»EPA
           United States
           Environmental Protection
           Agency
Office of Research and
Development
Washington, DC 20460
EPA/540/R-05/002
March 2005
           Innovative Technology
           Verification Report
           Technologies for Monitoring
           and Measurement of Dioxin
           and Dioxin-like Compounds
           in Soil and Sediment

           Wako Pure Chemical Industries, Ltd
           Dioxin ELISA Kit

-------

-------
                                      EPA/540/R-05/002
                                      March 2005
       Innovative Technology
          Verification  Report

   Wako Pure Chemical  Industries, Ltd.
Dioxin ELISA Kit Wako (for environmental)
                     Prepared by

                      Battelle
                    505 King Avenue
                  Columbus, Ohio 43201

                  Contract No. 68-C-00-185
                    Stephen Billets
                Environmental Sciences Division
               National Exposure Research Laboratory
               Office of Research and Development
               U.S. Environmental Protection Agency
                   Las Vegas, NV 89119

-------
                                               Notice
This document was prepared for the U.S. Environmental Protection Agency (EPA) Superfund Innovative Technology
Evaluation Program under Contract No. 68-C-OO-l 85. The document has met the EPA's requirements for peer and
administrative review and has been approved for publication. Mention of corporation names, trade names, or
commercial products does not constitute endorsement or recommendation for use.

-------
                                               Foreword
The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the nation's natural
resources. Under the mandate of national environmental laws, the Agency strives to formulate and implement actions
leading to a compatible balance between human activities and the ability of natural systems to support and nurture
life. To meet this mandate, the EPA's Office of Research and Development (ORD) provides data and scientific
support that can be used to solve environmental problems, build the scientific knowledge base needed to manage
ecological resources wisely, understand how pollutants affect public health, and prevent or reduce environmental
risks.

The National Exposure Research Laboratory is the Agency's center for investigation of technical and management
approaches for identifying and quantifying risks to human health and the environment. Goals of the Laboratory's
research program are to (1) develop and evaluate methods and technologies for characterizing and monitoring air, soil,
and water; (2) support regulatory and policy decisions; and (3) provide the scientific support needed to ensure
effective implementation of environmental regulations and strategies.

The EPA's Superfund Innovative Technology Evaluation (SITE) Program evaluates technologies designed for
characterization and remediation of contaminated Superfund and Resource Conservation and Recovery Act (RCRA)
sites. The SITE Program was created to provide reliable cost and performance  data in order to speed the acceptance
and use of innovative remediation, characterization,  and monitoring technologies by the regulatory and user
community.

Effective monitoring and measurement technologies are needed to assess the degree of contamination at a site,
provide data that can be used to determine the risk to public health  or the environment, and monitor the success or
failure of a remediation process. One component of the EPA SITE Program, the Monitoring and Measurement
Technology (MMT) Program, demonstrates and evaluates innovative technologies to meet these needs.

Candidate technologies can originate within the federal government or the private sector. Through the SITE Program,
developers are given the opportunity to conduct a rigorous demonstration of their technologies under actual field
conditions. By completing the demonstration and distributing the results, the Agency establishes a baseline for
acceptance and use of these technologies. The MMT Program is managed by the ORD and the Environmental
Sciences Division in Las Vegas, Nevada.

                                                   Gary Foley, Ph.D.
                                                   Director
                                                   National Exposure Research Laboratory
                                                   Office of Research and Development
                                                    in

-------
                                                Abstract
A demonstration of technologies for determining the presence of dioxin and dioxin-like compounds in soil and
sediment was conducted under the U.S. Environmental Protection Agency's (EPA's) Superfund Innovative
Technology Evaluation Program in Saginaw, Michigan, at Green Point Environmental Learning Center from April 26
to May 5, 2004. This innovative technology verification report describes the objectives and the results of that
demonstration, and serves to verify the performance and cost of the Wako Pure Chemical Industries, Ltd. Dioxin
ELISA Kit Wako (for environmental). Four other technologies were evaluated as part of this demonstration, and
separate reports have been prepared for each technology. The objectives of the demonstration included evaluating the
technology's accuracy, precision, sensitivity, sample throughput, tendency for matrix effects, and cost. The test also
included an assessment of how well the technology's results compared to those generated by established laboratory
methods using high-resolution mass spectrometry (HRMS). The demonstration objectives were accomplished by
evaluating the results generated by the technology from 209 soil, sediment, and extract samples. The test samples
included performance evaluation (PE) samples (i.e., contaminant concentrations were certified or the samples were
spiked with known contaminants) and environmental samples collected from 10 different sampling locations.

The Wako Pure Chemical Industries, Ltd. Dioxin  ELISA Kit Wako (for environmental) is an immunoassay
technology that reports total dioxin/furan concentration in a sample. The sample units are in pg/g 2,3,7,8-
tetrachlorodibenzo-/>-dioxin (TCDD) equivalents  (EQ). While the kit is most reactive to 2,3,7,8-TCDD, it is
responsive to all dioxin/furans at some level. As part of the PE, the technology results were compared to toxicity
equivalent (TEQ) results generated by a reference laboratory, AXYS Analytical Services, using EPA Method 1613B,
which involves the use of HRMS. It should be noted that the results generated by this technology may not directly
correlate to HRMS TEQD/F in all cases because it is known that the congener responses and cross-reactivity of the kit
are not identical to the toxicity equivalency factors that are used to convert congener HRMS concentration values to
TEQD/F. The effect of cross-reactivities may contribute to this technology's reporting results that are biased high or
low compared to HRMS TEQD/F results. Therefore, this technology should not be viewed as producing an equivalent
measurement value to HRMS TEQD/F, but as a screening value to approximate HRMS TEQD/F concentration. It has
been suggested that correlation between the Wako and HRMS TEQ could be improved by first characterizing a site
and calibrating the Wako results to HRMS results. Subsequent analysis using the Wako kit for samples obtained from
this site may then show better correlation with the HRMS TEQ result. This approach was not evaluated during this
demonstration.

A summary of the performance of the Dioxin ELISA Kit Wako (for environmental) is as follows: The Wako results
were biased both positively and negatively relative to the certified and reference laboratory results. No statistically
significant matrix effects were observed by sample type (PE vs. environmental vs. extract), matrix type (soil vs.
sediment vs. extract), or polynuclear aromatic hydrocarbon concentration. Wako completed all 209 sample analyses in
the field within  a nine-day period. The technology's estimated method detection limit (83 to 201 pg/g 2,3,7,8-TCDD
EQ) was significantly higher than was reported by the developer (20 pg/g 2,3,7,8-TCDD EQ), but PE samples with
TEQ concentrations in the precisely appropriate range for evaluation of this technology's detection limit were not
available, so these calculated  values should be considered a rough estimate. The kit had a false positive rate of 10%
and a false negative rate of 13% around 20 pg/g TEQ. The kit had the same false positive rate around 50 pg/g (10%),
but less false negatives (8%). These data suggest that the Wako kit could be an effective screening tool for
determining sample results above and below 20 pg/g TEQ, and even more effective as  a screen for samples above and
below 50 pg/g TEQ, particularly considering that the cost to analyze the 209 demonstration samples was significantly
less than that of the reference laboratory ($150,294 vs. $213,580). All samples were analyzed on-site in 9  days (in
comparison to the reference laboratory,  which took 8 months to report all results).

                                                    iv

-------
                                              Contents
Chapter                                                                                          Page

Notice	ii
Foreword  	  iii
Abstract 	iv
Abbreviations, Acronyms, and Symbols	ix
Acknowledgments	xii

1   Introduction	  1
    1.1   Description of the SITE MMT Program  	  1
    1.2   Scope of This Demonstration	 3
         1.2.1   Organization of Demonstration	 4
         1.2.2   Sample Descriptions and Experimental Design	 4
         1.2.3   Overview of Field Demonstration	 5

2   Description of Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit	 6
    2.1   Company History  	 6
    2.2   Product History	 6
    2.3   Technology Description 	 6
    2.4   Developer Contact Information 	 9
    2.5   Product Information  	 9

3   Demonstration and Environmental Site Descriptions  	  10
    3.1   Demonstration Site Description and Selection Process	  10
    3.2   Description of Sampling Locations  	  11
         3.2.1   Soil Sampling Locations 	  11
         3.2.2   Sediment Sampling Sites	  13

4   Demonstration Approach 	  15
    4.1   Demonstration Objectives 	  15
    4.2   Toxicity Equivalents	  15
    4.3   Overview of Demonstration Samples	  17
         4.3.1   PE Samples  	  17
         4.3.2   Environmental Samples	 21
         4.3.3   Extracts  	 22
    4.4   Sample Handling	 24
    4.5   Pre-Demonstration Study	 25
    4.6   Execution of Field Demonstration  	 25
    4.7   Assessment of Primary and Secondary Objectives 	 26
         4.7.1   Primary Objective PI: Accuracy	 26
         4.7.2   Primary Objective P2: Precision	 26
         4.7.3   Primary Objective P3: Comparability	 27
         4.7.4   Primary Objective P4: Estimated Method Detection Limit 	 27
         4.7.5   Primary Objective P5: False Positive/False Negative Results	 28

-------
                                       Contents (continued)
         4.7.6   Primary Objective P6: Matrix Effects	 28
         4.7.7   Primary Objective P7: Technology Costs	 28
         4.7.8   Secondary Objective SI: Skill Level of Operator	 29
         4.7.9   Secondary Objective S2: Health and Safety Aspects 	 29
         4.7.10  Secondary Objective S3: Portability	 29
         4.7.11  Secondary Objective S4: Sample Throughput  	 29

5   Confirmatory Process 	 30
    5.1   Traditional Methods for Measurement of Dioxin and Dioxin-Like
         Compounds in Soil and Sediment  	 30
         5.1.1   High-Resolution Mass Spectrometry  	 30
         5.1.2   Low-Resolution Mass Spectrometry	 30
         5.1.3   PCB Methods	 31
         5.1.4   Reference Method Selection	 31
    5.2   Characterization of Environmental Samples 	 31
         5.2.1   Dioxins and Furans 	 31
         5.2.2   PCBs  	 32
         5.2.3   PAHs  	 32
    5.3   Reference Laboratory Selection	 32
    5.4   Reference Laboratory Sample Preparation and Analytical Methods	 33
         5.4.1   Dioxin/Furan Analysis	 33
         5.4.2   PCB Analysis	 33
         5.4.3   TEQ Calculations 	 33

6   Assessment of Reference Method Data Quality  	 35
    6.1   QA Audits	 35
    6.2   QC Results  	 36
         6.2.1   Holding Times and Storage Conditions 	 36
         6.2.2   Chain of Custody	 36
         6.2.3   Standard Concentrations	 36
         6.2.4   Initial and Continuing Calibration  	 36
         6.2.5   Column Performance and Instrument Resolution  	 37
         6.2.6   Method Blanks	 37
         6.2.7   Internal Standard Recovery	 37
         6.2.8   Laboratory Control Spikes	 37
         6.2.9   Sample Batch Duplicates  	 37
    6.3   Evaluation of Primary Objective PI: Accuracy	 37
    6.4   Evaluation of Primary Objective P2: Precision	 38
    6.5   Comparability to Characterization Data	 39
    6.6   Performance Summary	 39

7   Performance of Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit 	 42
    7.1   Evaluation of Dioxin ELISA Kit Performance	 42
         7.1.1   Evaluation of Primary Objective PI: Accuracy  	 42
         7.1.2   Evaluation of Primary Objective P2: Precision  	 43
         7.1.3   Evaluation of Primary Objective P3: Comparability 	 43
         7.1.4   Evaluation of Primary Objective P4: Estimated Method Detection Limit	 46
         7.1.5   Evaluation of Primary Objective P5: False Positive/False Negative Results	 46
                                                   VI

-------
                                       Contents (continued)
         7.1.6   Evaluation of Primary Objective P6: Matrix Effects  .................................... 47
         7.1.7   Evaluation of Primary Objective P7: Technology Costs  ................................. 48
    7.2   Observer Report: Evaluation of Secondary Objectives ......................................... 48
         7.2.1   Evaluation of Secondary Objective SI: Skill Level of Operator ........................... 49
         7.2.2   Evaluation of Secondary Objective S2: Health and Safety Aspects ......................... 49
         7.2.3   Evaluation of Secondary Objective S3: Portability  ..................................... 50
         7.2.4   Evaluation of Secondary Objective S4: Throughput  .................................... 50
         7.2.5   Miscellaneous Observer Notes  [[[ 51

8   Economic Analysis [[[ 52
    8.1   Issues and Assumptions  [[[ 52
         8.1.1   Capital Equipment Cost [[[ 52
         8.1.2   Cost of Supplies [[[ 52
         8.1.3   Support Equipment Cost [[[ 52
         8.1.4   Labor Cost  [[[ 53
         8.1.5   Investigation-Derived Waste Disposal Cost ........................................... 53
         8.1.6   Costs Not Included  [[[ 53
    8.2   Dioxin ELISA Kit Costs [[[ 54
         8.2.1   Capital Equipment Cost [[[ 54
         8.2.2   Cost of Supplies [[[ 54
         8.2.3   Support Equipment Cost [[[ 54
         8.2.4   Labor Cost  [[[ 56
         8.2.5   Investigation-Derived Waste Disposal Cost ........................................... 56
         8.2.6   Summary of Dioxin ELISA Kit Costs  ............................................... 56

-------
                                       Contents (continued)
                                                Figures

1-1    Representative dioxin, furan, and polychlorinated biphenyl structure	 3
2-1    Dioxin ELISA Kit Wako (for environmental) extraction procedure	 7
2-2    Dioxin ELISA Kit Wako (for environmental) assay procedure  	 9
2-3    Dioxin ELISA Kit Wako (for environmental) in operation during the field demonstration  	 9
6-1    Comparison of reference laboratory and characterization D/F data for environmental samples	 41


                                                 Tables

2-1    Dioxin ELISA Kit Wako (for environmental) Cross-Reactivity	 8
3-1    Summary of Environmental Sampling Locations 	 12
4-1    World Health Organization Toxicity Equivalency Factor Values	 16
4-2    Distribution of Samples for the Evaluation of Performance Objectives	 18
4-3    Number and Type of Samples Analyzed in the Demonstration  	 18
4-4    Summary of Performance Evaluation Samples	 19
4-5    Characterization and Homogenization Analysis Results for Environmental Samples	 23
4-6    Distribution of Extract Samples	 24
5-1    Calibration Range of HRMS Dioxin/Furan Method	 30
5-2    Calibration Range of LRMS Dioxin/Furan Method 	 30
6-1    Objective PI Accuracy - Percent Recovery	 38
6-2    Evaluation of Interferences 	 39
6-3a   Objective P2 Precision - Relative Standard Deviation	 40
6-3b   Objective P2 Precision - Relative Standard Deviation (By Sample Type)	 41
6-4    Reference Method Performance Summary - Primary Objectives	 41
7-1    Objective PI Accuracy - Percent Recovery	 42
7-2a   Objective P2 Precision - Relative Standard Deviation (All Samples) 	 44
7-2b   Objective P2 Precision - Relative Standard Deviation (By Sample Type)	 45
7-3    Objective P3 - Comparability Summary Statistics of RPD	 45
7-4    Objective P3 - Comparability Using Interval Assessment 	 45
7-5    Objective P3 - Comparability for Blank Samples	 46
7-6    Objective P4 - Estimated Method Detection Limit	 47
7-7    Objective P5 - False Positive/False Negative Results	 47
7-8    Objective P6 - Matrix Effects Using RSD as a Description of Precision by Matrix Type	 48
7-9    Objective P6 - Matrix Effects Using RSD as a Description of Precision by PAH
       Concentration Levels (Environmental Samples Only)  	 48
7-10   Objective P6 - Matrix Effects Using PE Materials	 49
8-1    Cost Summary 	 55
8-2    Reference Method Cost Summary	 57
9-1    Wako Pure  Chemical Industries, Ltd. Dioxin ELISA Kit Wako (for environmental)
       Performance Summary - Primary Objectives  	 59
9-2    Wako Pure  Chemical Industries, Ltd. Dioxin ELISA Kit Wako (for environmental)
       Performance Summary - Secondary Objectives 	 60
                                                  Vlll

-------
                          Abbreviations, Acronyms, and Symbols
Ah
ANOVA
ASE
ATSDR
CIL
CoA
COC
CRM
DER
D/F
DMSO
DNA
DNR
D/QAPP
ELC
ELISA
EMDL
EMPC
EPA
ERA
EQ
FDSC
g
GC
HPLC/GPC
HRGC
HRMS
i.d.
IDW
ITVR
kg
L
LRMS
aryl hydrocarbon
analysis of variance
accelerated solvent extraction
Agency for Toxic Substances and Disease Registry
Cambridge Isotope Laboratories
Certificate of Analysis
chain of custody
certified reference material
data evaluation report
dioxin/furan
dimethyl sulfoxide
deoxyribonucleic acid
Department of Natural Resources
demonstration and quality assurance project plan
Environmental Learning Center
enzyme-linked immunosorbent assay
estimated method detection limit
estimated maximum possible concentration
Environmental Protection Agency
Environmental Resource Associates
equivalent
Food and  Drug Safety Center
gram
gas chromatography
high-performance liquid chromatography/gel permeation chromatography
high-resolution capillary gas chromatography
high-resolution mass spectrometry
internal diameter
investigation-derived waste
innovative technology verification report
kilogram
liter
low-resolution mass spectrometry
                                                IX

-------
                   Abbreviations, Acronyms, and Symbols (Continued)
m
MDEQ
MDL
mg
mL
MMT
MS
NERL
ng
NIST
NOAA
ORD
PAH
PC
PCB
PCDD/F
PCP
PE
Pg
POD-conjugate
ppb
ppm
ppt
psi
QA/QC
RM
RPD
RSD
SDL
SIM
SITE
SOP
SRM
TCDD
TEF
TEQ
micrometer
meter
Michigan Department of Environmental Quality
method detection limit
milligram
milliliter
Monitoring and Measurement Technology
mass spectrometry
National Exposure Research Laboratory
nanogram
National Institute for Standards and Technology
National Oceanic and Atmospheric Administration
Office of Research and Development
polynuclear aromatic hydrocarbons
positive  control
polychlorinated biphenyl
polychlorinated dibenzo-p-dioxin/dibenzofuran
pentachlorophenol
performance evaluation
picogram
peroxidase conjugated with a dioxin analog
parts per billion; nanogram/g; ng/g
parts per million; microgram/g; |ig/g
parts per trillion; picogram/g; pg/g
pound per square inch
quality assurance/quality control
reference material
relative percent difference
relative standard deviation
sample-specific detection limit
selected ion monitoring
Superfund Innovative Technology Evaluation
standard operating procedure
Standard Reference Material®
tetrachlorodibenzo-/?-dioxin
toxicity equivalency factor
toxicity equivalent

-------
                   Abbreviations, Acronyms, and Symbols (Continued)

