»EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
EPA/540/5-91/006a
August 1991
Superfund
Technology Evaluation
Report: Design and
Development of a
Pilot-Scale Debris
Decontamination System
Volume I
-------
-------
EPA/540/5-91/006a
August 1991
TECHNOLOGY EVALUATION REPORT:
DESIGN AND DEVELOPMENT OF A PILOT-SCALE
DEBRIS DECONTAMINATION SYSTEM
VOLUME I
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
Printed on Recycled Paper
-------
NOTICE
The information in this document has been funded by the United States
Environmental Protection Agency (EPA) under Contract No. 68-03-3413 and the
Superfund Innovative Technology Evaluation (SITE) Program. This document has
been subjected to the Agency's administrative and peer review and has been
approved for publication as an EPA document. Mention of trade names or com-
mercial products does not constitute endorsement or recommendation for use.
-------
FOREWORD
The Superfund Innovative Technology Evaluation (SITE) Program was author-
ized in the 1986 Superfund Amendments. The program is a joint effort between EPA's
Office of Research and Development (ORD) and Office of Solid Waste and Emergency
Response (OSWER). The purpose of the program is to assist the development of
hazardous waste treatment technologies necessary to implement new cleanup stan-
dards which require greater reliance on permanent remedies. This is accomplished
through technology demonstrations which are designed to provide engineering and
cost data on selected technologies.
This project consists of design, development, and field demonstrations under
the SITE Program of an EPA-developed hydromechanical debris washing technology
designed for decontamination of debris at Superfund sites. Demonstrations were
conducted at a PCB-contaminated debris site in Detroit, Michigan; a PCB-con-
taminated transformer site in Hopkinsville, Kentucky; and a herbicide-contaminated
drum site near Chickamauga, Georgia. The demonstration effort was directed toward
obtaining information on performance of the technology for assessing its use at
uncontrolled hazardous waste sites. Volume I of this Technology Evaluation Report
describes the development, demonstration, and evaluation of the debris decon-
tamination system. Volume II contains copies of the analytical data submitted by the
various laboratories involved in the project.
A limited number of copies of this report will be available at no charge from
EPA's Center for Environmental Research Information, 26 West Martin Luther King
Drive, Cincinnati, Ohio 45268. Requests should include the EPA document number
found on the report's front cover. When the limited supply is exhausted, additional
copies can be purchased from the National Technical Information Service, Ravens-
worth Building, Springfield, Virginia 22161, (703) 487-4600. Reference copies will be
available at EPA libraries in their Hazardous Waste Collection. You can also call the
SITE Clearinghouse hotline at 1-800-424-9346 or 202-382-3000 in Washington, D.C., to
inquire about the availability of other reports.
E. Timothy Oppelt, Director
Risk Reduction Engineering Laboratory
in
-------
ABSTRACT
In support of EPA's SITE Program, this report documents the development,
demonstration, and evaluation of an EPA-developed technology, a hydromechanical
debris cleaning system, for decontamination of debris at Superfund sites. Although
most of the debris at Superfund Sites has no potential for reuse, decontaminated de-
bris could either be returned to the site as "clean fill" or in the case of metallic debris,
sold to a metal smelter.
During Phase I of this project, an innovative approach for decontaminating de-
bris--a hydromechanical cleaning system-was developed and evaluated. A bench-
scale portable module consisting of an enclosure for washing debris and a closed-loop
cleaning solution purification system was tested. Based on bench-scale results, a
pilot-scale Experimental Debris Decontamination Module (EDDM) was developed. The
EDDM was designed and assembled on a 48-foot semitrailer and field-tested at the
Carter Industrial Superfund Site in Detroit, Michigan. Field testing results indicated an
average of 70 percent removal of the PCBs.
Phase II was directed toward developing debris washing into a proven technolo-
gy for removing various contaminants from debris found at hazardous waste sites.
During Phase II, tests were performed in a controlled environment to optimize the
process. A 20-gallon debris-washing unit with a spray tank and a wash tank was de-
signed and fabricated for bench-scale studies. Results of bench-scale testing showed
removal efficiencies of 95 to 99 percent for all contaminants studied (oil and grease,
representative PCB's, and pesticides).
Subsequently, also as a part of Phase II, a transportable, pilot-scale, 300-gallon
Debris Washing System (DWS) was designed and fabricated. The DWS entails the
application of an aqueous solution during a high-pressure spray cycle, followed by a
turbulent wash cycle. The aqueous cleaning solution is recovered and reconditioned
for reuse concurrently with the actual debris-cleaning process, which minimizes the
quantity of process water required to clean the debris. The DWS was field-tested at a
PCB-contaminated (transformer casings) site in Hopkinsville, Kentucky, and at an her-
bicide-contaminated (drums) site near Chickamauga, Georgia.
IV
-------
CONTENTS
Notice
Foreword
Abstract
Figures
Tables
Acronyms/Abbreviations
Acknowledgments
.->
1. Executive Summary
1.1 Introduction
1.2 Objectives
1.3 Phase I: Development and testing of experimental modules
1.4 Phase II: Design, construction, and demonstration of a
transportable debris-washing system
1.5 Design, fabrication, and demonstration of pilot-scale DWS
1.6 Demonstration at the Shaver's Farm drum disposal site
1.7 Conclusions
n
iii
iv
vii
viii
x
xi
1
1
2
3
4
5
6
2. Introduction
2.1 Background
2.2 SITE Program
2.3 Program objectives
2.4 Purpose of this report
2.5 Report organization
3. Phase I: Development and Testing of an Experimental Debris-
Decontamination Module
3.1 Bench-scale experiments
3.2 Field Demonstration of EDDM at the Carter Industrial Site
7
7
9
9
10
11
11
12
-------
CONTENTS (continued)
Page
4. Phase II: Design, Construction, and Demonstration of a Transportable
Debris-Washing System ' 20
4.1 Bench-scale experiments and results 21
4.2 Design, fabrication, and initial testing of pilot-scale DWS 28
4.3 Field Demonstration of DWS at the Gray PCB Site 28
4.4 Field Demonstration of DWS at Shaver's Farm Site 38
5. Quality Assurance/Quality Control Analyses 52
5.1 Demonstration of DWS at Gray PCB Site 52
5.2 Sampling and analysis 52
5.3 Summary of QA/QC procedures used by Hayden Environmental Group 54
5.4 Demonstration of DWS at Shaver's Farm Site 57
5.5 Sampling and analysis 57
5.6 Summary of QA/QC procedures used by Radian Corporation 59
6. Cost of Demonstrations 64
6.1 Capital equipment cost 64
6.2 Cost of demonstration 64
7. Conclusions and Recommendations 67
7.1 Conclusions 67
7.2 Recommendations 68
8. Full-Scale Debris Washing System: Conceptual Design 69
References 71
Appendices
A QA/QC Data Gray PCB Site Demonstration 72
B QA/QC Data Shaver's Farm Site Demonstration 75
VI
-------
FIGURES
Number
1
2
3
4
5
6
7
8
9
Amount of Oil and Grease on Metal Surface After Completion of
Cleaning Cycle
Carter Industrial Site Map, Detroit, Michigan
TO
Schematic of Pilot-Scale Experimental Debris Decontamination
Module
Bench-Scale Debris Washing System
Schematic of Pilot-Scale Debris Washing System
The Assembled Pilot-Scale DWS at a Hazardous Waste Site
Enclosure for the Pilot-Scale DWS at the Gray Site
Aerial Photograph of Shaver's Farm Site
Schematic Diagram of a Full-Scale DWS
14
15
17
23
29
33
34
39
70
VII
-------
TABLES
Number
1
2
3
4
6
7
8
9
10
11
12
Summary of Results for Oil/Grease and Total Suspended
Solids Analysis
Analytical Results Obtained During Field Demonstration of
EDDM at Carter Industrial Superfund Site
A ConTparison of the Cleaning Capabilities of Surfactant
Solutions Based on Removal of Oil and Grease
Summary of Bench-Scale Results of Controlled Debris Analyzed
for PCBs and Pesticides (Trial 1)
Summary of Bench-Scale Results on Controlled Debris Analyzed
for PCBs and Pesticides (Trial 2)
Summary of Bench-Scale Results of Controlled Debris Analyzed
for Lead (Trial 3)
Results of Surface Wipe Samples Analyzed for Oil and Grease
During Warehouse Testing
Operating Parameters Matrix for the DWS Demonstration
Analytical Results Obtained During Field Demonstration of DWS
at Gray PCB Site
Analytical Results of Process Water Generated During Gray PCB
Site Demonstration
Bench-Scale Results: Analyses for Benzonitrile
Bench-Scale Results: Analyses for Dicamba
Page
13
18
24
25
26
27
30
32
36
37
41
41
VIII
-------
TABLES (Continued)
Number .Page
13 Method Validation: Results of Dicamba Analyses of Gauze and
Metal Strip Spiked with Dicamba 42
14 Method Validation: Results of Benzonitrile Analysis of Gauze
Spiked with Benzonitrile 42
15 Results of Surface Wipe Samples Analyzed for Benzonitrile, 2,4-
Dichlorophenol, 2,6,-Dichlorophenol, and 1,2,4-Trichlorobenzene
During Field Demonstration of DWS at Shaver's Farm Site 45
16 Results of Metal Strip Samples Analyzed for Benzonitrile, 2,4-
Dichlorophenol, 2,6-Dichlorophenol, and 1,2,4-Trichlorobenzene
During Field Demonstration of DWS at Shaver's Farm Site 46
17 Results of Surface Wipe Samples Analyzed for Dicamba, 2,4-D and
2,4,5-T During Field Demonstration of DWS at Shaver's Farm
Site 47
18 Results of Metal Strip Samples Analyzed for Dicamba, 2,4-D, and
2,4,5-T During Field Demonstration of DWS at Shaver's Farm
Site 48
19 Results of Metal Strip Samples Analyzed for Dioxins and Furans
During Field Demonstration of DWS at Shaver's Farm Site 49
20 Analytical Results for Process Water Generated During Shaver's
Farm Site Demonstration 51
is;
21 QA Objectives for Precision, Accuracy, Completeness, and Method
Detection Limit (Gray PCB Site Demonstration) 55
22 Methods Used to Quantitate Selected Methods (Gray PCB Site
Demonstration) 57
23 QA Objectives for Precision, Accuracy, Completeness, and Method
Detection Limit (Shaver's Farm Site Demonstration) 60
24 Summary of Pilot-Scale DWS Demonstration Costs 65
IX
-------
ACRONYMS/ABBREVIATIONS
BOAT - Best Demonstrated Available Technology :
CERCLA Comprehensive Environmental Response Compensation and Liability
Act
DWS - Debris Washing System
EDDM - Experimental Debris Decontamination Module
EPA - United States Environmental Protection Agency
GC/MS - Gas Chromatograph/Mass Spectrometer
MS -- Matrix spike
MSD - Matrix spike duplicate
NPL -- National Priorities List . |
ORD - Office of Research and Development '
OSWER - Office of Soilid Waste and Emergency Response i
PCB -- Polychlorinated biphenyl
PRP - Potential Responsible Party
QAPjP - Quality Assurance Project Plan
RCRA - Resource Conservation and Recovery Act
RPD ~ Relative percent difference
RSD - Relative Standard Deviation
RREL - Risk Reduction Engineering Laboratory
SARA -- Superfund Amendments and Reauthorization Act
SITE ~ Superfund Innovative Technology Evaluation
TSCA - Toxic Substance Control Act
-------
ACKNOWLEDGMENTS
This report was prepared under the direction and coordination of Naomi
Barkley, EPA SITE Project Manager in the Risk Reduction Engineering Laboratory,
Cincinnati, Ohio. Technical peer review was provided by David Smith and Herbert
Pahren (RREL, Cincinnati, Ohio).
This report was prepared for EPA's SITE Program by Majid Dosani and Dr.
Michael Taylor of IT Environmental Program, Inc. (ITEP) under Contract No. 68-03-
3413. Other ITEP staff contributing to the project were John Wentz, Avinash Patkar,
Timothy Kling, Dave Elstun, and Richard Webb. 'T>
The SITE demonstration of this technology was conducted in cooperation with
Ralph Dolhoff, On-Scene Coordinator, EPA Region V, and Charles Eger, On-Scene
Coordinator, EPA Region IV.
Analytical Services were provided by Radian Corporation (Austin, Texas), PCS
(Dayton, Ohio), and IT Analytical Services (Cincinnati, Ohio).
Xi
-------
-------
SECTION 1
EXECUTIVE SUMMARY
1.1 Introduction
Numerous sites in the United States are contaminated with hazardous waste,
and the cleanup of these sites is the top environmental priority of the decade.
Currently, more than 1200 sites are included in the National Priorities List (NPL), and
many more have been proposed for inclusion on the list.
A typical hazardous waste site contains one or more toxic organic or inorganic
chemical residues. These residues are frequently intermingled with remnants of razed
structures (e.g., wood, steel, concrete block, bricks) as well as contaminated soil,
gravel, concrete, and perhaps metallic debris (e.g., machinery and equipment,
transformer casings, drums, and miscellaneous scrap metal). Decontamination of
these materials is important as a means of preventing the spread of contamination
offsite and reducing exposure levels to future users of the buildings or equipment. To
date, no generally applicable decontamination technique has been developed for the
removal of contaminants from structures or debris. Currently, large pieces of equip-
ment are typically decontaminated by steam-cleaning, and contaminated buildings and
structures are frequently torn down and disposed of in a hazardous waste landfill or
incinerator instead of being decontaminated.
Most contaminated debris at hazardous waste sites has no potential for reuse
and therefore cleanup of the site typically entails removal and transportation of the
debris for off-site disposal in a RCRA Subtitle C hazardous waste landfill or incinerator.
These latter options are costly and entail the risk of spreading the contamination well
beyond the borders of the site. Methods are needed for the onsite decontamination of
debris to reduce the risk of spreading contamination offsite and to permit debris dis-
posal in an economical yet environmentally safe manner.
1.2 Objectives
The project consisted of two phases. The objectives of Phase I were as fol-
lows:
-------
To evaluate a hydromechanical cleaning system, an innovative approach
for decontaminating debris. !
To conduct bench-scale testing with a portable module for the decontam-
ination of debris. ;
Based on bench-scale results, to develop a pilot-scale Experimental De-
bris Decontamination Module (EDDM). i
i
To field-test the EDDM at a hazardous waste site.
The objectives of Phase II were as follows: ;
To continue development of the EDDM into a proven technology for
removing various contaminants from debris found on hazardous waste
sites.
To conduct bench-scale tests to optimize the process.
To design and construct a transportable pilot-scale debris washing sys-
tem (DWS).
To field-test the pilot-scale DWS at two hazardous waste sites where
various types of debris are present.
To prepare a conceptual design of a full-scale debris washing system.
1.3 Phase I: Development and Testing of Experimental Modules
During Phase I of the project, a hydromechanical cleaning system, an innovative
approach to decontaminating debris, was developed and evaluated. A;bench-scale,
portable module consisting of an enclosure where debris was placed and a closed-
loop solvent-delivery system was tested. Based on the bench-scale results, a pilot-
scale EDDM was developed and field-tested.
A 300-gal-capacity pilot-scale EDDM was designed, assembled, installed (on a
48-ft semitrailer), and tested at the Carter Industrial Superfund Site in Detroit, Ml. This
site contained large quantities of different types of PCB-contaminated debris, including
scrap metal, 55-gallon metal drums, tools, equipment, and some furniture items.
Two 200-lb batches of metallic debris were cleaned in the system. Before and
after treatment, surface-wipe samples were obtained to determine the contaminant
removal efficiency of the system. The percentage reduction of PCBs achieved during
-------
cleaning ranged from 33 to 87 percent (average reduction of 58 percent) for Batch 1
and from 66 to 99 percent (average reduction of 81 percent) for Batch 2.
The surfactant solution in the EDDM was sampled twice during the actual
cleaning process, and PCB concentrations of 928 and 420 /xg/L were found. Upon
completion of the debris-washing experiment, the cleaning solution was pumped
through a series of particulate filters and finally through activated carbon. The PCB
concentration was reduced to 5.4 jug/L during this treatment. Most municipalities
allow water containing a PCB concentration of < 1 /ng/l. to be sewered, and this level
was achieved by recycling the process water through carbon a second time.
1.4 Phase II: Design, Construction, and Demonstration of a Transportable
Debris-Washing System
Phase II of this project was directed toward further development of debris
washing into a proven technology for removing various contaminants from debris
found on hazardous waste sites in preparation for a full-scale demonstration at
Superfund and other hazardous waste sites. An initial series of bench-scale tests were
performed in a controlled environment to optimize the newly-designed washing
system. After the bench-scale evaluation, a transportable pilot-scale version of the
debris washing system (DWS) was designed, constructed, and demonstrated at actual
hazardous waste sites.
Based on experience gained during the Carter site field test, a bench-scale (20
gal of surfactant solution capacity) debris washing unit was designed, constructed,
and assembled. This system consisted of a spray tank, wash tank, oil-water
separator, and ancillary equipment (i.e., heater, pumps, strainers, metal tray, etc.).
This bench-scale DWS was developed to determine the ability of the system to remove
contaminants from debris and to facilitate selection of the most efficient surfactant
solution.
During these bench-scale experiments, surface-wipe samples of the six pieces
of control debris were taken before and after treatment and-ajialyzed for oil and
grease. Based on the results, a nonionic surfactant solution was selected as the
solution best suited for cleaning oily metal parts and debris.
