC/HC/oxygen systems:
i. Oxidation of Chloromethane / Methane / Oxygen Mixtures
ii. Oxidation of Methylene Chloride / Methane / Oxygen Mixtures
iii. Pyrolysis and Oxidation of Chloroform and Chloroform / Methane Mixtures
iv. Pyrolysis and oxidation of 1,1,1, Trichloroethane (HIT)
Recent research is on the mixed reaction system: CH2C12/CH4/Toluene/O2. Data
is typically collected at one atm absolute pressure between the temperatures
of 680 to 840 C (953 to 1113 K), at residence times from 0.2 to 2.0 seconds.
LAMINAR FLAT FLAME STUDIES
Experimental and modeling work in our laboratories has provided the foundation
for a novel hypothesis to explain ClHC-induced hydrocarbon/air flame inhibi-
tion. A major component of the hypothesis is that combustion-generated hydro-
gen chloride (HC1) inhibits the conversion of carbon monoxide (CO) to carbon
dioxide (CO2) by consuming hydroxyl radical (OH). The work performed with the
flat flame is designed to further test this hypothesis.
The experimental research with the laminar flow flat flame has been conducted
with the methane/methylchloride/air systems at both fuel-lean and fuel-rich
feeds. Experimental data from stable species probe sampling and GC analyses
have been obtained from these flames. Results show that the molar CO/CO2 ratio
increases as the chlorine content of the feed increases (for constant 0),
which is consistent with the inhibition hypothesis.
Rate-of-production (pathway) analyses of the modeling results, all of which
utilize the mechanisms developed in this and the previous, projects as de-
scribed below, have been performed. These results support the hypothesis that
OH depletion, and hence flame inhibition, occurs primarily via reaction with
flame-generated HC1. In addition, especially in the fuel rich flames, H is
also consumed by HC1 to form H2, which also consumes OH. This represents a
fundamental shift of the product slate from CO2 to H2O.
Relative OH concentration profiles have been generated for one each of the
fuel-lean and rich flames listed above (no CH3C1 and CH3C1/CH4 = 0.25) by
laser induced fluorescence (LIF) in these flames. The process required to
convert raw data signals to relative concentrations is difficult, requiring
consideration of the rotational partition functions and collisional quenching
rates, both of which are temperature dependent. We have found that accurate
and spatially precise thermocouple temperature profiles are difficult to make
in this atmospheric pressure flame. We are working to improve the reliability
209
-------
of the temperature data. In addition, work is continuing in regard to a
correct spectral interpretation of the LIF data.
THERMOCHEMICAL DATABASES DEVELOPMENT
Thermodynamic databases, developed at NJIT for use in reaction mechanisms to
model oxidation and pyrolysis, now include:
i) Chloro and dichloro methane plus chloro and dichloro ethane oxidation in
hydrogen, methane and or ethane co-fired fuels.
ii) Chloroform and Carbon Tetrachloride in H2, CH4 and/or C2H6 co-fired fuels.
iii) Highly Chlorinated Ethanes - tri chloro to hexachloro ethanes in hydro-
gen, methane and / or ethane co-fired fuels
iv) Chlorobenzene and benzene in H2, methane and / or ethane co-fired fuels.
v) Dibenzofurans and dioxins, Chlorinated Dibenzodioxins and Furans.
The NJIT Chlorocarbon Thermochemical Property Database has been significantly
improved. The Group additivity technique has been modified to include interac-
tion groups for both substituents on aromatic rings as well as multiple chlo-
rines on adjacent carbons of chloroalkanes and chloroalkenes. The improvement
in our group additivity scheme with the use of only four interaction groups is
also exhibited in the entropies and heat capacities (temperatures from 298 to
10,000 K). A second improvement is in the database for PCDDs and PCDFs. Here
we have developed a database complete enough to allow equilibrium and mecha-
nism calculations for PCDD/Fs.
NJIT DETAILED ELEMENTARY REACTION MECHANISMS
Our reaction mechanisms, each ranging from 200-300 elementary reactions and
60-80 species, developed and validated against NJIT lab flow reactor experi-
ments plus literature data where available, applicable for use in modeling
oxidation, pyrolysis, and combustion/incineration processes. Our mechanisms
are based upon the fundamental principles of transition state theory, molecu-
lar thermodynamics, thermochemical kinetics, group additivity, and Quantum
Rice-Ramsberger-Kassel (QRRK) theory. A key goal in our developments is the
applicability of the mechanism beyond range of its experimental calibration.
We have developed detailed mechanisms for:
i) Pyrolysis and Oxidation of Chloromethane and/or Methylene Chloride/Hydrogen
/ Methane / Oxygen Mixtures
ii) Pyrolysis and Oxidation of chloroform and Carbon Tetrachloride - hydrogen
/ Methane Mixtures
iii) Pyrolysis and oxidation of Tri, tetra, penta and hexachloro ethanes.
iv) A Benzene / Chlorobenzene oxidation and pyrolysis. This mechanism is
still in validation and expansion / refinement, as we continue to learn about
the pyrolysis and oxidation reactions of aromatics such as benzene.
These mechanisms are primarily for combustion applications. They have pressure
dependent rate constants, which, though published for 1 atmosphere combustion
conditions, are known at different pressures. One limitation of the mechanisms
is the lack of molecular weight growth above C2 species. (One exception here
is the benzene / Chlorobenzene system which is still under development).
We have completed a significant number of modifications and extensions to our
210
-------
benzene and chlorobenzene oxidation and pyrolysis reaction mechanism. The
current mechanism is now validated against experimental reaction data for
chlorobenzene pyrolysis and oxidation from our own laboratory and from four
other laboratories.
The mechanism generates model predictions which are in excellent agreement
with data on CO production from anisole pyrolysis reported by Lin. Similar
excellant agreement is obtained from our model as compared with the data of
Mulder and Louw at two temperatures for both product formation and anisole
decay;. The model also accurately predicts the results of Mackie et al. on
anisole pyrolysis and oxidation. The second exciting development is that the
first draft of a dioxin formation mechanism is complete and in validation;
though, it is too early to show data trends with confidence. At levels of
precursors around 1 ppm, however, the mechanism predicts levels of PCDD/Fs
several orders of magnitude lower. In this mechanism, we have a number of
routes to formation; however, we need additional routes to destruction.
Our C1HC mechanisms have been used by the MIT Group (A. Sarofim, J. Longwell)
in a collaborative effort to model stable species measured in an MIT bench
scale turbulent combustor. This work has addressed both the C1HC inhibition
issue as well as the effects of incomplete micromixing on PIC formation.
COLLABORATION WITH EPA RESEARCH TRIANGLE PARK GROUP
During the current reporting period, our collaboration with the EPA-RTP com-
bustion group began with an on-site meeting. An overall strategy was de-
veloped and experimental and modeling plans were made. The emphasis of the
research will be placed on the afterburner (secondary combustion chamber) of
the RTP incinerator simulator.
Phase I will employ a Pyretron burner with downstream choke in the afterburn-
er, while phase II will use a Variable Swirl burner and no choke. Each phase
will begin with cold flow velocity pitot tube measurements. Then SO2 tracer
studies will be performed under various combustion conditions; e.g., with and
without flow from the primary chamber (kiln). These tracer data will be
forwarded to NJIT for residence time distribution analysis. A reactor model
based on a complex combination of "ideal reactors [perfectly stirred reactor
(PSR) and plug flow reactor (PFR)] has been created for this purpose. Ideally,
the afterburner would behave as a PSR followed by a PFR. The departure from
this ideal will be determined from the tracer work. This model is currently
operational. RTP reports that cold flow and tracer studies for Phase I will
begin early in the 4th quarter of this year.
Phase III involves an extensive series of afterburner combustion tests.
Baseline runs will involve a natural gas fuel. Then chlorinated dopants will
be added (CH2C12, CHC13, and monochlorobenzene). Continuous emission monitor-
ing will be performed (CO, CO2, O2, etc.) as.well as other analyses. Staged
combustion experiments are also planned in which fuel-rich combustion is
maintained in the first zone, with added air/fuel/steam injected into the
second stage. Species and temperature data will be forwarded to NJIT for
detailed modeling using a generalized reactor engineering model developed
expressly to model the EPA afterburner. This model employs the CHEMKIN pack-
age and will utilize the NJIT mechanisms discussed above. At a later date, a
linear eddy turbulent mixing model developed at MIT will be used in modeling
the afterburner. -
211
-------
RELEASE OF CHLORINATED ORGANIC COMPOUNDS FROM A CONTAMINATED ESTUARINE
SEDIMENT
Peter L. Gess, MarkT. Prytula and Spyros G. Pavlostathis
School of Civil Engineering
Georgia Institute of Technology, Atlanta, GA 30332
(404) 894-9367
INTRODUCTION
In spite of major efforts to protect the nation's water bodies from toxic chemicals,
contamination of aquatic ecosystems still persists. Contaminated bottom sediments, previously
thought of as natural sinks for contaminants in aquatic ecosystems, are now viewed as a major
source of contamination. Due to the persistence of the sediment contaminants, emphasis has
recently been placed on the fate and transport of in-place pollutants. Although research on
contaminated sediments has intensified in the last decade, we still lack fundamental knowledge of
the processes and process interactions that govern the partitioning and transformation of sediment-
bound toxic chemicals.