TEQD/F               total toxicity equivalents of dioxins/furans
TEQPCB               total toxicity equivalents of World Health Organization dioxin-like polychlorinated
                     biphenyls
TOC                 total organic carbon
total TEQ             total toxicity equivalents including the sum of the dioxin/furan and World Health
                     Organization dioxin-like polychlorinated biphenyls
WHO                World Health Organization
                                                XI

-------
                                         Acknowledgments

This report was prepared for the U.S. Environmental Protection Agency (EPA) Superfund Innovative Technology
Evaluation (SITE) Program under the direction and coordination of Stephen Billets of the EPA's National Exposure
Research Laboratory (NERL)—Environmental Sciences Division in Las Vegas, Nevada. George Brilis and
Brian Schumacher of the EPA NERL reviewed and commented on the report. The EPA's NERL thanks Michael Jury,
Al Taylor, and Sue Kaelber-Matlock of the Michigan Department of Environmental Quality (MDEQ) and
Becky Goche and Doug Spencer of the U.S. Fish and Wildlife Service for their support in conducting the field
demonstration. We appreciate the support of the Dioxin SITE Demonstration Panel for their technical input to the
demonstration/quality assurance project plan.  In particular, we recognize Andy Beliveau, Nardina Turner,
Greg Rudloff, Allen Debus, Craig Smith, David Williams, Dwain Winters, Jon Josephs, Bob Mouringhan,
Terry Smith, and Joe Ferrario of the U.S. EPA. Thanks also go to EPA Region 2, EPA Region 3, EPA Region 4, EPA
Region 5, EPA Region 7, and the MDEQ for collecting and supplying environmental samples for inclusion in the
demonstration. Andy Beliveau, Allen Debus, and Nardina Turner served as EPA reviewers of this report. Michael
Jury (MDEQ), Sue Kaelber-Matlock (MDEQ), Jim Sanborn (California-EPA), and Jeffrey Archer (U.S. Food and
Drug Administration) served as additional reviewers of this report. Computer Sciences Corporation provided a
technical editing review of the report. This report was prepared for the EPA by Battelle. Special acknowledgment is
given to Amy Dindal, who was the Battelle Project Manager, and to Josh Finegold, Nicole Iroz-Elardo, Mark Misita,
Tim Pivetz, Mary Schrock, Rachel Sell, Bea Weaver, and Zack Willenberg for their contributions to the preparation
of this report.
                                                  Xll

-------
                                              Chapter 1
                                             Introduction
The U.S. Environmental Protection Agency (EPA),
Office of Research and Development (ORD), National
Exposure Research Laboratory (NERL) contracted with
Battelle (Columbus, Ohio) to conduct a demonstration of
monitoring and measurement technologies for dioxin
and dioxin-like compounds in soil and sediment. A field
demonstration was conducted as part of the EPA
Superfund Innovative Technology Evaluation (SITE)
Monitoring and Measurement Technology (MMT)
Program. The purpose of this demonstration was to
obtain reliable performance and cost data on the
technologies to provide (1) potential users with a better
understanding of the technologies' performance and
operating costs under well-defined field conditions and
(2) the technology developers with documented results
that will help promote the acceptance and use of their
technologies.

This innovative technology verification report (ITVR)
describes the SITE MMT Program and the scope of this
demonstration (Chapter 1);  a description of the Wako
Pure Chemical Industries, Ltd. Dioxin Enzyme-Linked
Immunosorbent Assay (ELISA) Kit Wako (for
environmental) (Chapter 2); the demonstration site and
the sampling locations (Chapter 3); the demonstration
approach (Chapter 4); the confirmatory process
(Chapter 5);  the assessment of reference method data
quality (Chapter 6); the performance of the technology
(Chapter 7);  the economic analysis for the technology
and reference method (Chapter 8); the demonstration
results in summary form (Chapter 9); and the references
used to prepare this report (Chapter 10). Appendix A
contains a verification statement; Appendix B contains
supplemental information provided by the developer;
Appendix C  is a summary of method blank and batch
duplicate data by the reference laboratory; and Appendix
D contains a one-to-one matching of the developer and
reference laboratory data.
1.1  Description of the SITE MMT Program
Performance verification of innovative environmental
technologies is an integral part of the regulatory and
research mission of the EPA. The SITE Program was
established by the EPA Office of Solid Waste and
Emergency Response and ORD under the Superfund
Amendments and Reauthorization Act of 1986. The
overall goal of the Program is to conduct performance
verification studies and to promote the acceptance of
innovative technologies that may be used to achieve
long-term protection of human health and the
environment. The program is designed to meet three
primary objectives: (1) identify and remove obstacles to
the development and commercial use of innovative
technologies, (2) demonstrate promising technologies
and gather reliable performance and cost information to
support site characterization and remediation activities,
and (3) develop procedures and policies that encourage
use of innovative technologies at Superfund sites as well
as at other waste sites or commercial facilities. The SITE
Program includes the following elements:

•  MMT Program—Evaluates technologies that sample,
   detect, monitor, or measure hazardous and toxic
   substances. These technologies are expected to
   provide better, faster, or more  cost-effective methods
   for producing real-time data during  site characteriza-
   tion and remediation efforts than conventional
   laboratory technologies.

•  Remediation Technology Program—Conducts
   demonstrations of innovative treatment technologies
   to provide reliable performance, cost, and
   applicability data for site cleanups.

•  Technology Transfer Program—Provides and
   disseminates technical information in the form of
   updates, brochures, and other publications that
   promote the SITE Program and participating

-------
   technologies. It also supports the technologies by
   offering technical assistance, training, and
   workshops.

The MMT Program's technology verification process is
designed to conduct demonstrations that will generate
high-quality data so that potential users have reliable
information regarding the technology performance and
cost. Four steps are inherent in the process: (1) needs
identification and technology selection, (2) demonstra-
tion planning and implementation, (3) report preparation,
and (4) information distribution. The first step of the
technology verification process begins with identifying
technology needs  of the EPA and regulated community.
The EPA Regional offices, the U.S. Department of
Energy, the U.S. Department of Defense, industry, and
state environmental regulatory agencies are asked to
identify technology needs for sampling, measurement,
and monitoring of environmental media. Once a need is
identified, a search is conducted to identify suitable
technologies that will address the need. The technology
search and identification process consists of examining
industry and trade publications, attending related
conferences, and exploring leads from technology
developers and industry experts. Selection of
technologies for field testing includes evaluation of the
candidate technologies based on several criteria. A
suitable technology for field testing

•  is designed for use  in the field or in a mobile
   laboratory,

•  is applicable to a variety of environmentally
   contaminated sites,

•  has potential for solving problems that current
   methods cannot satisfactorily address,

•  has estimated costs that are lower than those of
   conventional methods,

•  is likely to achieve  equivalent or better results than
   current methods in  areas such as data quality and
   turnaround time,

•  uses techniques that are easier or safer than current
   methods, and

•  is commercially available.

Once candidate technologies are identified, developers
are asked to participate in a developer conference. This
conference gives the developers an opportunity to
describe their technologies' performance and to learn
about the MMT Program.

The second step of the technology verification process is
to plan and implement a demonstration that will generate
representative, high-quality data to assist potential users
in selecting a technology. Demonstration planning
activities include a pre-demonstration sampling and
analysis investigation that assesses existing conditions at
the proposed demonstration site or sites. The objectives
of the pre-demonstration investigation are to (1) confirm
available information on applicable physical, chemical,
and biological characteristics of contaminated  media at
the sites to justify selection of site areas for the demon-
stration; (2) provide the technology developers with an
opportunity to evaluate the areas, analyze representative
samples, and identify logistical requirements; (3) assess
the overall logistical and quality assurance requirements
for conducting the demonstration; and (4) select and
provide the reference laboratory with an opportunity to
identify any matrix-specific analytical problems
associated with the contaminated media and to propose
appropriate solutions. Information generated through the
pre-demonstration investigation is used to develop the
final demonstration  design and to confirm the nature and
source of samples that will be used in the demonstration.

Demonstration planning activities also include
preparation of a demonstration plan that describes the
procedures to verify the performance and cost  of each
technology. The demonstration plan incorporates
information generated during the pre-demonstration
investigation as well as input from technology devel-
opers, demonstration site representatives, and technical
peer reviewers. The demonstration plan also incorporates
the quality assurance (QA)/quality control (QC)
elements needed to produce data of sufficient quality to
document the performance and cost of each technology.

During the demonstration, each technology is evaluated
independently and, when possible and appropriate, is
compared to a reference technology. The performance
and cost of one technology are not compared to those of
another technology evaluated in the demonstration.
Rather, demonstration data are used to evaluate the
individual performance, cost, advantages, limitations,
and field applicability of each technology.

-------
As part of the third step of the technology verification
process, the EPA publishes a verification statement
(Appendix A) and a detailed evaluation of each
technology in an ITVR. To ensure its quality, the ITVR
is published only after comments from the technology
developer and external peer reviewers are satisfactorily
addressed. All demonstration data used to evaluate each
technology are summarized in a data evaluation report
(DER) that constitutes a complete record of the
demonstration. The DER includes audit reports, observer
reports, completed data validation checklists, certificates
of analysis, and the data packages (i.e., raw data) from
the reference laboratory. The DER is not published as an
EPA document, but a copy may be obtained from the
EPA project manager.

The fourth step  of the verification process is to distribute
demonstration information. To benefit technology
developers and potential technology users, the  EPA
makes presentations, publishes and distributes fact
sheets, newsletters, bulletins, and ITVRs through direct
mailings and on the Internet. Information on the SITE
Program is available on the EPA  ORD Web site
(http://www.epa.gov/ORD/SITE). Additionally, a
Visitor's Day, which is held in conjunction with the
demonstration, allows the developers to showcase their
technologies and gives potential users the opportunity to
have a firsthand look at the technologies in operation.

1.2     Scope of This Demonstration
Polychlorinated dibenzo-p-dioxins and polychlorinated
dibenzofurans, commonly referred to collectively as
"dioxins," are of significant concern in site remediation
projects and human health assessments because they are
highly toxic. Dioxins and furans  are halogenated
aromatic hydrocarbons and are similar in structure as
shown in Figure 1-1. They have similar chemical and
physical properties. Chlorinated dioxins and furans are
technically referred to as polychlorinated dibenzo-/?-
dioxins (PCDD) and polychlorinated dibenzofurans
(PCDF). For the purposes of this document, they will be
referred to simply as "dioxins," "PCDD/F," or "D/F."
Dioxins and furans are not intentionally produced in
most chemical processes. However, they can be
synthesized directly and are commonly generated as  by-
products of various combustion and chemical
processes.(1) They are colorless crystals or solids with
high melting points, very low water solubility, high fat
solubility, and low volatility. Dioxins and furans are
extremely stable under most environmental conditions,
making them persistent once released in the
environment. Because they are fat soluble, they also tend
to bioaccumulate.

There are 75 individual chlorinated dioxins and 135
individual chlorinated furans. Each individual dioxin and
furan is referred to as a congener. The properties of each
congener vary according to the number of chlorine
atoms present and the position where the chlorines are
attached. The congeners with chlorines attached at a
minimum in the 2,3,7, and 8 positions are considered
most toxic. A total of seven dioxin and 10 furan
congeners contain chlorines in the 2, 3, 7, 8 positions
and, of these, 2,3,7,8-tetrachlorodibenzo-/?-dioxin
(2,3,7,8-TCDD) is one of the most toxic and serves as
the marker compound for this class.

Certain polychlorinated biphenyls (PCBs) have
structural and conformational similarities to dioxin
compounds (Figure 1-1) and are therefore expected to
exhibit toxicological similarities to dioxins as well.
Currently only 12 of the total 209 PCB congeners are
thought to have "dioxin-like" toxicity. These 12 are
PCBs with four or more chlorines with just one or no
substitution in the ortho position, and which assume a
flat configuration with rings in the same plane. These
        Cl
Cl
         2,3,7,8-Tetrachlorodibenzo-p-dioxin
        Cl
Cl
          2,3,7,8-Tetrachlorodibenzofuran

            CL
            Cl                   Cl
          3,3',4,4',5,5'-Hexachlorobiphenyl

        Figure 1-1. Representative dioxin,
        furan, and polychlorinated biphenyl
        structure.

-------
"dioxin-like" PCBs are often refered to as non-ortho and
mono-ortho substituted coplanar PCBs.

Conventional analytical methods for determining
concentrations of dioxin and dioxin-like compounds are
time-consuming and costly. For example, EPA standard
methods require solvent extraction of the sample,
processing the  extract through multiple cleanup
columns, and analyzing the cleaned fraction by gas
chromatography (GC)/high-resolution mass
spectrometry (HRMS). The use of a simple, rapid, cost-
effective analytical method would allow field personnel
to quickly assess the extent of contamination at a site
and could be used to direct or monitor remediation or
risk assessment activities. This data could be used to
provide immediate feedback on potential health risks
associated with the site and permit the development of a
more focused and cost-effective sampling strategy. At
this time, more affordable and quicker analytical
techniques will not replace HRMS. However, before
adopting an alternative to traditional laboratory-based
methods, a thorough assessment of how commercially
available technologies compare to conventional
laboratory-based analytical methods using certified,
spiked, and environmental samples is warranted. A
summary of the demonstration activities to evaluate
measurement technologies for dioxin and dioxin-like
compounds in  soil and sediment is provided below. The
experimental design and demonstration approach are
described in greater detail in Chapter 4 and was
published in the Demonstration and Quality Assurance
Project Plan (D/QAPP).(2)

1.2.1   Organization of Demonstration
The key organizations and personnel involved in the
demonstration, including the roles and responsibilities of
each, are fully  described in the D/QAPP.(2) EPA/NERL
had overall responsibility for this project. The EPA
reviewed and concurred with all project deliverables
including the D/QAPP and the ITVRs, provided
oversight during the demonstration, and participated in
the Visitor's Day. Battelle served as the verification
testing organization for EPA/NERL. Battelie's
responsibilities included developing and implementing
all elements of the D/QAPP; scheduling and
coordinating the activities of all demonstration
participants; coordinating the collection of environ-
mental samples; serving as the characterization
laboratory by performing the homogenization of the
environmental samples and confirming the efficacy of
the homogenization and approximate sample
concentrations; conducting the demonstration by
implementing the D/QAPP; summarizing, evaluating,
interpreting, and documenting demonstration data for
inclusion in this report; and preparing draft and final
versions of each developer's ITVR. The developers were
five companies who submitted technologies for
evaluation during this demonstration. The responsi-
bilities of the developers included providing input to,
reviewing, and concurring with the D/QAPP; providing
personnel and supplies as needed for the demonstration;
operating their technology during the demonstration; and
reviewing and commenting on their technology's ITVR.
AXYS Analytical Services, Ltd. was selected to serve as
the reference analytical laboratory. AXYS analyzed
each demonstration sample by EPA Method 1613B(3)
and EPA Method 1668A(4) according to the statement of
work provided in the D/QAPP. The Michigan
Department of Environmental Quality (MDEQ) hosted
the demonstration, coordinated the activities of and
participated in Visitor's Day, and collected and provided
some of the environmental samples that were used in the
demonstration. The Dioxin SITE Demonstration Panel
served as technical advisors and observers of the
demonstration activities. Panel membership, which is
outlined in the D/QAPP, included representation from
EPA Regions 1, 2, 3, 4, 5, 7, and 9; EPA Program
Offices; the MDEQ; and the U.S. Fish and Wildlife
Services. Members of the panel participated in five
conference  calls with the EPA, Battelle, AXYS, and the
developers. The panel contributed to the experimental
design and  D/QAPP development; logistics for the
demonstration, including site selection, sample
collection, reference laboratory selection; and data
analysis and technology evaluation procedures. As an
example of the significant impact the panel had on the
demonstration, it was the EPA members of the panel
who suggested expanding the scope of the project from
focusing exclusively on dioxins and furans, to also
include PCBs and the generation of characterization data
for polynuclear aromatic hydrocarbons (PAHs).

1.2.2  Sample Descriptions and Experimental
       Design
Soil and sediment samples with a variety of distinguish-
ing characteristics such as high levels of PCBs and

-------
PAHs were analyzed by each participant. Samples were
collected from a variety of dioxin-contaminated soil and
sediment sampling locations around the country.
Samples were identified and supplied through EPA
Regions 2, 3, 4, 5, and 7 and the MDEQ. The samples
were homogenized and characterized by the character-
ization laboratory prior to use in the demonstration to
ensure a variety of homogeneous, environmentally
derived samples with concentrations over a large
dynamic range (< 50 to > 10,000 picogram/gram  [pg/g])
were included. The environmental samples comprised
128 of the 209 samples included in the demonstration
(61%). Performance evaluation (PE) samples were
obtained from five commercial sources. PE samples
consisted of known quantities of dioxin and dioxin-like
compounds. Fifty-eight of the 209 demonstration
samples (28%) were PE samples.  A suite of solvent
extracts was included in the demonstration to minimize
the impact of sample homogenization and to provide a
uniform matrix for evaluation.  A total of 23 extracts
(11% of the total number of samples) was included in
the demonstration. The demonstration samples are
described in greater detail in Section 4.3.

1.2.3  Overview of Field Demonstration
All technology developers participated in a pre-
demonstration study where a representative subset of the
demonstration samples was analyzed.  The pre-
demonstration results indicated that the Wako Pure
Chemical Industries, Ltd. technology was suitable for
participation in the demonstration. The demonstration of
technologies for the measurement of dioxin and dioxin-
like compounds was conducted at the Green Point
Environmental Learning Center in Saginaw, Michigan,
from April 26 to May 5, 2004.  Five technologies,
including immunoassay test kits and aryl hydrocarbon
(Ah)-receptor binding technologies, participated in the
demonstration. The operating procedures for the
participating technologies are described in the D/QAPP.

The technologies were operated by the developers.
Because the sample throughput of the technologies
varied widely, it was at the discretion of the developers
how many of the 209 demonstration samples were
analyzed in the field. Results from the demonstration
samples, in comparison with results generated by AXYS
using standard analytical methods, were used to evaluate
the analytical performance of the technologies, including
the parameters of accuracy, precision, and comparability.
Observations from the field demonstration were used to
assess sample throughput, ease of use, health and safety
aspects, and the field portability of each technology.  The
PE of the Wako Pure Chemical Industries, Ltd. Dioxin
ELISA Kit Wako (for environmental) is presented in this
ITVR. Separate ITVRs have been published for the other
four participating technologies.

-------
                                            Chapter 2
          Description of Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit
This technology description is based on information
provided by Wako and only editorial changes were made
to ensure document consistency. Actual cost and
performance data, as reported and observed during the
demonstration, will be provided later in this document.
The Dioxin ELISA Kit Wako (for environmental) from
Wako Pure Chemical Industries, Ltd. was developed to
screen minute amounts of dioxin. With amicroplate
reader, samples can be assayed simultaneously for
PCDD/Fs.