As part of the continuing investigation into the performance of the DWS, the
representative pieces of debris were spiked with a mixture of spiking material (used
motor oil, grease, topsoil, and sand) containing representative contaminants (DDT,
lindane, PCBs, and lead sulfate) and washed in the DWS with the selected surfactant
solution. Three trials were performed. Surface wipe samples of debris from the first
two trials were analyzed for PCBs, lindane, and DDT; the surface wipe samples from
the third trial were analyzed for total lead.
-------
The average overall reductions of PCBs and pesticides achieved during Trials 1
and 2 were greater than 99 and 98 percent, respectively. The overall reduction of lead
was greater than 98 percent. '
After completion of the bench-scale debris-washing experiments, the cleaning
solution was neutralized to a pH of 8 and then pumped through a series of particulate
filters and finally through activated carbon. During this treatment, 1:he PiCB, lindane,
and DDT concentrations were reduced to <2.0, 0.03, and 0.33 /ug/L, respectively.
The concentration of lead was reduced to 0.2 mg/L after treatment.
1.5 Design, Fabrication, and Demonstration of Pilot-Scale DWS
Based on the results obtained from bench-scale studies, a 300-gal capacity
pilot-scale DWS was designed and constructed. The pilot-scale DWS was assembled
in a warehouse in Cincinnati Ohio, and several tests were conducted. After the ware-
house testing, the DWS was disassembled, loaded onto a 48-foot semitrailer, and
transported to the Gray PCB site in Hopkinsville, Kentucky, which was selected for the
field demonstration. The entire DWS was reassembled on a 25-ft X 24-ft concrete pad.
A temporary enclosure (approximately 25 ft high) was built on the concrete pad to
enclose the DWS and to protect the equipment and the surfactant solution from rain
and cold weather. The Gray PCB site contained between 70 and 80 burned-out
transformer casings and other large amounts of scrap metal. The demonstration took
place in December 1989, and ambient temperatures were at or below freezing during
the entire operation. :
Before the cleaning process began, the transformer casings (ranging from 5 gal
to 100 gal in size) were cut in half with a metal-cutting partner saw. A pretreatment
sample was obtained from one-half of each of the transformer casings by a surface-
wipe technique. The transformer halves were placed into a basket and lowered into
the spray tank, which was equipped with multiple water jets that blast loosely adhered
contaminants and dirt from the debris. After the spray cycle, the basket of debris was
removed and transferred to the wash tank, where the debris was washed with a high-
turbulence wash. Each batch of debris was cleaned for a period of 1 hr in the spray
tank and 1 hr in the wash tank. During both the spray and wash cycles, a portion of
the cleaning solution was cycled through a closed-loop system in which the oil/PCB-
contaminated cleaning solution was passed through an oil/water separator, and the
cleaned solution was then recycled into the DWS. After the wash cycle, the basket
containing the debris was returned to the spray tank, where it was rinsed with fresh
water. !
Upon completion of the cleaning process, posttreatment surface: wipe samples
were obtained from each of the transformer pieces to assess the post-decontamina-
tion PCB levels. The before-treatment concentrations ranged from 0.1 to 98 Mg/100
-------
cm2. The after-treatment analyses showed that all the cleaned transformers had a
PCB concentration lower than the acceptable level of 10 /ng/100 cm2.
After treatment of all the transformers at the site, the surfactant solution and the
rinse water were placed in the water treatment system, where they were passed
through a series of particulate filters, then through an activated-carbon drum, and
finally through an ion-exchange column. The before- and after-treatment water
samples were collected and analyzed for PCBs and selected metals (cadmium, cop-
per, chromium, lead, nickel, and arsenic).
The water treatment system reduced the PCB concentration in the water to
below the detection limit. The concentrations of each of the metals (except arsenic)
were reduced to the allowable discharge levels set by the city of Hopkinsville. Upon
receipt of the analytical results, the treated water was pumped into a plastic-covered
10,000-yd3 pile of contaminated soil at the site.
During this site cleanup, 75 transformers (approximately 5000 Ib) were cleaned
in the DWS. All of them are now considered clean and acceptable for sale to scrap
metal dealers or to a smelter for reuse.
1.6 Demonstration at the Shaver's Farm Drum Disposal Site
In August 1990, a second demonstration of the DWS was conducted at the
Shaver's Farm drum disposal site near Chickamauga, Georgia, where 55-gal drums
containing varying amounts of a herbicide, Dicamba (2-methoxy-3,6-dichlorobenzoic
acid), and benzonitrile (a precursor in the manufacture of Dicamba) were buried. EPA
Region IV had excavated more than 4000 drums from one location on this 5-acre site
when this demonstration occurred.
The pilot-scale system was transported to this site on a 48-ft semitrailer and
assembled on a 25 x 24 ft concrete pad. The temporary enclosure used at the Gray
site was reassembled to protect the equipment from rain. Ambient temperature at the
site during the demonstration ranged from 75° to 105°F.
The 55-gal herbicide-contaminated drums were cut into four sections, and pre-
treatment surface-wipe samples were obtained from each section. The drum pieces
were first placed in the spray tank of the DWS for 1 hr of surfactant spraying, then in
the wash tank for an additional hour of surfactant washing, and finally in the spray
tank for 30 min of water rinsing. The drum pieces were then allowed to air-dry before
the posttreatment surface-wipe samples were taken. Ten batches of 1 to 2 drums per
batch were treated during this demonstration.
Pretreatment concentrations of benzonitrile in surface wipe samples ranged
from 8 to 47,000 /xg/100 cm2 and averaged 4556 M9/100 cm2; posttreatment samples
-------
ranged from below detection limit to 117 /*g/100 cm2 and averaged 10 /ig/100 cm2.
Pretreatment Dicamba values ranged from below detection limit to 180 Mg/100 cm2
and averaged 23 Mg/100 cm2; posttreatment concentrations ranged from below detec-
tion limit to 5.2 /ig/100 cm2 and averaged 1 Mg/100 cm2. :
All Superfund site activities described in this document were governed by EPA-
approved Health and Safety and Quality Assurance Plans.
1.7 Conclusions
Field-test results obtained with the pilot-scale DWS during demonstrations at
two Region IV hazardous waste sites showed the unit to be both transportable and
rugged. Extreme high and low temperatures had little effect on the operation of the
equipment. The system successfully removed PCBs from transformer casing surfaces
and herbicides, pesticides, dioxins, and furan residues from drum surfaces.
The cleaning solution was recovered, reconditioned, and reused during the
actual debris-cleaning process; this minimized the quantity of process water required
for the decontamination procedure. The water treatment system was effective in re-
ducing contaminant concentrations, with the exception of arsenic and possibly
Dicamba, to below the detection limit.
Planned progression of this EPA-developed technology includes design, devel-
opment, and demonstration of a full-scale, transportable version of the DWS unit.
-------
SECTION 2
INTRODUCTION
2.1 Background
Congress enacted the Comprehensive Environmental Response, Compensation
and Liability Act (CERCLA) of 1980 to address past hazardous waste disposal
practices and the environmental and human health effects of those practices. In the
reauthorization of CERCLA, called the Superfund Amendments and Reauthorization
Act of 1986 (SARA), Congress expressed concern over the use of land-based disposal
and containment technology to mitigate releases of hazardous substances at hazard-
ous waste sites.
In response to SARA, the EPA's Office of Solid Waste and Emergency
Response (OSWER) and Office of Research and Development (ORD) established a
formal program to promote the development and use of innovative technologies to
clean up Superfund sites across the country. This program is called the Superfund
Innovative Technology Evaluation (SITE) Program.
2.2 SITE Program
The overall goal of the SITE Program is to "carry out a program of research,
evaluation, testing, development and demonstration of alternative or innovative
treatment technologies... which may be utilized in response actions to achieve more
permanent protection of human health and welfare and the environment." Specifically,
the program's goal is to maximize the use of alternatives to land disposal in cleaning '
up Superfund sites by encouraging the development and demonstration of new,
innovative treatment and monitoring technologies.
The SITE Program comprises four major elements:
Demonstration Program
Emerging Technologies Program
Measurement and Monitoring Technologies Program
Technology Information Services
-------
The Demonstration Program is one of the most important aspects of the SITE
Program which evaluates field- or pilot-scale technologies that can be scaled up for
commercial use. The Demonstration Program is the primary focus of the SITE
Program because the technologies evaluated are close to being available for remedia-
tion of Superfund sites. The main objective of the Demonstration Program is to : ,^
develop extensive performance engineering and cost information for new technologies.
With this information, potential users can make informed decisions on whether to use
these technologies to remediate hazardous waste sites. |
I
i
The results of the demonstration identify possible limitations of the technology,
the potential need for pre- and post-processing of wastes, the types; of wastes and
media to which the process can be applied, the potential operating problems, and the
approximate operating costs. The demonstrations also permit evaluation of long-term
risks. Demonstrations usually occur at Superfund sites or under conditions that
duplicate or closely simulate actual wastes and conditions found at Superfund sites to
ensure the reliability of the information collected and acceptability of. the data by users.
Technologies are selected for the SITE Demonstration Program through annual
requests for proposal. Proposals are reviewed by EPA to determine the ;technologies
with the most promise for use at Superfund sites. To qualify for the program, a new
technology must have been developed to pilot or full scale and musrt offer some
advantage over existing technologies. Mobile technologies are of particular interest.
Once EPA has accepted a proposal, the Agency and the developer work with
the EPA Regional Offices and State agencies to identify a site containing; wastes
suitable for testing the capabilities of the technology. The developer is responsible for
demonstrating the technology at the selected site, and is expected to pay the costs to
transport, operate, and remove the equipment. The EPA is responsible for project
planning, sampling and analysis, quality assurance and quality control, preparing re-
ports, and disseminating information. :
The Emerging Technologies Program focuses on conceptually proven, but un-
tried technologies. These technologies are in an early stage of development involving
laboratory or pilot testing. Successful technologies are encouraged to advance to the
Demonstration Program.
The Monitoring and Measurement Technologies Program identifies existing
technologies that can improve field monitoring and site characterizations. It supports
the development and demonstration of new technologies that provide faster, more
cost-effective real-time data on contamination and cleanup levels. Finally, it formulates
the protocols and Standard Operating Procedures for demonstrated methods and
equipment. !
8
-------
An Applications Analysis Report and Technology Evaluation Report are pub-
lished at the conclusion of each demonstration. Research reports on emerging tech-
nology projects are also produced. Results and status updates are distributed to the
user community - EPA Regions, state agencies, remediation contractors, and respon-
sible parties - through many media and activities.
2.3 Program Objectives
The following were the objectives of this two-phased SITE Project for decontam-
ination of debris, using the debris washing system. The objectives of Phase I were as
follows:
To evaluate a hydromechanical cleaning system, an innovative approach
for decontaminating debris.
To conduct bench-scale testing with a portable module for the decon-
tamination of debris.
Based on bench-scale results, to develop a pilot-scale Experimental De-
bris Decontamination Module (EDDM).
To field-test the EDDM at a hazardous waste site.
The objectives of Phase II were as follows:
To continue development of the EDDM into a proven technology for
removing various contaminants from debris found on hazardous waste
sites.
To conduct bench-scale tests to optimize the process.
To design and construct a transportable pilot-scale debris washing
system (DWS). HI
To field-test the pilot-scale DWS at two hazardous waste sites where
various types of debris are present.
2.4
To prepare a conceptual design of a full-scale debris washing system.
Purpose of This Report
The Technology Evaluation Report provides a comprehensive description of the
demonstration and its results. This report is intended for engineers performing a
detailed evaluation of the technology for a specific site and waste situation. The
-------
purpose of these technical evaluations is to obtain a detailed understanding of the
performance of the technology during the demonstration and to ascertain the
advantages, risks, and costs of the technology for the given application. This infor-
mation is used to produce conceptual designs in sufficient detail to enable the
preparation of preliminary cost estimates for the demonstrated technology.
2.5 Report Organization
This Technology Evaluation Report is presented in two volumes. Volume I
describes the design and development of the pilot-scale debris decontamination
system and presents the results of the demonstrations conducted at three hazardous
waste sites. Section 3 discusses Phase I of the project, which included the
development and testing of the EDDM. Section 4 presents information on Phase II,
which entailed the design, construction and field demonstrations of a transportable
debris washing system. Section 5 contains a summary of the Qualify Assurance
aspects of the analyses performed during the field testing of the pilot-scale DWS.
Section 6 contains a summary of cost of demonstrations. Section 7 presents conclu-
sions and recommendations. Section 8 addresses the conceptual design of a full-
scale debris washing system.
Volume II contains copies of the analytical data submitted by the various
laboratories involved in the project. Both volumes are available from NTIS.
10
-------
SECTION 3
PHASE I: DEVELOPMENT AND TESTING OF AN EXPERIMENTAL DEBRIS-
DECONTAMINATION MODULE
During Phase I of the project, a hydromechanical cleaning system, an innovative
approach to decontaminating debris, was developed and evaluated. A bench-scale
portable module consisting of an enclosure for placement of debris and a closed-loop
solvent-delivery system was tested for decontamination of debris at hazardous waste
sites. Based on the bench-scale results, a pilot-scale Experimental Debris Decontami-
nation Module (EDDM) was then developed and field-tested.
The purpose of Phase I was to develop and prove the concept of this innovative
approach to cleaning contaminated debris. This section presents the development,
procedure, and results of the bench- and field-testing of the EDDM.
3.1 Bench-Scale Experiments
Before the details of the pilot-scale design of the EDDM were finalized, a bench-
scale experiment was performed to determine the best commercially available cleaning
solution possible for cleaning PCB-contaminated debris and metal parts in the field.
Four cleaning solutions were selected for the experiment: tap water, 10 percent sulfur-
ic acid, and the two detergents BB-100 and Power Clean (BB-100 and Power Clean
are nonionic, biodegradable industrial degreasers).
The experimental procedure involved the application of measured quantities of
used motor oil, grease, and soil to rusted iron parts to simulate the kind of grime likely
to be encountered on oily, PCB-contaminated metal parts and debris in the field. Ex-
periments were performed in a 10-gallon hydromechanical cleaning unit, that contains
an axial flow pump, a propeller shaft, a propeller, a pressure chamber, and a calm
fluid section. Three tests were performed with each cleaning solution, and a fresh set
of oil/grease-contaminated metal parts were used in each test. For consistency, each
set of contaminated parts was matched closely with regard to the size, shape, and
type of metal. The parts were also arranged in the same order in the parts-washer
basket during washing.
11
-------
Upon completion of each test, two aliquots of cleaning solution were collected;
one aliquot was submitted for oil and grease analysis and the other,, for total suspend-
ed solids analysis. Two surface-wipe samples from selected metal parts were also
collected for oil and grease analysis to determine the level of oil/grease remaining on
the metal surfaces after treatment in the parts washer. The skimmer oil from each of
the three runs was mixed together for oil and grease analysis. All the samples were
analyzed at PCS, Inc. (now called the Hayden Environmental Group, Inc.) in Dayton,
Ohio.
3.1.1 Results of Bench-Scale Studies
Table 1 summarizes the results of the oil/grease and total suspended solids
analyses. The analytical results obtained for the wipe samples indicate that the
amount of oil and grease remaining on the metal surfaces was significantly higher after
cleaning with tap water and 10 percent sulfuric acid and comparatively lower after
cleaning with BB-100 and Power Clean. This demonstrates the poor cleaning perform-
ance of water and sulfuric acid. Moreover, the handling of 10 percent sulfuric acid
was difficult, and the acid had a corroding effect on the hydromechanical cleaning
equipment. Hence, it was concluded that water and sulfuric acid should not be con-
sidered as potential cleaning solutions for oily PCB-contaminated debris.
Based on the results of the surface-wipe testing listed in Table 1, a BB-100
solution appears to clean better than a Power Clean solution. This is also shown
graphically in Figure 1, which plots the results of wipe samples (in milligrams of oil and
grease per square centimeter) for each run. The results tend to indicate that BB-100
removed solids from metal surfaces more effectively than did Power Clean. The data
also show that upon completion of the third run, the BB-100 solution still had more
cleaning capacity to remove dirt from metal parts than did Power Clean. Hence, of the
four cleaning solutions tried, BB-100 was selected as the cleaning solution best suited
for cleaning oily PCB-contaminated metal parts and debris in the field.
3.2 Field Demonstration of EDDM at the Carter Industrial Site
3.2.1 Site Description
Carter Industrial, a Superfund site, located at 4690 Humboldt Avenue, near
Interstate Highway 98 in Detroit, Michigan, was made available by Region V for dem-
onstration of the EDDM. The site consists of a U-shaped area into which Humboldt
Avenue enters at the center. The site contains two buildings plus an incinerator and a
smelter. A 20-ft x 20-ft concrete pad is located in the area where Humboldt Avenue
dead ends at the site, and a small flat dirt area is located adjacent to the concrete
pad. A vacant lot is located 500 ft south of the site at the corner of Humboldt and
Forest Avenues. Figure 2 presents a site diagram of the Carter Site. Previous opera-
tions at Carter may have involved draining PCB-contaminated oil from transformers
and burning materials in onsite incinerators. The site contains large quantities of
12
-------
TABLE 1. SUMMARY OF RESULTS FOR OIL/GREASE AND TOTAL
SUSPENDED SOLIDS ANALYSIS
Experimental
Run No.