Recent investigations and monitoring of the Bayou d' Inde, a tributary of the Calcasieu River
in Southwestern Louisiana, have shown that high-molecular-weight chlorinated organic compounds
(benzenes and butadiene) and metals (chromium, copper, lead, mercury, nickel) persist in the
sediments. In addition, seafood species were found to be contaminated with hexachlorobenzene
and hexachloro-1,3-butadiene and the sediments produced chronic toxicity to aquatic life (1). A
research project was initiated to investigate the mechanisms and transformations related to the fate
of the halogenated organic compounds found in the sediments of Bayou d' Inde. This report deals
with desorption kinetics and the equilibria of sediment-bound chlorinated organic compounds.
METHODOLOGY
Two sampling stations were located in the Bayou d' Inde, one in an industrial canal
connected to the Bayou d' Inde whereas a fourth station was located in the Calcasieu River above
a salt water barrier. Sediment and water samples were collected in April 1993 and analyzed for
pH, alkalinity, major anions (Cl~, S04", N03", P04"), volatile and dissolved solids, volatile and
extractable chlorinated organic compounds, total organic carbon, and dissolved organic and
inorganic carbon. In situ, aqueous parameters ~ such as pH, temperature, conductivity, salinity,
and dissolved oxygen — were also measured at a depth of 1 m.
Standard methodologies were used for most of the analyses (2, 3, 4, 5). Sediment samples
were first extracted using the Soxhlet technique in a 1:1 (v/v) acetone:hexane solvent mixture, the
extract was then concentrated and the solvent was changed to isooctane. The sediment extract
was then analyzed by gas chromatography (electron capture detector). Liquid/liquid extraction with
isooctane was used to concentrate the aqueous contaminants. Tribromobenzene was used as an
internal standard.
Desorption experiments were performed in batch reactors. Sediment samples were placed
in a series of Teflon centrifuge tubes and then diluted with deionized water allowing for a
headspace. The final solids concentration was ca. 3% (w/v). Sodium azide was added to each
tube (2 g/L final concentration). The tubes were then capped and continuously rotated on a
mechanical tumbler (1 rpm). Three tubes at a time were periodically removed and centrifuged at
12,000 rpm for 25 min. The supernatant was liquid/liquid extracted and then analyzed by gas
212
-------
chromatography. After centrifugation, Soxhlet extraction of the sediment pellet was periodically
performed for mass balance calculations. Similar batch desorption experiments were also
performed using 2-L wide-mouth glass jars.
Batch equilibrium desorption experiments were performed as follows. Differing masses of
sediment were added to twelve 160-ml serum bottles, then diluted with deionized water (sodium
azide included) allowing for headspace. Triplicate serum bottles were prepared with the following
solids concentrations: 0.6, 1.5, 2.8 and 5.7% (w/v). The serum bottles were placed on a
mechanical tumbler, and rotated continuously at 1 rpm. After 40 days, the serum bottles were
removed, their contents placed in Teflon tubes and centrifuged at 12,000 rpm for 25 min. The
supernatant was removed, liquid/liquid extracted and then analyzed by gas chromatography. The
sediment pellet of one replicate from each solids concentration was Soxhlet extracted and the
extract was then analyzed by gas chromatography.
Finally, a differential reactor (mini-column) was set-up to simulate continuous-flow
conditions. Deionized water containing 2 g/L of sodium azide was pumped at a flow rate of 3.1
mL/d through a stainless steel column containing 0.4414 g of sediment. The eluent was directly
collected in a 20-mL glass syringe and periodically analyzed by liquid/liquid extraction and gas
chromatography.
RESULTS
System Characterization
The in-situ measurements resulted in the following ranges: pH, 6 to 7.3; temperature, 19 to
29°C; conductivity, 40 to 4,630 //mho/cm; salinity, 0 to 2.5%o; dissolved oxygen, 5.7 to 5.9'mg/L.
Analyses of water samples performed in the laboratory resulted in the following parameter value
ranges: pH, 6.6 to 7.6; conductivity, 44 to 4,150 //mho/cm; total suspended solids, 6 to 20 mg/L-
dissolved organic carbon, 10.6 to 12.7 mg/L; sulfate, 4 to 76 mg/L; chloride, 3 to 1220 mg/L-
hexachlorp-1,3-butadiene (HCBD), 0.4 to 2 fjg/L; hexachlorobenzene (HCB), 1 to 2 jjg/L. Parameter
values for the sediments ranged as follows: pH, 6.6 to 7.4; total organic carbon, 2 to 4.8% dry
mass; total sulfide, 0.2 to 1.6 mg/g dry mass. The distribution of sediment particles was as
follows: sand, 66 to 71%; silt, 12 to 19%; clay, 4 to 5%; and colloids, 9 to 12%. The following
chlorinated organic compounds were identified in sediment extracts from samples collected in the
Bayou d' |nde and the industrial canal (mg/kg dry weight basis): dichlorobenzenes, 3.9;
trichlorobenzenes, 2.6 to 27.5; tetrachlorobenzenes, 0.5 to 10.4; pentachlorobenzene (PeCB) 3 4
to 20.1; HCB, 8.9 to 123; and HCBD,, 5.6 to 36. Therefore, HCBD and chlorobenzenes with high
chlorine content are the predominant sediment contaminants.
Release of Sediment-Bound Contaminants
Five chlorinated organic compounds were consistently found in the aqueous phase during
the desorption experiments: HCBD, HCB, PeCB, 1,2,4,5-tetrachlorobenzene (1245-TeCB), and
1,2,4-trichlorobenzene (124-TrCB). The desorption pattern of the chlorinated organic compounds
was biphasic: a fast first release, followed by a much slower rate of release (Fig. 1). With the
exception of 124-TrCB which resulted in ca. 3% desorption within 62 d, the other four
contaminants showed less than 0.5% desorption, based on the time zero contaminant content of
the solid phase. The carbon attributed to the chlorinated organic contaminants accounted for less
than 0.03% of the total DOC released within 62 d of incubation. The rate and extent of
contaminant desorption from sediment samples taken from different stations was inversely
correlated to the organic carbon content of the samples.
213
-------
• HO
• HOD
10
20 30 40 50
Time (d)
60
O
o-
1.0000
0.9995
0.9990
0.9985
0.9980
0.9975
0.9970
0.9965
(
N
-
™
-
—
)
\
T _
i
I
I
10 20 30 40 5
Time (d)
Figure 1. Desorption pattern of sediment-
bound chlorinated organic compounds.
Rgure 2. Desorption pattern of HCB (data
points and 95% confidence intervals) and
simulation based on the two-site model
(continuous line).
The desorption isotherms for ail five contaminants fitted the linear model with a desorption-
resistant contaminant solid-phase concentration (qr, //g/kg), as follows: q = qr + Kf Ce where: q
» total solid-phase contaminant concentration, //g/kg; Kf = partition coefficient, ml/g; and Ce =
liquid-phase contaminant concentration, fjg/m\. The partition coefficients (Kf) pertain only to the
contaminant portion that can freely achieve equilibrium. Based on the two-site model, Kf refers to
the equilibrium achieved at the fast sites. The overall partition coefficient is expressed as follows:
Kp » Kf -f Ks/ where Ks refers to the partition coefficient describing the desorption equilibrium at
the slow sites. Based on the experimental data, a fit of the two-site model resulted in the following
ranges for the K_' values (in ml/g): 124-TrCb, 276 to 410; 1245-TeCB, 2610 to 5600; PeCB, 3330
to 3930; HCB, 5190 to 5750; and HCBD, 2210 to 2520. Figure 2 shows the two-site model fit to
the desorption data of HCB. When the partition coefficients were normalized to the organic carbon
content of the samples, the following partition coefficient values resulted (log Koc): 124-TrCb, 3.76
to 3.93; 1245-TeCB, 4.73 to 5.07; PeCB, 4.84 to 4.91; HCB, 5.03 to 5.08; and HCBD, 4.66 to
4.72. These values compare favorably with literature reported K00 values (6). The first-order
constant describing the microscopic net desorption rate from the slow sites was estimated and
varied as follows (in d'1): 124-TrCB, 0.061 to 0.080; 1245-TeCB, 0.013 to 0.018; PeCB, 0.001 to
0,008; HCB, 0.002 to 0.008; and HCBD, 0.0129 to 0.0135.
When the continuous-flow mini-column was used, the extent of contaminant desorption
was greater than that previously attained by the batch desorption experiments. Because of the
high operating pressure (ca. 4,500 psig), it is likely that the eluent was in contact with sites that
were more secluded under the conditions of all previous batch experiments. After about 3600 pore
volumes of deionized water had passed through the sediment sample (in about 57 days), the
highest extent of desorption was observed for HCBD (ca. 14%). The extent of desorption was
higher for the less hydrophobia contaminants (Fig. 3). The fraction of the solid-phase contaminant
which desorbed was linearly correlated with the logK00 of the contaminants (Fig. 4). The values of
the first-order desorption constant from the slow sites were estimated as follows (in d"1): PeCB,
0.0017; HCB, 0.000013; and HCBD, 0.0485.