2.1   Company History
1922  The Chemicals Department of Takeda Chobei
      Shoten (the present Takeda Chemical Industries,
      Ltd.) separated and established as an independent
      company, Takeda Pure Chemicals Ltd. in June.
1935  Tokyo Office opened
1940  Osaka Plant opened
1944  Tokyo Plant opened
1947  Company name changed to Wako Pure Chemical
      Industries, Ltd.
1952  Head office moved to the current address
1964  Tokyo Plant moved to the current site
1967  Tokyo and Osaka Research Laboratories
      completed
1968  Harima Plant opened
      Paris Liaison Office opened (until 1970)
1970  Dusseldorf Liaison Office opened (until 1983)
1972  Head Office building completed
1974  Wako Chemicals GmbH established in
      Dusseldorf under 100% capital ownership
1978  Dallas Liaison Office opened (until 1989)
1981  Wako Chemicals USA, Inc. founded in Dallas,
      Texas, under 100% capital ownership
1983  Uedesheim Plant, Wako  Chemicals GmbH,
      completed
      Neuss Liaison Office opened
1988  Mie Plant opened
1989  Tobu Distribution Center opened
      Wako Chemicals USA, Inc. moved to Richmond,
      Virginia
      Richmond Liaison Office opened
1990  Miyazaki Plant opened
      Richmond Plant, Wako Chemicals USA, Inc.
      completed
1991  Seibu Distribution Center opened
      Matumoto Plant opened
1994  Hafen Plant, Wako Chemicals GmbH completed
1995  All domestic plants completed registration of
      ISO-9002 certification
1998  Wako obtained ISO-9002 certificate of
      registration for the total quality system of the
      domestic operation
2001  Wako obtained Environment Management
      System ISO-14001 certificate of registration
      Tokyo and Osaka plants were approved as
      official testing laboratories based on ISO/IEC-
      17025

2.2    Product History
The kit was developed by Wako and the Food and Drug
Safety Center (FDSC) Research Institute in Japan. FDSC
received research funding from the Ministry of Health,
Labour and Welfare to develop the kit. FDSC developed
the monoclonal antibody used for the  analysis of
dioxins, and Wako developed the Dioxin ELISA Kit
Wako (for environmental) using the monoclonal
antibody.

2.3    Technology Description
The extraction procedure is described in Figure 2-1. The
process involves accelerated solvent extraction (ASE)
followed by cleanup.  Once extracted,  a monoclonal
antibody specific to dioxin is mixed with a sample
solution or the positive control (PC) provided with the
            Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                  6

-------
                                                 of
                            t»9i!«
                            __,  _
                         C    Cxlne-lkm
        10 g

               eilraetion

                       Oven temp,: 200 "C
            1.500psi      Flush;
         2Smifs,         Purge.  120 see,
          with Tifroo¥sip*Ly

(1) (Multilayer silica g®l column ctiromaitographf
               (20- to 30-ntesh) onto the Presep         gel
          with TiirtoVap'SOCl

(2) Pfritatescfanine IramshtSlzecJ sfSiea feS c:0lumrs
  Cleanup; 2 mL of     followed by 1 ml of a 1:9 mwtum of
        1 inL
      dry with nitrc^en concetitrator

(3)       with           (SO pi)

|4) MiK the          and Itie ELISA kit buffer soiytkwi
  |10 (ji * 90 jjL )
               Figure 2-1. Dioxin ELISA Kit Wako (for environmental) extraction procedure.
Dioxin ELISA Kit Wako (for environmental).
Peroxidase conjugated with a dioxin analog
(POD-conjugate) is then added, reacting with a primary
antibody to dioxin in the sample. The mixture is added
to a microplate coated with a secondary antibody that
captures the antibody-POD-conjugate and incubated at
2° to 8°C for 18 to 20 hours. After washing the resultant
microplate with a buffer, the antibody-POD-conjugate
complex formed on the plate  is reacted with substrate for
peroxidase. The reaction is stopped by adding stop
solution, and the microplate reader reads the signal.
The Dioxin ELISA Kit Wako (for environmental)
contains the secondary antibody microplate, the PC,
buffers A and B, the primary antibody, peroxidase
conjugate (lyophilized), wash solution concentrate,
substrate, citrate buffer, stop  solution, and a plate seal.

The monoclonal antibodies used for the Dioxin ELISA
Kit Wako (for environmental) indicate cross-reactivity
nearly equal to the positive control
(2,7,8-trichlorodibenzo [1,4]  dioxin-1-yl) acrylic acid)
and 2,3,7,8-TCDD. It is possible to find dioxin
concentrations as the amount equivalent to
2,3,7,8-TCDD toxicity equivalent (TEQ). (See
Table 2-1.)
                  The Dioxin ELISA Kit Wako (for environmental)
                  sensitivity is claimed by Wako to be from 1.6 to 100 pgs
                  per assay, and 96 samples can be assayed in two days.
                  The assay procedure is summarized in Figure 2-2.

                  Wako reported its data in pg/g 2,3,7,8-TCDD equivalent
                  (EQ). Although the Wako units specifically include
                  2,3,7,8-TCDD and the kit is most sensitive to this
                  congener, the technology represents a total D/F toxicity
                  equivalent value. Wako reported results in the
                  approximate range of 20 to 2,000 pg/g. Concentrations
                  measured to be below or above these concentrations
                  were reported semiquantitatively (e.g.,<20 pg/g or
                  >2,000pg/g). Wako notes that results can be reported
                  below 20 pg/g by taking a larger sample size (than  10 g)
                  and can be reported above 2,000 pg/g by performing
                  dilutions. This is the developer method that was
                  implemented during the field demonstration. A photo of
                  the technology in operation during the field
                  demonstration is presented in Figure  2-3. Wako provided
                  supplemental information about the performance of their
                  technology during the demonstration and it is presented
                  in Appendix B.
              Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
Table 2-1. Dioxin ELISA Kit Wako (for environmental) Cross-Reactivity
No.
1
2
o
J
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Class.
Dioxins












Furans









Coplanar-PCB













Brominated dioxin



Brominated furan
Brominated/
chlorinated dioxin

Number of
Chlorines
Bichloride
Trichloride
Tetrachloride


Pentachloride

Hexachloride



Heptachloride
Octachloride
Tetrachloride
Pentachloride

Hexachloride



Heptachloride

Octachloride
Tetrachloride

Pentachloride




Hexachloride



Heptachloride


Tetrabromide
Pentabromide
Hexabromide
Octabromide
Pentachloride
Monobromine
trichloride
Monobromine
tetrachloride
Article name
2,7-DiCDD
2,3,7-TnCDD
1,2,3,4-TeCDD
1,3,6,8-TeCDD
2,3,7,8-TeCDD
1, 2,4,6,8/1 ,2,4,7,9-PeCDD
1,2,3,7,8-PeCDD
1,2,3,4,6,7-HxCDD
1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD
1,2,3,4,6,7,8-HpCDD
1,2,3,4,6,7,8,9-OCDD
2,3,7,8-TeCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,6,7,8-HxCDF
1,2,3,7,8,9-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDF
1,2,3,4,7,8,9-HpCDF
1,2,3,4,6,7,8,9-OCDF
3,3',4,4'-TeCB(#77)
3,4,4',5-TeCB(#81)
2,3,3',4,4'-PeCB(#105)
2,3,4,4',5-PeCB(#114)
2,3',4,4',5-PeCB(#118)
2',3,4,4',5-PeCB(#123)
3,3',4,4',5-PeCB(#126)
2,3,3',4,4'5-HxCB(#156)
2,3,3',4,4',5'-HxCB(#157)
2,3',4,4',5,5'-HxCB(#167)
3,3',4,4',5,5'-HxCB(#169)
2,2',3,3',4,4',5-HpCB(#170)
2,2',3,4,4',5,5'-HpCB(#180)
2,3,3',4,4'5,5'-HpCB(#189)
2,3,7,8-TeBrDD
1,2,3,7,8-PeBrDD
1,2,3,6,7,8-HxBrDD
1,2,3,4,6,7,8,9-OBrDD
2,3,4,7,8-PeBrDF
2-Br-3,7,8-TriCDD
l-Br-2,3,7,8-TeCDD
Toxicity
Equivalency
Factor




1.0

1.0

0.1
0.1
0.1
0.01
0.0001
0.1
0.05
0.5
0.1
0.1
0.1
0.1
0.01
0.01
0.0001
0.0001
0.0001
0.0001
0.0005
0.0001
0.0001
0.1
0.0005
0.0005
0.00001
0.01
0.0001
0.00001
0.0001







ELISA Kit Cross-
Reactivity Data
0.05
0.16
0.0002
0.0004
1.00
0.008
0.48
0.003
0.07
0.04
0.06
0.006
0.00004
0.14
0.03
0.17
0.02
0.08
0.07
0.06
0.002
0.002
0.00004
0.0008
0.0009
0.00003
0.00004
0.00005
0.00004
0.0005
0.00001
0.00002
0.00003
0.00003
0.00002
0.00002
0.00006
0.39
0.04
0.003
0.0007
0.11
0.40
0.19
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
               POD-Conjugate
                Reaction of POD-
                conjugate and
                unreacted   *"""
                antibody
                            Sample or PC
                          Add to microplate well
                          with secondary
                          Antibody fixation plate

                         jl,..Ji,..,n,	r
            Fix reactants to
            microplate well
              React the POD with the substrate for coloring. Then
              measure the color densitv with the microolate reader.

     Figure 2-2.  Dioxin ELISA Kit Wako (for
     environmental) assay procedure.
 Figure 2-3. Dioxin ELISA Kit Wako (for
 environmental) in operation during the field
 demonstration.
2.4     Developer Contact Information
Additional information about this technology can be
obtained by contacting:

Wako Pure Chemical Industries, Ltd.
1-2 Doshomachi 3-Chome Chuo-ku
Osaka 540-8605 Japan
Telephone: +81-6-6203-3841
Web site: http://www.wako-chem.co.jp
E-mail:  cservice@wako-chem.co.jp

U.S. Subsidiary:
        Wako Chemicals USA, Inc.
        1600 Bellwood Road
        Richmond, Virginia 23237-1326 USA
        Telephone: (804) 271-7677
        Web site: http://www.wakousa.com
        E-mail: emmy@wakousa.com
2.5   Product Information
Code No.: 295-59601
Product Name: Dioxin ELISA Kit Wako (for
environmental)
Quantity: for 96 tests
              Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                       9

-------
                                              Chapter 3
                     Demonstration and Environmental Site Descriptions
This chapter describes the demonstration site, the
sampling locations, and why each was selected.

3.1    Demonstration Site Description and
       Selection Process
This section describes the site selected for hosting the
demonstration, along with the selection rationale and
criteria. Several candidate host sites were considered.
The candidate sites were required to meet certain
selection criteria, including necessary approvals,
support, and access to the demonstration site; enough
space and power to host the technology developers, the
technical support team,  and other participants; and
various levels of dioxin-contaminated soil and/or
sediment that could be analyzed as part of the
demonstration.  Historically, these demonstrations are
conducted at sites known to be contaminated with the
analytes of interest. The visibility afforded the sites is  a
valuable way of keeping the local community informed
of new technologies and to help promote the EPA's
commitment to promote and advance science and
communication.

After review of the information available, the site
selected for the  demonstration was the Green Point
Environmental Learning Center (ELC) site, located
within the city of Saginaw, Michigan. The Saginaw
city-owned, 76-acre Green Point ELC, formerly known
as the Green Point Nature Center, is managed by the
Shiawassee National Wildlife Refuge. The Green Point
ELC is situated  within the Tittabawassee River flood
plain. The MDEQ found higher than normal levels of
dioxins in soil and sediment samples taken from the
flood plain of the Tittabawassee River. The flood plain is
not heavily laden with PCBs; however, low levels of
PCBs have been detected in some areas. Soil samples
taken from areas outside the flood plain were at typical
background levels. The  source of the contamination was
speculated to be attributed to legacy contamination from
chemical manufacturing.

To summarize, Green Point ELC was selected as the
demonstration site based on the following criteria:

•   Access and Cooperation of the State and Local
    Community—Representatives from the MDEQ,
    EPA Region 5, and the local U.S. Fish and Wildlife
    Services supported the demonstration by providing
    site access for the demonstration, logistical support
    for the demonstration, and supported a Visitor's Day
    during the demonstration.

    Space Requirements and Feasibility—The demon-
    stration took place in the parking lot adjacent to the
    Green Point ELC, not directly on an area of
    contamination.  The site had electrical power and
    adequate space to house the trailers and mobile labs
    that were used for the demonstration. Furthermore,
    the site was close to an international airport. The
    weather in Michigan at the time of the demonstration
    was unpredictable; however, all participants were
    provided heated containment (a mobile laboratory or
    construction trailer).

•   Site Diversity—The area encompassing the Green
    Point site had different levels and types of dioxin
    contamination in both the soil and sediment that
    were used to evaluate the performance of the
    technologies.

The demonstration was conducted at the Green Point
ELC over a 10-day period from April 26 to May 5, 2004.
All technologies were operated inside trailers equipped
with fume hoods or inside mobile laboratories. As such,
the ambient weather conditions during the demonstration
had little  impact on the operation of the technologies,
since all of the work spaces were climate-controlled with
heat and air conditioning. The outdoor weather
conditions were generally cool and rainy, but the
                                                    10

-------
developers kept their working environment at
comfortable temperatures (16 to 18°C).  The low
temperature over the 10-day demonstration period was
2°C, the high temperature was 26°C, and the average
temperature was 9°C. Precipitation fell  on eight of the
10 days, usually in the form of rain, but  occasionally as
sleet or snow flurries, depending on the  temperature. The
largest amount of precipitation on a given demonstration
day was 0.50 inches.

3.2    Description of Sampling Locations
This section  provides an overview of the 10 sampling
sites and methods of selection. Table 3-1 summarizes
each of the locations, what type of sample (soil or
sediment) was provided, the number of samples
submitted from each location, and the number of
samples included in the demonstration from each
location. Samples were collected from multiple
sampling sites so that a wide variety of matrix conditions
could be used to evaluate the performance of the
technologies in addressing monitoring needs at a diverse
range of Superfund sites.

Samples consisted of either soil or sediment and are
described below based on this distinction. It should be
noted that it was not an objective of the  demonstration to
accurately characterize the concentration of dioxins,
furans, and PCBs from a specific sampling site. It was,
however, an  objective to ensure comparability between
technology samples and the reference laboratory
samples. This was accomplished by homogenizing each
matrix, such  that all sub-samples of a given matrix had
consistent contaminant concentrations. As a result,
homogenized samples were not necessarily
representative of original concentrations at the site.

3.2.1  Soil Sampling Locations
This section  provides descriptions of each of the soil
sampling locations, including how the sites became
contaminated and approximate dioxin concentrations, as
well as the type and concentrations of other major
constituents, where known [such as PCBs, penta-
chlorophenol (PCP), and PAHs]. This information was
provided by the site owners/sample providers (e.g., the
EPA, EPA contractors, and the MDEQ).
3.2.1.1 Warren County, North Carolina
Five areas of the Warren County PCB Landfill in North
Carolina, a site with both PCB and dioxin
contamination, were sampled. Dioxin concentrations in
the landfill soils range approximately from 475 to
700 pg/g, and PCB concentrations are greater than
100 parts per million (ppm). The Warren County PCB
Landfill contains soil that was contaminated by the
illegal spraying of waste transformer oil containing
PCBs from over 210 miles of highway shoulders. Over
30,000 gallons of contaminated  oil were disposed of in
14 North Carolina counties. The landfill is located on a
142-acre tract of land. The EPA permitted the landfill
under the Toxic Substances Control Act. Between
September and November 1982, approximately 40,000
cubic yards (equivalent to 60,000 tons) of PCB-
contaminated  soil were removed and hauled to the newly
constructed landfill located in Warren County, North
Carolina. The landfill is equipped with both polyvinyl
chloride and clay caps and liners. It also has a dual
leachate collection system. The material in the landfill is
solely from the contaminated roadsides. The landfill was
never operated as a commercial  facility. The remedial
action was funded by the EPA and the State of North
Carolina. The site was deleted from the National
Priorities List on March 7, 1986.

3.2.1.2 Tittabawassee River Flood Plain
The MDEQ sampled the Tittabawassee River flood plain
soils from three sites in the flood plain. The source of the
contamination was speculated to be attributed to legacy
contamination from chemical manufacturing. Two
samples were  collected from two locations at Imerman
Park in Saginaw Township. The first sample was taken
near the boat launch, and the second sample was taken in
a grassy area near the  river bank. Previous analysis from
these areas of this park indicated a range of PCDD/F
concentrations from 600 to 2,500 pg/g. Total PCBs from
these previous measurements were in the low parts-per-
trillion range.  Two samples were collected from two
locations at Freeland Festival Park in Freeland, MI. The
first sample was taken above the river bank, and the
second sample was taken near a brushy forested area
within the park complex. Previous PCDD/F
concentrations were from 300 to 3,400 pg/g, and total
PCBs were in the low part per trillion (ppt) range. The
final two samples were collected from Department of
Natural Resources (DNR)-owned property in Saginaw,
                                                   11

-------
Table 3-1.  Summary of Environmental Sampling Locations
Sample Type
Soil
Sediment
Sampling Location
Warren County, North Carolina
Tittabawassee River, Michigan
Midland, Michigan
Winona Post, Missouri
Solutia, West Virginia
Newark Bay, New Jersey
Raritan Bay, New Jersey
Tittabawassee River, Michigan
Saginaw River, Michigan
Brunswick, Georgia
Total
Number of Samples
Submitted for Consideration
5
6
6
6
6
6
6
6
6
5
58
Included in Demonstration
3
3
4
3
3
4
3
3
3
3
32
which was formerly a farming area located almost at the
end of the Tittabawassee River where it meets the
Shiawassee River to form the Saginaw River. Previous
PCDD/F concentrations ranged from 450 to 1,150 pg/g.
Total PCBs were not previously analyzed, but
concentrations were expected to be less than 1 ppm. The
DNR property is approximately a 10-minute walk from
where the demonstration was conducted at the Green
Point ELC.

3.2.1.3 Midland, Michigan
Soil samples were collected by the MDEQ from various
locations in Midland, Michigan. The soil type and nature
of dioxin contamination are different in the Midland
residential area than it is on the Tittabawassee River
flood plain, but it is from the same suspected source
(legacy contamination from chemical manufacturing).
Samples were collected in various locations around
Midland. Estimated TEQ concentrations ranged from
10 pg/g to 1,000 pg/g.

3.2.1.4 Winona Post
The Winona Post site in Winona, Missouri, was a
Superfund cleanup of a wood treatment facility.
Contaminants at the site included PCP, dioxin, diesel
fuel, and PAHs. Over a period of at least 40 years, these
contaminants were deposited into an on-site drainage
ditch and sinkhole.  Areas of contaminant deposition
(approximately 8,500 cubic yards of soils/sludge) were
excavated in late 200 I/early 2002. This material was
placed into an approximate 2!/2-acre  treatment cell
located on facility property. During  2002/2003, material
at the treatment cell was treated through addition of
amendments (high-ammonia fertilizer and manure) and
tilling.  Final concentrations achieved in the treatment
cell averaged 26 milligram per kilogram (mg/kg) for
PCP and from 8,000 to 10,000 for pg/g dioxin
equivalents. Samples obtained for this study from this
site were obtained from the treatment cell after these
concentrations had been achieved.