1
2
3
1
2
3
1
1
2
2
3
3
Total of 1,2, & 3
Sample Type3
Cleaning solution
Cleaning solution
Cleaning solution
Cleaning solution
Cleaning solution
Cleaning solution
Wipe No. 1
Wipe No. 2
Wipe No. 1
Wipe No. 2
Wipe No. 1
Wipe No. 2
Oil from skimmer
Analysis
Oil and grease,
mg/liter
Oil and grease,
mg/iiter
Oil and grease,
mg/liter
Total suspended
solids, mg/liter
Total suspended
solids, mg/liter
Total suspended
solids, mg/liter
Oil and grease,
mg/cm2
Oil and grease,
mg/cm2
Oil and grease,
mg/cm2
Oil and grease,
mg/cm2
Oil and grease,
mg/cm2
Oil and grease,
mg/cm2
Oil and grease,
mg/liter
Cleaning Solution
Water
42
151
241
5
7
15
1.77
1.82
10.54
4.40
2.43
NAb
NA
Sulfurlc Acid Cone.
10%,wUvol.
161
143
138
128
255
148
1.48
1.75
0.7
4.8
3.81
3.27
1540
BB-100
Cone. 15%, v/v
7
182
319
600
904
1000
0.32
0.25
0.15
V
0.42
0.26
0.33
3380
Power Clean
Cone. 1 :6 Ratio
1670
1470
2440
206
576
484
0.50
0.49
0.43
0.48
0.34
0.61
3900
3 All samples are posttreatment samples.
b Not Analyzed.
-------
CM
0.6
0.5
0
8 0.4
Is
lo.3-
c
o
o
ca
£ 0.2
TJ
CO
(D
f
< 0.0
Power Clean
BB-100
2
Run No.
Figure 1. Amount of oil and grease on metal surface after completion
of cleaning cycle.
14
-------
BUILDING
N
ALLEY
BUILDING
INCINERATOR
AUTO SALVAGE
YARD
BUILDING
a
HUMBOLDT AVE.
SMELTER
D
BUILDING
Figure 2. Carter Industrial site map, Detroit, Michigan.
-------
different types of PCB-contaminated debris, including large quantities of scrap metal,
55-galIon metal drums, tools, equipment, and some furniture items.
3.2.2 Process Operation of EDDM During Field Testing
A 300-gallon-capacity EDDM was assembled on a 48-foot trailer in a warehouse
in Cincinnati, Ohio, and was transported to the Carter Site, where it was field-tested.
The EDDM consisted of a hydromechanical cleaning tank, a chip-removal system, and
an oil/water separator.
The hydromechanical cleaning tank (300-gallon capacity) is an off-the-shelf,
patented device consisting of an axial flow pump, a propeller shaft, propeller, pressure
chamber, and calm fluid section. During washing, the unit removes oil from the parts.
The oil is suspended in the cleaning fluid and carried into the calmi fluid section of the
tank, where the oil is allowed to accumulate on the surface. A skimmer is provided to
collect oil in the calm fluid section of the tank. The oil is wiped off the skimmer wheel
by the wiper blades and drained to a container for disposal.
The chip removal system is a mobile, self-contained, portable unit consisting of
a centrifugal pump and a perforated, stainless steel strainer basket. During washing,
the chip removal system automatically removes chips that accumulate from parts
washing. Figure 3 represents a flow diagram of the pilot-scale module.
Two 200-lb batches of metallic debris were cleaned with a solution of BB-100
surfactant in the hydromechanical cleaning unit. Before the cleaning process was
initiated, five individual pieces of metal from each batch were sampled for PCBs by a
surface-wipe technique (1). The debris items were placed in a basket, the basket was
transferred into the EDDM, and the cleaning process was initiated. Each batch of
debris was cleaned for a period of 2 hours. The cleaning solution was cycled through
a continuous closed-loop system, where the oil/PCB-contaminated wash solution was
passed through an oil/water separator, and the clean solution was then recycled into
the module. At the completion of the cleaning process, five additional wipe samples
were obtained from the same pieces of metallic debris to assess the post-decontami-
nation level of PCBs. The surface-wiping procedure was carried out as described in
the Field Manual for Grid Sampling of PCB Spill Sites to Verify Cleanup (1), which en-
tails the use of a hexane-soaked cotton gauze pad to wipe a 100-cm2 area on the
surface of the object being sampled. A detailed description of surface-wipe sampling
is presented in Section 5.
3.2.3 Results of Field Demonstration
Table 2 shows the quantity of PCBs on the surface of each piece of metal be-
fore and after cleaning. The percentage reduction of PCBs achieved during cleaning
ranges from 33 to 87 percent (average reduction of 58 percent) for Batch 1 and from
16
-------
Contaminated.
Debris
Clean Debris
OIL HYDROMECHANICAL CHIP
COLLECTION CLEANING TANK REMOVAL
SYSTEM
OIL/WATER
SEPARATOR
DISPOSAL
CARTRIDGE
FILTERS
SOLUTION
TREATMENT
Figure 3. Schematic of pilot-scale experimental debris decontamination module.
-------
TABLE 2. ANALYTICAL RESULTS OBTAINED DURING FIELD DEMONSTRATION OF EDDM
AT CARTER INDUSTRIAL SUPERFUND SITE
00
Batch Number
1
2
Sample ID Number
Pretreatment
9/7-DET-B1-S1
9/7-DET-B1-S2
9/7-DET-B1-S3
9/7-DET-B1-S4
9/7-DET-B1-S5
FB-1-9/8
9/8-DET-B2-S1
9/8-DET-B2-S2
9/8-DET-B2-S3
9/8-DET-B2-S4
9/8-DET-B2-S5
FB-2-9/8
Posttreatment
9/7-DET-B1-S7
9/7-DET-B1-S8
9/7-DET-B1-S9
9/7-DET-B1-S10
9/7-DET-B1-S11
9/8-DET-B2-S6
9/8-DET-B2-S7
9/8-DET-B2-S8
9/8-DET-B2-S9
9/8-DET-B2-S10
PCB Concentration on Surfaces (p.g/100cm2)
Pretreatment
134
490
1280
73
203
Field Blank: <1 .0
8
6090
374
96
1690
Field Blank: 1.0
Posttreatment
50
178
856
43
23
13
1800
128
10
18
% Reduction
63
64
33
41
87
Avg: 58
-63
70
66
90
99
Avg: 81
-------
66 to 99 percent (average reduction of 81 percent) for Batch 2. In the case of Sample
1 in Batch 2, however, the PCB concentration apparently increased after the cleaning
process, probably because the posttreatment wipe sample in this case was obtained
from a location on the debris surface that was initially more heavily contaminated with
PCBs.
The results seem to indicate that more effective removal of PCBs was achieved
during the cleaning of Batch 2 than during the cleaning of Batch 1. Better cleaning
results for Batch 2 may be explained by the following. In the case of Batch 2, the
basket containing the debris was removed from the washer after 1 hour of cleaning,
and the parts were manually rearranged so that all the sides of debris were exposed
to the cleaning solution with the same force of the turbo washer. The basket was then
lowered back into the washer and cleaning was continued for 1 more hour. In the
case of Batch 1, however, the cleaning process was continued for 2 hours without the
debris in the basket being rearranged.
The surfactant solution in the Turbowasher was sampled twice during the actual
cleaning process, and PCB concentrations of 928 and 420 p.g/L were found. After
completion of the debris-washing experiment, the cleaning solution was pumped
through a series of particulate filters and finally through activated carbon. The PCB
concentration was reduced to 5.4 /zg/L during this treatment Most municipalities
allow water containing a PCB concentration of < 1 /xg/L to be sewered, and this level
was achieved by recycling the process water through carbon a second time.
The copies of bench-scale and pilot-scale analytical data provided by the PCS,
Inc., Laboratory are included in Volume II of this report.
19
-------
SECTION 4
PHASE II: DESIGN, CONSTRUCTION, AND DEMONSTRATION
OF A TRANSPORTABLE DEBRIS-WASHING SYSTEM
Phase II of this project was directed toward further development of debris wash-
ing into a proven technology for removing various contaminants from debris found on
hazardous waste sites in preparation for full-scale demonstrations at Superfund and
other hazardous waste sites. An initial series of bench-scale tests were performed in a
controlled environment for further optimization of the newly designed equipment and
washing process. After bench-scale evaluation, a transportable pilot-scale version of
the debris washing system (DWS) was designed, constructed, and demonstrated at
actual hazardous waste sites.
The knowledge and experience acquired during the debris decontamination
demonstration at the Carter site were used in the design and construction of a bench-
scale debris washing unit. This unit was a smaller version of the pilot-scale unit that
was designed and constructed later in the project. The bench-scale DWS was
developed to assess the ability of the system to remove contaminants from debris and
to facilitate selection of the most efficient surfactant solution.
The bench-scale system (20-gallon size) consisted of a spray tank, a wash
tank, an oil/water separator, and ancillary equipment (such as heaters, pumps,
strainers, metal tray, etc.). Although the bench-scale DWS and the pilot-scale DWS
developed in Phase II were substantially different from the EDDM used in Phase I, the
concept of the system remained the same. Some of the major changes: that were
introduced in the development of DWS compared with EDDM are as follows:
1) A high-pressure spray tank was added to the system for the removal of
excessive or firmly attached gross contamination from the debris prior to
its cleaning in the wash tank. Experience gained at the Carter site indi-
cated that removing gross contamination before placing debris in the
washer would increase the cleaning efficiency of the system.
2) The wash tank was modified to give high turbulence to increase surface-
cleaning performance.
20
-------
3) The EDDM was a single-step process (only a wash step), whereas the
DWS was a three-step process (a high-pressure spray step followed by a
high-turbulence wash step, and finally a water-rinse step).
4) The DWS was assembled on a concrete pad on the ground, whereas the
EDDM was mounted on a trailer.
Details of the bench-scale and pilot-scale DWS are presented in this section.
4.1 Bench-Scale Experiments and Results
The best cleaning solution tested in Phase I, BB-100, was used at the Carter
Industrial Site during the pilot-scale demonstration. Although BB-100 was found to be
an effective cleaning solution, a decision was made to test other nonionic surfactants
that might be more effective and/or more economical than BB-100.
An extensive survey was made of several different types of commercially avail-
able cleaning solutions. Four additional nonionic, nontoxic, low foaming, metal-clean-
ing surfactant solutions (BG-5, MC-2000, LF-330, and L-422) were selected for an
experimental evaluation with BB-100 to determine their capacity to solubilize and re-
move contaminants from the surface of the debris. The extracting power of these
undissociated surfactants is related to the presence of a proper balance of hydrophilic
and lipophilic groups within the molecule, which enable surfactants to interact with
both polar and nonpolar substances. Unlike the anionic and cationic types, the effec-
tiveness of nonionic surfactants are not susceptible to moderate pH changes and the
presence of electrolytes. Surfactants with a proper balance in their hydrophilic and
lipophilic affinities are effective emulsifying agents because they tend to concentrate at
the oil/water interface.
During the bench-scale experiments, an attempt was made to select debris that
was representative of the types of material that might be found on a hazardous waste
site. For purposes of consistency with regard to the shape, size, type of metal, and
sampling procedure, a set of representative or typical debris was selected, which
included three rusted metal plates, a brick, a concrete block, and a piece of plastic.
These six pieces of representative debris were used in each of the bench-scale tests.
Automotive parts obtained from an auto salvage yard were also included with each
batch of representative debris. Cleanliness of these salvage yard items was visually
checked at the conclusion of each trial run to determine the surfactant's performance
on aged greasy and oily debris.
Prior to each bench-scale experiment, the six pieces of control debris were
"contaminated" by dipping them into a spiking material consisting of known amount of
used motor oil, grease, topsoil, and sand. The pieces of "contaminated" debris were
then arranged on a metal tray with some assorted parts from an auto salvage yard.
21
-------
The tray containing the debris was inserted in the spray tank and subjected to a high-
pressure spray of surfactant solution for 15 minutes. At the end of spray cycle, the
tray was transferred to the high-turbulence wash tank, where the debris was washed
for 30 minutes with a solution of the same surfactant as that in the spray tank. After
the wash cycle was completed, the tray was removed from the wash tank and the
debris was allowed to air-dry. The bench-scale DWS is shown in Figure 4-
Surface-wipe samples were obtained from the six pieces of control debris
before and after treatment. These wipe samples were analyzed for oil and grease,
and the results are summarized in Table 3. Based on the results of the wipe testing,
L-422 was selected as the solution best suited for cleaning oily metal parts and debris.
As part of the continuing investigation into the performance of the DWS, repre-
sentative pieces of debris were spiked with a mixture of spiking material (used motor
oil, grease, topsoil, and sand) containing representative contaminants (DDT, lindane,
PCB, and lead sulfate) and washed in the DWS with L-422 as the cleaning solution.
Three trials were performed. Surface-wipe samples from debris from the first two trials
were analyzed for PCB, lindane, and DDT, and the surface-wipe samples from the
third trial were analyzed for total lead.
Tables 4 and 5 summarize the quantities of PCBs and pesticides on the surface
of each piece of debris before and after cleaning. Table 6 summarizes the quantities
of lead found before and after treatment. The average overall percentage reductions
of PCBs and pesticides achieved during Trials 1 and 2 were greater than 99 and 98
percent respectively. The overall percentage reduction of lead was greater than 98
percent.
After the completion of the bench-scale debris-washing experiments, the
cleaning solution was neutralized to a pH of 8 and then pumped through a series of
particulate filters and finally through activated carbon. During this treatment, the PCB,
lindane, and DDT concentrations were reduced to <2.0, 0.03, and 0.33 /.ig/L, respec-
tively. The concentration of lead was reduced to 0.2 mg/L after treatment. During the
water treatment, it was noticed that a gel-like precipitate was formed when the L-422
cleaning solution was neutralized to pH 8. This precipitate quickly plugged the particu-
late filters and had the potential for clogging the activated carbon drums. As a result,
BG-5, which performed almost as well as L-422 and did not form any precipitate when
neutralized, was selected as the cleaning solution for the pilot-scale study.
After BG-5 was selected as the optimal cleaning solution, experiments were
conducted with 3 percent and 5 percent concentration solutions of BG-5 to determine
the optimum concentration. Again, oily debris from a salvage yard was cleaned;
based on the visual inspection, the 5 percent BG-5 solution obviously performed better
than the 3 percent solution. The optimum temperature recommended by the
22
-------
Figure 4. Bench-scale Debris Washing System.
23
-------
TABLE 3. A COMPARISON OF THE CLEANING CAPABILITIES OF
SURFACTANT SOLUTIONS BASED ON REMOVAL OF OIL AND GREASE
Sample ID
Number
C1 -1-6/20
C2-1-6/20
C3-1-6/20
C4-1-6/20
C5-1-6/20
C6-1-6/20
C1 -2-6/22
C2-2-6/22
C3-2-6/22
C4-2-6/22
C5-2-6/22
C6-2-6/22
01 -5-6/30
C2-5-6/30
C3-5-6/30
C4-5-6/30
C5-5-6/30
C6-5-6/30
C1 -6-6/30
C2-6-6/30
C3-6-6/30
C4-6-6/30
C5-6-6/30
| C6-6-6/30
C1 -7-7/6
C2-7-7/6
C3-7-7/6
C4-7-7/6
C5-7-7/6
| C6-7-7/6
Pretreatment
(g/100cm2)
6.7131
5.2936
5.5088
5.4138
4.7310
5.4889
4.1462
4.0274
4.0025
6.7795
7.3356
4.2899
5.2878
5.0433
5.8143
6.0277
4.1388
4.1278
4.5929
4.9409
5.3973
4.9976
3.9820
4.9440
5.1850
4.8263
5.1807
5.8047
5.0469
4.7127
Posttreatment
(g/100cm2)
0.0071
0.0171
0.0111
0.0060
0.0074
0.0107
0.0313
0.0155
0.0322
0.0057
0.0162
0.0144
0.0688
0.0878
0.0811
0.0533
0.0262
0.0667
0.0221
0.0349
0.1315
0.1498
0.0067
0.0477
0.0017
0.0028
0.0011
0.0023
0.0015
0.0019
Percent
Reduction
99.89
99.68
99.80
99.89
99.84
99.80
99.25
99.61
99.19
99.91
99.78
99.66
98.70
98.26
98.61
99.11
99.37
98.38
99.52
99.29
97.56
97.00
99.83
99.03
99.97
99.94
99.98
99.96
99.97
99.96
Avg. Percent
Reduction
99.82
99.57
98.74
98.71
I
99.96
24
-------
TABLE 4. SUMMARY OF BENCH-SCALE RESULTS OF CONTROLLED
DEBRIS ANALYZED FOR PCBs AND PESTICIDES (TRIAL 1)
Sample ID
Number
r
CM1-SP1-7/21
I
r
CM2-SP1-7/21
I
r
CM3-SP1-7/21
L
r
CB1-SP1-7/21
I
r
CB2-SP1-7/21
L
r
CC1-SP1-7/21
L
r
CP1-SP1-7/21
I
Contaminant
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Lindane
4, 4' ODD
4, 4' DDT
PCB-1260
Pretreatment
(Hg/100cm2)
13,800
1010
6710
3550
12,500
1020
7610
3230
12,300
1020
7800
2990
14,600
1220
7640
2570
12,900
1170
10,100
3360
14,000
1240
10,200
3410
9370
952
7120
2500
g
Posttreatment
(H9/100cm2)
0.75
3.8 U
5.0 U
2.0 U
0.7
3.8 U
5.67
2.0 U
0.7
3.8 U
5.0 U
2.0 U
5.8
3.8 U
11.6
20.3
130
4.9
360
90.4
11.1
3.8 U
28.3
15.3
1.1
3.8 U
12.6
23.4
Percent
Reduction
99.99
599.62
599.93
599.94
99.99
599.63
99.93
599.94
99.99
599.63
599.93
599.93
99.96
599.69
99.85
99.21
98.99
99.58
96.43
97.31
99.92
599.69
99.72
99.55
99.99
599.60
99.82
99.06
Average Overall Performance
Average
Performance
>99.87
**QQ QQ
2.3O.OO
>99.72
>99.62
>99.39
aU indicates that the target compound was not detected at this level.