214
-------
• 1.Z4-TiC»
' 1.Z.4.S-T.C1
' P«C1
' HCi
' HOO
800 1600 2400 3200 4000
Pore Volumes
o 0.20
8 0.15
to
to
.Ł 0.10
o
O
0.05 -
0.00
• T.C*
Bra
A MS
• HCt
3.6 3.8 4.0
4.2 4.4 4.6
Log Koc
4.8 5.0 5.2
Figure 3. Continuous-flow desorplion of
sediment-bound chlorinated organic
compounds (mini-column).
Figure 4. Correlation between extent of
continuous-flow desorption (mini-column) and
K
of sediment-bound chlorinated organic
contaminants (HCBD was not included in the
linear regression analysis).
CONCLUSIONS
In a period of several weeks, 3% or less of the sediment-bound contaminants desorbed
under batch conditions. Continuous-flow simulation increased the overall extent of desorption, but
it did not exceeded 15% for a similar period of time. The desorption of the sediment contaminants
followed a biphasic pattern, which was described by a two-site, first-order linear model. A very
small fraction of the contaminant mass is associated with fast desorbing sites, most of the
contaminant mass undergoes a very slow desorption. Resistance to desorption is primarily due to
the high hydrophobicity of the contaminants and the relatively high organic carbon content of the
sediments.
REFERENCES
1: Cunningham P. et al. Toxics Study of the Lower Calcasieu River. PB90 226150/AS,
National Technical Information Service, Springfield, VA, 1990.
2.
3.
4.
5.
6.
American Public Health Association. Standard Methods for the Examination of Water and
Wastewater, 17th ed., APHA-AWWA-WPCF, Washington, DC, 1989.
Page, A. L. (ed.). Methods of Soil Analysis - Part 2: Chemical and Microbiological
Properties, 2nd ed., American Society of Agronomy, Inc., and Soil Science Society of
America, Inc., Madison, Wl, 1982.
U. S. Environmental Protection Agency. Test Methods for Evaluating Solid Waste SW-
846, 3rd ed., EPA/OSWER, Washington, DC., 1986.
American Society for Testing and Materials. Annual Book of ASTM Standards. Philadelphia
PA., 1990. '
Gess, P.L. Release of Chlorinated Organic Compounds from a Contaminated Estuarine
Sediment. MS Thesis, Georgia Institute of Technology, 1994.
215
-------
FLUME STUDIES ON THE DETACHMENT OF KAOLINITE CLAY
AND ASSOCIATED CONTAMINANTS FROM A COARSE SEDIMENT
T.W. Sturm. K. Dennett, A. Amirtharajah, and T. Mahmood
School of Civil Engineering
Georgia Institute of Technology
Atlanta. GA 30332
INTRODUCTION
Waste discharges from point sources and nonpoint pollution originating from
agricultural and urban runoff have led to the storage of pollutants such as pesticides and
heavy metals within the sediment bed of many water bodies. These pollutants are often
adsorbed to clay particles in fine sediments which also provide cohesiveness of the sediment
bed. The prediction of erosion or resuspension of these clay particles and associated
contaminants at the sediment-water interface in rivers, lakes, and estuaries is vital to the
hazard assessment and remediation of contaminated aquatic sediments.
Previous relationships developed for predicting erosion and resuspension of fine
sediments are highly empirical. They depend on empirical coefficients which vary with the
particular sediment-water system of interest, and which are often estimated by ad hoc
laboratory procedures that do not necessarily reproduce the hydrodynamic and physicochemical
processes occurring at the sediment-water interface. Most numerical models of the fate and
transport of fine sediments and associated contaminants depend on these highly tenuous
empirical relationships for the erosion and resuspension process.
Sediment pick-up functions for noncohesive sediments have long been formulated as a
function of the excess shear stress with respect to a critical shear stress which is required
for initiation of motion (1). This approach has enjoyed considerable success for noncohesive
sediments because the primary force resisting initiation of motion is the submerged weight of
individual sediment grains so that the critical shear stress depends primarily on grain size.
In fine sediments with a significant clay fraction, however, interparticle physicochemical
forces which cause cohesion also resist initiation of motion, and these forces have proved
much more difficult to quantify. These cohesive sediment forces depend on many factors,
which include amount ana type of clay mineral, degree of consolidation, type and
concentration of ions in the water, and the presence of natural organic matter (2).
Other mechanisms besides erosion and resuspension exist for contaminant transfer at
the sediment-water interface including such processes as bioturbation and diffusion.
However, Sheng (3) estimates that sediment resuspension fluxes associated with episodic or
diurnal events can be 10 to 100 times the diffusive flux. On the other hand, estimates of
erosion or resuspension by empirical formulas differ by as much as two orders of magnitude
(3).
Recent advances in estimating the fundamental forces required for detachment of
colloidal particles from larger sediment grains offer a new approach for quantifying the
forces resisting erosion/resuspension of cohesive sediments (4). Theoretical expressions
have been developed to estimate the magnitude of interparticle forces, including
electrostatic and van der Waals forces, which control particle detachment (4). Ongoing
research sponsored by the EPA Hazardous Substance Research Center/South and Southwest has the
goal of applying calculations of interparticle forces to the problem of estimating erosion
and resuspension rates of cohesive bottom sediments and associated contaminants as a function
of ionic strength and concentration of natural organic matter. Experimental studies are
being conducted in a laboratory flume in which the flow field is similar to the actual shear
flow experienced by field sediments. The results of these interparticle force calculations
and experimental flume studies are applicable to numerical models used in the evaluation of
the transport of contaminated bottom sediments in lakes and rivers due to natural erosion and
resuspension, resuspension due to dredging activities, and erosion of contaminated particles
as affected by salinity changes in estuaries.
216
-------
METHODOLOGY
The initial experimental approach to the problem has been to attach kaolinite clay
particles to a fine gravel which forms the movable but stable bed of a recirculating, tilting
flume, arid to measure the detachment of the clay particles as a function of applied shear
stress. A definition sketch of the flume is shown in Figure 1. It is approximately 6 m long
by 38 cm wide. The flow rate is measured by a calibrated bend meter, and uniform flow is
obtained by adjusting the tailgate. Extensive tests were conducted in the flume to establish
depth-discharge relationships for uniform, subcritical flow at slopes of 0.002, 0.003, and
0.004 with channel aspect ratios (width/depth) ranging from 5 to 10. In addition, velocity
profiles were measured using a Pitot tube connected to a differential pressure transducer,
amplifier, and voltmeter which were controlled by a microcomputer.
The measurement of center!ine velocity profiles showed that the boundary layer was
fully developed at the flume test section. Measured center!ine velocity profiles were used
to predict shear stress in a central core of the flow at the test section by measuring the
slope of the semi logarithmic velocity profile near the bed. The shear stresses measured from
the velocity profiles are compared with those obtained from the uniform flow formula in
Figure 2.
Several series of particle erosion tests in the flume have been conducted. In a
sample tray placed at the centerline of the test section of the flume, the large, noncohesive
sediment grains (fine gravel) comprising the bed are coated with a kaolinite clay suspension.
The fine gravel has a mean grain size of 3.3 mm, while the kaolinite clay has a mean particle
size of approximately 1.5 /urn. Uniform flow conditions are established, and the plume of
eroded/detached clay particles is measured as a function of time using a Chemtrac continuous
particle monitor and a Hach turbidimeter. The particle monitor measures fluctuations in
light intensity transmitted through a sample stream flowing through a transparent tube.
Continuous samples are also collected which can be analyzed for particle number concentration
and size distribution (0.5 to 150 jum) with a Brinkmann particle analyzer. A separate sample
stream is-collected for filtration analysis of total mass eroded.
RESULTS •
Typical results for a series of flume experiments are shown in Figures 3 and 4.
Figure 3 shows an increase in total mass of kaolinite clay eroded with increasing bed shear
stress. In this series of experiments, the shear stress varied from 0.9 to 1.6 N/m2. The
total mass eroded in Figure 3 was determined from three separate measurements which included
filtered mass, particle number concentration from the Brinkmann particle analyzer, and the
particle monitor output. Although it appears from these results that a critical shear stress
exists for zero mass eroded, additional experiments are needed to pinpoint its value.
In Figure 4, the cumulative mass erosion curves as determined from the particle number
concentration data are shown as a function of time for the conditions of minimum and maximum
bed shear, stress in Figure 3. The slope of the erosion curves indicates the erosion or
detachment rate per unit surface area of the contaminated bed sample. The erosion rate
decreases'with time as the total mass eroded is approached. Both the total mass eroded and
the initial erosion rate increase with increasing bed shear stress. Additional flume
experiments are planned to determine the effects of ionic strength of the eroding water and
adsorbed natural organic matter on the erosion rate of kaolinite clay as well as field
samples of contaminated sediment. These erosion rates will be parameterized based on
calculations of the microscopic forces which are required to detach colloidal-sized particles
from a sediment bed (5).
CONCLUSIONS
Erosion or detachment of kaolinite clay particles from much larger, noncohesive
sediment grains depends on excess shear stress with respect to a critical shear stress. Both
the total mass eroded and the initial time rate of erosion increase with increasing shear
stress. Future experiments should be directed toward characterizing the erosion and
transport of cohesive sediments in terms of the microscopic interparticle forces which vary
with ionic strength, type of clay, and concentration of natural organic matter.