3.2.1.5  Solutia
The chemical production facility at the Solutia site in
Nitro, West Virginia, is located along the eastern bank of
the Kanawha River, in Putnam County, West Virginia.
The site has been used for chemical production since the
early 1910s. The initial production facility was
developed by the U.S. government for the production of
military munitions during the World War I era between
1918 and 1921. The facility was then purchased by a
small private chemical company, which began manu-
facturing chloride, phosphate, and phenol compounds at
the site. A major chemical manufacturer purchased the
facility in  1929 from Rubber Services Company. The
company continued to expand operations and accelerated
its growth in the 1940s. A variety of raw materials has
been used at the facility over the years, including
inorganic compounds, organic solvents, and other
organic compounds, including Agent Orange. Agent
Orange is a mixture of chemicals containing equal
amounts of two herbicides: 2,4-D (2,4 dichlorophenoxy-
acetic acid) and 2,4,5-T (2,4,5 trichlorophenoxyacetic
acid). Manufacture of this chemical herbicide began at
the site in  1948 and ceased in 1969. The source of the
                                                    12

-------
dioxin contamination in the site soils was associated
with the manufacture of 2,4,5-T, where dioxins are an
unintentional by-product. The site has a dioxin profile
from ppt to low parts per billion (ppb) range. No PCBs
or PAHs were identified in the soil.

3.2.2  Sediment Sampling Sites
This section provides descriptions of each of the
sediment sites that includes how the sites became
contaminated and approximate dioxin concentrations, as
well as the type and concentrations of other major
constituents (such as PCBs, PCP, and PAHs). This infor-
mation was provided from site owners/samples providers
(e.g., the EPA, EPA contractors, and the MDEQ).

3.2.2.1 New York/New Jersey Harbors
Dredged materials from the New York and New Jersey
harbors were provided as samples for the demonstration.
The U.S. Army Corps of Engineers, New York District,
and EPA Region 2 are responsible for managing dredged
materials from the New York and New Jersey harbors.
Dioxin levels affect the disposal options for dredged
material. Dredged materials are naturally occurring
bottom sediments, but some in this area have been
contaminated with dioxins and other compounds by
municipal or industrial wastes or by runoff from
terrestrial sources such as urban areas or agricultural
lands.

3.2.2.1.1 Newark Bay
Surrounded by manufacturing industries, Newark Bay is
a highly contaminated area with numerous sources
(sewage treatment plants, National Pollutant Discharge
Elimination System discharges, and nonpoint sources).
This bay is downstream from a dioxin Superfund site
that contains some of the highest dioxin concentrations
in the United States and also is downstream from a
mercury Superfund site. The dioxin concentration in the
area sampled for this demonstration was approximately
450 pg/g. Average PCB concentrations ranged from 300
to 740 ppb. Fine-grained sediments make up 50% to
90% of the dredged material. Average total organic
carbon (TOC) was about 4%.

3.2.2.1.2 RaritanBay
Surrounded by industry and residential discharges,
Raritan Bay has dioxin contamination in the area, but it
is not to the degree of Newark Bay. No major Superfund
sites are located in the vicinity. Dioxin concentration
should be significantly less than in Newark Bay. PCB
concentrations are around 250 ppb. The fine-grained
sediment and TOC values were similar to percentages in
Newark Bay.

3.2.2.2  Tittabawassee River
The first Tittabawassee River location was
approximately %-mile upstream of the Bob Caldwell
Boat launch in Midland, Michigan. The sediments are
dark gray, fine sand with some silt.  The  estimated TEQ
concentration was 260 pg/g; however, concentrations as
high as 2,100 pg/g TEQ have  been found in this area.
The second site was on the Tittabawassee River
approximately 100 yards downstream from old Smith's
Crossing Bridge in Midland, Michigan.  The sediment
was brown and sandy with organic material. The
estimated TEQ concentration  was 870 pg/g; but, again,
concentrations as high  as 2,100 pg/g TEQ are possible in
the area. The third site  was on Tittabawassee River at the
Emerson Park Golfside Boat Launch. The sediment was
gray black silty sand, with many leaves and high organic
matter. The estimated TEQ concentration was < 5 pg/g.
The fourth site was on the Tittabawassee River adjacent
to Imerman Park in Saginaw County across from the
fishing dock. The sediment was sand with some silt. The
estimated TEQ concentration  was between 100 and
2,000 pg/g TEQ. The fifth site was on the Tittabawassee
River approximately 1 mile downstream of Center Road
Boat launch in Saginaw Township. The sediment
consisted of sand and gravel with some  shells and not
much organic matter. The estimated TEQ concentration
was between 100 and 1,000 pg/g TEQ. The sixth site
also was on the Tittabawassee River across from the
Center Road Boat launch. The sediment was fine sand
with high organic matter. The estimated TEQ
concentration was 1,000 pg/g  TEQ.  The source of the
contamination was speculated to be attributed to legacy
contamination from chemical  manufacturing.

3.2.2.3  Saginaw River
Saginaw River samples were collected at six locations.
The first sampling location was in the Saginaw River
just downstream of Green Point Island. Samples were
collected near the middle of the river in  about 21 feet of
water. The sample was granular with some organic
material. The estimated TEQ concentration was 100 ppt.
Another Saginaw River sample was taken upstream of
Genesee Bridge on the right side of the river. The sample
                                                   13

-------
was a brown fine sand from about 15 feet of water. The
estimated TEQ concentration was 100 ppt. The third
location was in the Saginaw River downstream of the
Saginaw wastewater treatment plant in about eight feet
of water. The sample was gray silty clay with an
unknown TEQ concentration. The fourth location was in
the Saginaw River in about eight feet of water. The
sample was a black sandy material. The estimated TEQ
concentration for this  location was unknown. The fifth
location was downstream of a petroleum pipeline
crossing upstream of the Detroit and Mackinaw railroad
bridge crossing. This location was selected because of its
proximity to a former  PCB dredging location. The
sediment sample consisted of dark black silt with some
sand. The estimated TEQ concentration was unknown,
but PCB concentrations are expected to be high. The
sixth and final sampling location was near the mouth of
the Saginaw River in about five feet of water. The
sediment was a mix of fine black silt and layers of sand
and shells. The estimated TEQ concentration for this
location was also unknown.

3.2.2.4 Brunswick Wood Preserving Site
The Brunswick Wood Preserving Superfund site is
located in Glynn County, Georgia, north of the city of
Brunswick. The site was originally located in the city of
Brunswick, but moved to its present location around
1958. The site is approximately 84 acres and is about
two-thirds of a mile long. Burnett Creek, a tidally
influenced stream, is located at the western corner of the
site. At several points, most, if not all, of the drainage
from the site flows into Burnett Creek. The site was first
operated by American Creosote Company, which
constructed the facility sometime between 1958 and
1960. The site was acquired by Escambia Treating
Company in 1969 from Georgia Creosoting Company
and the Brunswick Creosoting Company. In 1985, a
corporate reorganization resulted in the purchase of the
facility by the Brunswick Wood Preserving Company,
which operated the site until it closed in early 1991.
Each of the three major wood-treating operations was
carried out at the facility: PCP, creosote, and
chromium-copper-arsenic (CCA). The site was listed on
the EPA's National Priorities List on April  1,  1997.

Sediment samples from the Brunswick Wood Preserving
site in Brunswick, Georgia, were collected from six
locations on the site, including areas thought to have
lower (< 300 pg/g TEQ) and higher (> 10,000 pg/g
TEQ) dioxin/furan concentrations. Due to the processes
that occurred on this site, the samples also contain
varying levels of PAHs and PCP, but they were not
expected to contain PCBs.
                                                    14

-------
                                               Chapter 4
                                     Demonstration Approach
This chapter discusses the demonstration objectives,
sample collection, sample homogenization, and
demonstration design.

4.1    Demonstration Objectives
The primary goal of the SITE MMT Program is to
develop reliable performance and cost data on innovative,
commercial-ready technologies. A SITE demonstration
must provide detailed and reliable performance and cost
data so that technology users have adequate information
to make sound decisions regarding comparability to
conventional methods. The demonstration had both
primary and secondary objectives. Primary objectives
were critical to the technology evaluation and required
the use of quantitative results to draw conclusions
regarding a technology's performance. Secondary
objectives pertained to information that is useful to know
about the technology but did not require the use of
quantitative results to draw conclusions regarding a
technology's performance.

The primary objectives for the demonstration of the
participating technologies were as follows:

PI. Determine the accuracy.
P2. Determine the precision.
P3. Determine the comparability of the technology to
    EPA standard methods.
P4. Determine the estimated method detection limit
    (EMDL).
P5. Determine the frequency of false positive and false
    negative results.
P6. Evaluate the impact of matrix effects on technology
    performance.
P7. Estimate costs associated with the operation  of the
    technology.

The secondary objectives for the  demonstration of the
participating technologies were as follows:
SI. Assess the skills and training required to properly
    operate the technology.
S2. Document health and safety aspects associated with
    the technology.
S3. Evaluate the portability of the technology.
S4. Determine the sample throughput.

Application of these objectives to the demonstration was
addressed based on input from the Dioxin SITE
Demonstration Panel members,(2) general user
expectations of field measurement technologies, the time
available to complete the demonstration, technology
capabilities that the developers participating in the
demonstration intend to highlight, and the historical
experimental components of former SITE Program
demonstrations to maintain consistency.

Note that this demonstration does not assess all
parameters that can affect performance of the
technologies in comparison to the reference methods
(i.e., not all compounds have been characterized in the
test samples, calibration of technologies results to
HRMS results on site-by-site basis was not evaluated,
etc.).  However, the demonstration as outlined below
was agreed upon by the Dioxin SITE Demonstration
Panel members to provide a reasonable evaluation of the
technologies.

4.2   Toxicity Equivalents
For risk assessment purposes, estimates of the toxicity of
samples that contain a mixture of dioxin, furan, and PCB
congeners are often expressed as TEQs. TEQs are
calculated by multiplying the concentration of each
congener with a toxicity equivalency factor (TEF),
according to the equation:

                 TEQ=Cc*TEF
where Cc is the concentration of the congener. The TEF
(see Table 4-1) provides an equivalency factor for each
                                                    15

-------
Table 4-1.  World Health Organization Toxicity Equivalency Factor Values
Compound00
PCDDs
2,3,7,8-TCDD
1,2,3,7,8-PeCDD

1,2,3,4,7,8-HxCDD
1,2,3,6,7,8-HxCDD
1,2,3,7,8,9-HxCDD

1,2,3,4,6,7,8-HpCDD
555555 f
OCDD
Dioxin-like PCBs
Coplanar
3,3',4,4'-TCB (PCB 77)
3,4,4',5-TCB (PCB 81)
3,3',4,4',5-PeCB (PCB 126)
3,3',4,4',5,5'-HxCB (PCB 169)




WHO TEF

1
1

0.1
0.1
0.1

0.01

0.0001


0.0001
0.0001
0.1
0.01




Compound
PCDFs
2,3,7,8-TCDF
1,2,3,7,8-PeCDF
2,3,4,7,8-PeCDF
1,2,3,4,7,8-HxCDF
1,2,3,7,8,9-HxCDF
1,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
1,2,3,4,6,7,8-HpCDF
555555 f
1,2,3,4,7,8,9-HpCDF
OCDF

mono-ortho
2,3,3',4,4'-PeCB (PCB 105)
2,3,4,4',5-PeCB (PCB 114)
2,3',4,4',5-PeCB (PCB 118)
2,3,4,4',5-PeCB (PCB 123)
2,3,3',4,4',5-HxCB (PCB 156)
2,3,3',4,4',5-HxCB (PCB 157)
2,3',4,4',5,5'-HxCB (PCB 167)
2,3,3',4,4'5,5'-HpCB (PCB 189)
WHO TEF

0.1
0.05
0.5
0.1
0.1
0.1
0.1
0.01
0.01
0.0001


0.0001
0.0005
0.0001
0.0001
0.0005
0.0005
0.00001
0.0001
   T = Tetra, Pe = Penta, Hx = Hexa, Hp = Hepta, O = Octa, CDD = chlorinated dibenzo-/>-dioxin, CDF = chlorinated
    dibenzofuran, CB = chlorinated biphenyl
congener's toxicity relative to the toxicity of 2,3,7,8-
TCDD.  The TEFs used in this demonstration were
determined by the World Health Organization (WHO)
for mammalian species.(5) The total TEQ from dioxin
and furans (TEQD/F) in a sample is calculated by adding
up all of the TEQ values from the individual dioxin and
furan congeners. The total TEQ contribution from PCBs
(referred to as TEQPCB) is calculated by summing up the
individual PCB TEQ values. The total TEQ in a sample
is the sum of the TEQD/F and TEQPCB values. TEQ
concentrations for soils and sediments are typically
reported in pg/g, which is equivalent to ppt.

Concentrations of dioxins, furans, and PCBs,
represented as total TEQ concentration, provide a
quantitative estimate of toxicity for all congeners
expressed as if the mixture were a TEQ mass of 2,3,7,8-
TCDD only. While the TEQ concept provides a way to
estimate potential health or ecological effects, the
limitations of this approach should be understood. The
WHO report noted that the TEF indicates an order of
magnitude estimate of the toxicity of a compound
relative to 2,3,7,8-TCDD.(5) Therefore, the accuracy of
the TEF factors could be affected by differences in
species, in the functional responses elicited by the
compounds, and in additive and nonadditive effects
when the congeners are present in complex mixtures.
The WHO(5) report concluded, however, that it is
unlikely that a significant error would be observed due
to these differences. The larger impact to the TEF
concept is the presence of Ah-receptor binding
compounds, such as PAHs (including naphthalenes,
anthracenes, and fluorenes) and brominated and
chloro/bromo-substituted analogues of PCDD/Fs, that
have not been assigned TEF values but which may
contribute to the total TEQ. This potentially can result
in an underestimation of TEQs in environmental samples
using the TEF approach.(5)

This demonstration was designed with these limitations
of the TEQ concept in mind.  The samples chosen
contained a variety of combinations of dioxins, furans,
                                                    16

-------
and PCBs and at a wide range of concentration levels.
Some samples were high in analytes with better
understood TEFs, while others were high in analytes
with TEFs that have more uncertainty. Some were high
in other Ah-receptor binding compounds such as PAHs,
while still others were free of these possible TEQ
contributing compounds. The purpose was to evaluate
each of the technologies under a variety of conditions
and assess the comparability of the TEQD/F and TEQPCB
values determined by the reference laboratory.

4.3    Overview of Demonstration Samples
The goal of the demonstration was to perform a detailed
evaluation of the overall performance of each
technology for use in the field or mobile environment.
The demonstration objectives were centered around
providing performance data that support action levels for
dioxin at contaminated sites. The Centers for Disease
Control's Agency for Toxic Substances and Disease
Registry (ATSDR) has established a decision framework
for sites that are contaminated with dioxin and dioxin-
like compounds.® If samples are determined to have
dioxin TEQ levels between 50 and 1,000 pg/g, the site
should be further evaluated; action is recommended for
levels above 1,000 pg/g (i.e.,  1 ppb) TEQ. A mix of PE
samples,  environmentally contaminated ("real-world")
samples,  and extracts were evaluated that bracket the
ATSDR guidance levels. Table 4-2 lists the primary and
secondary performance objectives for this demonstration
and which sample types were used in each evaluation.
The PE samples were used primarily to determine the
accuracy of the technology and consisted of purchased
soil and sediment standard reference materials with
certified concentrations of known contaminants and
newly prepared spiked samples. The PE samples also
were used to evaluate precision, comparability, EMDL,
false positive/negative results, and matrix effects.
Environmentally contaminated samples were  collected
from dioxin-contaminated sites around the country and
were used to evaluate the precision, comparability, false
positive/negative results, and matrix effects. Extracts,
prepared in toluene, which was the solvent used by the
reference laboratory, were used to evaluate precision,
EMDL, and matrix effects. All samples were used to
evaluate qualitative performance objectives such as
technology cost, the required skill level of the operator,
health and safety aspects, portability, and sample
throughput. Table 4-3 shows the number of each sample
type included in the experimental design. The following
sections describe each sample type in greater detail.

4.3.1  PE Samples
PE standard reference materials are available through
Cambridge Isotope Laboratories (CIL) Andover,
Massachusetts), LGC Promochem (United Kingdom),
Wellington Laboratories (U.S. distributor TerraChem,
Shawnee Mission, Kansas), the  National Institute of
Standards and Technology (NIST) (Gaithersburg,
Maryland), and Environmental Resource Associates
(ERA, Arvada, Colorado). All of these sources were
utilized to obtain PE samples for use in this
demonstration, and Table 4-4 summarizes the PE
samples that were included.  PE samples consisted of
three types of samples: (1) reference materials (RMs) or
certified samples, which included soil and/or sediment
samples with certified concentrations of dioxin, furan,
and/or PCBs; (2) spiked samples, which included a
certified dioxin, furan, PCB, and PAH-clean matrix
spiked with known levels of dioxin and/or other
contaminants; and (3) blank samples that were certified
to have levels of dioxins, furans, WHO PCBs, and PAHs
that were non-detectable or were considerably lower
than the detection capabilities of developer technologies.
The PE samples were selected based on availability and
on the correlation of the PE composition as it related to
the environmental samples that were chosen for the
demonstration (e.g., the PE sample had a similar
congener pattern to one or more of the environmental
sites).

Table 4-4 indicates a correlation between the
composition of the PE sample and the samples from the
environmental sites, where applicable. The certified
samples only required transfer from the original jar to
the demonstration sample jar. The spiked samples were
shipped to the characterization laboratory in bulk
quantities so each had to be aliquoted in 50-g quantities.
Additional details about each source of PE  sample are
provided further in this section.
                                                    17

-------
Table 4-2. Distribution of Samples for the Evaluation of Performance Objectives
Performance Objective
PI: Accuracy
P2: Precision
P3: Comparability
P4: EMDL
P5 : False positive/negative results
P6: Matrix effects
P7: Cost
SI: Skill level of operator
S2: Health and safety
S3: Portability
S4: Sample throughput
Sample Type Used in
Evaluation
PE
PE, environmental, extracts
PE, environmental, extracts
PE, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
Table 4-3.  Number and Type of Samples Analyzed in the Demonstration
Sample Type
PE
Environmental
Extracts
Total number of samples per technolosv
No. of Samples
58
128
23
209
4.3.1.1 Cambridge Isotopes Laboratories
Two RMs were obtained from CIL for use in this
demonstration. RM 5183 is a soil sample that was
collected from a location in Texas with the intended
purpose of serving as an uncontaminated soil for use as a
spiking material. The soil was sieved to achieve uniform
particle size and homogenized to within 5% using a
disodium fluorescein indicator.  Samples were then
sterilized three times for two hours at 121°C and
15 pounds per square inch (psi). Analytical results
indicated that the soil had low levels of D/F and PCBs.