25
-------
TABLE 5. SUMMARY OF BENCH-SCALE RESULTS OF CONTROLLED
DEBRIS ANALYZED FOR PCBs AND PESTICIDES (TRIAL 2)
Sample ID
Number
CM 1-7/24
CM2-7/24
CM3-7/24
I
CB1-7/24
CC1-7/24
CP1-7/24
Contaminant
Lindane
4, 4' DDT
PCB-1260
Lindane
4, 4' DDT
PCB-1260
Lindane
4, 4' DDT
PCB-1260
Lindane
4, 4' DDT
PCB-1260
Lindane
4. 4' DDT
PCB-1260
Lindane
4. 4' DDT
PCB-1260
Pretreatment
(Hg/100cm2)
11,800
9320
1770
8180
7540
1780
6150
5840
1450
5810
5660
1220
6440
6610
1390
10,300
8400
1620
g
Posttreatment
(u,g/100cm2)
0.13 U
2.32
2.0 U
0.31 U
4.8
2.79
0.41
2.61
2.0 U
3.49
10.5
4.1
397
389
66.1
52
223
35
Percent
Reduction
100
99.97
>99.89
100
99.94
99.84
99.99
99.95
>99.86
99.94
99.81
99.66
93.83
94.11
95.24
99.49
97.34
97.84
Average Overall Performance
Average
Performance
>99.91
99.80
94.39
98.22
>98.08
'U Indicates that the target compound was not detected at this level.
26
-------
TABLE 6. SUMMARY OF BENCH-SCALE RESULTS OF CONTROLLED DEBRIS
ANALYZED FOR LEAD (TRIAL 3)
Metal
Cppcrete
Block
Plastic
Sample ID
Number
CM1-SP3-7/28
CM2-SP3-7/28
CM3-SP3-7/28
CB1-SP3-7/28
CC1-SP3-7/28
CP1-SP3-7/28
Contaminant
Lead
Lead
Lead
Lead
Lead
Lead
Pretreatment
(ug/100cm2)
876
414
450
508
414
446
Posttreatment
(ug/100cm2)
6.0
6.0
<3.0
<3.0
<3.0
<3.0
Percent
Reduction °"
99.31
98.55
>99.33
>99.41
>99.27
>99.33
Average Performance on All Materials Tested
Average
Performance
>99.06
-.. f \v.\y, *.
.. " >. "" V *
>98.08
27
-------
manufacturer of BG-5 was 140°F; therefore, all the bench-scale and subsequent pilot-
scale testings were conducted at 140°F.
4.2 Design, Fabrication, and Initial Testing of Pilot-Scale DWS
Based on the results obtained from bench-scale studies, a pilot-scale DWS was
designed and constructed. The pilot-scale DWS consists of a 300-gallon spray tank; a
300-galIon wash tank; a surfactant holding tank; a rinse-water holding tank; an
oil/water separator; and a solution treatment system comprised of a diatomaceous
earth filter, a series of activated-carbon columns, and an ion exchange column. The
system also includes other ancillary equipment, such as a heater for the 300-gallon
spray tank (to heat the cleaning solution), a stirrer motor, a metal basket, particulate
filters, and pumps (centrifugal and air-diaphragm). Figure 5 presents; a schematic of
the pilot-scale DWS.
The pilot-scale system was scaled up to be 15 times larger than the bench-
scale system. Some of the engineering considerations involved in the design of the
pilot-scale DWS were a scale-up factor, spray characteristics and distribution, spray
angle and mean droplet diameter, flow and pressure balance, solution residence time,
pump performance curves, materials and codes, and compatibility of construction
materials with cleaning solution.
The pilot-scale system was assembled in a warehouse located in Cincinnati,
Ohio. Several tests were conducted with pieces of oil/grease-coatecl metallic objects
found in the warehouse. Surface-wipe samples were obtained before and after the
debris was washed in the pilot-scale system and were analyzed for oil and grease.
Table 7 summarizes the results of this testing. The warehouse testing also involved
the optimization of test parameters such as duration of spray cycle, (duration of wash
cycle, and temperature of the cleaning solution. Based on the results and a visual
inspection of the washed debris, the system was determined to be effective in
removing oil and grease from the surface of these objects.
4.3 Field Demonstration of DWS at the Gray PCB Site
4.3.7 Site Description
The Gray PCB site is approximately 25 acres in size and located in Hopkinsville,
Kentucky. From 1968 to 1987, a metal reclaiming facility was operated at this site.
Operations included the open burning of electrical transformers to recover copper for
resale. The soil where the transformers were burned is contaminated with lead and
PCBs. On March 19, 1987, representatives from the Kentucky Department of Environ-
mental Protection conducted an inspection at the site and observed the following con-
ditions: 1) the facility was no longer in operation; 2) approximately 70 to 80 burned-
out transformers were still on site, along with other large amounts of materials such as
28
-------
Contaminated
Debris
Clean Debris
Step 1 - Spray Cycle
Step 2 - Wash Cycle
IB Step 3- Rinse Cycle
DE Filter
Water Treatment Step
Pump
Contaminated
Debris
Surfactant
Tank
Activated Carbon
Oil/Water Separator
Treated Water
Storage Tank
rfffffffffffffffffffffffffffffffffffffffffSSft.
Cartridge
Filters
Figure 5. Schematic of pilot-scale Debris Washing System.
-------
TABLE 7. RESULTS OF SURFACE WIPE SAMPLES ANALYZED FOR OIL
AND GREASE DURING WAREHOUSE TESTING
Sample ID
Number
D1 -1-1 0/5
D2-1-10/5
D3-1-10/5
D4-1-10/5
D5-1-10/5
D1 -2-1 0/9
D2-2-10/9
D3-2-10/9
D4-2-10/9
D5-2-10/9
D1 -3-1 0/10
D2-3-10/10
D3-3-10/10
D4-3-10/10
D5-3-10/10
Pretreatment
(mg/100cnf)
142
431
690
94
486
312
422
283
583
2602
174
602
286
227
73
Posttreatment
(mg/1 00 cm2 )
47.1
41.2
76.1
76.2
67.3
32.2
53.2
39.1
55.2
233.0
48.0
62.7
55.4
61.2
55.9
Percent
Reduction
66.7
90.4
89.0
19.0
86.1
90.0
87.4
86.2
90.5
91.0
72.4
90.0
80.6
73.0
23.3
Avg. Percent
Reduction
i
: 70.2
89.0
i
67.9
30
-------
asbestos-covered pipes, automobiles, and miscellaneous scrap metal; and 3) multiple
burn areas and two sink holes were noted.
4.3.2 Operating Parameters
\
During the demonstration, the operating parameters (such as surfactant con-
centration and temperature, flowrate and pressure during wash and spray cycle, and
washing time) were established by ITEP based on the results of bench-scale testing
and on the engineering design of the pilot-scale DWS.
The optimum cleaning solution concentration of 5 percent was recommended
by the manufacturer of the surfactant solution. Prior to initiation of the process opera-
tion, a 5 percent surfactant solution was prepared in the detergent holding tank by
adding 15 gallons of surfactant to 285 gallons of water. The mixture was agitated and
heated to 140°F. An additional 300-gallon of 5 percent surfactant solution was also
prepared in the oil/water separator. Since the same cleaning solution was reused
during the entire demonstration, the solution mixture was recharged after every five
batches of debris washing by adding approximately 5 to 10 gallons of fresh surfactant
solution in both tanks (surfactant holding tank and oil/water separator). The amount
of fresh surfactant added to these tanks was based upon the visual observation of the
dirt in the solution mixture. Although this was a crude way to maintain the concen-
tration of the cleaning solution, this method was found to be most practical during the
demonstration.
The optimum solution temperature of 140° F was recommended by the
manufacturer of surfacant solution. During the first batch of debris washing, it was
noted that the temperature of the cleaning solution dropped 15° to 20 °F due to heat
loss in pipes, tank, and contact with the debris. To avoid this loss in temperature, for
all the subsequent batches of debris washing, the cleaning solution was heated to
about 160°F at the beginning of each new batch. Hence, by doing so, the
temperature of the cleaning solution was maintained around 140°F during the washing
process.
The flow rates and pressure of the cleaning solution during spray, wash and
rinse cycles in the respective tanks remain within ±10 percent of the desired values.
The operating parameters matrix for the DWS demonstration are summarized in Ta-
ble 8.
4.3.3 Process Operation of DWS During Field Testing
After the warehouse testing, the DWS was disassembled and loaded onto a
48-foot semitrailer. The system and ancillary equipment were transported to the Gray
PCB site on December 5, 1989. The entire DWS was reassembled on a 25-ft x 24-ft
concrete pad that had been poured on the site prior to the arrival of the equipment. A
31
-------
temporary enclosure (approximately 25 ft high) was built on the concrete pad to en-
close the DWS and to protect the equipment and the surfactant solution from rain and
cold weather. During the course of this demonstration, the ambient temperature at the
site ranged from -30 °F to 50 °F. Figures 6 and 7 show the assembled DWS at a haz-
ardous waste site and the DWS enclosure at the Gray PCB site.
TABLE 8. OPERATING PARAMETERS MATRIX FOR DWS DEMONSTRATIONS
Operating parameters
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Concentration of surfactant solution
Temperature of surfactant solution
Flow rate during spray/rinse cycles in spray tank
Pressure during spray/rinse cycles in spray tank
Flow rate during wash cycle in wash tank
Pressure during wash cycle in wash tank
Flow rate during solution recycle in oil/water
separator
Time of spray cycle
Time of wash cycle
Time of rinse cycle
Optimum
value
5% (v/v)
140°F
100 gpm
60 psi
350 gpm
40 psi
5 gpm
60 min
60 min
30 min
Before the cleaning process was begun, the transformer casings (ranging from 5
to 100 gallons in size) were cut in half with a circular metal cutting saw. A pretreat-
ment sample was obtained from each half of the transformer casings by a surface-
wipe technique (1).
The transformer halves were placed in the metal basket, and lowered into the
spray tank. The tank was equipped with multiple water jets to blast loosely adhered
contaminants and dirt from the debris. The cleaning solution (BG-5), which had a con-
centration of 5 percent and a temperature of 140°F, was pumped from the surfactant
holding tank into the spray tank and then recycled back into the holding tank. After
the spray cycle, the entire quantity of the cleaning solution in the surfactant holding
tank was pumped into the wash tank. The basket was removed from the spray tank
and transferred to the wash tank, where the debris was subjected to a high-turbulence
wash. Each batch of debris was cleaned for 1 hour in the spray tank and 1 hour in
the wash tank. During both spray and wash cycles, a portion of the cleaning solution
was cycled through a closed-loop system in which the oil/PCB-contaminated cleaning
32
-------
Figure 6. The assembled pilot-scale DWS at a hazardous waste site.
33
-------
Figure 7. Enclosure for the pilot-scale DWS at the Gray site.
34
-------
solution was passed through an oil/water separator and then through a diatomaceous
earth filter. The clean solution was then recycled into the DWS. After the wash cycle,
the basket containing the debris was returned to the spray tank, where it was rinsed
with fresh water.
Upon completion of the cleaning process, posttreatment wipe samples were
obtained from each of the transformer pieces to assess the post-decontamination lev-
els of PCBs. The Quality Assurance Project Plan was strictly followed during sampling.
The surface-wiping procedure was carried out as described in the Field Manual for
Grid Sampling of PCB Spill Sites to Verify Cleanup (1). (A detailed description of
surface-wipe sampling is presented in Section 5.)
4.3.4 Results of Field Demonstration
Table 9 summarizes the average concentrations of PCBs on the internal surfaces
of the transformer casings before and after cleaning. The before-treatment concen-
tration ranged from 0.1 to 98 /xg/100 cm2. The posttreatment analysis showed that all
but seven of the cleaned transformer pieces had a PCB concentration of less than the
acceptable level of 10 Mg/100 cm2. The seven transformer pieces with a concentra-
tion greater than the acceptable level were again washed in the DWS, and posttreat-
ment samples were again obtained and analyzed. The PCB concentration in these
seven samples after the second wash was below the detection limit of 0.1 /*g/
100 cm2. The Quality Assurance/Quality Control of analyses performed on wipes is
discussed in Section 5.
After all transformers at the site were treated, the surfactant solution and the rinse
water were neutralized to a pH of around 8 with concentrated sulfuric acid. The
neutralized surfactant solution and rinse water were treated in the water-treatment sys-
tem by passing the solutions through a series of particulate filters, then through
activated-carbon drums, and finally through an ion-exchange column. The treated
water was stored in a 1000-gallon polyethylene tank pending analysis. The before-
and after-treatment water samples were collected and analyzed for PCBs and selected
metals (cadmium, copper, chromium, lead, nickel, and arsenic).
The PCB concentration in the water was reduced by the treatment system to
below the detection limit. The concentrations of each of the six metals (except
arsenic) Were reduced to the allowable discharge levels set by the city of Hopkinsville.
Table 10 summarizes the results of pre- and posttreatment water samples. Upon re-
ceipt of the analytical results of the water, the stored treated water was pumped into a
plastic-covered, 10,000-yd3 pile of contaminated soil at the site through a 3/4-in. inci-
sion in the plastic covering at the top of the soil pile with a rubber hose. After all of
the water was pumped into the contaminated soil, the hose was removed and the inci-
sion was covered with a duct tape.
35
-------
TABLE 9. ANALYTICAL RESULTS OBTAINED DURING FIELD
DEMONSTRATION OF DWS AT GRAY PCB SITE
Batch Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Average PCB Concentration on Surfaces (ng/100 cm2)
Before Cleaning After Cleaning
Average
19.7(Na = 10)
9.9 (N= 6)
6.6 (N= 4)
4.1 (N = 6)
4.0 (N= 8)
2.0 (N= 4)
2.8 (N = 2)
23.5 (N = 5)
8.3 (N= 4)
5.2 (N= 4)
9.4 (N= 4)
48.8 (N = 4)
12.3 (N = 2)
16.7(N = 2)
18.5 (N = 4)
11.3 (N = 2)
24.8 (N = 4)
8.4 (N- 5)
8.3 (N= 4)
24.0 (N = 3)
18.6 (N - 8)
25.0 (N = 4)
8.6 (N= 4)
6.8 (N = 8)
Range
<0.1-94.0
4.8-17.0
5.0-9.9
<0.1-12.0
<0.1 -28.0
<0.1-7.8
1.4-4.3
<0.1-70.0
2.9-23.0
<0.1 -9.7
<0.1-17.0
2.3-98.0
9.6-15.0
8.7-25.0
8.1-27.0
8.6-14.0
1.1-80.0
<0.1-19.0
<0.1-18.0
13.0-45.0
<0.1 -44.0
12.0-35.0
1.5-18.0
<0.1 -31.0
Average
1.5(N = 10)
1.5 (N = 6)
1.4 (N =4)
0.8 (N = 6)
<0.1(N = 8)
2.9 (N= 4)
3.9 (N= 2)
1.3 (N -5)
3.1 (N=4)
1.9 (N =4)
3.0 (N= 4)
1.1 (N=4)
5.1 (N = 2)
<0.1 (N = 2)
<0.1 (N = 4)
2.0 (N= 2)
2.2 (N= 4)
3.4 (N = 5)
3.2 (N= 4)
3.3 (N= 3)
0.4 (N= 8)
<0.1 (N = 4)
<0.1 (N = 4)
0.3 (N= 8)
Range
<0.1-9.7
<0.1 -4.7
<0.1-3.3
<0.1-4.1
<0.1-<0.1
<0.1-10.0
<0.1 -7.7
<0.1-3.8
1.5-4.9
<0.1-2.8
<0.1-9.5
<0.1-3.2
<0.1-10.0
<0.1 -<0.1
<0.1 -<0.1
1.5-2.5
<0.1-8.4
<0.1-7.4
<0.1 - 5.3
<0.1-9.8
<0.1 - 2.1
<0.1-<0.1
<0.1-<0.1
<0.1-1.4
Average
Percentage
Removed
92
85
79
80
>98
b
b
94
63
63
67
98
59
>99
>99
82
91
60
61
86
98
>99
>99
96
a N Indicates the number of samples.
b The distribution of PCB contamination on the surfaces of these transformers is obviously not uniform and therefore in some
cases a meaningful comparison of post-treatment PCB levels with pre-treatment levels cannot be achieved.