217
-------
REFERENCES
1.
2.
3.
4.
5.
Vanoni. V. (ed.). Sedimentation Engineering.
of Civil Engineers. New York, 1977.
ASCE Manual No. 54. American Society
Mehta, A.J., Hayter. E.J., Parker, W.G., Krone, R.B., and Teeter, A.M. Cohesive
Sediment Transport, I: Process Description. ASCE Journal of Hydraulic Engineering.
115 (8): 1076-1093. 1989.
Sheng. Y.P. Task Committee on Transport Processes at the Interface of Water Column
and Bottom Sediment. In: Proceedings of the 1990 National Conference on Hydraulic
Engineering, ASCE, New York. 1990. p. 47.
Amirtharajah, A. and Raveendran, P. Detachment of Colloids from Sediments and Sand
Grains. Colloids and Surfaces. 73: 211-227. 1993.
Raveendran, P. Mechanisms of Particle Detachment During Filter Backwashing. Ph.D.
Thesis presented to the School of Civil Engineering, Georgia Institute of Technology,
Atlanta, Georgia, November. 1993.
FOR MORE INFORMATION
Contact T.W. Sturm (Tel. 404-894-2218) or A. Amirtharajah (Tel. 404-853-0628) at
School of Civil Engineering. Georgia Institute of Technology, Atlanta, GA 30332.
1.1
m
1 1
!NV.
1 1 ^
0! !
i i , —
1 1 s~~
\ V
i i
e 1.2m ,
CONTAMINATED
SAMPLE
30 em
K^.uL.i^| 10 cm
^^^M W
38«n ^^^|
, , ^~T
*-3m >
< 6m ^
PLAN
PROFILE
Figure 1. Plan and Profile of Flume.
218
-------
10
Width/Depth
8=0.002 a 8=0.003 A 8=0.004
Figure 2. Comparison of Shear Stress, T, Determined From Velocity Profiles and
Uniform Flow Formula, yDS (Y=specific weight of water. D=depth of flow, S=bed slope).
2000
73
03
-a
o
UJ
1500-
1000-
03
500-
0.0
0.5 ' 1,0 1.5
Shear Stress, N/m^2
2.0
Filtered mass a Particle number • Particle monitor
Figure 3. Total Mass of Kaolinite Clay Eroded From Sample Tray.
219
-------
10
20 30 40
Time, sec
50
60
1.0N/m~2 -B- 1.6N/m~2
Figure 4. Cumulative Mass of Eroded Kaolinite Clay as a Function of Time for
Shear Stresses of 1.0 and 1.6 N/m2.
220
-------
THE U.S. EPA INCINERATION RESEARCH FACILITY
J. W. Lee and S. Venkatesh
Acurex Environmental Corporation
3900 NCTR Road, Bldg. 45
Jefferson, Arkansas 72079
(501) 541-0004
R. C. Thurnau
Waste Minimization, Destruction, and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, OH 45268
(513) 569-7692
The U.S. EPA's Incineration Research Facility (ERF) located in Jefferson, Arkansas, houses
bench-scale and pilot-scale incinerators including a rotary kiln system (RKS) and a thermal
treatability unit (TTU). The RKS consists of a 2.0 MMBTU/hr rotary kiln primary chamber, a gas-
fired i.5 MMBTU/hr afterburner, and primary and secondary air pollution control systems (APCS).
It is capable of incinerating solid and liquid wastes. In 1993 the existing APCS was enhanced by the
installation of a flue gas reheater and fabric filter baghouse. The waste handling capabilities of the
IRF were also unproved by the addition of a scrubber/waste liquid containment tank-system with a
total storage capacity of 18000 gallons. The TTU is a small commercial pathological waste
incinerator converted to allow thermal treatment of liquid and solid wastes in a semi-batch
continuous mode. Quartz trays (8-in x 10-in x 2-in) containing up to 2 Ib of test materials each can
be subjected to thermal treatment conditions for 20 mm to 2 hr.
Three major test programs were completed in 1993 at the IRF using the RKS. The
first was hi support of the Office of Emergency and Remedial Response's (OERR) Superfund site
remediation program, and was an evaluation of the decontamination of soils contaminated with
organic constituents in the rotary kiln operating at low to moderate temperatures associated with
thermal desorption. The second was a major third-party test program in which a series of tests with
simulated low level mixed waste were conducted for the Westinghouse Savannah River Company,
the operating contractor for the Department of Energy Savannah River Plant and Savannah River
Technical Center. The third test program of 1993 was in support of EPA Region III, and the U.S.
Army Corp of Engineers to evaluate the incinerability of contaminated soil and a fluff waste from
the M.W. Manufacturing Superftmd site in Pennsylvania.
A current test program is a demonstration which is, the incineration of the Russian
Federation's ballistic missile propellant components comprised of unsymmetrical dimethyl hydrazine
(UDMH) as the fuel, and dinitrogen tetroxide as the oxidizer (N2O4). This program is supported by
the Defense Nuclear Agency as part of the Strategic Arms Reduction Treaty (START) negotiations,
and will demonstrate that incineration of UDMH and N2O4 can be achieved to levels acceptable to
both Russian and U.S. environmental regulations. An ongoing test program using the TTU was
initiated in FY93 to evaluate the effectiveness of silica/alumina/clay based additives as sorbents for
metal capture.
Results from the previous year's tests and current year activities will be highlighted hi the
poster presentation. For more information contact R. C. Thurnau, the EPA project officer at the
above address.
221
-------
FOUR YEAR STUDY OF ASBESTOS
IN NEW JERSEY PUBLIC SCHOOLS
by
John R. Kominsky
Environmental Quality Management, Inc.
1310 Kemper Meadow Drive
Cincinnati, Ohio 45240
(513) 825-7500
and
.. Thomas J. Powers
Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
EPA Contract No. 68-D2-0058
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7550
Introduction
The U.S. Environmental Protection Agency (EPA) recommends a pro-active,
in-place management program whenever asbestos-containing material is present
in buildings. Asbestos removal is required only when necessary to prevent
significant public exposure to airborne asbestos during building demolition or
renovation activities. The ultimate goal of every asbestos abatement project
is to eliminate, or reduce to the extent possible, the actual or potential
hazard airborne asbestos may present to building occupants. If all safeguards
are not properly applied, asbestos removals may actually elevate airborne
levels of asbestos in a building.
In 1988, EPA-RREL and NJDOH-EHS conducted a study to document Asbestos
Hazard Emergency Response Act (AHERA) air-sampling practices during final
clearance and to measure final clearance levels of airborne asbestos at 20
projects involving removal of asbestos-containing material (ACM) in 17 New
Jersey schools. In 1990, EPA/NJDOH conducted a study at the same 17 schools
to measure airborne asbestos levels 2 years after the abatements in 1988.
Although the 1990 study provided data regarding the residual levels of
asbestos 2 years after abatement, the extent to which these data represented
conditions of actual occupancy remained uncertain. In 1991, EPA/NJDOH
measured airborne asbestos levels under occupied conditions at the 17 schools
3 years after abatement. In 1992, EPA/NJDOH conducted a final study at the 17
schools to measure airborne asbestos levels under actual occupied conditions 4
years after abatement.
222
-------
Conclusions
1) Overall, when all of the 20 sites were considered collectively, there
was no apparent trend toward progressively increasing airborne asbestos
concentrations 2 to 4 years after the 1988 abatements.
2) Response actions conducted by the schools in 1991 and 1992 demonstrated
that elevated airborne asbestos levels can be reduced to acceptable
levels (i.e., <0.02 s/cm).
3) Asbestos-containing debris from the 1988 abatement and from
postabatement operations and maintenance (O&M) activities may have
contributed to the elevated airborne asbestos levels (>0.02 s/cm )
present in 1991 and/or 1992 at nine sites.
4) Errors in the Asbestos Management Plans relating to material
identification or material location were documented at 13 of the 17
schools.
5) When the AHERA Z-test is used to clear an abatement project, it is
generally more appropriate to utilize the outdoor samples as the
, reference point than the perimeter samples collected inside the
building.
223
-------
UPDATE OF THE EPA DEVELOPED FULL SCALE DEBRIS WASHING SYSTEM
Naomi P. Barkley
U. S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Drive
Cincinnati, OH 45268
(513)569-7854
The automated, transportable (mounted on three semi-trailers), self-contained debris washing
system (DWS) developed by IT Corporation for EPA's Risk Reduction Engineering Laboratory is ready
for deployment to a hazardous waste site for an initial demonstration of the new system. Depending
upon the debris characteristics, the DWS can decontaminate between 50 to 120 tons per day.
Experience gained under actual field conditions when the pilot-scale DWS was demonstrated under
the Superfund Innovative Technology Evaluation (SITE) Program confirmed that the DWS technology
had definite promise for addressing decontamination of metallic debris at hazardous waste sites.
Results of the pilot-scale demonstration indicated areas where improvements could be made in the
system and led to the design and development of the full-scale unit.
The full-scale system includes a 12-ft long by 7-ft wide by 8-ft deep cleaning chamber, where the
wash, spray and rinse process steps occur. Unit operations require loading a heavy-duty basket with
approximately 2 tons of crane-lifted metallic debris, positioned on cradles, in the cleaning chamber.