RM 5184 is a heavily contaminated soil sample with
relatively high levels of D/F and PCBs. According to
the Certificate of Analysis (CoA),  approximately 75 kg
of contaminated sediment were obtained from an EPA
Superfund site in Massachusetts that was known to
contain considerable contamination from PCBs and
other chemical pollutants. The sediment was sieved to
achieve uniform particle size and homogenized to within
5% using a disodium fluorescein indicator. Samples
were then sterilized three times for two hours at 121°C
and 15 psi.

RM 5183 and RM 5184 are newly available from CIL.
For both RM 5183 and RM 5184, certified analytical
values are provided for the D/F and the 12 WHO PCB
congeners. The samples were included in an inter-
national interlaboratory study conducted by CIL and
Cerilliant Corporation. More than 20 laboratories
participated in analysis of the D/Fs; up to 20 laboratories
participated in the analysis of the PCBs. Participating
                                                    18

-------
Table 4-4. Summary of Performance Evaluation Samples
Sample
Type
ID
PE#1
PE#2
PE#3
PE#4
PE#5
PE#6
PE#7
PE#8
PE#9
PE#10
PE#11
PE#12
Source
CIL
LGC
Promochem
Wellington
CIL
NIST
ERA
ERA
ERA
ERA
ERA
ERA
ERA
PE Type
Certified
Certified
Certified
Certified
Certified
Spiked
Spiked
Spiked
Spiked
Spiked
Spiked
Organic,
Semivolatile,
Blank Soil
Product
No.
RM5183
CRM529
WMS-01
RM5184
SRM 1944
custom
custom
custom
custom
custom
custom
056
(lot 56011)
Certified Concentration
TEQD/F
(Pg/g)
3.9
6,583
62
171
251
11
33
NS
NS
NS
11
0.046
TEQPCB
(Pg/g)
5.0
424C
10.5
941
41C
NSf
NS
NS
11
1121
3,760C
0.01
PAH
(mg/kg)
0.18
NAd
NA
27
2.4e
O.33
<0.33
61g
<0.33
<0.33
<0.33
<0.33
Total Number ofPE samples
Correlation
to Environ.
Sample
Type D)a
6
5
6
2,8,9
3,4
10
10
5,7
1
1
1
not
applicable

No. of
Replicates
Per
Sample
7b
4
7b
4
4
4
4
4
4
4
4
8
58
  Environmental Sample IDs are provided in Table 4-5.
  Seven replicates were analyzed for EMDL evaluation.
  Little or no certified PCB data were available; mean of reference laboratory measurements was used.
  NA = no data available.
  Approximate concentration of 2-methyl naphthalene, acenaphthene, and fluorene, which were the only PAHs that were included in the
  analysis.
  NS = not spiked.
  Each of the 18 target PAHs was spiked at levels that ranged from 1 to 10 mg/kg. (See Section 5.2.3 for the list of 18 PAHs.)
laboratories used a variety of sample preparation and
analytical techniques.

4.3.1.2 LGC Promochem
Certified reference material (CRM) 529 was obtained
from LGC Promochem. The following description is
taken from the reference material report that
accompanied CRM 529. The soil for CRM 529 was
collected in Europe from a site where chloro-organic and
other compounds had been in large-scale production for
several decades, but where production had ceased more
than five years before sampling. The site had been
contaminated during long-term production of
trichlorophenoxyacetic acid.  An area of sandy soil was
excavated to a depth of several meters. Several hundred
kilograms of this mixed soil were air-dried at about 15°C
for three months. After removal of stones and other
foreign matter by sieving, the remaining material was
sterilized in air at 120°C for 2 hours, thoroughly mixed,
and ground in an Alpine air jet mill to a particle size of
< 63 micrometers (j-im). The material was homogenized
once more in a Turbula mixer and packaged in 50  g
quantities. The final mean moisture content at the time
of bottling was found to be 1.5%. According to the CoA,
certified values are provided for five dioxin congeners,
seven furan congeners, three chlorobenzene compounds,
and three chlorophenol compounds.  No PCBs were
reported with certified values on the CoA, so the mean
concentration determined by the reference laboratory
was used as the certified value.

4.3.1.3 Wellington
PE sample WMS-01 was obtained from TerraChem, the
U.S. distributor for Wellington, an Ontario-based
company. As described in the CoA, WMS-0 lisa
homogeneous lake sediment that was naturally
contaminated (and not fortified). The crude, untreated
sediment used to prepare WMS-01 was collected from
Lake Ontario. The sediment obtained was subsequently
air-dried; crushed to break up agglomerates; air-dried
again; and then sieved, milled, and re-sieved (100%
< 75 |im). The sediment was then subsampled into 25-g
aliquots. The demonstration samples for only the
Wellington PE samples were 25 g rather than 50 g based
                                                     19

-------
on the package size available from Wellington. Certified
values for the 17 D/F congeners and the 12 WHO PCB
congeners are provided on the CoA.

4.3.1.4 National Institute for Standards and
       Technology
Standard Reference Material® (SRM) 1944 was
purchased through NIST. As described in the CoA,
SRM 1944 is a mixture of marine sediment collected
from six sites in the vicinity of New York Bay and
Newark Bay in October 1994.  Site selection was based
on contaminant levels measured in previous samples
from these sites and was intended to provide relatively
high concentrations for a variety of chemical classes of
contaminants. The sediment was collected using an
epoxy-coated modified Van Veen-type grab sampler
designed to sample the sediment to a depth of
10 centimeters (cm). A total of approximately 2,100 kg
of wet sediment was collected from the six sites. The
sediment was freeze-dried, sieved (nominally 61 to
250 |im), homogenized in a cone blender, radiation
sterilized, then packaged in 50-g quantities. Certified
values are provided on the CoA for the 17 D/F
congeners, 30 PCB congeners, 24 PAHs, four
chlorinated pesticides, 36 metals, and TOC. Since only
three WHO PCBs were reported out of the 30 PCB
congeners, the mean concentration of the reference
laboratory measurements was used as the certified value
so that the TEQPCB concentration would not be
underestimated when compared to the developer
technologies.

4.3.1.5 Environmental Resource Associates
ERA synthesized PE samples for this demonstration.
ERA spiked blank, uncontaminated soil to pre-
determined levels of D/Fs, PCBs, and/or PAHs. Spiked
PE samples were prepared to include additional
concentration ranges and compositions that were not
covered with the commercially available certified
materials. The organic semivolatile soil blank (ERA
Product #056, Lot 56011) is atopsoil that was obtained
from a nursery and processed according to ERA
specifications by a geochemical laboratory. The particle
size distribution of the soil was -20/+60 mesh. The soil
was processed and blended with a sandy loam soil to
create a blank soil with the following make-up: 4.1%
clay, 4.5% silt, 91.2% sand, and 0.2% organic material.
Initially, ERA was required to certify that the blank soil
matrix to be used as the blank and for the preparation of
the spiked PE samples was "clean" relative to the list of
required target analytes. This was accomplished through
a combination of ERA-conducted analyses (PAHs,
pesticides, semivolatile organic compounds, Aroclors
which are trade mixtures of PCB congeners) and
subcontracted analytical verification (D/F and PCBs).
The subcontracted analyses were performed by Alta
Analytical Perspectives, LLC, in Wilmington, North
Carolina. The Alta Analytical Certificate of Results and
the ERA Certification sheets for the organic semivolatile
soil blank indicated that trace levels of the octa-dioxins
and several WHO PCB congeners were  detected, but the
total TEQ (combined D/F and PCBs) was less than
0.06 pg/g. The level of PAHs, pesticides, Aroclors, and
semi-volatile organic compounds in the  soil was
determined to be < 0.33 pg/g. The TEQ  level was
considerably below the detection capabilities of the
participating technologies, so the organic semivolatile
soil blank was considered adequately clean for use in
this demonstration.

The manufacturing techniques that ERA used to prepare
the PE samples for this demonstration were consistent
with those used for typical semivolatile  soil products by
ERA. These techniques have been validated through
hundreds of round robin performance test studies over
ERA's more than 25 years in business.  The D/F stock
solutions used in the manufacture of these PE samples
were purchases from CIL. The PCB and PAH stock
solutions were purchased from ChemService. For each
PE sample, a spiking concentrate was prepared by
combining appropriate weight/volume aliquots of stock
materials required for that PE sample. Typically,
additional solvent was added to this concentrate to yield
sufficient volume of solution, appropriate for the mass of
soil to be spiked. Based on a soil mass of 1,600 g, the
volume of spike concentrate was approximately 10 to
30 milliliter (mL). For each PE sample, the blank soil
matrix was weighed into a 2-liter (L) wide mouth glass
jar, the spike concentrate was distributed onto the soil,
and the soil was allowed to air-dry for 30 to 60 minutes.
The PE samples were then capped and mixed in a rotary
tumbler for 30 minutes. Each PE sample was certified as
the concentration of target analytes present in the blank
matrix, plus the amount added during manufacture,
based on volumetric and gravimetric measurements.
CoAs were provided by ERA for all six  ERA-provided
PE samples. The certified values provided by ERA were
different from the commercially available certified
                                                   20

-------
samples since the data were not based on analytically
derived results. Further confirmation of the concen-
trations was conducted by the reference laboratory.

4.3.2  Environmental Samples
Handling of the environmental samples is described in
this section. Note that once the environmental samples
were collected, they were dried and homogenized as best
as possible to eliminate variability introduced by sample
homogeneity. As such, the effect of moisture on the
sample analysis was not investigated.

4.3.2.1 Environmental Sample Collection
Samples were collected by the EPA, an EPA contractor,
or the MDEQ and shipped to the characterization
laboratory. When determining whether a soil or sediment
site had appropriate dioxin contamination, a guideline
concentration range of < 50 pg/g to 5,000 pg/g was used.

Once necessary approvals and sampling locations had
been secured, sample containers were shipped to site
personnel.  Each  site providing samples received
1-gallon containers (Environmental Sampling Supply,
Oakland, California, Part number 3785-1051, wide-
mouth, 128-ounce high-density polyethylene round
packer) for collecting five or six samples.

Instructions for sample collection, as well as how the
containers were to be labeled and returned, were
included in a cover letter with the sample containers that
were shipped to each site.  Personnel collecting the
samples were instructed to label two containers
containing the same sample as "1 of 2" and "2 of 2" and
to attach a description or label to each container with a
description of the sample, including where the sample
was collected and the estimated concentrations of dioxin
and any other anticipated contamination (e.g., PCBs,
PAHs, PCP). Final  instructions to sample providers
indicated that collected samples were to be shipped back
to the characterization laboratory using the provided
coolers. Federal Express labels that included an account
number and the shipping address were enclosed in each
shipment.

Sample providers also were asked to provide any
information about the possible source of contamination
or any historical data and other information, such as
descriptions of the sites, for inclusion in the
demonstration and quality assurance project plan
(D/QAPP).(2)

4.3.2.2 Homogenization of Environmental Samples
If the material had very high moisture content, the jar
contents were allowed to settle, and the water was
poured off. Extremely wet material was poured through
fine mesh nylon material to remove water. After water
removal, the material was transferred to a Pyrex™ pan
and mixed. After thorough mixing, an aliquot was stored
in a pre-cleaned jar as a sample of "unhomogenized"
material and was frozen.1 The remaining bulk sample
was mixed and folded bottom to top three times. This
material was split equally among multiple pans. In each
pan, the material was spread out to cover the entire
bottom of the pan to an equal depth of approximately
0.5 inches. The pans were placed in an oven at 35 °C and
held there until the samples were visibly dry. This
process took from 24 to 72 hours, depending on the
sample moisture. The trays were removed from the oven
and allowed to rise to room temperature by sitting in a
fume hood for approximately 2 hours. Approximately
500 g of material were put in a blender and blended for
2 minutes. The blender sides were scraped with a spatula
and the sample blended for a second 2-minute period.
The sample was sieved [USA Standard testing, No.  10,
2.00-millimeter opening] and the fine material placed in
a tray. Rocks and particles that were retained on the
sieve were placed in a pan. This process was repeated
until all of the sediment or soil were blended and sieved.
The blended and sieved sediment or soil in the tray was
mixed well, and four aliquots of 100- to 300-g each were
put into clean jars (short, wide-mouth 4-ounce,
Environmental Sampling Supply,  Oakland, California,
Part number 0125-0055) to be used for the
characterization analyses. The remaining sediment or
soil was placed in a clean jar, and the particles that were
retained on the sieve were disposed of. The jars of
homogenized sediment and soil were stored frozen
(approximately -20°C), unless the samples were being
used over a period of several days, at which time they
were temporarily stored at room temperature.
  Ideally, the samples would have been stored at 4° ± 2°C;
  but, due to the large volume of buckets and jars that needed
  to be stored, the most adequate available storage at the
  characterization laboratory was a walk-in freezer that was at
  approximately minus 20°C.
                                                    21

-------
4.3.2.3 Selection of Environmental Samples
Once homogenized, the environmental samples were
characterized for dioxin/furans (EPA Method 1613B(3)),
PCBs, low-resolution mass spectrometry (LRMS)
modified EPA Method 1668A(4), and 18 target PAHs
{National Oceanic and Atmospheric Administration
(NOAA]) method(7)] to establish the basic composition
of the samples. (Characterization analyses are described
in Chapter 5.) Because the soil and sediment samples
were dried and homogenized, they were indistinguish-
able. As such, the soil and sediment samples were jointly
referred to as "environmental" samples, with no
distinction made between soil or sediment other than
during the matrix effects evaluations, as described in
Section 4.7.6. Environmental samples were selected for
inclusion in the demonstration based on the preliminary
characterization data. The number and type of samples
from each sampling location included in the
demonstration are presented in Table 4-5.

Four aliquots of the homogenized material and one
aliquot of unhomogenized material were analyzed. Two
criteria had to be met for the environmental sample to be
considered for inclusion in the demonstration.  The first
criterion was that the relative standard deviation (RSD)
of the total D/F TEQ values from the four aliquots had to
be less than 20% for samples with total TEQ values
> 50 pg/g; RSD values up to 30% were considered
acceptable if the concentration was < 50 pg/g TEQ. The
second criterion was that no single RSD for an
individual congener could be greater than 30%. If both
of these criteria were met, the sample met the
homogenization criteria and was considered for
inclusion in the demonstration. If either of these criteria
was not met, options for the sample included
(a) discarding it and not considering it for use in the
demonstration, (b) reanalyzing it to determine if the data
outside the homogenization criteria were due to
analytical issues, or (c) rehomogenizing and reanalyzing
it. Of these options, (a) and (b) were utilized, but (c) was
not because an  adequate number of environmental
samples were selected using criteria  (a) and (b). The
average D/F concentration and RSDs for the
homogenization analyses of environmental samples are
shown in Table 4-5. The composition of two particular
Saginaw River samples was of interest for inclusion in
the demonstration because of their concentration and
unique congener pattern, but the homogenization criteria
were slightly exceeded (i.e., 28% and 34% RSD, for
Saginaw River Sample #2 and Saginaw River Sample
#3, respectively). Since multiple replicates of every
sample were analyzed, those samples were included in
the study because of their unique nature but are flagged
as slightly exceeding the homogenization criteria. A
correlation of environmental samples to PE samples,
similar to that presented in Table 4-4, is presented in
Table 4-5.

4.3.3  Extracts
A summary of the extract samples is provided in
Table 4-6. The  purpose of the extract samples was to
evaluate detection and measurement performance
independent of the sample extraction method. As shown
in Table 4-6, two environmental samples (both
sediments) were extracted using Soxhlet extraction with
toluene. These  extractions were performed by AXYS
Analytical Services consistent with the procedures to
extract the demonstration samples for reference
analyses.(2) The environmental sample extracts
represented a 10-g sediment sample extraction and were
reported in pg/mL, which was calculated by the
following equation:
pg/mL =
           Wgsamples)x(lOgaUquot)
            (300 mL extraction volume)
where DF = dilution factor.

Total extract volume per 10-g aliquot was 300 mL, but
the sample extracts were concentrated and provided to
the developers as 10-mL extracts, so a 3 Ox dilution
factor is included. The extracts were not processed
through any cleanup steps, but were derived from
sediment samples that also were included in the suite of
environmental samples. All environmental sample
extractions were prepared in the same solvent (toluene).
The extract samples also included three toluene-spiked
solutions that were not extractions of actual
environmental samples. Because adequate
homogenization at trace quantities was difficult to
achieve, one set of extract samples was spiked at low
levels (approximately 0.5 pg/mL of 2,3,7,8-TCDD) and
used as part of the EMDL evaluation.
                                                   22

-------
Table 4-5. Characterization and Homogenization Analysis Results for Environmental Samples
Sample
Type ID
Env Site #1
Env Site #2
Env Site #3
Env Site #4
Env Site #5
Env Site #6
Env Site #7
Env Site #8
Env Site #9
Env Site
#10
Environmental
Site Location
Warren County,
North Carolina
Tittabawassee
River, Michigan
Newark Bay,
New Jersey
Raritan Bay, New
Jersey
Winona Post,
Missouri
Tittabawassee
River, Michigan
Brunswick,
Georgia
Saginaw River,
Michigan
Midland,
Michigan
Solutia, West
Virginia
Soil or
Sediment
soil
soil
sediment
sediment
soil
sediment
sediment
sediment
soil
soil
Sample
No.
1
2
3
1
2
3
1
2
3
4
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
4
1
2
3
Average Total
TEQD/F
Concentration
(Pg/g)
274
5,065
11,789
42
435
808
16
62
45
32
12
14
13
3,831
11,071
11,739
1
55
16
69
65
14,500
921
1,083
204
239
184
149
25
48
1,833
3,257
RSD (%)
11
7
3
23b
5
10
26b
14
26b
6
2
3
7
1
2
1
23b
7
26b
8
1
2
9
28C
34C
5
5
7
10
10
19
11
Average RSDfor all environmental samples used in demonstration
Total number of environmental samples
No. of Replicates
per Sample
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Correlation
with PE Sample
Type D)a
9, 10, 11
4
5
5
2,8
1,3
8
4
4
6,7
77%
72S
a PE Sample IDs are provided in Table 4-4.
b RSD values up to 30% were allowed for samples where the characterization analyses determined concentration to be < 50 pg/g total TEQD/F.
c RSD value slightly exceeded the homogeneity criteria, but samples were included in the demonstration because they were samples of interest.
                                                          23

-------
Table 4-6.  Distribution of Extract Samples
Sample Type ID
Extract* 1
Extract #2
Extract #3
Extract #4
Extract #5
Sample ID
Environmental #6, Sample
TR#2
Environmental #7,
Sample #1
Spike #r
Spike #2a
Spike #3a
Sample Description
Soxhlet extraction in toluene; no
cleanup
Soxhlet extraction in toluene; no
cleanup
0.5pg/mL
2,3,7,8-TCDD
100 pg/mL 2,3,7,8-TCDD
1,000 pg/mL each WHO PCB
(TEQ-11)
10,000 pg/mL each WHO PCB
(TEQ~1,000)C
Total number of extracts
No. of Replicates per Sample
4
4
7b
4
4
23
a Prepared in toluene.
b Seven replicates were analyzed for EMDL evaluation.
c This extract was spiked with only PCBs but a low-level (approximately 0.3 pg/mL) 2,3,7,8-TCDD contamination was confirmed by the
  reference laboratory.
4.4     Sample Handling
In preparation for the demonstration, the bulk
homogenized samples were split into jars for
distribution. Each 4-ounce, amber, wide-mouth glass
sample jar (Environmental Sampling Supply, Oakland,
California, Part number 0125-0055) contained approxi-
mately 50 g of sample. Seven sets of samples were
prepared for five developers, the reference laboratory,
and one archived set. A minimum of four replicate splits
of each sample was prepared for each participant, for a
total of at least 28 aliquots prepared for each sample.
The purchased PE samples (i.e., standard reference
materials and spiked materials) were transferred from
their original packaging to the jars to be used in the
demonstration for the environmental samples, making
the environmental and PE samples visually
indistinguishable.