36
-------
CO
-J
TABLE 10. ANALYTICAL RESULTS OF PROCESS WATER GENERATED
DURING GRAY PCB SITE DEMONSTRATION
Sample ID Number
Pretreatment
Pre-Ows-1/18a
Pre-Det-1/18a
Pre-Rins-1/18a
Pre-Owws-1/18
Pre-Det-1/18a
Pre-Rinse-1/18
Posttreatment
Post1-Det-1/18
Post3-Det-1/19
Post2-Rins-1/19
Post2-Det-1/18
Post1-Rins-1/19
__
""
Analyses
PCB fug/U
Cleaning Solution
Cleaning Solution
Rinse Water
METALS (tig/L)
Cleaning Solution
Copper
Lead
Cadmium
Chromium
Arsenic
Rinse Water
Copper
Lead
Cadmium
Chromium
Arsenic
Pretreatment
12/7.8a'b
5.4/5.4a-b
18/9.1a-b
915/10273
31160/156073
80/1 38a
6150/44593
3870/37703
614
13210
45
2056
1940
Posttreatment
4.2b
1.3b
<1.0
723
17400
63
4860
3720
103
3834
<10
882
3650
Duplicate analyses.
b Estimated concentration due to matrix interferences.
-------
Finally, the equipment was decontaminated with a high-pressure wash. The wash
water generated during decontamination was collected and treated in the water
treatment system. The system and the enclosure were then disassembled and loaded
into the semitrailer for transport back to Cincinnati, Ohio.
During this site cleanup, 75 transformers (approximately 5000 Ib) were washed in
the DWS. A total of 1000 gallons of process water was used in the demonstration of
the DWS. All of these transformers are now considered to be clean and can be sold
to scrap metal dealers or to a smelter for reuse.
The copies of bench- and pilot-scale analytical data provided by the Hayden
Environmental Group, Inc., are included in Volume II of this report.
4.4 Demonstration of DWS at Shaver's Farm Site
4.4.1 Site Description
.t
The Shaver's Farm Drum Disposal Site consists of a 5-acre tract of land in Walker
County, Georgia. Drums containing potentially hazardous wastes (Dicamba,
benzonitrile, 2,4-D, and 2,4,5-T) and nonregulated latex-related wastes (butadiene and
styrene) were allegedly buried on site by a construction and waste company in the
early 1970's. At least three Potentially Responsible Parties (PRPs) contracted with this
firm to transport and dispose of their manufacturing residues. The drum disposal area
slopes nearly 25 degrees at the base of a small ridge and drains into a series of small
sinkholes situated approximately 400 to 450 feet downgradient.
i .. * . ''-
To date, EPA Region IV has excavated more than 4000 drums from one location
on this site. The drums contained residues from the production of benzonitrile, a
pesticide, and waste from the production of Dicamba, a herbicide. Prior to the initia-
tion of excavation activities, no waste materials were evident at land surface. Available
information suggests that there is at least one more location on site where drums are
buried. It is estimated by EPA Region IV that the number of drums at this second
location might range'between 6,000 and 12,000. Figure 8 presents an aerial photo-
graph of the site.
4.4.2 Bench-Scale Experiments and Results
Before final arrangements were made for the field demonstration of DWS at the
Shaver's Farm Site in Georgia, a series of bench-scale experiments were conducted to
determine the capability of the DWS to remove the primary contaminants (Dicamba
and benzonitrile) found at the site.
38
-------
CO
CO
ITEP Pilot-Scale
Debris Washing System
Figure 8. Aerial photograph of Shaver's Farm site.
-------
A spike solution was prepared by adding a mixture of benzonitrile (12 g) and
Dicamba (3 g) in acetone (81.5 g). To simulate the actual site scenario, a used, rust-
ed 55-galIon drum was cut into small pieces and using a brush, each piece was evenly
coated with the mixture of Dicamba and benzonitrile. The pieces were allowed to dry
for about 12 hours. The spiked debris was cleaned in the bench-scale DWS. The
surface-wipe samples taken before and after cleaning were obtained and analyzed for
Dicamba and benzonitrile. The results of bench-scale experiments for benzonitrile and
Dicamba are summarized in Tables 11 and 12, respectively.
The benzonitrile concentration was reduced from an initial average of 1660
100 cm2 (range of 54 to 4600 /ig/100 cm2) to below the detection limit (5 HQ/
100 cm2). In the Dicamba analysis, an unknown interference caused all the analytical
results (before and after treatment) to show Dicamba concentration of below the de-
tection limit (0.3 /xg/100 cm2). Although Dicamba analytical results were not useful,
the results for benzonitrile were certainly conclusive. As described below, the validity
of the analytical method for Dicamba was verified by analyses of matrix spike/matrix
spike duplicates.
The validity of the method used to analyze Dicamba and benzonitrile was verified
by spiking four samples (two wipes and two metal strips) with a known quantity of
Dicamba and two wipe samples with a known quantity of benzonitrile and submitting
them to the laboratory for Dicamba and benzonitrile analyses. The results showed
that, in the case of Dicamba, the recovery was 122 percent in the wipe samples and
104 percent in the metal strip samples; the benzonitrile recovery was 94 percent in the
wipe sample. Tables 13 and 14 present the percent recoveries of Dicamba and ben-
zonitrile.
The contaminated cleaning solution generated during bench-scale experiments
was treated with activated carbon to reduce the contaminant concentrations to accept-
able levels. Samples of wash solution before and after carbon treatment were ana-
lyzed for benzonitrile and Dicamba. The concentration of benzonitrile in the washing
solution was reduced to below the detection limit (5 /xg/L) from an initial concentration
of 92 p.g/L. The Dicamba concentration in the solution was reduced from 12,100 to
347 /zg/L The results of water analysis for benzonitrile and Dicamba are , also shown
in Tables 11 and 12, respectively.
i
4.4.3 Process Operation of DWS During Field Testing
After the bench-scale testing, the DWS and the ancillary equipment were trans-
ported to the Shaver's Farm site on a 48-foot semitrailer. The entire DWS was assem-
bled on a 25-ft x 25-ft concrete pad that was poured on the site before the equipment
arrived. The temporary enclosure used during the demonstration in Hopkinsville was
used again at the Shaver's Farm site. The enclosure was built on the concrete pad to
enclose the DWS and to protect the equipment and the surfactant solution from rain
40
-------
TABLE 11. BENCH-SCALE RESULTS: ANALYSES FOR BENZONITRILE
Wipes (u.g/100cm2)
Process Water (ng/L)
Sample ID Number
Pretreatment
P3-1-4/10
P3-2-4/10
P3-3-4/11
P3-4-4/11
PRE-W-4/13
Posttreatment
PT3-1-4/10
PT3-2-4/10
PT3-3-4/1 1
PT3-4-4/1 1
POST-W-4/13
Benzonitrile
Pretreatment
54
92
4600
1900
92
Posttreatment
NDa (<5.0)b
ND (<5.0)
ND (<5.0)
ND (<5.0)
ND (<5.0)
a Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
TABLE 12. BENCH-SCALE RESULTS: ANALYSES FOR DICAMBA
Wipes (u.g/100cm2)
-. .
.--.)-
'*
Process Water (ng/L)
Sample ID Number
Pretreatment
P1 -1-4/10
P2-1-4/10
P1 -2-4/10
" P2-2-4/10
P1 -3-4/11
P2-3-4/11
P1 -4-4/11
P2-4-4/11
PRE-W-4/12
Posttreatment
PT1 -1-4/10 .:
PT2-1-4/10
PT1 -2-4/11
PT2-2-4/11
PT1 -3-4/11
PT2-3-4/11
PT1 -4-4/11
PT2-4-4/11
Post-W-4/12
Dicamba
Pretreatment
<0.3
<0.3
<0.3
-------
TABLE 13. METHOD VALIDATION: RESULTS OF DICAMBA ANALYSES OF GAUZE AND
METAL STRIP SPIKED WITH DICAMBA
Sample ID
WL-1-6/14
M-H-1-6/14
Sample Type
Wipe
Metal Strips
Quantity
Spiked, jig
25,000
60,000
Analytical Results, [ig
Sample 2
Sample 1 (duplicate)
32,000 29,000
63,000 62,000
Average, jig
30,500
62,500
Recovery, %
122
104
TABLE 14. METHOD VALIDATION: RESULT OF BENZONITRILE ANALYSIS OF GAUZE
SPIKED WITH BENZONITRILE
Sample ID
Wipe MS/MSD
Sample Type
Wipe
Quantity
Spiked, jig
1,000
Analytical Results, jig
Sample 2
Sample 1 (duplicate)
843 1,029
Average, ng
936
Recovery, %
94
-------
and hot weather. During the course of this demonstration, the ambient temperature
ranged from 75 to 105 °F.
Before starting the cleaning process, the contaminated 55-gal drums were cut into
four pieces with a circular metal cutting saw. A pretreatment sample was obtained
from each of these four pieces by the surface-wipe technique. For corroboration of
the results of the surface-wipe tests and to test whether contaminants were imbedded
in the surfaces of the drums, a nibbler was used to take four or five metal strips (ap-
proximate size 6 cm x 3 cm) from one of the four drum pieces before the cleaning
process was initiated.
The drum pieces were placed directly into the spray tank and the cleaning
process was begun. The basket was not used because it was easier to place the
drum pieces in the spray tank manually than to use the forklift and the basket. After
completion of the cleaning process, posttreatment wipe samples were obtained from
each of the drum pieces to assess the post-decontamination levels of herbicides and
pesticides. Posttreatment metal strip samples were also taken from the same piece of
drum where the pretreatment sample was obtained. The Quality Assurance Project
Plan (QAPjP) was strictly followed during sampling.
Ten batches of drums were treated during this demonstration. After completion
of the treatment, the cleaning solution and rinse water resulting from the decontamina-
tion of the debris in the DWS were treated in the water-treatment system. The treated
water was stored in the 1000-gallon polyethylene tank pending analysis. In this dem-
onstration, however, an ion exchange column was not used in the water-treatment
system because the drums did not have any heavy metal contamination. The entire
debris-cleaning procedure and the water-treatment process used at the Shaver's Farm
site were similar to those used during the Gray PCB site demonstration, which were
described in detail in Subsection 4.3.
The surface wipes, the metal strips, and the water samples obtained during the
demonstration were sent to Radian Corporation (Austin, TX) for analyses. All these
samples were analyzed for four semivolatiles (benzonitrile, 2,4-dichlorophenol, 2,6-
dichlorophenol, and 1,2,4,-trichlorobenzene), 3-herbicides (Dicamba, 2,4-D, and'
2,4,5-T), dioxins, and furans. Although the primary contaminants suspected on the
Shaver's Farm site were benzonitrile and Dicamba, analyses of the remaining three
semivolatiles, and two herbicides were performed at the request of EPA Headquarters
in Washington, D.C.
For PCBs, the EPA has established a cleanup standard of 10 /*g/100 cm2 for
surfaces. Because neither Dicamba nor benzonitrile is a frequently occurring contami
nant on a hazardous waste site, the cleanup criteria for surfaces has not yet been
established for these two contaminants. For the Shaver's Farm site, EPA has set a
cleanup target concentration of 25 ppm in soil for both Dicamba and benzonitrile.
43
-------
Based on the surface cleanup standards for PCBs and the soil standards for Dicamba
and benzonitrile, it is assumed that when EPA sets up the cleanup criterion for surface
concentration of Dicamba and benzonitrile at a later date, it will not exceed 10 M9/100
cm2. Hence, during the field demonstration of DWS at Shaver's Farm, it was decided
that all wipe samples to be analyzed for Dicamba and benzonitrile will have a detection
limit of 5 M9/100 cm2 or less.
4.4.4 Results of Field Demonstration
Table 15 summarizes the surface-wipe concentrations of semivolatiles on the
internal surfaces of the drums before and after cleaning. The before-treatment con-
centration of benzonitrile ranged from 8 to 47,000 /xg/100 cm2 (average, 4556
Atg/100 cm2), and the posttreatment concentration ranged from below the detection
limit to 117 M9/100 cm2 (average 10 M9/100 cm2). The pretreatment concentrations of
2,4-dichlorophenol in all the samples except two samples from Batch 2 were below the
detection limit. The posttreatment concentrations of 2,4-dichlorophenol in all samples
were below the detection limit. The remaining two semivolatiles (2,6-dichlorophenol
and 1,2,4-trichlorobenzene) were not detected in any of the pre- or the posttreatment
samples.
Table 16 summarizes the concentrations of semivolatiles in the metal strips. The
pretreatment concentration of benzonitrile ranged from below the detection limit to
190 ng/g (average 42 M9/9), and the posttreatment concentration ranged from below
the detection limit to 0.89 iug/g (average 0.35 /ng/g). The concentrations of pre- and
posttreatment samples analyzed for 2,4-dichlorophenol, 2,6-dichlorophenol, and 1,2,4-
trichlorobenzene were below the detection limit with the exception of the sample from
Batch 3 (analyzed for 2,4-dichlorophenol), which had a pretreatment concentration of
26 Aig/g and a posttreatment concentration of 1.9 Atg/g.
Table 17 summarizes the results of surface-wipe samples analyzed for herbicides
(Dicamba, 2,4-D, and 2,4,5-T). The pretreatment concentrations of Dicamba ranged
from below the detection limit to 180 M9/"IOO cm2 (average 23 M9/100 cm2) and
posttreatment concentration ranged from below the detection limit to 5.7 ng/ 100 cm
(average 1 M9/100 cm2). The remaining two herbicides (2,4-D and 2,4,5-T) were not
detected in any of the pre- or the posttreatment samples.
Table 18 presents the results of metal strip samples analyzed for Dicamba, 2,4-D,
and 2,4,5-T. The pretreatment concentrations of Dicamba ranged from 0.057 to
82 /xg/g (average 16-jig/g), and the posttreatment concentrations ranged from below
the detection limit to 13 /zg/g (average 2 /xg/g). Concentrations of 2,4-D, and 2,4,5-T
were not detected in any of the pre- or the posttreatment samples.
In addition to the semivolatiles and herbicides analyses, two metal strip samples
were analyzed for dioxins and furans. Table 19 summarizes the results of these two
44
-------
TABLE 15. RESULTS OF SURFACE WIPE SAMPLES ANALYZED FOR BENZONITRILE,
2,4-DICHLOROPHENOL, 2,6-DICHLOROPHENOL, AND 1,2,4-TRICHLOROBENZENE
DURING FIELD DEMONSTRATION OF DWS AT SHAVER'S FARM SITE
(|ig/100cm2)
Batch Number
1
2
3
4
5
6
7
8
9
10
Sample ID Number
Pretreatment
P1 -1-7/1 7
P2-1-7/17
P1 -2-7/1 8
P2-2-7/18
P1 -3-7/1 8
P2-3-7/18
P1 -4-7/1 9
P4-4-7/19
P1 -5-7/1 9
P2-5-7/19
P3-6-7/20
P4-6-7/20
P1 -7-8/2
P2-7-8/2
P3-8-8/2
P3-9-8/2
P4-9-8/2
P1 -10-8/2
P2-1 0-8/2
Posttreatrnent
PT1-1-7/17
PT2-1-7/17
PT1 -2-7/20
PT2-2-7/20
PT1 -3-7/20
PT2-3-7/20
PT1 -4-7/20
PT4-4-7/20
PT1 -5-7/20
PT2-5-7/20
PT3-6-7/20
PT4-6-7/20
PT1 -7-8/2
PT2-7-8/2
PT3-8-8/2
PT3-9-8/3
PT4-9%3
PT1 -10-8/3
PT2-1 0-8/3
Benzonltrile
Pretreatment
180a(50)b
130a(50)
125
90
43
28
4400
2700
47000
22000
10a(5)
8a(5)
200
320
1400
3000
3500
22a (5)
1400
Posttreatrnent
NDC
ND
117
7.8a(5)
ND
ND
ND
ND
10a(5)
7.9a (5)
ND
ND
ND
10a(5)
28
ND
7a(5)
ND
. ND
2,4-Dlchlorophenol
Pretreatment
ND (50)
ND (50)
34
43
ND
ND
NAd
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Posttreatrnent
ND
ND
ND
ND
16a(5)
14a (5)
NA
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
2,6-Dichlorophenol
Pretreatment
ND (50)
ND (50)
ND
ND
ND
ND
NA
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Posttreatrnent
ND
ND
ND
ND
ND
ND
NA
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
1,2,4-Trlchlorobenzene
Pretreatment
ND (50)
ND (50)
ND
ND
ND
ND
NA
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Posttreatrnent
ND
ND
ND
ND
ND
ND
NA
NA
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
01
a Estimated result less than 5 times detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
c None detected in excess of the minimum detectable concentration of 5 ng/1 OOcm^ unless otherwise specified.
d Not analyzed.