The hydraulically-operated cover of the cleaning chamber is closed and 2000 gallons of 160° F
detergent solution is introduced, the basket is rotated and the turbulent wash cycle is started. After the
wash cycle, the detergent solution is pumped back into the detergent holding tank and the spray cycle
starts. During this cycle detergent solution is delivered at 500 gal/min through spray nozzles located
inside the basket. A portion of the cleaning solution is cycled through a closed-loop oil-water separator
system, and then recycled back into the detergent holding tank
The rinse cycle is then instituted where clean water is applied through the spray nozzles to remove
residual-contaminated liquid from the debris surfaces. Time required for each cycle varies depending
on type of site, contamination, debris and other variables.
For more information: Michael Taylor or Majid Dosani, IT Corporation, 11499 Chester Road
Cincinnati, OH 45246, (513) 782-4700.
224
-------
AN EVALUATION OF DRINKING WATER SAMPLES PREPARED USING
ULTRAFILTRATION AND UV-TIO. TECHNOLOGIES
Kathleen S. Patterson, Lucille M. Garner,
John C. Ireland and Bradford L. Smith
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Drinking Water Research Division
Cincinnati, Ohio 45268
(513/569-7947, 569-7417, 569-7413, 569-7238)
Susan Richardson
Environmental Research Laboratory
U.S. Environmental Protection Agency
Athens, Georgia 30605
(706/546-3199)
In the United States approximately 25 million people receive their
drinking water from small systems (serving fewer than 3,300 people). These
systems are the most frequent violators of federal drinking water regulations.
As the number of regulations increase, compliance will become even more
difficult for small systems to achieve. Package plants, water treatment units
assembled in a factory that arrive at a site virtually ready to use, are
currently being evaluated as feasible treatment options for small systems.
In the present study, drinking water was prepared using a combination of
ultrafiltration and photocatalysis, both applicable to use in package plants.
Previous results indicate that ultrafiltration followed by chlorination
removed microbiological contaminants. However, based on the levels of total
organic halide (TOX) and mutagenicity observed, the levels of disinfection by-
products (DBP) in the finished water appeared to be essentially the same as in
chlorinated raw water. Consequently photocatalysis, using UV light and
titanium dioxide (TiO?), was employed. This process has been shown to cause
the oxidative destruction of many organic compounds.
At the Test and Evaluation Facility, Mill Creek water was treated using
ultrafiltration then recirculated through the UV-Ti02 unit until a 10%
reduction in total organic carbon (TOC) was produced. A portion of the water
treated by ultrafiltratiori and UV-TiO, was subsequently chlorinated. Samples
of raw water, raw water + chlorine (CM, and ultrafiltered water with and
without C12 were also prepared. TOC, TOX, and microbiological contaminants
were determined for each sample. Concentrates of the organic compounds in
each water sample were prepared by XAD resin chromatography. These
concentrates were assayed for mutagenicity in the Ames assay and used for the
identification of DBP by gas chromatography/mass spectrometry.
(This abstract does not reflect EPA policy).
For more information contact Kathleen S. Patterson, USEPA, RREL, DWRD,
SFEB, 26 W. M. L. King Drive, Cincinnati, Ohio 45268, 513/569-7947.
225
-------
RREL SITE REMEDIATION TECHNICAL SUPPORT PROGRAM
Benjamin L. Blaney, Chief
Technical Support Branch
Super-fund Technology Demonstration Division
Risk Reduction Engineering Laboratory
U. S. Environmental Protection Agency
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
(513) 569-7406
The Risk Reduction Engineering Laboratory provides a variety of technical
support to Regional Remedial Project Managers and other site remediation
managers and contractors. This poster will highlight some of these support
activities in the areas of site-specific assistance and technology transfer,
including:
- treatability study assistance
- site-type specific remediation guidance documents
- technology specific documents
- computerized databases
Forms for ordering technology transfer documents will be available.
For more information: Benjamin L. Blaney.
226
-------
THE ENVIRONMENTAL PROTECTION AGENCY'S
INNOVATIVE TECHNOLOGY PROGRAM
Norman J. Kulujian
Technical Liaison
U.S. Environmental Protection Agency
Office of Research and Development
841 Chestnut Street
Philadelphia, Pennsylvania 19107
Telephone: 215-597-1113
FAX: 215-597-3150
President Clinton announced a proposal in the State of the Union speech in February 1993, aimed
at protecting the environment while strengthening America's industrial base. The Environmental
Technology Initiative (ETI), which began this year, is a long term program that will allocate $1.85
billion dollars over the next nine years to develop more advanced process changes and treatment and
control technologies that yield environmental benefits and increase economic growth. Although the ETI
is a complete environmental technology program that considers all the factors necessary to
commercialize and sell innovative technologies, this poster session focuses on the Risk Reduction
Engineering Laboratory's research, development, and demonstration projects.
EPA's Innovative Technology Council (ITC) is an internal Agency advisory and advocacy group
dedicated to fostering the development, commercialization, and use of innovative technologies. The ITC
selected four areas to fund projects in 1994. EPA has solicited partners from outside the Agency who
will leverage funding and provide expertise in the following areas.
1. Clean Technology for Small Businesses
This area attempts to bring the benefits of pollution prevention to small companies by having EPA and
selected partners act as a convener and educator. Projects fall into three categories: technical
assistance; joint research with industry; and research on safer chemicals, products, materials, and
systems.
2. Improving Global Competitiveness of U.S. Environmental
Technologies
The U.S. Technology for International Environmental Solutions (U.S. TIES) initiative promotes technical
assistance, training and information dissemination, in-country demonstrations, evaluation and testing,
feasibility studies, market and needs assessments, and international standards development.
3. Innovative Technologies Development
The Innovative Technologies Development initiative is directed toward developing, testing, evaluating,
and bringing into the marketplace technologies that first, address the most critical needs of EPA
programs, and second, exhibit the greatest potential for important technological breakthroughs.
227
-------
4. Gaps. Barriers, and Incentives
This initiative is designed to explore the reasons why innovative technologies are not pursued by
investors, venture capitalists, and the regulated industry purchasers. This area intends to identify the
barriers, changes, and incentives that will promote the commercialization and sale of innovative
technologies.
The Risk Reduction Engineering Laboratory's innovative technology projects and partners will be
identified in handout material. In addition, information and Agency contacts will be available for the
complete ETI program.
For information on the Risk Reduction Engineering Laboratory's programs, contact:
Stephen C. James
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
26 West Martin Luther King Drive
Cincinnati, OH 45268
513-569-7877 (Telephone)
513-569-7680 (FAX)
For information on the ETI (EPA's total innovative technology program), contact:
Penny M. Hanson
Office of Environmental Engineering and Technology Demonstration
Mail Code 8301
U.S. Environmental Protection Agency
401 M Street, S.W.
Washington D.C. 20460
202-260-4073 (Telephone)
202-260-3861 (FAX)
For information on the ITC (EPA's innovative technology strategy), contact:
Norman J. Kulujian
228
-------
EVALUATION OF IN-VESSEL COMPOSTER DESIGNS
FOR HAZARDOUS WASTE TREATMENT
John A. Glaser1, Majid A. Dosani2, Carl L. Potter1
Srinivas Krishnan2, Timothy Deets2, and E. Radha Krishnan2
United States Environmental Protection Agency
Risk Reduction Engineering Laboratory
26 W. Martin Luther King Dr.
Cincinnati, Ohio 45268
2IT Corporation
11499 Chester Rd.
Cincinnati, Ohio 45246
ABSTRACT
Composting techniques have been successful in treating municipal solid waste. Recent
efforts by the U.S. Army to treat soils contaminated with munition chemicals have identified
a new application of composting to the treatment of solid materials contaminated with
hazardous waste. The munition composting has been conducted hi field situations using open
composting conditions. The mass balance of the munition chemicals in such application sis
not easy to track. Our research is designed to assess the potential of composting as applied to
hazardous waste constituents. We have conducted a developmental program to evaluate the
performance of several in-vessel composting bioreactors to ensure optimal control of the
composting process and gain information about the mass balance of the targeted contaminant.
As such the reactors are designed as diagnostic tools to evaluate composting conditions for
specific waste materials in contaminated solid matrices. These reactor systems are designed
to mimic composting conditions with instrumented control of process factors. The reactor
designs will be discussed with initial performance data.
229
-------
ACOUSTIC LOCATION OF LEAKS IN PRESSURIZED
UNDERGROUND PETROLEUM PIPELINES
Robert W. Hillger and Anthony N. Tafuri
U.S. Environmental Protection Agency
Releases Control Branch
2890 Woodbridge Ave
Edison, NJ 08837-3679
908-321-6639
ABSTRACT
Millions of underground storage tanks (USTs) are used to store petroleum and other chemicals.
The pressurized underground pipelines associated with USTs containing petroleum motor fuels are
typically 2 in. in diameter and 50 to 200 ft in length. These pipelines typically operate at pressures of 20
to 30 psl. Longer lines, with diameters up to 4 in., are found in some high-volume facilities. When a
leak is detected, the first step in the remediation process is to find its location. Passive-acoustic
measurements, combined with advanced signal-processing techniques, provide a nondestructive method
of leak location that is accurate and relatively simple, and that can be applied to a wide variety of
pipelines and pipeline products.