The samples were randomized in two ways. First, the
order in which the filled jars were distributed was
randomized. All jars had two labels. The label on the top
of the jar was the analysis order and contained sample
numbers 1 through 209. A second label placed on the
side of the jar contained a coded identifier including a
series of 10 numbers coded to include the site, replicate,
developer, and matrix. All samples believed to have at
least one D/F or PCB congener greater than 10,000 pg/g
were marked with an asterisk for safety purposes. This
was consistent for both the developer and reference
laboratory samples. The developer was given the option
of knowing which environmental site the samples came
from and whether the sample was a soil or sediment.
Wako elected to have the samples identified as soil or
sediment. As described in the D/QAPP, AXYS was
informed of which environmental site that the samples
came from so it could use congener profiles and dilution
schemes determined during the pre-demonstration phase
as a guide, along with the concentration range data that
was provided in the D/QAPP. This information was
supplied to the reference laboratory with the samples,
along with which samples contained high (i.e., a sample
with at least one congener with concentration
> 120,000 pg/g) or ultrahigh (i.e., a sample with at least
one congener with concentration > 1,200,000 pg/g) PCB
levels.  Using this information, AXYS regrouped the
samples in batches so that, to the extent possible,
samples from the same site would be analyzed within the
same analytical batch. Because an analytical laboratory
might know at least what site samples came from, and
because it is reasonable from an analytical standpoint to
group samples that might require similar dilution
schemes and which have similar congener patterns in an
analytical batch, this approach was an acceptable
deviation from the original intention of having the
samples run by the reference  laboratory completely blind
and in the prescribed analytical order. Wako analyzed
the samples in the order received. The extracts were the
first 23 samples in the Wako  analysis order.
                                                    24

-------
The environmental samples were stored at room
temperature until homogenized. After homogenization
and prior to distribution during the demonstration, the
samples were stored in a walk-in freezer (approximately
-20 °C) at the characterization laboratory. At the
demonstration site, the samples were stored at ambient
temperature. After the demonstration analyses were
completed, the samples were stored at the
characterization laboratory in the walk-in freezer until
the conclusion of the project.

4.5    Pre-Demonstration Study
Prior to the demonstration, pre-demonstration samples
were sent to Wako for evaluation in its laboratory. The
pre-demonstration study comprised 15 samples,
including PE samples, environmental samples, and
extracts. The samples selected for the pre-demonstration
study covered a wide range of concentrations and
included a representative of each environmental site
analyzed during the demonstration.

The pre-demonstration study was conducted in two
phases. In Phase 1, Wako was sent six soil/sediment
samples with the corresponding D/F, PCB, and PAH
characterization data to perform a self-evaluation of the
Dioxin ELISA Kit Wako (for environmental). In Phase
2, seven additional soil/sediment samples and two
extracts were sent to Wako for blind evaluation. AXYS
analyzed all 15 pre-demonstration samples blindly. The
Wako pre-demonstration results were paired with the
AXYS results and returned to Wako so they could use
the HRMS pre-demonstration sample data to refine the
performance of the test kit prior to participating in the
field demonstration. Results for the pre-demonstration
study can be found in the data evaluation report, which
can be obtained by contacting the EPA program manager
for this demonstration. The results confirmed that Wako
was a viable candidate to continue in the demonstration
process.

4.6    Execution of Field Demonstration
Wako arrived on-site on Thursday, April 22, and spent
several hours on four consecutive days setting up two
facilities (a mobile laboratory and a support trailer). The
demonstration officially commenced on Monday, April
26 after 1.5 hours of safety and logistical training.
During this meeting, the health and safety plan was
reviewed to ensure participants understood the safety
requirements for the demonstration.  Logistics, such as
how samples would be distributed and results reported,
were also reviewed during this meeting. After the safety
and site-specific training meeting and prior to samples
being received by the developers, each trailer and mobile
laboratory was surface wipe sampled on the floor to the
entrance of the developer work area to establish the
background level of D/F and PCB contamination. The
wipe sampling procedure was followed as described in
the D/QAPP. Following demobilization by the
developers, all of the trailers and mobile laboratories
were cleaned and surface wipe sampled. Analysis of the
pre- and post-deployment wipe samples indicated that all
trailers and mobile laboratories met the acceptable
clearance criteria that were outlined in the D/QAPP.
Only one fume hood had to be re-cleaned and re-
sampled before receiving final clearance.

Ideally, all 209 demonstration samples would have been
analyzed on-site, but sample throughput of some of the
technologies participating in the demonstration would
require three weeks or more in the field to analyze 209
samples.  Consequently, it was decided, as reported in
the D/QAPP,  that the number of samples to be analyzed
in the field by each developer would be determined at
the discretion of the developer.

Wako received its first batch of samples by midmorning
on April 26. Wako completed analysis of all 209 demon-
stration samples (23 extracts and 186 soil/sediment
samples) in 9 working days (on May 4). It should be
noted that the morning of April 28 was dedicated to a
Visitor's Day, so minimal work on sample analyses was
performed. Once the complete data set was submitted,
Wako was offered the opportunity to re-analyze any
samples before reporting final results. Wako elected to
re-analyze 36 samples. The samples for re-analysis were
received by Wako representatives on June 18. Wako
reported the reanalysis results on August 4. It was at the
developer's discretion whether to keep the originally
reported results or to replace with re-analysis data.  Of
the 36 reanalysis samples, 12 of the original sample
results were retained and 24 were reported with new
results based on the re-analysis.
                                                   25

-------
4.7    Assessment of Primary and Secondary
       Objectives
The purpose of this section is to describe how the
primary and secondary objectives are assessed, as
presented in Chapters 6 and 7.

Wako reported its data in pg/g 2,3,7,8-TCDD equivalent
(EQ). The Wako results were compared to the reference
laboratory results and certified values relative to TEQD/F.
Although the Wako units specifically include 2,3,7,8-
TCDD and the kit is most sensitive to this congener, the
technology represents a total D/F toxicity equivalent
value, so it  was compared to the HRMS total TEQD/F
concentration in all cases.

Wako reported results in the approximate range of 20 to
2,000 pg/g. Concentrations measured to be below or
above these concentrations were reported
semiquantitatively (e.g.,<20 pg/g or >2,000 pg/g).
Treatment of semiquantitative data is described in each
data analysis section. Wako notes that results can be
reported below 20 pg/g by taking a larger sample size
(than  10 g)  and can be reported above 2,000 pg/g by
performing dilutions.

4.7.1   Primary Objective PI: Accuracy
The determination of accuracy was based on agreement
with certified or spiked levels of PE samples. PE
samples containing concentrations from across the
analytical range of interest were analyzed. Measure-
ments from the 58 PE samples were evaluated to
determine whether there was a statistically significant
difference between the measurements and the certified
value  or spiked level. Percent recovery values relative to
the certified or spiked concentrations were also
calculated.  The PE samples were analyzed by the
laboratory reference method for confirmation of certified
and spiked  values.

To evaluate accuracy, the mean of replicate results from
the field technology measurement was compared to the
certified or spiked value of the PE samples to calculate
percent recovery. The equation used was:

                R=  C/CRxWO%

where C is the average concentration value (in pg/g
2,3,7,8-TCDD EQ) calculated from the technology
replicate measurements and CR is the certified value (in
pg/g TEQD/F). Nondetects and values reported as ">
(value)" were not included in the accuracy assessment.
Mean concentration values were determined when at
least three of four replicates were reported as actual
values [i.e., were not reported as, "< (value)" or ">
(value)"].  The mean, median, minimum, and maximum
R values are reported as an assessment of overall
accuracy. An ideal R value would be 100%.

4.7.2  Primary Objective P2: Precision
To evaluate precision, all samples (including PE,
environmental, and extract samples) were analyzed in at
least quadruplicate. Seven replicates of three different
samples were analyzed to evaluate EMDLs.

Precision was evaluated at both low and high
concentration levels and across different matrices. The
statistic used to evaluate precision was RSD. The
equation used to  calculate standard deviation (SD)
between replicate measurements was:
          SD =
where SD is the standard deviation and C is the average
measurement. Both are reported in pg/g 2,3,7,8-TCDD
EQ.

The equation used to calculate RSD between replicate
measurements was:
                RSD=
                       SD
                        C
xlOO%
RSD, reported in percent, was calculated if detectable
concentrations were reported for at least three replicates.
The mean, median, minimum, and maximum RSD
values are reported as an assessment of overall precision.

Low RSD values (< 20%) indicated high precision. For a
given set of replicate samples, the RSD of results was
compared with that of the laboratory reference method's
results to determine whether the reference method is
more precise than the technology or vice versa for a
particular sample set. The mean RSD for all samples  was
calculated to determine an overall precision estimate.
                                                    26

-------
4.7.3  Primary Objective P3: Comparability
Data comparability was maximized by using the
homogenization procedures and applying criteria for
acceptable results prior to a sample being included in the
demonstration. (See Section 4.3.2.3 for additional
information.)

Technology results reported by Wako Pure Chemical
Industries, Ltd. were compared to the corresponding
reference laboratory results by calculating relative
percent difference (RPD). The equation for RPD,
reported in percent, is as follows:
        RPD  =
   K-MD)
average (MR,MD)
xlOO%
where MR is the reference laboratory measurement (in
pg/g) and MD is the developer measurement (in pg/g
2,3,7,8-TCDD EQ). Nondetects were not included in this
evaluation.  For PE samples, TEQD/F RPD calculations
were only performed for the analyte classes that the PE
sample contained. For example, PE sample #9 was only
spiked with PCBs. Consequently, TEQD/F RPD
calculations were not performed.

The absolute value  of the difference between the
reference and developer measurements in the equation
above was not taken so that the RPD would indicate
whether the technology measurements were greater than
the reference laboratory measurements (negative RPD
values) or less than the reference laboratory
measurements (positive RPD). Because negative values
for RPD could be obtained with this approach, the
median RPD of all individual RPDs was calculated
rather than the average RPD in calculation of
comparability between the Wako results and reference
laboratory measurements. The median, minimum, and
maximum RPD values were reported as an  assessment of
overall comparability. RPD values between positive and
negative 25% indicated good agreement between the two
measurements.

As another  measure of comparability, the reference data
were grouped into four TEQ concentration ranges. The
ranges were < 50 pg/g, 50 to 500 pg/g, 500 to 5,000
pg/g, and > 5,000 pg/g.  The intervals were determined
by the Demonstration Panel and were based on current
guidance for cleanup levels. The percentage of developer
results that agreed with those ranges of values was
reported. It should be noted that developer results that
correctly agreed with the reference laboratory result, but
fell in a different interval, were counted as "in
agreement." For example, a Wako result as reported < 51
pg/g 2,3,7,8-TCDD EQ (which fell into the second
interval) agreed with a reference laboratory result
reported as 25.8 pg/g TEQD/F (which fell into the first
interval). The same rule applied to concentrations that
were reported as > 2,000 pg/g 2,3,7,8-TCDD EQ.

The accuracy of reporting blank samples was assessed.
The blanks included eight replicate samples that
contained levels of D/Fs that were below the reporting
limits of the developer technology but contained levels
that could be detected by the reference methods (see
Table 4-4). If the reference laboratory result was in the
nondetect interval reported by the developer technology
reporting limit, this result was considered  accurately
reported by the developer. The accuracy of the blank
samples was reported in terms  of % agreement. Ideal %
agreement values would be 100%.

4.7.4  Primary Objective P4: Estimated Method
       Detection Limit
The method detection limit (MDL) calculation procedure
described in the demonstration plan was 40 CFR
Part 136, Appendix B,  Revision 1.11. This procedure is
based on an assumption that the replicates are
homogeneous enough to allow proper measurement of
the analytical precision and that the concentration is in
the appropriate range for evaluation of the technology's
sensitivity. For this evaluation, Wako analyzed seven
aliquots each of a low-level PE soil, PE sediment, and a
toluene-spiked extract. MDL-designated samples are
indicated in Tables 4-4 and 4-6. PE samples in the
precisely appropriate range  for evaluation  of this
technology's detection limit were not available. The
developer reported nondetect values  for some of the
replicates, so provisions had to be made for the treatment
of nondetects. As such, the results from these samples
were used to calculate  an estimated MDL  (EMDL) for
the technology.

A Student's t-value and the standard deviation of seven
replicates were used to calculate the EMDL in pg/g TEQ
as shown in the following equation:
                                                    27

-------
            EMDL=  t
                       (n-l, l-ro=0.99
(SD)
where  t(n_1:1_m=a99) = Student's t-value appropriate for a
99 percent confidence level and a standard deviation
estimate with n-l degrees of freedom. Nondetect values
were assigned the reported value (i.e., "< 1" was
assigned as value of 1), half of the reported value (i.e.,
"< 1" was assigned 0.5), or zero. The various treatments
of nondetect values were performed to see the impact
that reduced statistical power (i.e., lower degrees of
freedom) had on the EMDL calculation.  The lower the
EMDL value, the more sensitive the technology is at
detecting contamination.

4.7.5  Primary Objective P5: False Positive/False
       Negative Results
The tendency for the Wako kit to return false positive
results (e.g., results reported above a specified level for
the field technology but below a specified level by the
reference laboratory) was evaluated. The frequency of
false positive results was reported as a fraction of results
available for false positive  analysis. Similarly, the
frequency of false negatives results was examined. For
this purpose, the results were evaluated for samples
reported as  having concentrations above and below 20
pg/g TEQ and 50 pg/g TEQ. As such, the samples that
were reported as < 20 (or 50) pg/g TEQ by the reference
laboratory but > 20 (or 50) pg/g TEQ by Wako were
considered  false positive. Conversely, those samples that
were reported as < 20 (or 50) pg/g TEQ by Wako, but
reported as  > 20 (or 50) pg/g by the reference laboratory,
were considered false negatives. In the case of
semiquantitative results (reported as < or >), if the
laboratory result was within the interval reported by the
developer, it was not considered a false positive or false
negative result. Ideal false positive and negative
percentages would be equal to zero.

4.7.6  Primary Objective P6: Matrix Effects
The likelihood of matrix-dependent effects on
performance was investigated by grouping the data by
matrix type (i.e., soil, sediment, extract), by sample type
(i.e., PE, environmental, and extract), by varying levels
of PAHs, by environmental site, and by known
interferences. Precision (RSD) data were summarized by
soil, sediment, and extract (matrix type); by
environmental, PE, and extract (sample type); and by
PAH concentration. Analysis of variance (ANOVA)
tests were performed to determine if there was a
dependence on matrix type or sample type. Only the
environmental samples were included in the matrix
effect assessment based on PAH concentration, because
only the environmental samples were analyzed for PAHs
during the characterization analysis (described in  Section
5.2.3). Some PAH data were available forthe PE
samples, but data were not available for all of the  same
analytes that were determined during the
characterization analysis. The environmental samples
were segregated into four ranges of total  PAH
concentrations: < 1,000 nanogram/g (ng/g), 1,000 to
10,000 ng/g, 10,000 to 100,000 ng/g, and > 100,000
ng/g. The precision (RSD) data were  summarized for
samples within these PAH concentration ranges.
ANOVA tests were used to determine if the summary
values for RSD were statistically different, indicating
performance dependent upon PAH concentration. For
the environmental site evaluation, the comparability
(RPD) values from each of the 10 environmental sites
were compared to see if the developer results were more
or less comparable to the reference laboratory for a
particular site. For known interferences, the developer's
reported results for PE samples were  summarized for
samples where the PE samples did not contain the target
analyte (e.g., did the developer report D/F detections for
a sample only spiked with PCBs).

This objective also evaluated whether performance was
affected by measurement location (i.e., in-field versus
laboratory conducted measurements), although this is
not a traditional matrix effect. However, Wako analyzed
all 209 samples during the field demonstration, and only
a few were rerun in their laboratories  and rereported, so
measurement location was not evaluated for Wako.

4.7.7  Primary Objective P7: Technology Costs
The full cost of each technology was documented and
compared to typical and actual costs for D/F and PCB
reference analytical methods. Cost inputs included
equipment, consumable materials, mobilization and
demobilization, and labor. The evaluation of this
objective is described in Chapter 8, Economic Analysis.
                                                    28

-------
4.7.8  Secondary Objective SI: Skill Level of
       Operator
Based on observations during the field demonstration,
the type of background and training required to properly
operate the Dioxin ELISA Kit Wako (for environmental)
was assessed and documented. The skill required of an
operator was also evaluated. The evaluation of this
secondary objective also included user-friendliness of
the technology.

4.7.9  Secondary Objective S2: Health and Safety
       Aspects
Health and safety issues, as well as the amount and type
of hazardous and nonhazardous waste generated, were
evaluated based on observer notes during the field
demonstration. This also included an assessment of the
personal protective equipment required to operate the
technology.

4.7.10 Secondary Objective S3: Portability
Observers documented whether the Dioxin ELISA Kit
Wako (for environmental) could be readily transported
to the field and how easy it was to operate in the field.
This included an assessment of what infrastructure
requirements were provided to Wako Pure Chemical
Industries, Ltd. (e.g., a mobile laboratory), and an
assessment of whether the infrastructure was adequate
(or more than adequate) for the technology's operation.
Limitations of operating the technology in the field are
also discussed.

4.7.11 Secondary Objective S4: Sample
       Throughput
Sample throughput was measured based on the observer
notes, which focused on the time-limiting steps of the
procedures, as well as the documentation of sample
custody. The number of hours Wako Pure Chemical
Industries, Ltd. worked in the field was documented
using attendance log sheets where Wako Pure Chemical
Industries, Ltd. recorded the time they arrived and
departed from the demonstration site.  Time was
removed for training and Visitor's Day activities. The
number of operators involved in the sample analyses
also was noted. Throughput of the developer technology
was compared to that of the reference laboratory.
                                                   29

-------
                                             Chapter 5
                                       Confirmatory Process
This chapter describes the characterization analyses and
the process for selecting the reference methods and the
reference laboratory.