-------
TABLE 16. RESULTS OF METAL STRIP SAMPLES ANALYZED FOR BENZONITRILE,
2,4-DICHLOROPHENOL, 2,6-DICHLOROPHENOL, AND 1,2,4-TRICHLOROBENZENE
DURING FIELD DEMONSTRATION OF DWS AT SHAVER'S FARM SITE
Batch Number
1
2
3
4
5
6
7
8
9
10
Sample ID Number
Pretreatment
P3-1-7/17
P3-2-7/18
P3-3-7/18
P5-4-7/19
P5-5-7/19
P5-6-7/20
P5-7-8/2
P5-8-8/2
P5-9-8/2
P5-1 0-8/2
Posttreatment
PT3-1-7/17
PT3-2-7/20
PT3-3-7/20
PT5-4-7/20
PT5-5-7/20
PT5-6-7/20
PT5-7-8/2
PT5-8-8/2
PT5-9-8/3
PT5-1 0-8/3
Benzonltrlle
Pretreatment
4.9
5.4a(1.2)
ND (0.15)
ND
190
1.0
140
16
61
ND(1.2)
Posttreatment
0.89
0.31 a (0.1 2)
ND (0.10)
ND
0.8a (0.22)
ND (0.08)
0.28a(0.16)
0.41 a (0.1 5)
0.1 8a (0.1 7)
ND(0.11)
2,4-Dlchlorophenol
3retreatment
0.33a (0.33)b
5.0a(1.2)
26
NAd
ND (0.28)
ND (0.085)
ND (0.19)
ND (0.14)
ND (0.10)
ND (0.12)
Posttreatment
NDC(0.15)
ND (0.12)
1.9
NA
ND (0.22)
ND (0.08)
ND(0.16)
ND (0.15)
ND (0.17)
ND(0.11)
2,6-Dlchlorophenol
Pretreatment
ND (0.15)
ND(1.2)
ND (0.15)
NA
ND (0.28)
ND (0.085)
ND(0.19)
ND (0.14)
ND(0.10)
ND (0.12)
Posttreatment
ND (0.15)
ND (0.12)
ND(0.10)
NA
ND (0.22)
ND (0.08)
ND(0.16)
ND (0.15)
ND (0.17)
ND(0.11)
1,2,4-Trlchlorobenzene
Pretreatment
ND (0.15)
ND(1.2)
ND (0.15)
NA
ND (0.28)
0.1 3a (0.085;
ND (0.19)
ND (0.14)
ND (0.10)
ND (0.12)
Posttreatment
ND (0.15)
ND (0.15)
ND(0.10)
NA
ND (0.22)
ND (0.080)
ND(0.16)
ND (0.15)
ND (0.17)
ND(0.11)
o>
a Estimated result less than 5 times detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
c None detected at specified detection limit. .
d Not analyzed.
-------
TABLE 17. RESULTS OF SURFACE WIPE SAMPLES ANALYZED FOR DICAMBA,
2,4-D, AND 2,4,5-T DURING FIELD DEMONSTRATION OF DWS
AT SHAVER'S FARM SITE
(H-g/IOOcm2)
Batch Number
4
5
6
7
8
9
10
Sample ID Number
Pretreatrnent
P2-4-7/19
P3-4-7/19
P3-5-7/19
P4-5-7/19
P1 -6-7/20
P2-6-7/20
P3-7-8/2
P4-7-8/2
P1 -8-8/2
P2-8-8/2
P1 -9-8/2
P2-9-8/2
P3-1 0-8/2
P4-1 0-8/2
Postireatment
PT2-4-7/20
PT3-4-7/20
PT3-5-7/20
PT4-5-7/20
PT1 -6-7/20
PT2-6-7/20
PT3-7-8/2
PT4-7-8/2
PT1 -8-8/2
PT2-8-8/2
PT1 -9-8/3
PT2-9-8/3
PT3-1 0-8/3
PT4-1 0-8/3
Dicamba
Pretreatrnent
1.9
3.4
ND
ND
ND (2.7)
ND (2.7)
7.3a (2.7)
15
55
13
1.7
ND (2.7)
41
180
Posttreatment
0.63a (0.27)b
ND
ND
2.6
ND
ND (2.7)
1.8
2.3
5.7a (2.7)
0.62a (0.27)
0.63a (0.27)
ND
0.30a (0.27)
0.34a (0.27)
2,4-D
Preireatment
NDC
NAd
ND
ND
ND (12)
ND(12)
ND
ND(12)
ND (12)
ND
ND
ND(12)
ND (12)
ND (12)
Posttreatment
ND
NA
ND
ND
ND
ND (12)
ND
ND
ND (12)
ND
ND
ND
ND
ND
2,4,5-T
Pretreatrnent
ND
NA
ND
ND
ND (2.0)
ND (2.0)
ND
ND (2.0)
ND (2.0)
ND
ND
ND (2.0)
ND (2.0)
ND (2.0)
Posttreatment
ND
NA
ND
ND
ND
ND (2.0)
ND
ND
ND (2.0)
ND
ND
ND
ND
ND
a Estimated result less than 5 times detection limit.
" Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
c None detected in excess of minimum detectable concentration of Dicamba at 0.27; 2,4-D
unless otherwise specified.
" Not analyzed.
at 1.2; and 2,4,5-T at 0.20
-------
TABLE 18 RESULTS OF METAL STRIP SAMPLES ANALYZED FOR DICAMBA, 2,4-D, AND 2,4,5-T
DURING FIELD DEMONSTRATION OF DWS AT SHAVER'S FARM SITE
frg/g)
Batch Number
2
3
4
6
7
9
10
Sample ID Number
Pretreatment
1P3-2-7/18
1P3-3-7/18
1P5-4-7/19
1P5-6-7/20
1P5-7-8/2
1P5-9-8/2
1P5-1 0-8/2
Posttreatment
:1PT3-2-7/20
1PT3-3-7/20
1PT5-4-7/20
1PT5-6-7/20
1PT5-7-8/2
1PT5-9-8/3
1PT5-1 0-8/3
Dlcamba
Pretreatment
0.37
82
0.057
25
2.3
2.3
0.40
Posttreatment
0.0303 (0.0097)b
13
0.0203 (0.0065)
ND (0.059)
0.023a (0.062)
0.11
0.01 7a (0.0062)
2,4-D
Pretreatment
NDC (0.34)
ND (0.36)
ND (0.029)
ND (0.26)
ND (0.50)
ND (0.32)
ND (0.24)
Posttreatment
ND (0.043)
ND (0.38)
ND (0.029)
ND (0.26)
ND (0.28)
ND (0.040)
ND (0.028)
2,4
Pretreatment
ND (0.0056)
ND (0.060)
ND (0.0048)
ND (0.044)
ND (0.084)
ND (0.054)
ND (0.040)
5-T
Posttreatment
ND (0.0072)
ND (0.064)
ND (0.0048)
ND (0.044)
ND (0.046)
ND (0.0066)
ND (0.0046)
a Estimated result less than 5 times detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
c None detected at specified detection limit.
-------
CD
TABLE 19. RESULTS OF METAL STRIP SAMPLES ANALYZED FOR
DIOXINS AND FURANS DURING FIELD DEMONSTRATION
OF DWS AT SHAVER'S FARM SITE
(ng/g)
Analyses
HpCDD
HpCDF
HxCDD
HxCDF
OCDD
OCDF
PeCDD
PeCDF
TCDD
2,3,7,8-TCDD
TCDF
Sample ID: 1P3-1-7/17
Pretreatment
1.2
ND(0.16)
2.8
ND (0.12)
4.1
ND(0.19)
ND(0.11)
ND (0.066)
ND(0.10)
ND(0.10)
ND (0.062)
Posttreatment
NDa (0.30)b
ND (0.23)
ND (0.24)
ND(0.11)
ND(0.66)
ND (0.36)
ND (0.14)
ND (0.085)
ND (0.13)
ND (0.13)
ND (0.079)
Sample ID: 1P5-8-8/2
Pretreatment
ND (0.22)
ND(0.17)
ND(0.18)
ND(0.11)
1.6° (0.45)
ND (0.22)
ND(0.10)
ND (0.065)
ND(0.10)
ND(0.10)
ND (0.065)
Posttreatment
ND(0.31)
ND (0.23)
0.71 c (0.27)
ND(0.17)
5.5
ND (0.34)
ND(0.16)
ND (0.092)
ND (0.15)
ND (0.15)
ND (0.089)
Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the
c Estimated result less than 5 times detection limit. " ; '
''- V1::.- ' . ' - ..." '- -- * ..'. V . , .''- '." .- -
anaiyte.
-------
samples. In the first sample, only three congeners were present in the pretreatment
sample. The pretreatment concentrations of these three congeners (HpCDD, HxCDD,
and OCDD) were 1.2, 2.8, and 4.1 ng/g, respectively. The posttreatrnent iconcentra-
tions of all congeners were below the detection limit. In the second sample, only
OCDD was identified. The pretreatment concentration was 1.6 ng/g (estimated value),
and the posttreatrnent concentration was 5.5 ng/g. In the case of Sample 2, the
OCDD concentration apparently increased after the cleaning process, probably be-
cause of interferences during analysis. :
The pre- and posttreatrnent samples of process water were analyzed for benzo-
nitrile, 2,4-dichlorophenol, 2,6-dichlorophenol, 1,2,4-trichlorobenzene, Dicamba, 2,4-D,
2,4,5-T, dioxins, and furans. The concentrations of all these compounds except ben-
zonitrile and dicamba were below detection limit in pre- and posttreatrnent samples.
The concentration of benzonitrile in the pretreatment water samples was 250 and
400 /ig/L (analyzed in duplicate), and the posttreatrnent concentration was below the
detection limit. The concentration of Dicamba in the pretreatment samples was 6800
and 6500 i*Q/L (analyzed in duplicate), and the posttreatment concentration was
<630 ng/L (estimated value). Table 20 summarizes the results of process water
samples.
The quality assurance/quality control of analyses performed on wipes, metal
strips, and water samples is discussed in Section 5. Copies of the bench-scale and
pilot-scale analytical data provided by the analytical laboratories are included in
Volume II of this report.
Because the concentration of Dicamba in the posttreatment water could not be
accurately measured, the treated water stored in the polyethylene holding tank was
pumped into an onsite water-treatment system for further treatment before its dis-
charge into a nearby creek.
Finally, the equipment was decontaminated with a high-pressure? wash. The wash
water generated during this decontamination was also collected and pumped into the
onsite water-treatment system. The system and the enclosure were disassembled and
loaded into the semitrailer for transport back to Cincinnati, Ohio.
50
-------
TABLE 20. ANALYTICAL RESULTS FOR PROCESS WATER GENERATED
DURING SHAVER'S FARM SITE DEMONSTRATION
Sample ID Number
Pre treatment
Det-Pre-8/3
Ows-Pre-8/3a
Det-Pre-8/3
Ows-Pre-8/3a
Det-Pre-8/3
Posttreatment
Post-8/3
Post-8/3
Post-8/3
Analyses
Semi-Volatiles ftig/L)
Benzonitrile
2,4-Dichlorophenol
2,6-Dichlorophenol
1 ,2,4-Trichlorobenzene
Herbicides (jug/1)
Dicamba
2,4-D
2.4,5-T
Dioxins/Furans (nq/L)
HpCDD
HpCDF
HxCDD
HxCDF
OCDD
OCDF
PeCDD
PeCDF
5 TCDD
2,3,7,8-TCDD
TCDF
Pretreatment
250/4003
ND (4.8)
ND (4.8)
ND (4.8)
6800/65003
NDb(1100)
ND (190)
NDb(10)
ND(7.4)
ND (8.4).
ND (5.0)
ND(23)
ND(10)
ND (4.8)
ND (3.0)
ND (4.8)
ND (4.8)
ND (2.9)
Posttreatment
NDb (4.8)c
ND(4.8)
ND(4.8)
ND(4.8)
630d(260)
ND(1100)
ND(190)
ND (4.8)
ND(3.0)
14d(3.6)
ND (2.0)
ND(11)
ND(5.7)
ND(2.1)
ND(1.3)
ND(2.1)
ND(2.1)
ND(1.2)
a Duplicate analyses.
b Not detected at specified detection limit.
c Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
d Estimated result less than 5 times detection limit.
-------
SECTION 5
QUALITY ASSURANCE/QUALITY CONTROL ANALYSES
The objective of this study was to design, develop, and evaluates a pilot-scale
debris decontamination system. The data generated consisted of the results of chemi-
cal analyses performed during the project. All data and observations were recorded in
permanent laboratory notebooks. This section presents the quality assurance/quality
control (QA/QC) analyses performed on surface-wipe, metal, and water samples taken
during the field demonstrations of the DWS.
5.1 Demonstration of DWS at Gray PCB Site
During the field demonstration of the DWS at the Gray PCB site, surface-wipe
samples of pieces of metallic debris (transformer casings) were taken to ascertain the
extent of PCB contamination on the surfaces of the debris. The contamination of a
piece of debris may vary at different points on its surface, and pieces of debris used in
the demonstration probably will exhibit varying degrees of contamination. Many fac-
tors can contribute to these varying degrees of contamination on a particular piece of
debris, such as the concentration of contaminants in the soil, the contact time (if the
piece was buried), the extent of surface oxidation, the porosity of the material, and
exposure to the elements. As a way of minimizing these differences, pretreatment and
posttreatment samples were taken directly adjacent to each other, but not at the same
location. Because the initial wiping of the surface of a sample removes the contamina-
tion, if one were to wipe the same surface after cleaning, the result obtained would be
biased low. Results were reported as the relative percent difference (RPD) between
the pretreatment and posttreatment levels of total PCBs. The PCB concentration in
the cleaning solution and rinse water after filtration/sorbent treatment indicated the ulti-
mate method at disposal.
5.2 Sampling and Analysis
5.2.1 Wipe Sampling Theory
Under the Toxic Substance Control Act (TSCA), the EPA has established the PCB
Spill Cleanup Policy for releases of materials containing PCBs at concentrations of 50
parts per million (ppm) or greater (2). The Policy states that high-contact solid
52
-------
surfaces should be cleaned to a level of 10 /ig/100 cm2. The method of determining
contamination on hard surfaces such as metal, wood, concrete, plastic, and glass is to
analyze surface-wipe samples (3).
Procedures for standard wipe tests vary, but generally include wiping a specific
area with an absorbent tissue or cotton gauze swatch that has been wetted with a
solvent (hexane). The EPA policy recommends that a standard-size template (10 cm
by 10 cm) be used to delineate the area of cleaning. The gauze pad or glass wool of
known size should be saturated with hexane. The hexane-prepared wipe is stored in a
sealed glass vial both before and after sample collection. The before and after
surface-wipe samples will indicate the efficiency of the cleaning process on the basis
of the percentage of PCBs removed from the debris surfaces. The EPA also requires
the collection and testing of field blanks.
The PCB concentrations in surface-wipe samples can be determined by first
extracting the wipe samples in accordance with Method 3550 and then quantitating
them in accordance with Method 8080, which involves injecting the sample extract into
a gas chromatograph (GC) set at the operating conditions specified in the Method and
equipped with an electron-capture detector (4). The result is reported in micrograms
PCB per wipe or micrograms per 100 cm2.
A QA plan for wipe samples is outlined in an EPA report (3): This plan includes
chain of custody, field blanks, check samples, replicate samples, spiked samples, etc.
Analyzing surface-wipe samples provides a systematic method of determining the
extent of PCB contamination or the effectiveness of the cleanup operation. The
method produces varying results, depending on the sample collector, the porosity of
the surfaces, and the details of the sampling procedure. This method, however, is
relatively quick and inexpensive, and it is easily adapted to statistical sampling. It cur-
rently affords the most reasonable way to sample surfaces for PCB contamination.
5.2.2 Surface-Wipe Sampling Procedures
The surface-wipe samples were obtained in accordance with procedures outlined
in the Field Manual for Grid Sampling of PCB Sites to Verify Cleanup (1). The fol-
lowing procedure was used to take surface-wipe samples from pieces of metallic de-
bris.
A wide-mouth jar was first filled with 3-in. by 3-in. cotton gauze pads and then
filled with hexane, which effectively saturated the pads. With gloved hands, the sam-
pler obtained the sample by thoroughly wiping a 100-cm2 area (delineated by a tem-
plate) with the moistened gauze pad, moving it from left to right and then from top to
bottom of the metallic debris surface. The sampler then folded the gauze pad with the
sampled side inward and placed it in the sample bottle, which was then capped, la-
beled, affixed with a yellow TSCA PCB mark, and placed in an ice chest (to keep the
53
-------
sample at 4°C). The sample collection data were entered into the field log book and
on the Chain-of-Custody Form. The template was then thoroughly rinsed with hexane
and wiped with a disposable wiping cloth. The rubber gloves worn by the sampler
when taking the wipe samples and the wiping cloth were discarded in a plastic bag for
disposal as PCB-contaminated materials. One, field blank and two additional wipes
[one for the matrix spike (MS) and one for the matrix spike duplicate (MSD)] were also
placed in sample containers at this time.
5.2.3 Process Water Sampling
Upon completion of the treatment of all transformers at the site, the contaminated
process water was treated in the treatment system. Prior to the treatment, two
baseline samples were collected from each tank, composited, and analyzed for PCBs
and selected metals. Subsequently, the contaminated water was treated, and two
samples (total of 1 gallon each) of treated water was collected during the process.
One of these two posttreatment samples was collected earlier during the treatment
process at 15-minute intervals and composited until about the first half of the total
process water was treated. The second posttreatment water sample was then collect-
ed of the remaining process water in the similar fashion. Both of these samples were
analyzed for PCBs and selected metals.
5.2.4 Sample Containers
The surface-wipe samples were collected in a 4-oz clean glass jar vyith a Teflon-
lined lid. The process water samples were collected in a 1-gallon amber glass bottle
equipped with a Teflon-lined cap. Plastic containers or lids were not used in the stor-
age of samples because the samples could become contaminated by the phthalate
esters and other hydrocarbons within the plastic.
5.3 Summary of QA/QC Procedures Used by Hayden Environmental Group
The Quality Assurance Objectives listed in the QAPjP for the project are shown in
Table 21. The QA data corresponding to each of the objectives shown in Table 21 are
discussed in the following subsections.
5.3.1 Potychlorinated Biphenyls (PCBs)
The analytical method used to quantitate PCBs in surface-wipe samples was
taken from the third edition of EPA's Test Methods for Evaluating Solid Waste (4). The
procedures outlined in SW-846 Method 3550 were used to extract PCBs from surface
wipes, and EPA Method 8080 was used to quantitate the PCBs in these! extracts.