Experiments were conducted at the UST Test Apparatus Pipeline in which three acoustic sensors
separated by a maximum distance of 38.1 m (125 ft) were used to monitor signals produced by 11.4,
5.7, and 3.8 L/h (3.0, 1.5, and 1.0 gal\h) leaks in the wall of a 5 cm (2 in.) diameter pressurized
petroleum pipeline. The line pressures and hole diameters used in the experiments ranged from 69 to
138 kpa (10 to 20 psi) and 0.4 to 0.7 mm (0.01 to .03 in.), respectively. Application of a leak location
algorithm based upon the technique of coherence function analysis resulted in mean difference between
predicted and actual leak locations of approximately 10 cm.
Spectra computed from leak-on and leak-off time series indicate that the majority of acoustic
energy received in the far field of the leak is concentrated in a frequency band from 1 to 4 kHz. The
strength of the signal within this band was found to be proportional to the leak flow rate and line
pressure. Energy propagation from leak to sensor was observed via three types of wave motion:
longitudinal waves in the product, and longitudinal and transverse waves in the steel. The similarity
between the measured wave speed and the nominal speed of sound in gasoline suggests that
longitudinal waves in the product dominate the spectrum of received acoustic energy. The effects of
multiple-mode wave propagation and the reflection of acoustic signals within the pipeline were observed
as non-random fluctuations in the measured phase difference between sensor pairs. Additional
experiments with smaller holes and higher pressures (138 to 345 kPa [20 to 50 psi]) are required to
determine the smallest leaks that can be located over distances of several hundred feet. The current
experiments indicate that improved phase-unwrapping algorithms and/or lower noise instrumentation are
required to optimize system performance.
230
-------
PLUNGING WATER JETS: EVALUATING AN INNOVATIVE.
HIGH CURRENT. DIVERSIONARY OIL BOOM
John S. Farlow
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
289C Woodbridge Avenue (MS-104)
Edison, NJ 08837
(908) 321-6635
John M. Cunningham
Office of Emer»r:ncy and Remedial Response (5202G)
U.S. Environmental Protection Agency
401 M Street, SW
Washington, D.C. 20460
(703)603-8707
INTRODUCTION
Typical curtain booms begin failing to contain spilled oil in currents not much above 1.1 miles per
hour, and many skimmers begin to fail in currents not much above 3.5 mph. However, many water
bodies, including streams, large rivers, and estuaries, have currents well in excess of 1.1 mph. The U.S.
Environmental Protection Agency's Risk Reduction Engineering Laboratory had previously tested an
innovative diversionary system based on vertical, plunging water jets and found it capable of moving
floating, spilled oil across small streams into areas of low current (such as the inside of river bends)
where conventional cleanup technology works well. The results of scaling the system up by a factor of
ten (e.g., from a 0.75" up to a 6.0" jet nozzle inside diameter) and evaluating it at three current speeds
from 0.6 to 4.2 miles per hour in the Detroit-St. Clair River system are summarized here. Even in the
presence of these larger turbulence scales and intensities, single plunging water jets were found to be
capable of diverting floating pollutants over thirty feet in the presence of medium river currents. Multiple
jets arranged en echelon were demonstrably capable of diverting material much farther.
A plunging water jet is created by directing a stream of high velocity water (about 35 feet per
second) downwards through a nozzle mounted three feet above the surface of floating spilled oil. The
vertical jet of water entrains air and carries it below the water surface to some equilibrium depth. The
formerly entrained, air bubbles floating back toward the surface create a rising, vertical water current that
becomes horizontal at the water surface. This resultant horizontal surface current moves floating matter
such as spilled oil radially out from the jet's point of impact (thus clearing a circular area centered on the
jet's point of impact). When the jet moves relative to the (undisturbed) water surface, the cleared
circular area is deformed into a parabola with the horizontal surface current flowing perpendicularly out
on both sides from the axis of the "wake" (the track of the jet across the river surface). The surface
current resulting from one or more water jets can be used to clear spilled oil and other floating
contaminants from a given area (such as a water intake or a diver entry point). Conversely, it can be
used to thicken floating contaminants in order to make various removal devices more effective and to
increase the effective sweep width of moving oil skimmers. In particular, since it is little affected by non-
breaking waves and floating debris, a jet can be used to divert spilled oil in streams of all sjzes and in
river currents well in excess of what a conventional boom can tolerate.
The plunging water jet concept was first envisaged in 1978 by Michael G. Johnson (Johnson's patent
is now held by Douglas Engineering of 181-C Mayhew Way, Walnut Creek, CA 94596). The U.S.
Environmental Protection Agency then began to sponsor a series of initial feasibility tests which
231
-------
culminated in 1983 in a 3/4 inch nozzle system demonstrated in small local streams and a local estuary.
Based on this work, the EPA published in 1984 a field manual for the fabrication and use of a 3/4 inch
diameter nozzle through which water flowed at about 80 gallons per minute (Ref. 1). This combination
yielded a diversion distance of about twelve feet in currents of two miles per hour. In another
application, three pairs of jets were also used to increase the effective sweep width of the US Coast
Guard's six mile per hour "Zero Relative Velocity" Skimmer. Concurrently, Dr. Richard Hires of Stevens
Institute of Technology reported in 1980 an analysis of his excellent and extensive series of observations
of 1/4, 3/8 and 1/2 inch diameter nozzles at the Davidson Laboratory's tow tank facility (2).
METHODOLOGY
The objective of the 1992 field work described here was to evaluate the effectiveness of plunging
water jets with much larger diameter nozzles and higher flow rates when used on simulated spills into a
large, higher current river system (the Detroit-St. Clair Rivers) where natural turbulence scales are greater
than in areas than hitherto examined. The nozzle diameters were increased from the former maximum
of 3/4 of an inch to six inches and the flow rates through the nozzles from 80 to 1714 gallons per
minute. Three river current speeds from 0.6 to 4.2 miles per hour were observed during this study.
The primary equipment consisted of a six-inch diameter, CCN150 submersible pump (rated at 2,250
gallons per minute under an eighty-foot head and driven by a portable, diesel-powered hydraulic system)
that fed river water through a six-inch diameter flexible hose (and a turbine flow meter with a pressure
gauge) to a manifold supplying four vertical, poly-vinyl chloride nozzles (of 2", 3", 4", and 6" inside
diameter) controlled by individual ball valves. The manifold was strapped to the underside of a guyed,
four-inch, steel, box beam suspended horizontally, at right angles outboard, from the side of a thirty-six
foot long work vessel. Flow rate through whichever nozzle's ball valve was open could be varied by
changing the throttle settings of the pump's hydraulic power pack. Small pine chips (intended for pet
bedding) and a powdered sorbent (made from cellulose fiber and expanded perlite, and treated to be
oleophilic and hydrophobic) were both used as floating tracers to outline the limits of the jet's parabolic
area of influence. Floating polypropylene lines supported at five-foot intervals by fluorescent painted
gallon jugs were used to provide both longitudinal and horizontal length scales. Times and distances
estimated visually both by an observer in a small boat and from photographic and video records were
used to calculate speeds. Flow rates and pressures were measured directly. Grouped averages of
some of the sixty-eight sets of field observations made are presented in the table. The table shows that
the effect of river speed on a single jet's diversion distance increases markedly as speed exceeds two
miles per hour. Jet velocity (that is, the flow of water through the nozzle) is also a very significant factor,
with diversion distance increasing as jet velocity increases.
FIELD DATA SUMMARY (grouped averages)
Nozzle
Diameter
(inches)
4
4
4
Jet
Flow Rate
(gpm)
1400
1400
f400
Jet
Velocity
(fps)
36.4
36.1
36.4
River
Speed
(mph)
0.7
1.6
4.2
Distance
Diverted
(ft)
38
35
13
232
-------
CONCLUSIONS
This evaluation shows that plunging water jets are very effective for diverting floating spilled oil and
other pollutants in large rivers.
A single, four-inch inside-diameter nozzle at flows above 1400 gallons per minute (gpm) can divert
floating materials about thirty-five feet in river currents of at least one and six tenths miles per hour. This
substantial improvement is three times greater than the previously demonstrated diversion distance of
about twelve feet by a three-quarter inch nozzle at 80 gpm in a similar current.
In a river current of four and a quarter miles per hour (under otherwise similar conditions), the
diversion distance was found to be thirteen feet. This smaller distance is also useful, especially when
considered in the light of the fact that a conventional boom would fail to divert any oil under similar
conditions. Because plunging water jets can be arranged to operate in echelon, multiple jets can
increase the diversion distance almost any amount. At this time, plunging water jets provide the only
known means of diverting floating, spilled oil in four mile per hour currents.
REFERENCES
1. Nash, J.H. and J.S. Farlow. Field Manual for Plunging Water Jet Use in Oil Spill Cleanup. EPA-
600/2-84-045. U.S. Environmental Protection
Agency, Cincinnati, Ohio, 1984. 20 pp.