5.1    Traditional Methods for Measurement of
       Dioxin and Dioxin-Like Compounds in
       Soil and Sediment
Traditional methods for analysis of dioxin and dioxin-
like compounds involve extensive sample preparation
and analysis  using expensive instrumentation resulting in
very accurate and high-quality, but costly, information.
The ability to use traditional methods for high-volume
sampling programs or screening of a contaminated site
often is limited by budgetary constraints. The cost of
these analyses can range approximately from $500 to
$ 1,100 per sample per method, depending on the method
selected, the  level of QA/QC incorporated into the
analyses, and the reporting requirements.

5.1.1  High-Resolution Mass Spectrometry
EPA Method 1613B(3) and SW-846 Method 8290(8) are
both appropriate for low and trace-level analysis of
dioxins and furans in a variety of matrices. They involve
matrix-specific extraction, analyte-specific cleanup, and
high-resolution capillary GC (HRGC)THRMS analysis.
The main differences between the two methods are that
EPA Method 1613B has an expanded calibration range
and requires  use of additional 13C12-labeled internal
standards resulting in more accurate identifications and
quantitations. The calibration  ranges for the HRMS
methods based on a typical 10-g sample  and
20-microliter (\\L) final sample volume are presented in
Table 5-1.
Table 5-1.  Calibration Range of HRMS
           Dioxin/Furan Method
Compound
Tetra
Compounds
Penta-Hepta
Compounds
Octa
Compounds
EPA Method
1613B
1-400 pg/g
5-2,000 pg/g
10-4,000 pg/g
SW-846 Method
8290
2-400 pg/g
5-1, 000 pg/g
10-2,000 pg/g
5.1.2  Low-Resolution Mass Spectrometry
SW-846 Method 8280 is appropriate for determining
dioxins and furans in samples with relatively high
concentrations, such as still bottoms, fuel oils, sludges,
fly ash, and contaminated soils and waters. This method
involves matrix specific extraction, analyte-specific
cleanup, and HRGC/LRMS analysis. The calibration
ranges in Table 5-2 are based on a typical 10-g sample
size and 100-(iL final volume.

Table 5-2.   Calibration Range of LRMS
            Dioxin/Furan Method
Compound
Tetra-Penta Compounds
Hexa-Hepta Compounds
Octa Compounds
SW-846 Method 8280
1,000-20,000 pg/g
2,500-50,000 pg/g
5,000-1 00,000 pg/g
                                                   30

-------
5.1.3  PCB Methods
There are more options for analysis of dioxin-like
compounds such as PCBs. EPA Method 1668A(4) is for
low- and trace-level analysis of PCBs. It involves matrix-
specific extraction, analyte-specific cleanup, and
HRGC/HRMS analysis. This method provides very
accurate determination of the WHO-designated dioxin-
like PCBs and can be used to determine all 209 PCB
congeners. Not all PCBs are determined individually
with this method because some are determined as sets of
coeluting congeners. The calibration range for PCBs
based on a typical 10-g sample and 20-(iL final sample
volume is from 0.4 to 4,000 pg/g. PCBs also can be
determined as specific congeners by GC/LRMS or as
Aroclors by GC/electron capture detection.

5.1.4  Reference Method Selection
Three EPA analytical methods for the quantification of
dioxins and furans were available: Method 1613B,
Method 8290, and Method 8280. Method 8280 is a
LRMS method that does not have adequate sensitivity
(i.e., the detection limits reported by the developers are
less than that of the LRMS method). Methods 1613B and
8290 are HRMS methods with lower detection limits.
Method 1613B includes more labeled internal standards
than Method 8290, which affords more accurate
congener quantification. Therefore, it was determined
that Method 1613B best met the needs of the demon-
stration, and it was selected as the dioxin/furan reference
method. Reference data of equal quality needed to be
generated to determine the PCB contribution to the TEQ,
since risk assessment is often based on TEQ values that
are not class-specific. As  such, the complementary
HRMS method for PCB TEQ  determinations, Method
1668A,(4) was selected as the reference method for PCBs.
Total TEQD/F concentrations were generated by Method
1613B, and total TEQPCB  concentrations were generated
by Method  1668A. These data were summed to derive a
total TEQ value for each sample.

5.2    Characterization of Environmental
       Samples
All of the homogenized environmental samples were
analyzed by the Battelle characterization laboratory to
determine which would be included in the demonstration.
The environmental samples were characterized for the
17 D/Fs by Method 1613B, the 12 WHO PCBs by
LRMS-modified Method  1668A, and 18 target PAHs by
the NOAA Status and Trends GC/Mass Spectrometry
(MS) method.(7)

5.2.1  Dioxins and Furans
Four aliquots of homogenized material and one
unhomogenized (i.e., "as received") aliquot were
prepared and analyzed for seventeen 2,3,7,8-substituted
dioxins and furans following procedures in EPA Method
1613B. The homogenized and unhomogenized aliquots
were each approximately 200 g. Depending on the
anticipated levels of dioxins from preliminary
information received from each sampling location,
approximately 1 to 10 g of material were taken for
analysis from each aliquot, and spiked with 13C12-labeled
internal standards, and extracted with methylene chloride
using ASE techniques. One method blank and one
laboratory control spike  were processed with the batch of
material from each site. The sample extracts were
processed through various cleanup techniques, which
included gel permeation chromatography or acid/base
washes, as well as acid/base silica and carbon cleanup
columns. As warranted,  based on sample  compositions,
some samples were put through additional acid silica
cleanup prior to the carbon column cleanup. Extracts
were spiked with 13C12-labeled recovery standards and
concentrated to a final volume of 20 to 50 \\L. Dilution
and reanalysis of the extracts were performed if high
levels of a particular congener were observed in the
initial analysis; however, extracts were not rigorously
evaluated to ensure that all peaks were below the peak
area of the highest calibration standard.

Each extract was analyzed by HRGC/HRMS  in the
selected ion monitoring  (SIM) mode at a resolution of
10,000 or greater. A DB-5 column was used for analysis
of the seventeen 2,3,7,8-PCDD/F congeners.  The
instrument was calibrated for PCDD/F at levels specified
in Method 1613B with one additional calibration
standard at concentrations equivalent to one-half the level
of Method 1613B's lowest calibration point.  Using a
DBS column, 2,3,7,8-TCDF is not separated from other
non-2,3,7,8-TCDF isomers. However, since the primary
objective was to determine adequacy of homogenization
and not congener quantification, it was determined that
sufficient information on precision could  be obtained
with the DBS analysis of 2,3,7,8-TCDF and no second
column confirmation of 2,3,7,8-TCDF was performed.
                                                   31

-------
PCDD/F data were reported as both concentration (pg/g
dry) and TEQs (pg TEQ/g dry).

5.2.2  PCBs
One aliquot of material from each sampling location was
prepared and analyzed for the 12 WHO-designated
dioxin-like PCBs by GC/LRMS. The LRMS PCB
analysis method is based on key components of the PCB
congener analysis approach described in EPA Method
1668A and the PCB homologue approach described in
EPA Method 680. Up to 30 g of sample were spiked with
surrogates and extracted with methylene chloride using
shaker table techniques. The mass of sample extracted
was determined based on information supplied to the
laboratory regarding possible contaminant concen-
trations. The extract was dried over anhydrous sodium
sulfate and concentrated. Extracts were processed
through alumina column cleanup, followed by high-
performance liquid chromatography/gel permeation
chromatography (HPLC/GPC). Additionally, sulfur was
removed using activated granular copper. The post-
HPLC extract was concentrated and fortified with
recovery internal standards. Extracts were concentrated
to a final volume between 500 (iL and 1 mL, depending
on the anticipated concentration of PCBs in the sample,
as reported by the sample providers. PCB congeners and
PCB homologues were separated via capillary GC on a
DB5-XLB column and identified and  quantified using
electron ionization MS. This method provides specific
procedures for the identification and measurement of the
selected PCBs in SIM mode.

5.2.3  PAHs
One aliquot of material from each sampling location was
analyzed for PAHs. The 18 target PAHs included:
•   naphthalene
•   2-methylnaphthalene
•   2-chloronaphthalene
•   acenaphthylene
•   acenaphthene
•   fluorene
•   phenanthrene
•   anthracene
•   fluoranthene
•   pyrene
•   benzo(a)anthracene
•   chrysene
•   benzo(b)fluoranthene
•  benzo(k)fluoranthene
•  benzo(a)pyrene
•  indeno(l,2,3-cd)pyrene
•  dibenzo(a,h)anthracene
•  benzo(g,h,i)perylene.

The method for the identification and quantification of
PAH in sediment and soil extracts by GC/MS was based
on the NOAA Status and Trends method(7) and, therefore,
certain criteria (i.e., initial calibrations and daily
verifications) are different from those defined in
traditional EPA methods 625 and 8270C. Up to 30 g of
sample were spiked with surrogates and extracted using
methylene chloride using shaker table techniques. The
mass of sample extracted was determined based on
information supplied to the characterization laboratory
regarding possible contaminant concentrations. The
extract was dried over anhydrous sodium sulfate and
concentrated. The extract was processed through an
alumina cleanup column followed by HPLC/GPC. The
post-HPLC extract was concentrated and fortified with
recovery internal standards. Extracts were concentrated
between 500 \\L and 1 mL,  depending on the anticipated
concentration of PCBs in the sample, as reported by the
sample providers. PAHs were separated by capillary GC
on a DB-5, 60-m column and were identified and
quantified using electron impact MS. Extracts were
analyzed in the SIM mode to achieve the lowest possible
detection limits.

5.3    Reference Laboratory Selection
Based on a preliminary evaluation of performance and
credibility, 10 laboratories were contacted and were  sent
a questionnaire geared toward understanding the
capabilities of the laboratories, their experience with
analyzing dioxin samples for EPA, and their ability to
meet the needs of this demonstration. Two laboratories
were selected for the next phase of the selection process
and were sent three blind audit samples. Each laboratory
went through a daylong audit that included a technical
systems audit and a quality systems audit. At each
laboratory, the  audit consisted of a short opening
conference; a full day of observation of laboratory
procedures, records, interviews with laboratory staff; and
a brief closing meeting. Auditors submitted followup
questions to each laboratory to address gaps  in the
observations.
                                                   32

-------
Criteria for final selection were based on the observations
of the auditors, the performance on the audit samples,
and cost. From this process, it was determined that
AXYS Analytical Services (Sidney, British Columbia,
Canada) would best meet the needs of this
demonstration.

5.4    Reference Laboratory Sample Prepara-
       tion and Analytical Methods
AXYS Analytical Services received all 209 samples on
April 27, 2004. To report final data, AXYS submitted
14 D/F and 14 PCB data packages from June 11 to
December 20, 2004.  The following sections briefly
describe the reference methods performed by AXYS.

5.4.1  Dioxin/Furan Analysis
All procedures were carried out according to protocols as
described in AXYS Summary Method Doc MSU-018
Rev 2 18-Mar-2004 [AXYS detailed Standard Operating
Procedure (SOP) MLA-017 Rev 9 May-2004], which is
based on EPA Method 1613B. AXYS modifications to
the method are summarized in the D/QAPP.(2) Briefly,
samples were spiked with a suite of isotopically labeled
surrogate standards prior to extraction, solvent extracted,
and cleaned up through a series of chromatographic
columns that included silica, Florisil, carbon/Celite, and
alumina columns. The extract was concentrated and
spiked with an isotopically labeled recovery (internal)
standard. Analysis was performed using an FiRMS
coupled to an FIRGC equipped with a DB-5 capillary
chromatography column [60 meters (m), 0.25-mm
internal diameter (i.d.), 0.1-|im film thickness]. A second
column, DB-225 (30 m, 0.25-mm i.d., 0.15-|im film
thickness), was used for confirmation of 2,3,7,8-TCDF
identification.  Samples that were known to contain
extremely high levels of PCDD/F were extracted without
the addition of the surrogate standard, split, then spiked
with the isotopically labeled surrogate standard prior to
cleanup. This approach allowed extraction of the
method-specified 10 g sample volume, and subsequent
sufficient dilution that high level analytes were brought
within the instrument calibrated linear range. While this
approach induces some uncertainty because the  actual
recovery of analytes from the extraction process is
unknown, it was decided by the demonstration panel that
in general analyte recovery through the extraction
procedures are known to be quite good and that the
uncertainty introduced by this approach would be less
than the uncertainty introduced by other approaches such
as extracting a significantly smaller sample size.

5.4.2  PCB Analysis
The method was carried out in accordance with the
protocols described in AXYS Summary Method Doc
MSU-020 Rev 3 24-Mar-2004 (AXYS detailed SOP
MLA-010 Rev 5 Sep-2003), which is based on EPA
Method 1668A, with changes through August 20, 2003.
AXYS modifications to the method are summarized in
the D/QAPP. Briefly, samples were spiked with
isotopically labeled surrogate standards, solvent
extracted, and cleaned up on a series of chromatographic
columns that included silica, Florisil, alumina,  and
carbon/Celite columns. The final extract was spiked with
isotopically labeled recovery (internal) standards prior to
instrumental analysis. The extract was analyzed by
HRMS coupled to an HRGC equipped with a DB-1
chromatography column (30 m, 0.25-mm i.d., 0.25-|im
film thickness). Because only the WHO-designated
dioxin-like PCBs were being analyzed for this program
and in order to better eliminate interferences, all samples
were analyzed using the DB-1 column, which is an
optional confirmatory column in Method 166 8A rather
than the standard SPB Octyl column. Samples that were
known to contain extremely high levels of PCBs were
extracted without the addition of the surrogate  standard,
split, then spiked with the isotopically labeled surrogate
standard prior to cleanup.  This approach allowed
extraction of the method-specified 10 g sample volume,
and subsequent sufficient dilution that high level analytes
were brought within the instrument calibrated linear
range. While this approach induces some uncertainty
because the actual recovery of analytes from the
extraction process is unknown, it was decided by the
demonstration panel that in general analyte recovery
through the extraction procedures are known to be quite
good and that the uncertainty introduced by this approach
would be less than the uncertainty introduced by other
approaches such as extracting a significantly smaller
sample size.

5.4.3  TEQ Calculations
For the reference laboratory data, D/F and PCB congener
concentrations were converted to TEQ and subsequently
summed to determine total TEQ, using the TEFs
established by WHO in 1998  (see Table 4-l).(5)
Detection limits were reported as sample-specific
                                                   33

-------
detection limits (SDLs). SDLs were determined from
2.5 times the noise in the chromatogram for D/F and 3.0
times the noise for PCBs, converted to an area, and then
converted to a concentration using the same calculation
procedure as for detected peaks. Any value that met all
quantification criteria (>SDL and isotope ratio) was
reported as a concentration. A "J" flag was applied to
any reported value between the SDL and the lowest level
calibration. The concentration of any detected congener
that did not meet all quantification criteria (such as
isotope ratio or peak shape) was reported but given a "K"
flag to indicate estimated maximum possible
concentration (EMPC).(8) TEQs were reported in two
ways to cover the range of possible TEQ values:

(1)    All nondetect and EMPC values were assigned a
       zero concentration in the TEQ calculation.

(2)    Nondetects were assigned a concentration of one-
       half the SDL. EMPCs were assigned a value
       equal to the EMPC.
In both cases, any total TEQ value that had 10%
contribution or more from J-flagged or K-flagged data
was flagged as J or K (or both) as appropriate.

TEQs were calculated both ways for all samples.  For
TEQD/F, 63% of the samples had the same TEQ value
based on the two different calculation methods, and the
average RPD was 8% (median = 0%). For TEQPCB, 65%
of the samples had the same TEQ value based on the two
different calculation methods, and the average RPD was
9% (median = 0%). Because overall there were little
differences between the two calculation methods, as
presented in Appendix D, TEQ values calculated by
option # 1 were used in comparison with the developer
technologies. On a case-by-case basis, developer results
were compared to TEQs calculated by option #2 above,
but no significant differences in comparability results
were observed so no additional data analysis results using
these TEQ values were presented.
                                                   34

-------
                                              Chapter 6
                         Assessment of Reference Method Data Quality
Ensuring reference method data quality is of paramount
importance to accurately assessing and evaluating each
of the innovative technologies. To ensure that the
reference method has generated accurate, defensible
data, a quality systems/technical audit of the reference
laboratory was performed during analysis of demon-
stration samples after the first batch of demonstration
sample analyses was complete.  The quality systems/
technical audit evaluated implementation of the
demonstration plan.  In addition, a full data package was
prepared by the reference laboratory for each sample
batch for both dioxin and dioxin-like PCB analyses.
Each data package was reviewed by both a QA specialist
and technical personnel with expertise in the reference
methods for agreement with the reference method as
described in the demonstration plan. Any issues
identified during the quality systems/technical audit and
the data package reviews were addressed by the
reference laboratory prior to acceptance of the data. In
this section, the reference laboratory performance on the
QC parameters is evaluated. In addition, the reference
data were statistically evaluated for the demonstration
primary objectives of accuracy and precision.

6.1    QA Audits
A quality systems/technical audit was conducted at the
reference laboratory, AXYS Analytical Services, Ltd. by
Battelle auditors on May 26, 2004, during the analysis of
demonstration samples. The purpose of the audit was to
verify AXYS  compliance with its internal quality system
and the D/QAPP.(2) The scope specifically included a
review of dioxin and PCB congener sample processing,
analysis, and data reduction; sample receipt, handling,
and tracking;  supporting laboratory systems; and
followup to observations and findings identified during
the independent laboratory assessment conducted by
Battelle on February 11, 2004, prior to contract award.
Checklists were prepared to guide the audit, which
consisted of a review of laboratory records and
documents, staff interviews, and direct observation.

The AXYS quality system is documented in a
comprehensive Quality Assurance/Quality Control
(QA/QC) manual and detailed SOPs. No major
problems or issues were noted during the audit. Two
findings were identified, one related to a backlog of
unfiled custody records and the  other related to the need
for performance criteria for the DB-1 column used for
the analysis of PCB congeners by HRMS. Both issues
were addressed satisfactorily by AXYS after the audit.
One laboratory practice that required procedural
modification was identified: the laboratory did not
subject all QC samples to the most rigorous cleanup
procedures that might be required for individual samples
within a batch. The AXYS management team agreed that
this procedure was incorrect.  As corrective action, the
QA manager provided written instructions regarding
cleanup of the quality control samples to the staff, and
the laboratory manager conducted follow up discussion
with the staff. Other isolated issues noted by the auditors
did not reflect systemic problems and were typical of
analytical laboratories  (e.g., occasional documentation
lapses or an untrackable balance weight).

The audit confirmed that the laboratory procedures
conformed to the SOPs and D/QAPP and that the  quality
system was implemented effectively. Samples were
processed and analyzed according to the  laboratory
SOPs and D/QAPP using the Soxhlet Dean Stark
extraction method. No substantial deviations were
noted. The audit verified the traceability of samples
within the laboratory, as well as the traceability of
standards, reagents, and solvents used in preparation,
and that the purity and reliability of the latter materials
were demonstrated through documented  quality checks.
                                                    35

-------
In addition, the audit confirmed that analytical
instruments and equipment were maintained and
calibrated according to manufacturers' specifications and
laboratory SOPs. Analytical staff members were
knowledgeable in their areas of expertise. QC samples
were processed and analyzed with each batch of
authentic samples as specified by the D/QAPP. QA/QC
procedures were implemented effectively, and corrective
action was taken to address specific QC failures. Data
verification, reporting, and validation procedures were
found to be rigorous and sufficient to ensure the
accuracy of the reported data. The auditors concluded
that AXYS is in compliance with the D/QAPP and its
SOPs, and that the data generated at the laboratory are of
sufficient and known quality to be used as a reference
method for this project.