54
-------
TABLE 21. QA OBJECTIVES FOR PRECISON, ACCURACY, COMPLETENESS, AND
METHOD DETECTION LIMIT (GRAY PCB SITE DEMONSTRATION)
Critical
Measurement
Sonication extraction
PCB concentration
Liquid-liquid
extraction
Oil and grease
Matrix Type
Surface wipes
Surface
wipes, water
Process
water
Surface wipes
Method
Reference
SW-846
Method 3550
SW-846
Method 8080
SW-846
Method 3510
SW-846
Method 9070
Measured Units
Wipes :|ig/100cm2
Water: u,g/ml
Milligrams
MDL
5 u.g/1 OOcm^
5 u,g/ml
0.2 mg
Precision3
±30
±30
±20
Accuracy'5
30-130
30-130
75-125
Completeness,
%
90
80
80
90
80
Reference
4
4
4
4
4
Ol
Ol
3 As Relative Percent Difference (RPD) of matrix spike duplicates
D As Percent Recovery Range of laboratory matrix spikes unless otherwise indicated
-------
5.3.1.1 Analyte Calibration
A five-point calibration was performed by using standard concentrations from 0.25
to 0.75 ng/AtL to generate response factors. These calibration-check standards were
reanalyzed after every 20 samples to verify that the response factor remained within
±15 percent of that generated from an average of the five-point standards.
5.3.1.2 Accuracy
The target accuracy range was 30 to 130 percent, as determined by the per-
centage of recovery of laboratory matrix spikes. The QAPjP states that matrix-spiked
samples are to be analyzed at a frequency of one for every 20 samples of each matrix
type analyzed. The 'surface-wipe samples were spiked with Aroclor 1242 in duplicate.
Summaries of spike recoveries for surface-wipe samples are shown in Table A-1 in
Appendix A. The mean percent recovery falls within the targeted limits of 30 to 130
percent.
5.3.1.3 Precision
The precision of the analytical method was evaluated by calculating the RPD for
the percentage recoveries of MS and MSD samples. The calculated RPD values
ranged from 3 to 29 percent. All values were within the 30 percent RPD required. The
precision results are summarized in Table A-1 in Appendix A.
5.3.1.4 Blanks
The field blanks from each batch were extracted and analyzed by Method 8080 in
the same manner as the other surface-wipe samples to check for background con-
tamination. The results of all the field blanks were below the detection limit of
<0.1 M9/100 cm2. Method blanks were also analyzed to check for any contamination.
During the analysis of the method blank, a purified solid matrix is carried through the
entire analytical scheme (extraction, concentration, and analysis). No contamination
was observed in any of the*method blanks. A summary of field and method blanks
results is presented in Table A-2 in Appendix A.
5.3.2 Metals
The analytical methods used to quantitate metals (copper, lead, cadmium, chro-
mium, and arsenic) in process water samples were taken from the third edition of
EPA's Test Methods for Evaluating Solid Waste (4) and are summarized in Table 22.
56
-------
TABLE 22. METHODS USED TO QUANTITATE SELECTED METHODS
(GRAY PCB SITE DEMONSTRATION)
Measure-
ment
Copper
Lead
Cadmium
Chromium
Arsenic
Matrix type
Water
Water
Water
Water
Water
Method reference Measured units
SW-846-7210 jzg/L
Method 220.1
SW-846-7420 HQ/L
Method 239.1
SW-846-7130 /zg/L
Method 213.1
SW-846-7190 nQ/L
Method 218.1
SW-846-7060 M9/L
Method 206.2
5.4 Demonstration of DWS at Shaver's Farm Site
During the field demonstration of the DWS at the Shaver's Farm site, surface-wipe
samples were taken from pieces of metallic debris (55-gal metallic drums) to ascertain
the extent of benzonitrile and Dicamba contamination on the surfaces of the debris.
Subsequent to cleaning the debris in the DWS, surface-wipe samples were again taken
and the efficacy of the cleaning process was judged on the basis of whether the
benzonitrile and Dicamba on the debris surfaces had been reduced to the "acceptable
level" of 5 /Ltg/100 cm2 in surface wipes and 5 /ng/g in metal strips.
In addition to surface-wipe sampling, pieces of the metallic strips were also
obtained before and after treatment and extracted to determine the amount of Dicam-
ba, benzonitrile, and (in two cases) PCDD and PCDF present in metallic strips.
The concentrations of Dicamba and benzonitrile in the cleaning solution and rinse
water after filtration/sorbent treatment will indicate the ultimate method of disposal.
5.5 Sampling and Analysis
5.5.1 Wipe Sampling Theory
The basic theory of surface-wipe sampling was explained earlier in Subsection
5.1.1.1. Dicamba and benzonitrile on the surface of metallic debris were sampled by
the surface-wipe technique. The concentrations of these contaminants were then
57
-------
quantitated by first extracting the wipe samples by Method 3540 (4). During the Sox-
hlet extraction procedure, the entire gauze wipe was put into the extraction solvent.
The Dicamba concentration was then quantitated by Method 8150, which involves in-
jecting an aliquot of the gauze extract into a GC set at the operating conditions speci-
fied in the method and equipped with an electron-capture detector (4). The benzoni-
trile concentration was quantitated in accordance with Method 8270, which involves
injecting an aliquot of the gauze extract into a gas chromatograph/rnass spectrometer
(GC/MS) set at the specified operating conditions (4). The analytical results are re-
ported in micrograms of contaminant per wipe or micrograms of contaminant per
100cm2. :
5.5.2 Wipe Sampling Procedure
The following procedure was used to collect surface-wipe samples from pieces of
debris. A wide-mouth jar was filled with a 1:1 mixture of pesticide-grade hexane and
acetone for effective saturation of the pads. This jar was kept tightly sea|ed until the
samples were taken. During sampling, individual gauze pads were removed with stain-
less steel forceps and handed to the sampler, who compressed each pad with his
gloved hand to remove excess solvent. The solvent-soaked gauze pad was then used
to thoroughly wipe a 100-cm2 area of the metallic debris surface (delineated by a tem-
plate), from left to right and then from the top to bottom to obtain the sample. The
sampler then folded the gauze pad with the sampled side inward and placed it in a
4-oz glass jar, which was then capped, labeled, and placed in an ice chest (to keep
the sample at 4°C). The sample collection data were entered into the field log book
and on the chain-of-custody form. The template was thoroughly rinsed with solvent
and wiped with a disposable wiping cloth. The sampler's rubber gloves and the wip-
ing cloth were then discarded in a plastic bag used for disposal of hazardous-waste-
contaminated materials.
When debris was covered with scale or caked-on deposits, the pretreatment sam-
pling was performed on an area that was free of surface deposits. Multiple surface
wipes were obtained to ensure that all material in the wipe area was removed from the
surface and transferred to the sample container.
5.5.3 Sampling of Metal Matrix
To corroborate the results of the surface-wipe tests and to demonstrate whether
contaminants were imbedded in the surfaces of metallic debris, the sampler obtained
one pretreatment and one posttreatment sample from each batch in the following man-
ner. A partner saw was used to remove a 10-in. by 10-in. piece of metal from the de-
bris item (e.g., a 55-gallon drum that had been removed from a burial trench) and
from the approximate center of the 10-in. by 10-in. piece (to avoid areas where heating
resulting from the sawing process may have dislodged the contaminants); approxi-
mately 100 g of metal strips were removed by power shears (no heat generation).
58
-------
These metal strips were immediately placed in a tightly capped jar, which was placed
in a 4°C sample cooler for shipment.
5.5.4 Process Water Sampling
Upon completion of the treatment of all batches of drums at the site, the con-
taminated process water was treated in the treatment system. Prior to the treatment,
two baseline samples were collected from each tank, composited, and analyzed for
Dicamba, benzonitrile, dioxins, and furans. During the subsequent treatment of the
contaminated water, two 1-gal samples of treated water were collected. One of these
two posttreatment samples was collected at 15-min intervals early in the treatment
process and composited. This composite sample represented roughly the first half of
the process water being treated. The second posttreatment water sample was
collected in a similar manner and represented the remaining half of the process water.
Both of these samples were analyzed for Dicamba, benzonitrile, dioxins, and furans.
5.5.5 Sample Containers
The surface-wipe samples and metal strip samples were collected in a 4-oz clean
glass jar with a Teflon-lined lid. The process water samples were collected in a 1-
gallon amber glass bottle with a Teflon-lined cap. Plastic containers or lids were not
used to store samples because the samples could become contaminated by the
phthalate esters and other hydrocarbons within the plastic.
5.6 Summary of QA/QC Procedures Used by Radian Corporation
The analytical method used to quantitate organics in surface-wipe, metal strip,
and process water samples was taken from the third edition of EPA's Test Methods
for Evaluating Solid Waste (4). Radian Corporation used the procedures outlined in
SW-846 Methods 3540 and 3510 to extract organics from surface wipes/metal strips
and process water, respectively. Method 3540 is a Soxhlet extraction process, where-
as Method 3510 is a liquid-to-liquid extraction process. Methods 8150, 8270, and
8280 from this same document were used to quantitate herbicides, semivolatiles, and
PCDD/PCDF, respectively, from extracts of surface wipes/metal strips and process
water.
The QA objectives listed in the QAPjP are presented in Table 23. When the
QAPjP for this project was written, the critical measurements were Dicamba and ben-
zonitrile. During the field demonstration of DWS at the Shaver's Farm site, EPA Head-
quarters in Washington, D.C., requested that analyses also be made for 2,4-dichloro-
phenol, 2,6-dichlorophenol, 1,2,4-trichlorobenzene, 2,4-D, and 2,4,5-T. Because it was
too late in the project to include these compounds in the QAPjP, Table 23 shows only
Dicamba and benzonitrile as the critical measurements. The QA data corresponding
59
-------
TABLE 23 QA OBJECTIVES FOR PRECISION, ACCURACY.COMPLETENESS, AND
METHOD DETECTION LIMIT (SHAVER'S FARM SITE DEMONSTRATION)
Critical
Measurement
Dicamba
Benzon'rtrile
Matrix Type
Surface
wipes, metal
strips
Surface
wipes, metal
strips
Process
water
Method
Reference
Extraction:
SW-846
Method 3540
Analysis:
SW-846
Method 8150
Extraction:
SW-846
Method 3540
Analysis:
SW-846
Method 8270
Extraction:
SW-846
Method 3510
Analysis:
SW-846
Method 81 50
Measured Units
Wipes: ng/100cm2
Strips: ug/g
Wipes: ng/100crc£
Strips: ug/g
Water: ug/L
MDL
5u.g/100cm2
5ug/g
5u.g/100cm2
5ng/g
WL
Preclslon3
±50
±50
±50
±50
±30
Accuracy*3
30-130
30-130
30-130
30-130
30-130
Completeness,
%
80
80
80
80
90
Reference
4
4
4
4
4
a As Relative Percent Difference (RPD) of matrix spike duplicates
b As Percent Recovery Range of laboratory matrix spikes unless otherwise indicated
-------
to each of the objectives shown in Table 23 are discussed in the following
subsections.
5.6.1 Semivolatiles
All semivolatile (benzonitrile, 2,4-dichlorophenol, 2,6-dichforophenol, and 1,2,4-
trichlorobenzene) extracts were analyzed using the procedures outlined in Method
8270 Of SW-846 (4).
5.6.1.1 Analyte Calibration
A five-point calibration was performed with Best Demonstrated Available
Technology (BOAT) standards at concentrations of 5, 10, 50, 100, and 200 Mg/Lfor
benzonitrile analysis and 20, 50, 80, 120, and 160 ng/Lfor 2,4-dichlorophenol, 2,6-
dichlorophenol, and 1,2,4-trichlorobenzene. Specific ion-response factors for the
calibration check compounds were verified to have less than 30 percent Relative
Standard Deviation (RSD) over the range calibrated. These calibration check com-
pounds were reanalyzed every 12 hours to verify that the response factor remained
within ±30 percent of that generated from an average of the five-point standard.
5.6.12 Accuracy
Accuracy was to be within the range of 30 to 130 percent, as determined by the
percentage recovery of the surrogates spiked into each sample. Surrogate recoveries
for pre- and posttreatment wipe samples are summarized in Tables B-1 and B-2 re-
spectively in Appendix B. The mean percentage recovery of all surrogates falls within
the limits of 30 to 130 percent, as specified in the QAPjP.
Surrogate recoveries for pre- and posttreatment metal strip samples are sum-
marized in Tables B-3 and B-4 respectively in Appendix B. The mean percentage
recovery of all surrogates falls within the limits of 30 to 130 percent specified in the
QAPjP.
. c.
5.6.1.3 Precision
Precision of the analytical method was evaluated by calculating the RPD for the
percentage recoveries of MS and MSD samples. Two sets of samples (MS and MSD)
were submitted to the laboratory for precision testing. Sample 1 was spiked with ben-
zonitrile, and Sample 2 was spiked with surrogate compounds. The average of the
RPD obtained for surface-wipes was well below the limit stipulated in the QAPjP (i.e.,
50 percent). The results for the precision for surface-wipes are summarized in Table
B-5 in Appendix B. The recovery of MS and MSD samples for metals could not be
calculated because the concentration in the native sample was very high.
61
-------
The matrix-spiked duplicates prepared for the purpose of assessing method
precision were analyzed at a frequency of 1 for every 20 samples analyzed for each
sample type.
5.6.1.4 Blanks
Three field blanks (wipes) were extracted and analyzed by Method 8270 in the
same manner as the other surface-wipe samples to check for background contamina-
tion. No target analytes were detected in any of the field blanks. In addition, reagent
blanks were analyzed to assess the purity of the reagents used for the analyses. No
contamination was found in any of the reagent blanks. Test results of the blanks for
surface-wipes and metal strips are shown in Tables B-6 and B-7 respectively in
Appendix B.
5.6.2 Herbicides
All herbicide extracts (Dicamba, 2,4-D, and 2,4,5-T) were analyzed in accordance
with the procedures outlined in Method 8150 of SW-846 (4).
5.6.2.1 Analyte Calibration
A six-point calibration was performed with standards at concentrations of 0.25,
0.5, 1, 2.5, 5, and 10 ppb for generating response factors. This also indicated an RSD
of less than 15 percent for all calibration check compounds. These calibration check
compounds were reanalyzed after every 10 samples to verify that the response factor
remained within ±15 percent of that generated from an average of the six-point stan-
dard.
5.6.2.2 Accuracy
Each sample was spiked with a surrogate (2,4-drchlorophenylacetic acid) before
analysis to monitor extraction efficiency. Surrogate recoveries for pre- and posttreat-
ment surface-wipe samples are summarized in Tables B-8 and B-9 respectively and
those for metal samples are shown in Tables B-10 and B-11, respectively, all in
Appendix B. The mean percentage recoveries of all surrogates were within acceptable
QA/QC limits.
5.6.2.3 Precision
The RPD was calculated for precision of the analytical method. The average of
the RPDs obtained for surface-wipes and metal strips was well below the 50 percent
limit specified in the QAPjP. The precision results for surface-wipes and metal strips
are summarized in Tables B-12 and B-13 respectively in Appendix B.
62
-------
5.6.2.4 Blanks
The field and reagent blanks for surface-wipes and metal strips were analyzed to
check for background contamination. No target analytes were detected in any of the
blanks. Analytical results of the blanks for surface-wipes and metal strips are pre-
sented in Tables B-14 and B-15 respectively in Appendix B.
5.6.3 Dioxins/Furans
All dioxin and furan extracts were analyzed in accordance with the procedures
outlined in Method 8280 of SW-846 (4).
5.6.3.1 Analyte Calibration
A five-point calibration of 0.2-, 0.5-, 1-, 2-, and 5-ppb standards was used to gen-
erate response factors. The calibration check compounds were reanalyzed eve.ry
12 hours to verify that the response factor remained within ±30 percent of that gener-
ated from an average of the five-point standard.
5.6.3.2 Accuracy
All samples were spiked with Carbon-13-labeled surrogate standards. Percentage
recoveries for each of the Carbon-13-labeled surrogates are shown in Table B-16 in
Appendix B. The mean percentage recovery of all surrogates ranges from 82 to 118.
5.6.3.3 Blanks
The reagent blanks for metal strips were analyzed to check for background
contamination. No target analytes were detected in any of the blanks. Results of
reagent blanks are shown in Table B-17 of Appendix B.
63
-------
SECTION 6
COST OF DEMONSTRATIONS
This section summarizes the capital equipment cost of pilot-scale DWS and the
cost of conducting the EPA SITE demonstrations of DWS at the Gray PCB site in
Hopkinsville, KY, and at the Shaver's Farm site near Chickamauga, GA.
6.1 Capital Equipment Cost
The cost for the design, engineering, equipment procurement, fabrication, and
installation of the pilot-scale DWS was approximately $75,000. The cost includes all
the subsystems and components installed and also the initial shakedown of the
system.