2. Hires, R.I. An Experimental Study of Currents Generated in the Receiving Waters by Plunging
Water Jets (Research report prepared for the New Jersey Department of Environmental
Protection, P.O. No. P-48063; SIT-DL-80-9-2136; DL Project 4759/858). Stevens Institute of
Technology, Hoboken, New Jersey, 1980. 42 pp.
FOR MORE INFORMATION
John S. Farlow
Releases Control Branch
Superfund Technology Demonstration Division
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency
2890 Woodbridge Avenue (MS-104)
Edison, NJ 08837-3679
(Phone: 908-321-6635)
233
-------
APPLICATION OF THE ELECTRON BEAM TREATMENT PROCESS TO
TREATMENT
_TI-SOURCE HAZARDOUS WASTE LEACHATE
William J. Cooper
David C. Kajdi
Charles N. Kurucz
Michael G. Nickel sen
. Thomas D. Waite
High Voltage Environmental Applications, Inc.
9562 Doral Boulevard
Miami, Florida 33178
(305) 593-5330
An area of considerable interest today is the on-site treatment of multi-source
hazardous waste leachates. One innovative treatment process being developed for the
destruction of hazardous organic chemicals on site is high energy electron beam irradiation.
We have previously reported that the process has been shown to be very effective in
destroying single solute contamination from aqueous waste streams. This paper describes a
natural progression of that work to results obtained from the treatment of multi-solute
leachates. The three studies to be presented describe the results obtained from: 1) an
aqueous solution leachate, 2) an aqueous solution containing a 1% (by volume) dense non-
aqueous phase liquid (DNAPL). and 3) an aqueous leachate with a light non-aqueous phase
liquid (LNAPL) layer. For all"three leachates it was possible to treat to the F039 multi-
source hazardous waste leachate standards for wastewater.
The process has several advantages when compared to other treatment processes. These
advantages are that complex mixtures can be treated simultaneously, and the process is not
adversely affected by dissolved solids such as iron and manganese. No air emissions are
produced and the process goes to completion at ambient temperatures. No organic sludge is
produced and the only by-products are low molecular weight acids and possibly ketones. The
process stream can be reinjected. discharged to the environment or treated in a biological
wastewater treatment system (POTW).
One of the most significant features of this process is that the presence of solids
does not affect the removal efficiency of the organic compounds. Therefore no pretreatment
1s required when suspended solids are present in the leachate or other influent streams.
For more information:
Mr. Franklin Alvarez, EPA Project Manager
Risk Reduction Engineering Laboratory
USEPA
26 West Martin Luther King Drive
Cincinnati, OH 45268
(513) 569-7631
234
-------
PHQTOLYSIS/BIODEGRADATION OF PCB CONTAMINATED SOILS
Kandi Brown, Doug Demott, Arie Groen and Duane Root
IT Corporation
312 Directors Drive
Knoxville, Tennessee 37923
(615)690-3211
Bench-scale tests were conducted to investigate the feasibility of a two-phase
detoxification process that has application to the treatment of soils contaminated with
polychlorinated biphenyls (PCBsK The first step in the process was to degrade the contaminants
by using ultraviolet (UV) radiation facilitated by the addition of a surfactant to mobilize the
contaminants. Biological degradation, the second step, was then used to further destroy organic
contaminants and detoxify the soil.
UV photolysis tests were conducted independently using a medium pressure mercury (Hg)
lamp, a 10 hertz (Hz) pulsed lamp and sunlight on soils with surfactant applications of 2 to 5
percent. Results from UV testing on a surface soil contaminated with about 10,000 parts per
million (ppm) PCBs (Aroclor 1248) and a pit soil containing about 200 ppm PCBs varied, but in
most cases showed minimal reduction of PCB concentration. The PCB reductions ranged from less
than 15 percent to a maximum of 52 percent. Tests using sunlight as a UV source showed no
detectable loss of PCBs after 25 days of exposure.
Biological treatment on surfactant/UV-treated and untreated soil was evaluated in two
bioslurry treatment experiments. The bioslurry experiments evaluated PCB degradation on
surfactant/UV-treated and untreated soils using cultures with and without inducer addition.
Bioslurry treatment did not provide significantly different results for the UV-treated surface soil
versus the untreated soil. Percent reductions of PCBs were highest for an untreated soil containing
350 ppm PCBs which gave 70, 20 and 30 percent reduction of the di-, tri- and tetra-PCBs,
respectively. In the enhanced bioslurry experiment, the addition of 1,000 ppm biphenyl stimulated
greater reduction in PCB concentrations on the same soil.
The results of bench-scale testing on degradation of PCBs using UV photolysis and
biological treatment will be presented. Experimental results will be detailed and data on
concentration of contaminant as a function of treatment time will be displayed along with
breakdown of PCB homolog data showing comparison of treatment affects on individual PCB
homologs for Aroclor 1248.
For more information contact Mr. Randy Parker, USEPA, Office of Research and
Development, 26 W. Martin Luther King Dr., Cincinnati, OH 45268, (513) 569-7271.
235
-------
USEPA - DOE Joint Assessment Program
Emma Lou George
USEPA, Risk Reduction Engineering Laboratory
Pollution Prevention Research Branch
Cincinnati, Ohio 45268
(513) 569-7578
Harry W. Edwards
Colorado State University
Department of Mechanical Engineering
Fort Collins, Colorado
(303) 491-5317
Charles J. Glaser
US Department of Energy
Office of Industrial Technologies
Washington, D.C. 20585
(202) 586-1298
The United States Environmental Protection Agency and the Department of Energy
have entered into an interagency agreement to join the EPA's Waste Minimization
Assessment Centers (WMAC) Program with DOE's Energy and Diagnostic Centers Program
(EADC), forming the Industrial Assessment Centers (LAC) Program, combining pollution
prevention assessments and energy audits into industrial assessments. These assessments are
performed by university staff and engineering students to qualifying small to medium sized
manufacturers in SIC codes 20-39. During 1994, six universities have been selected to launch
the pilot combined program. These are Colorado State University, University of Tennessee,
University of Massachusetts at Amherst, University of Wisconsin at Milwaukee, University
of Texas at College Station and Oregon State University. Each of these Centers will
perform ten industrial assessments in the pilot year.
Colorado State University and the University of Tennessee have performed both
pollution prevention and energy assessments through WMAC and EADC programs. These
two experienced schools performed test assessments in FY93 to examine different
approaches to conducting these industrial evaluations. One of these test assessments is
presented in this poster.
This assessment was performed by Colorado State University at a plant that
manufactures telescopic sights and mounting hardware. From this assessment, one major
pollution prevention opportunity is phase separation of the spent cutting fluid by acid
treatment. A major energy conservation opportunity is reducing compressed air leaks.
For further information, please contact the EPA project officer, Emma Lou George,
Phone (513) 569-7578 or Fax (513) 569-7111.
236
-------
DECISION-SUPPORT SOFTWARE FOR SOIL VAPOR EXTRACTION
TECHNOLOGY APPLICATION: HYPERVENTILATE
Chi-Yuan Fan and Anthony N. Tafuri
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Superfund Technology Demonstration Division
2890 Woodbridge Avenue
Edison, N.J. 08837
(908) 321-6635
ABSTRACT
Soil vapor extraction (SVE) is an in situ corrective action technology that can remove volatile organic
compounds (VOC) and selected residua! petroleum hydrocarbons from unsaturated soils. The EPA
Office of Underground Storage Tanks and Shell Oil Company entered into a Cooperative Research and
Development Agreement under the Federal Technology Transfer Act to develop a decision-support
computer software which has been termed "Hyperventilate". The computer program was designed to
assist regulators and underground storage tanks (UST) owners in evaluating whether SVE is an appro-
priate cleanup technology for use at leaking UST sites. It is an interactive, software guidance system for
evaluating the feasibility of using SVE at a specific site based on site and contaminant characteristics.
The objective of Hyperventilate is to guide the user through a systematic, iterative evaluation to
determine the feasibility of using SVE at a given site. The software utilizes data provided by the user to
develop a rough approximation of the system's desired and maximum removal rates. It can provide two
estimates of the minimum number of vapor extraction wells needed to achieve remediation. The first
estimate is developed by simply comparing the anticipated extraction well radius of influence with the
radial area of contamination. The second estimate is based on a calculation of the volume of air that
needs to be extracted from the soil to remove residual contamination. The user is ultimately responsible
for deciding if the estimates generated by the software are technically and economically practical for a
particular site. Hyperventilate is primarily a software tool for evaluating SVE as a remediation alternative;
it is not intended to be a detailed SVE modeling or design tool.
A report entitled, "Decision-Support Software for Soil Vapor Extraction Technology Application:
Hyperventilate," (EPA/600/R-93/028) provides guidance in evaluating the use of the IBM-compatible
version bf Hyperventilate. In addition, an overview of SVE principles and procedures is presented along
with the basic model principles and a sensitivity analysis of the software. A sample application is also
presented by using data from an actual UST site. The case study demonstrates how to estimate and
determine input parameters, goes through the steps involved in deriving estimates to evaluate whether
SVE is appropriate, and discusses interpretation of the case study results.
237
-------
REMOVAL OF ORGANIC COMPOUNDS FROM DRINKING
WATER USING MEMBRANE TECHNOLOGY
Carol Ann Fronk
Research Chemical Engineer
Systems and Field Evaluation Branch
Drinking Water Research Division
26 W. Martin Luther King Drive
Cincinnati, Ohio 45268
513/569-7592
Today there is a great deal of public concern regarding drinking water.
The public is not only concerned with taste-and-odor causing compounds but is
also concerned with many cancer-causing compounds that increasingly find their
way into the nation's drinking water. The Systems and Field Evaluation Branch
of the Drinking Water Research Division (DWRD) is responsible for evaluating
various technologies that may be feasible for meeting Maximum Contaminant
Levels. Several membrane processes are under investigation to determine their
effectiveness in removing synthetic organic contaminants as well as inorganics
from drinking water.
Reverse osmosis is the reverse of the natural process of osmosis.
Reverse osmosis can be described as the flow of water (or some other solvent)
across a boundary from a more concentrated side to the less concentrated side,
as the result of the application of pressure to the more concentrated side.
Reverse osmosis membranes have shown promise in two DWRD studies. In a
laboratory setting, these membranes were used to treat multisolute, aqueous
solutions of alkanes, alkenes, aromatics and pesticides. At a field-scale
research site in Suffolk County, New York, removal of agricultural
contaminants was accomplished through the use of membranes. Both laboratory
and field scale studies indicated that reverse osmosis shows promise for
removal of organic and inorganic contaminants from drinking water.
Pervaporation is yet another membrane process whereby a liquid mixture is
in direct contact with a membrane. On the other side of the membrane is a
vacuum which pulls the permeate out of the liquid mixture. In this case the
organic contaminants might be contained in a very small, highly concentrated
volume. There is interest in this technology for drinking water purposes
because of it's potential for side-stream-reduction. Therefore future studies
will address pervaporation from both an organic' s removal perspective as well
as a pollution prevention perspective.
For more information contact Carol Ann Fronk, USEPA, RREL, DWRD, SFEB,
26 W. M. L. King Drive, Cincinnati, Ohio 45268, 513/569-7592.
238
-------
PACKAGED HATER TREATMENT PLANT OPERATION AND
FIELD DATA DOCUMENTATION PROJECT
Susan Campbell and Benjamin W. Lykins, Jr.
U.S. Environmental Protection Agency
Risk Reduction Engineering Laboratory
Drinking Water Research Division
Cincinnati, Ohio 45268
(513/569-7426,- 569-7460)
The "Packaged Water Treatment Plant Operation and Field Data
Documentation Project" (the project) was a cooperative effort between the
American Water Works Association (AWWA) and the United States Environmental
Protection Agency (EPA), Drinking Water Research Division (DWRD).
In order to study the problems facing small communities, both today and
in the near future, as they try to comply with new water regulations, this
project was designed to examine existing, in place water treatment systems to
determine their capability in meeting current and future regulations. The
primary goal was to gather reliable and verifiable data regarding the long-
term cost and performance of package water treatment plants being operated by
small systems.
The project evaluated 48 small system package plants. The project used
the EPA definition of a small water system as under 1000 connections or fewer
than 3,300 people served. To allow for the identification of possible
differences in treatment approaches caused by differences in water quality,
funding availability and economic support, between the smallest and the
largest of these small water systems, they were divided into three population
ranges 25-500, 501-1,000, and 1,001 - 3,300.
For more information contact Susan Campbell, USEPA, RREL, DWRD, SFEB,
26 W. M. L. King Drive, Cincinnati, Ohio 45268, 513/569-7426.
239
-------
RISK REDUCTION LABORATORY (RREL),
DRINKING WATER TECHNOLOGY ACTIVITIES - U.S. EPA
Clois J. Slocum and Walter A. Feige
U.S. Environmental Protection Agency
26 W. Martin Luther King Drive
Cincinnati, OH 45268
Tel. 513/569-7496 and 513/569-7281
This poster conveys the activity of the Drinking Water
Research Division (DWRD). DWRD, of Risk Reduction Engineering
Laboratory (RREL) plans, coordinates and conducts a national
program to provide the technology necessary to help prepare the
primary and secondary regulations for drinking water. The
Division has the responsibility for a program that integrates
chemistry, engineering, microbiology, and cost-effective
techniques for assuring the delivery of safe drinking water to
reduce the risk of chemically and microbiologically induced
health effects to the public.
The Division conducts both intramural and extramural
research to establish practices for the control and removal of
contaminants, and for the prevention of water quality
deterioration during storage and distribution in the most
economical manner. The Division operates its own inhouse pilot
facilities, chemistry, and microbiology laboratories, as part of
the intramural program.
Major areas of research include disinfectants, disinfection
by-products, corrosion control, small system technologies,
distribution systems, bacteria, viruses and protozoa.
The Division works closely with municipal and county
activities that treat raw water and operate water supply
facilities, as well as with universities, professional, and
private organizations that conduct control technology activities.
The Division maintains technical liaison with the Office of
Ground Water and Drinking Water (OGWDW) in the Office of Water
(OW) to ensure timely completion of the regulatory and compliance
requirements of the Safe Drinking Water Act (SDWA).
For More Information: Walter A. Feige, U.S. EPA, 26 W.
Martin Luther King Drive, Cincinnati, OH 45268, Tel. 513/569-
7496.
240
-------
Pollution Prevention Research Branch Program
Emma Lou George and Harry M. Freeman
Pollution Prevention Research Branch
Waste Minimization, Destruction and Disposal Research Division
Risk Reduction Engineering Laboratory
Cincinnati, Ohio 45268
(513)569-7578
KEY ISSUE: How should consumer and industrial products and services be designed*
manufactured^ and used for a minimal effect on the environment?
Since 1988, the Pollution Prevention Research Branch (PPRB) at USEPA's Risk
Reduction Engineering Laboratory in Cincinnati, Ohio, has supported projects and provided
technical assistance to encourage the development and adoption of technologies, products,
and pollution prevention (P2) techniques to reduce environmental pollution. To date,
approximately $10 million dollars have been expended to support this program. The
program focuses on five major areas which include:
Cleaner Production Technologies: To develop, demonstrate and evaluate innovative
processes for reducing pollution through source reduction.
Cleaner Products: To support the design and development of products whose
manufacture, use, recycle and disposal represent reduced impacts on the environment
Tools to Support Pollution Prevention: To develop life cycle assessment and
measurement methodologies to guide pollution prevention strategies.
Pollution Prevention Assessments: To identify P2 opportunities in a variety of industries
and to transfer that technical information to assist others in implementing pollution
prevention.
Cooperative P2 Projects with Other Federal Agencies: To promote pollution prevention
in other Federal Agencies through cooperative research, training, and demonstration
projects.
This poster displays personnel and major areas of research through pictures and
narrative. Handouts will be available describing the ongoing projects and Branch
publications. Members of the Branch will be available periodically for discussion.
For further information, please contact Emma Lou George, Phone (513) 569-7578,
Fax (513) 569-7111 or Harry Freeman, Phone (513) 569-7215.
241
-------
ABSTRACT
USING DATA QUALITY OBJECTIVES (DQOs) TO SEARCH FOR HOT SPOTS
Esperanza Piano Renard1, John Warren2 and Carolyn Offutt3
U.S. Environmental Protection Agency
401 M St, S.W.
Washington, D.C. 20460
1908-321-4355
2202-260-9464
3703-603-8797
Each year the U. S. Environmental Protection Agency and the regulated community spend
approximately $5 billion collecting environmental data for scientific research, regulatory decision making,
and regulatory compliance and most importantly in Superfund cleanup and remediation of hazardous
sites. The goal of EPA and the regulated community is to minimize expenditures related to data
collection by eliminating unnecessary, duplicative or overly precise data and at the same time collect
data of sufficient quality to support defensible decision making. These goals can be accomplished by
ascertaining the type, quality and quantity of data necessary to address the problem before the study
begins.
The focus of this Poster presentation is to demonstrate the development and implementation of a
quality system through the DQO process developed specifically for Superfund. The Data Quality
Objective (DQO) Process for Superfund was developed by the Quality Assurance Management Staff for
the Office of Solid Waste and Emergency Response's Hazardous Site Control Division, with input from
the DQO Workgroup comprised of the 10 Regions and Headquarters' Environmental Services and Waste
Management Divisions.
The DQO process is a scientific and legally defensible data collection planning process to help
users decide what type, quality, and quantity of data will be sufficient for environmental decision making.
The DQO process provides a logical framework for planning multiple field investigations, thereby
improving cross-program response planning and allowing optimal cross program data useability, which
are Important goals of the Superfund Accelerated Cleanup Model (SACM).
The poster will apply the DQO process to the Leadbury Superfund Site. The DQO process will
demonstrate its usefulness as a tool and an effective means by which managers and technical staff may
plan and design a more efficient and timely sampling and analysis program that is consistent with the
Integrated site assessment and accelerated response activities of the Superfund Accelerated Cleanup
Model (SACM).
The poster will look at the problem of searching for a highly contaminated area (hot spot) within a
much larger area of minimal contamination. The poster shows how the initial assumptions on the size
and shape of the hotspot are-crucial to the design of the data collection activity. Different scenarios are
explored and some recommendations on optimal sampling schemes will be made.
242
oU.S. GOVERNMENT PRINTING OFFICE:199A-550-001/80334
------- |