6.2 Cost of Demonstration
The pilot-scale DWS was demonstrated at the Gray PCB site and at the Shaver's
Farm site. The costs (rounded to the nearest $100) for each field demonstration are
summarized in Table 24. The costs of the demonstration (as shown in Table 24) may
not be representative of any actual site operation because the pilot-scale DWS
represents an experimental system which is highly labor-intensive with a relatively low
processing rate. However, the operation cost could be greatly reduced when the
economy of a semiautomatic, large-scale system is considered. The field activities
conducted during both demonstrations included:
Site preparation
Mobilization
Equipment setup
Operations/test runs
Sample collections
Chemical analyses
Demobilization
64
-------
TABLE 24. SUMMARY OF PILOT-SCALE DWS DEMONSTRATION COSTS
O)
Ol
Category
Labor (including per diem)
Equipment and supplies
Equipment transportation
Equipment rental
Travel expenses for crew
Subcontractor
Chemical analysis
Total
Gray PCB Site Demonstration
Cost, $
47,000
16,000
1,500
15,000
4,000
19,500
19,000
122,000
Cost, %
39
13
1
12
3
16
16
100
Shaver's Farm Site
Demonstration
Cost, $
35,000
12,000
2,000
8,000
10,000
12,000
61,000
140,000
Cost, %
25
9
1
6
7
9
43
100
-------
6.2.1 Cost of Demonstration at Gray PCS Site
The cost incurred during the demonstration of the pilot-scale DWS at the Gray
PCB site was approximately $122,000. All the field activties lasted for 33 days. The
crew included one project engineer and two technicians working an average of 10
hours per day during the course of the entire demonstration.
The rental equipment included a forklift, generator, air compressor, and 48-ft
semitrailer. The subcontractor costs included a concrete pad, temporary enclosure,
and site security.
6.2.2 Cost of Demonstration at Shaver's Farm Site
The cost incurred during the second demonstration of the pilot-scale DWS at
the Shaver's Farm site was approximately $140,000. All the field activities at this site
lasted for 17 days. The crew included one project engineer and three technicians
working an average of 10 hours per day during the course of the entire demonstration.
The rental equipment included a forklift, generator, air compressor, and 48-ft
semitrailer. The subcontractor costs included a concrete pad and temporary
enclosure.
66
-------
SECTION 7
CONCLUSIONS AND RECOMMENDATIONS
7.1 Conclusions
On the basis of the bench-scale and pilot-scale demonstrations performed in
Phases 1 and 2, the following conclusions have been drawn:
1) The DWS and ancillary equipment proved to be portable and rugged.
2) Overall performance of the DWS was exceptional. Only minor, readily
resolvable difficulties and problems were encountered, as is the case
with any system startup.
3) The DWS was successfully employed to remove RGBs from the surfaces
of transformer casings found at a hazardous waste site.
4) The DWS was also effective in removing herbicides, pesticides, dioxins,
and furan residues from the surfaces of contaminated drums found at a
hazardous waste site.
5) Significant quantities of metals were removed during the pilot-scale test-
ing, which suggests that the system may also be applicable for treating
debris contaminated with metals.
6) The cleaning solution was successfully recovered and reconditioned for
reuse concurrently with the actual debris-cleaning process, which mini-
mizes the quantity of process water required to clean the debris.
7) The two demonstrations of DWS were carried out during the months of
December and June, when ambient temperatures ranged from -30° to
50 °F and from 75° to 105°F, respectively. The effectiveness of the sys-
tem's operation during these extreme temperatures was indicative of its
durability and versatility.
67
-------
8) The water treatment system was effective in reducing the concentrations
of all the organic and metal contaminants (except arsenic and Dicamba)
to below the detection limit.
7.2 Recommendations
The following recommendations have been made concerning the field-scale
DWS;
1) Additional bench- or pilot-scale studies should be performed to improve
the efficiency of water treatment systems.
2) The labor intensity of the system should be reduced by modifying it with
materials-handling equipment and automation.
3) Alternative equipment (preferably fully automated) should be identified to
cut the large and bulky pieces of debris before cleaning in the DWS.
4) Design of a full-scale, fully portable, self-contained DWS should be initiat-
ed for the removal of various contaminants from debris found on hazard-
ous waste sites.
5) The utility of the DWS in "treatment trains" in which two or more process-
es are used for sequential treatment of contaminated wastes should be
investigated.
68
-------
SECTION 8
FULL-SCALE DEBRIS WASHING SYSTEM: CONCEPTUAL DESIGN
This section describes the conceptual design of a full-scale DWS, which evolved
from the bench- and pilot-scale work. The lessons learned from these two stages are
incorporated into the design, and the elements that worked well have been retained. It
may be possible to modify an existing commercial system; however, the design de-
scribed here is conceptual design.
Figure 9 is a schematic block diagram of the full-scale DWS. The debris will be
loaded in a basket, lifted by a monorail crane, and lowered into the wash tank. The
wash tank will be sealed and filled with hot detergent solution from the detergent tank.
The debris will be washed by the solution recycled by the wash pump. A small bleed
stream will be sent to the water treatment system
At the end of the wash cycle, the basket will be transferred to the spray/rinse
tank, which may have the ability to rotate the basket or the sprays. The batch will be
sprayed with the hot detergent solution and rinsed with cold water. Again, a bleed
stream will be sent to the water treatment system.
The unit will be equipped with a control panel and a hot oil heating system.
The heating system can be an electric or oil-fired unit.
69
-------
-j
O
Contaminated
Debris
Monorail/
Crane
Spray/Rinse
Tank
Debris
Return
Pump
Oil/Water
Separator
Bleed
Pump
Filters and
Carbon Beds
T
Figure 9. Schematic diagram of a full-scale DWS.
-------
REFERENCES
1. Kelso, A.L., M.D. Erickson, and D.C. Cox. 1986. Field Manual for Grid Sam-
pling of PCB Spill Sites to Verify Cleanup. EPA-560/5-86/017. Prepared by
Midwest Research Institute and Washington Consulting Group for the U.S. Envi-
ronmental Protection Agency, Washington, D.C.
2. PCB Spill Cleanup Policy. 40 Code of Federal Regulations 761.125, Subpart G.
April 12, 1987.
3. Boomer, B. A., et al. 1985. Verification of PCB (Polychlorinited Biphenyl) Spill
Cleanup by Sampling and Analysis. EPA-560/5-85/026. Prepared by Midwest
Research Institute for the U.S. Environmental Protection Agency, Washington,
D.C.
4. U.S. Environmental Protection Agency. Test Methods for Evaluating Solid
Waste. 1986. Volume IB. Laboratory Manual Physical/Chemical Methods.
SW-846, Third Edition. Office of Solid Waste, Washington D.C.
71
-------
APPENDIX A
QA/QC DATA
GRAY PCB SITE DEMONSTRATION
72
-------
TABLE A-1 SUMMARY OF MATRIX SPIKE RECOVERIES
(METHOD 8080 - WIPE SAMPLES)
Sample ID: (MS/MSD)
9001557/9001558
9001534/9001535
9001284/9001285
9001264/9001265
9000739/9000740
9000664/9000665
9000709/9000710
9000677/9000678
9000689/9000690
Matrix Spike
(C1)
90
67
62
60
93
73
67
64
67
Matrix Spike
Duplicate (C2)
106
89
62
57
77
77
80
86
69
Precision3
16%
28%
0%
5%
19%
5%
18%
. 29%
3%
Range*5
b
b
b
b
b
b
b
b
b
a Precision estimated by calculation of relative percent difference (RPD) using the following equation:
|c,-c2|
X 100
b Range established in QAPjP as 30%
-------
TABLE A-2. FIELD AND REAGENT BLANKS (METHOD 8080 - WIPES SAMPLES)
(ng/lOQcm2)
Batch No.
1
2
3
4
5
6
7
8
9
10
11
12
13
15
16
18
19
20
21
22
23
24
Field Blank
^0*1
^0* I
^0» I
^0» 1
<0.1
^0» I
^0» I
^0» I
-------
APPENDIX B
QA/QC DATA
SHAVER'S FARM SITE DEMONSTRATION
75
-------
TABLE B-1. SUMMARY OF SURROGATE RECOVERIES OF PRETREATMENT
WIPE SAMPLES (METHOD 8270)
(PERCENT)
Surrogate
2-Fluorobiphenyl
2-Fluorophenol
Nitrobenzene-dS
PhenoPdS
Terphenyl-d14
2,4,6-Tribromophenol
Number of
Data Points
18
19
19
18
19
18
Mean Parc&nt
Recovery
71
67
65
71
100
67
Ranqe
9-114
4-102
8-94
0-99
15-184
4-130
Standard
Deviation
34
27
31
29
43
36
TABLE B-2. SUMMARY OF SURROGATE RECOVERIES OF POSTTREATMENT
WIPE SAMPLES (METHOD 8270)
(PERCENT)
Surrogate
2-Fluorobiphenyl
2-Fluorophenol
Nitrobenzene-d5
Phenol-d5
Terphenyl-d14
2,4,6-Tribromophenol
Number of
Data Points
17
17
17
17
17
17
Mean Percent
Recovery
94
73
80
74
107
78
Range
71-113
43-99
52-101
47-100
65-166
23-138
Standard
Deviation
14
16
14
16
30
22
-------
TABLE B-3. SUMMARY OF SURROGATE RECOVERIES OF PRETREATMENT
METAL STRIP SAMPLES (METHOD 8270)
(PERCENT)
Surroqate
2-Fluorbbiphenyl
2-Fluorophenol
Nitrobenzene-d5
Phenol-d5
Teiphenyl-d14
2,4,6-Tribromophenol
Number of
Data Points
9
9
9
9
. 9
9
Mean Percent
Recovery
74
58
60
78
108
55
Range
16-120
12-85
10-115
9-116
22-187
10-120
Standard
Deviation
40
23
34
33
50
34
TABLE B-4. SUMMARY OF SURROGATE RECOVERIES OF POSTTREATMENT
METAL STRIP SAMPLES (METHOD 8270)
(PERCENT)
Surrogate
2-FluorobTphenyl
2-Fluorophenol
Nitrobenzene-d5
Phenol-d5
Terphenyl-d14
2,4, 6-Tribromophenol
Number of
Data Points
9
9
9
9
9
9
Mean Percent
Recovery
100
62
81
72
11.8
62
Range
78-116
46-91
57-92
56-90
80-199
39-75
Standard
Deviation
12
17
11
12
41
10
-------
TABLE B-5. SUMMARY OF MATRIX SPIKE RECOVERIES
(METHOD 8270 WIPE SAMPLES)
Sample ID: MS &MSD-8/2
Benzonitrile
Sample ID: MS & MSD-7/20
Acenaphthene
4-Chloro-3-methylphenol
2-Chlorophenol
1,4-Dichlorobenzene
2,4-Dinitrotoluene
N-Nitrosodipropylamine
4-Nitrophenol
Pentachlorophenol
Phenol
Pyrene
1 ,2,4-Trichlorobenzene
Matrix Spike
(C1)
118
65
64
58
41
74
37
51
84
60
75
65
Matrix Spike
Duplicate (C2)
150
65
82
89
81
66
54
55
86
84
78
90
Precision3
24%
0%
25%
42%
66%
11%
37%
8%
2%
33%
4%
32%
Rangeb
50%
b
b
b
b
a Precision estimated by calculation of relative percent difference (RPD) using the following equation:
RPD =
b Range established in QAPjP as 50%
X 100
-------
TABLE B-6. FIELD AND REAGENT BLANKS (METHOD 8270 - WIPE SAMPLES)
Benzonitrile
2,4-Dichlorophenol
2,6-Dichlorophenol
1 ,2,4-Trichlorobenzene
Field Blank
(Batch #2)
NDa (5.0)b
ND (5.0)
ND (5.0)
ND (5.0)
Field Blank
(Batch #6)
ND (5.0)
ND (5.0)
ND (5.0)
ND (5.0)
Field Blank
(Batch #10)
ND (5.0)
ND (5.0)
ND (5.0)
ND (5.0)
Reagent Blank
ND (5.0)
ND (5.0)
ND (5.0)
ND (5.0)
Reagent Blank
ND (5.0)
ND (5.0)
ND (5.0)
ND (5.0)
Reagent Blank
ND (5.0)
ND (5.0)
ND (5.0)
ND (5.0)
a Not detected at specified detection limit.
k Numbers in parenthesis indicate the minimum detectable concentration of the anaiyte.
TABLE B-7. REAGENT BLANK (METHOD 8270 - METAL STRIP SAMPLES)
Benzonitrile
2,4-Dichlorophenol
2,6-Dichlorophenol
1 ,2,4-Trichlorobenzene
Reagent Blank
NDa (0.50)b
ND (0.50)
ND (0.50)
ND (0.50)
a, ^Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration
of the anaiyte.
-------
TABLE B-8. SUMMARY OF SURROGATE RECOVERIES OF PRETREATMENT
WIPE SAMPLES (METHOD 8150)
(PERCENT)
Surrogate
2,4-Dichlorophenylacetic acid
Number of
Data Points
13
Mean Percent
Recovery
99
Ranqe
33-132
Standard
Deviation
26
TABLE B-9. SUMMARY OF SURROGATE RECOVERIES OF POSTTREATMENT
- WIPE SAMPLES (METHOD 8150)
(PERCENT)
Surrogate
2,4-Dichlorophenylacetic acid
Number of
Data Points
13
Mean Percent
Recovery
104
Range
98-107
Standard
Deviation
3
00
o
TABLE B-10. SUMMARY OF SURROGATE RECOVERIES OF PRETREATMENT
METAL STRIP SAMPLES (METHOD 8150)
(PERCENT)
Surroqate
2,4-Dichlorophenylacetic acid
Number of
Data Points
7
Mean Percent
Recovery
103
Range
90-125
Standard
Deviation
15
TABLE B-11. SUMMARY OF SURROGATE RECOVERIES OF POSTTREATMENT
METAL STRIP SAMPLES (METHOD 8150)
(PERCENT)
Surrogate
2,4-Dichlorophenylacetic acid
Number of
Data Points
7
Mean Percent
Recovery
106
Range
90-122
Standard
Deviation
11
-------
O)
TABLE B-12. SUMMARY OF MATRIX SPIKE RECOVERIES
(METHOD 8150 - WIPE SAMPLES)
Sample ID: MS1-7/20
Dicamba
Sample ID: MS1-8/2
Dicamba
Matrix Spike
(C1)
90
90
Matrix Spike
Duplicate (C2)
95
82
Precision3
5%
9%
Range'3
50%
50%
3 Precision estimated by calculation of relative percent difference (RPD) using the following equation
RPD= cvTcT; x 100
2
b Range established in QAPjP as 50%
TABLE B-13. SUMMARY OF MATRIX SPIKE RECOVERIES
(METHOD 8150 - METAL STRIP SAMPLES)
Dicamba
a Precision estimated by calci
Matrix Spike
(C1)
137
lation of relative
RPD =
Matrix Spike
Duplicate (C2)
150
percent c
Ci-C2
C1 + C2
ifference (
X
Precision3
9%
Range'5
50%
RPD) using the following equation:
100
b Range established in QAPjP as 50%
-------
TABLE B-14. FIELD AND REAGENT BLANKS (METHOD 8150 - WIPE SAMPLES)
Dicamba
2,4-D
2,4,5-T
Field Blank
(Batch #4)
NDa (0.27)b
ND{1.2)
ND (0.20)
Field Blank
(Batch #8)
ND (0.27)
ND(1.2)
ND (0.20)
Field Blank
(Batch #9)
ND (0.27)
ND(1.2)
ND (0.20)
Reagent Blank
ND (0.27)
ND(1.2)
ND (0.20)
Reagent Blank
ND (0.27)
ND(1.2)
ND (0.20)
a Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
00
ro
TABLE B-15. REAGENT BLANK (METHOD 8150 METAL STRIP SAMPLES)
Dicamba
2,4-D
2,4,5-T
Reagent Blank
NDa (0.27)b
ND(1.2)
ND (0.20)
Reagent Blank
ND (0.27)
ND(1.2)
ND (0.20)
a Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration of the analyte.
-------
Surrogate
13G-1,2I3,4.6,7,8-HpCDD
13C-1,2.3>4,6.7,8-HpCDF
^c-iAWtS-HxCDD
13C-1,2,3,6.7,8-HxCDF
13C-1,2,3,7I8-PeCDD
13C-1I2,3,7,8-PeCDF
13C-2.3,7,8-TCDD
13C-2,3,7,8-TCDF
13C-OCDD
13C-OCDF
Number of
Data Points
4
4
4
4
4
4
4
4
4
"
Mean Percent
Recovery
93
82
93
102
100
98
118
QC
85
OA
84
Range
- '" . . .. .
86-101
74-91
89-99
97-1 07
97-1 03
94-102
113-124
69-107
67-104
Standard
Deviation
"
6
7
/
4
A
*T
1ft
IO
16
-------
TABLE B-17. REAGENT BLANK (METHOD 8280 - METAL STRIP SAMPLES)
(ng/g)
HpCDD
HpCDF
HxCDD
HxCDF
OCDD
OCDF
PeCDD
PeCDF
TCDD
2,3,7.8-TCDD
TCDF
Reagent Blank
NDa (0.70)b
ND (0.50)
ND (0.36)
ND (0.21)
ND (2.9)
ND(0.97)
ND(0.18)
ND(0.10)
ND(0.16)
ND(0.16)
ND (0.095)
0
I
Not detected at specified detection limit.
b Numbers in parenthesis indicate the minimum detectable concentration
of the analyte.
-------
-------
Environmental Protection
Agency
Information
Cincinnati OH 45268-1072
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
Official Business
Penalty for Private Use, $300
Please make all necessary changes on the above label,
detach or copy, and return to the address in the upper
left-hand corner
If you do not wisn to receive these reports CHECK HERE D,
detach, or copy this cover and return to the address in the
upper left-hand corner
EPA/540/5-91/006a
------- |