In addition, each data package was reviewed by both a
QA specialist and technical personnel with expertise in
the reference methods for agreement with the reference
method as described in the demonstration plan. Check-
lists were prepared to guide the  data package review.
This review included an evaluation of data package
documentation such as chain-of-custody (COC) and
record completeness, adherence to method prescribed
holding times and storage conditions, standard spiking
concentrations, initial and continuing calibrations meet-
ing established criteria, GC column performance, HRMS
instrument resolution, method blanks, lab control spikes
(ongoing precision and recovery samples), sample
duplicates, internal standard recovery, transcription of
raw data into the final data spreadsheets, calculation of
TEQs, and data flag accuracy. Any issues identified
during the data package reviews were addressed by the
reference laboratory prior to acceptance of the data. All
of the audit reports and responses are included in the
DER.

6.2    QC Results
Each data package was reviewed for agreement with the
reference method as  described in the demonstration plan.
This section summarizes the evaluation of the reference
method quality control data.

6.2.1  Holding Times and Storage Conditions
All demonstration samples were stored frozen (< -10°C)
upon receipt and were analyzed within the method
holding time of one year.
6.2.2  Chain of Custody
All sample identifications were tracked from sample
login to preparation of record sheets, to instrument
analysis sheets, to the final  report summary sheets and
found to be consistent throughout.  One COC with an
incomplete signature and one discrepancy in date of
receipt between the COC and sample log-in were
identified during the Battelle audit and were corrected
before the data packages with these affected items were
accepted as final.

6.2.3  Standard Concentrations
The concentration of all calibration and spiking
standards was verified.

6.2.4  Initial and Continuing Calibration
All initial calibrations met the criteria for response factor
RSD and minimal signal-to-noise ratio requirements for
the lowest calibration point.

Continuing calibrations were performed at the beginning
and end of every 12-hour analysis period with one minor
exception for dioxin/furan sample batch WG13551,
which contained five samples from Environmental Site
#1 (North Carolina) and 12 samples from Environmental
Site #5 (Winona Post).  On one analysis day, a high-
level sample analyzed just prior to the ending calibration
verification caused the verification to fail. In this
instance, the verification was repeated just outside of the
12-hour period.  The repeat calibration verification met
the acceptance criteria and was considered to show
acceptable instrument performance in the preceding
analytical period; therefore, the data were accepted.

Continuing calibration results were within the criteria
stated in Table 9-2 (D/F) and Table 9-4 (PCB) of the
D/QAPP, with one exception.  For PCB sample batch
WG12108, which contained nine samples from
Environmental Site #3 (Newark Bay) and 12 samples
from Environmental Site #4 (Raritan Bay), isotopically
labeled PCB 169 was above the acceptable range during
one calibration verification on May 15, 2004.  The
acceptance range included in the D/QAPP is tighter than
the acceptance range in  Method 1668A Table  6.
Because the result for labeled PCB 169 was within the
Method 1668A acceptance  limits, the data were
accepted.
                                                    36

-------
The minimum signal-to-noise criteria for analytes in the
calibration verification solution were met in all
instances.

6.2.5  Column Performance and Instrument
       Resolution
Column performance was checked at the beginning of
each 12-hour analytical period and met method criteria.

Instrument resolution was documented at the beginning
and end of each 12-hour period with one exception. In
PCB sample batch WG13554, which contained five PE
samples and 15 extract samples, on one analysis day
(September 17, 2004), the ending resolution
documentation was conducted at 12 hours and
54 minutes.  However,  as this resolution documentation
met all criteria, it was considered representative of
acceptable instrument performance during the analytical
period, and the data were accepted.

6.2.6  Method Blanks
Method blanks were analyzed with each sample batch to
verify that laboratory procedures did not introduce
significant contamination. A summary of the method
blank data is presented in Appendix C. There were many
instances for both D/F and PCB data where analyte
concentrations in the method blank exceeded the target
criteria in the D/QAPP. Samples from this
demonstration, which had very high D/F and PCB
concentrations, contributed to the difficulty in achieving
method blank criteria in spite of steps the reference
laboratory took to minimize contamination (such as
proofing the glassware  before use in each analytical
batch). In many instances, the concentrations of D/F and
PCBs in the samples exceeded 20 times the
concentrations in the blanks.  For all instances, the
sample results were unaffected because the method
blank TEQ concentration was compared to the sample
TEQ concentrations to  ensure that background
contamination did not significantly impact sample
results.

6.2.7  Internal Standard Recovery
Internal standard recoveries were generally within the
D/QAPP criteria. D/QAPP criteria were tighter than the
standard EPA method criteria; in instances where
internal standard recoveries were outside of the D/QAPP
criteria, but within the standard EPA method criteria,
results were accepted. In several instances, the dioxin
cleanup standard recoveries were affected by
interferences. As the cleanup standard is not used for
quantification of native analytes, these data were
accepted. Any samples affected by internal standard
recoveries outside of the D/QAPP and outside of the
EPA method criteria were evaluated for possible impact
on total TEQ and for comparability with replicates
processed during the program before being accepted.

6.2.8  Laboratory Control Spikes
One laboratory control spike (ongoing precision and
recovery sample), which consisted of native analytes
spiked into a reference matrix (sand), was processed
with each analytical batch to assess accuracy.  Recovery
of spiked analytes was within the D/QAPP criteria in
Table 9-2 for all analytes in all laboratory control spike
samples.

6.2.9  Sample Batch Duplicates
A summary of the duplicate data is presented in
Appendix C. One sample was prepared in duplicate in
most sample batches; four batches were reported without
a duplicate. Three of 14 dioxin sample batches and 5 of
14 PCB sample batches did not meet criteria of <20%
RPD between duplicates. Data where duplicates did not
meet D/QAPP criteria were evaluated on an individual
basis.

6.3    Evaluation of Primary Objective PI:
       Accuracy
Accuracy was assessed through the analysis of PE
samples consisting of certified standard reference
materials, certified spikes, and certified blanks.  A
summary of reference method percent recovery (R)
values is presented in Table 6-1. The Rvalues are
presented for TEQPCB, TEQD/F, and total TEQ. The
minimum, maximum, mean, and median R values are
presented for each set of TEQ results. The reference
method values were in best agreement with the certified
values for the TEQPCB results, with a mean R value of
96%. The mean R values for TEQD/F and total TEQ were
                                                   37

-------
Table 6-1.  Objective PI Accuracy - Percent Recovery
PE Sample
ID
1
2
3
4
5
6
7
8
9
10
11
12
PE Sample
Description
Cambridge 5183
LCG CRM-529
Wellington WMS-01
Cambridge 5 1 84
NIST 1944
ERA TCDD 10
ERA TCDD 30
ERA PAH
ERAPCB 100
ERAPCB 10000
ERA Aroclor
ERA Blank
All PE Samples
% Recovery
TEQPrR
81
100
93
120
102
NA
NA
NA
96
95
82
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
81
120
96
96
TEQn;F
111
106
106
106
91
79
77
NA
NA
NA
324
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
77
324
106
125
Total TEQ
94
106
105
118
93
79
77
NA
95
95
83
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
10
77
118
94
94
 NA = not applicable; insufficient data were reported to determine R or the sample was not spiked with those analytes.
125% and 94%, respectively.  The mean and median R
values for the TEQPCB and total TEQ were identical.  The
mean and median R values for TEQD/F were not similar
and were largely influenced by the TEQD/F recovery for
ERA Aroclor of 324%. The ERA Aroclor-certified
TEQD/F values were based on TCDD and TCDF only,
whereas the reference method TEQD/F values were based
on contributions from all 2,3,7,8-substituted D/F
analytes. The Rvalues presented in Table 6-1 indicate
that the reference method reported data that were on
average between 94 and 125% of the certified values of
the PE samples.

The effect of known interferences on reference method
TEQs is listed in Table 6-2. D/F and PCB TEQs were
not affected by PAH as evidenced through the analysis
of ERA PAH standard reference material. D/F and PCB
TEQs were not affected by each other as evidenced by
spikes that contained only one set of analytes having
negligible influence on the TEQ of the other analyte set.

6.4    Evaluation of Primary Objective P2:
       Precision
The 209 samples included in the demonstration
consisted of replicates of 49 discrete samples. There
were four replicates of each sample except for PE
sample Cambridge 5183 (7 replicates), ERA blank
reference material (8 replicates), Wellington WMS-01
standard reference material (7 replicates), and 0.5 pg/mL
2,3,7,8-TCDD extract (7 replicates).  Reference method
data were obtained for all 209 samples; however, TEQD/F
and total TEQ data for samples Ref 197 (ERA PCB 100)
and Ref 202 (LCG CRM-529) were omitted as outliers
as it appeared that these two samples were switched
during preparation after observing results of the
replicates and evaluating the congener profiles of these
two samples.

A summary of the reference method replicate RSD
values is presented in Tables 6-3a and 6-3b. The RSD
values are presented for TEQPCB, TEQD/F, and total TEQ
in Table 6-3a, and a summary by sample type is
presented in Table 6-3b, along with the minimum
R value, the maximum R value, and the mean R value
for each set of TEQ results and sample types. In terms of
sample type, the reference method had the most precise
data for the environmental sample TEQD/F results, with a
mean RSD value of 12%. This was followed closely by
environmental sample TEQPCB and total TEQ results,
which both had mean RSDs of 13%.  In terms of TEQ
values, the reference method had the  most precise data
for the total TEQ values, with a mean overall RSD of
13%.  Overall RSD values ranged from 1% to 119%.
Precision was significantly worse for certified blanks
and blank samples (e.g., samples that contained spikes of
only one analyte set and were blank for the other
                                                   38

-------
Table 6-2.  Evaluation of Interferences
PE Material with Known Interference
ERA PAH
ERAPCB 100
ERAPCB 10000
ERATCDD 10
ERA TCDD 30
Mean TEQ (pg/g)
0.195(D/F+PCB)
0.073 (D/F)
0.220 (D/F)
0.025 (PCB)
0.036 (PCB)
analytes) as might be expected due to the very low levels
detected in these samples.

6.5    Comparability to Characterization Data
To assess comparability, reference laboratory D/F data
for environmental samples were plotted against the
characterization data that was generated by Battelle prior
to the demonstration. Characterization data were
obtained as part of the process to verify homogenization
of candidate soil and sediment samples as described in
Chapter 5 and reported in Table 4-5. It should be noted
that second column confirmations of 2,3,7,8-TCDF
results were not performed during characterization
analyses; therefore, characterization TEQs are biased
high for samples where  a large concentrations of non-
2,3,7,8-TCDF coeluted with 2,3,7,8-TCDF on the DB-5
column.  Characterization samples also were not
rigorously evaluated to ensure that high concentration
extracts were diluted sufficiently so that all peak areas
were less than the peak areas of the highest calibration
standard. In spite of these differences between reference
and characterization analyses, the results had fairly high
correlation (R2 = 0.9899) as demonstrated in Figure 6-1.

6.6    Performance  Summary
This section provides a performance summary of the
reference method by summarizing the evaluation of the
applicable primary objectives of this demonstration
(accuracy, precision, and cost) in Table 6-4. A total of
209 samples was analyzed for seventeen
2,3,7,8-substituted D/F and 12 PCBs over an eight-
month time frame (April 27 to December 20, 2004).
Valid results were obtained for all 209 PCB analyses,
while 207 valid results were obtained for D/F. The D/F
and total TEQ results for samples Ref 197 (ERA PCB
100) and Ref 202 (LCG CRM-529) were omitted as
outliers because it appeared that these two samples were
switched during preparation after observing results of
the replicates and evaluating the congener profiles of
these two samples. The demonstration sample set
provided particular challenges to the reference
laboratory in that there was a considerable range of
sample concentrations for D/F and PCB. This caused
some difficulty in striving for low MDLs in the presence
of high-level samples. The range of concentrations in the
demonstration sample set also required the laboratory to
modify standard procedures, which contributed to
increased cost and turnaround time delay. For example,
an automated sample cleanup system could not be used
due to carryover from high- level samples; instead, more
labor-intensive manual cleanup procedures were used;
glassware required extra cleaning and proofing before
being reused; cleanup columns sometimes became
overloaded from  interferences and high-level samples,
causing low recoveries so that samples had to be
re-extracted or cleanup fractions had to be analyzed for
the lost analytes;  and method blanks often showed trace
levels of contamination, triggering the repeat of low-
level samples.

Because the reference method was not to be altered
significantly for this demonstration, the reference
laboratory was limited in its ability to adapt the trace-
level analysis to higher level samples. In spite of these
challenges, the quality of the data generated met the
project goals. The main effect of the  difficulties
associated with these samples was on schedule and cost.
                                                    39

-------
Table 6-3a. Objective P2 Precision - Relative Standard Deviation
Sample Type
Environmental
Extract
PE
Sample ID
Brunswick #1
Brunswick #2
Brunswick #3
Midland #1
Midland #2
Midland #3
Midland #4
NC PCB Site #1
NC PCB Site #2
NC PCB Site #3
Newark B ay #1
Newark Bay #2
Newark Bay #3
Newark Bay #4
RaritanBay #1
Raritan Bay #2
Raritan Bay #3
Saginaw River #1
Saginaw River #2
Saginaw River #3
Solutia #1
Solutia #2
Solutia #3
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #3
Titta. River Sed #1
Titta. River Sed #2
Titta. River Sed #3
WinonaPost#l
Winona Post #2
Winona Post #3
Envir Extract #1
Envir Extract #2
Spike #1
Spike #2
Spike #3
Cambridge 5183
Cambridge 5184
ERA Aroclor
ERA Blank
ERA PAH
ERA PCB 100
ERA PCB 10000
ERATCDD 10
ERA TCDD 30
LCG CRM-529
NIST 1944
Wellington WMS-01
RSD for TEQPCB
(%)
8
o
3
5
4
10
4
77
21
21
25
7
2
6
1
6
o
3
o
3
8
7
60
36
4
11
7
9
12
19
14
13
13
4
9
71
83
119
1
4
7
o
J
44
62
83
4
7
60
39
14
4
5
RSD for TEQD/F
(%)
6
16
8
9
6
6
9
15
2
12
28
22
6
12
5
2
5
25
19
19
13
7
5
6
10
26
27
37
9
2
9
4
50
2
6
5
13
19
4
6
65
27
65 a
91
5
6
2a
9
o
J
RSD for Total TEQ
(%)
6
16
8
9
6
6
10
20
21
24
25
20
6
11
4
1
4
23
18
19
13
7
5
5
10
26
26
37
8
2
9
4
50
2
9
3
4
9
2
43
61
30
3
7
5
6
1
7
3
       Does not include sample excluded due to sample preparation error.
                                                     40

-------
Table 6-3b. Objective P2 Precision - Relative Standard Deviation (By Sample Type)
Sample Type
Environmental
Extract
PE
Overall
RSD for TEQprR (%)
N
32
5
12
49
MIN
1
1
3
1
MAX
77
119
83
119
MED
8
71
11
8
MEAN
13
56
28
21
RSD for TEQn;F (%)
N
32
5
12
49
MIN
2
2
2
2
MAX
37
50
91
91
MED
9
6
7
9
MEAN
12
15
25
16
RSD for Total TEQ (%)
N
32
5
12
49
MIN
1
2
1
1
MAX
37
50
61
61
MED
10
4
7
8
MEAN
13
14
15
13
                 16000 -,
                        0
  5000         10000         15000
Characterization Data (TEQ D/F pg/g)
20000
        Figure 6-1. Comparison of reference laboratory and characterization D/F data for environmental
        samples.
Table 6-4. Reference Method Performance Summary - Primary Objectives
Objective
P 1 : Accuracy
P2: Precision
P7: Cost
Performance
Statistic
Number of data points
Median Recovery (%)
Mean Recovery (%)
Number of data points
Median RSD (%)
Mean RSD (%)
TEQPCB
8
96
96
49
8
21
TEQD/F
8
106
125
49
9
16
Total TEQ
10
94
94
49
8
13
209 samples were analyzed for 17 D/F and 12 PCBs. Total cost was $398,029.
D/F cost was $213,580 ($1,022 per sample) and PCB cost was $184,449 ($883 per sample).
                                               41

-------
                                              Chapter 7
         Performance of Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit
7.1    Evaluation of Dioxin ELISA Kit
       Performance
The Wako Pure Chemical Industries, Ltd. Dioxin ELISA
Kit Wako (for environmental) is an immunoassay
technology that reports total dioxin/furan concentration
in a sample. It should be noted that the results generated
by this technology may not directly correlate to HRMS
TEQD/F in all cases because it is known that the congener
responses and cross-reactivity of the kit are not identical
to the TEFs that are used to convert congener HRMS
concentration values to TEQD/F. The effect of
cross-reactivities may contribute to this technology's
reporting results that are biased high or low compared to
HRMS TEQD/F results.  Therefore, this technology
should not be viewed as producing an equivalent
measurement value to HRMS TEQD/F but as a screening
value to approximate HRMS TEQD/F concentration. It
has been suggested that correlation between the Wako
and HRMS TEQ could be improved by first
characterizing a site and calibrating the Wako results to
HRMS results. Subsequent analysis using the Wako kit
for samples obtained from this site may then show  better
correlation with the HRMS TEQ result. This approach
was not evaluated during this demonstration.

The following sections describe the performance of the
Dioxin ELISA Kit Wako (for environmental), according
to the primary objectives for this demonstration.  The
developer and reference laboratory data are presented in
Appendix D. The statistical methods used to evaluate the
primary objectives are described in Section 4.7. Detailed
data evaluation records can be found in the DER.

7.1.1  Evaluation of Primary Objective PI:
       Accuracy
A summary of the Dioxin ELISA Kit Wako (for
environmental) percent recovery (R) values is presented
in Table 7-1. The description of how Rvalues were
calculated is presented in Section 4.7.1. The Rvalues are
calculated by comparing Wako's 2,3,7,8-TCDD EQ
values with the certified TEQD/F values. The minimum R
value, the maximum R value, median R value, and the
mean Rvalue were 10%, 1,574%, 253%, and 443%,
respectively. As presented in Table 7-1, only five R
values could be calculated from the 12 PE samples
because many of the PE samples were reported as
nondetects by the Wako kit. Reporting these samples as
nondetects in some cases was accurate. For example,
ERA PCB 100, ERA PCB 10000, ERA Aroclor, and

 Table 7-1. Objective PI Accuracy - Percent Recovery
PE Sample
ID
1
2
3
4
5
6
7
8
9
10
11
12
PE Sample
Description
Cambridge 5 183 *
LCG CRM-529
Wellington
WMS-01
Cambridge 5184
NIST 1944
ERATCDD 10*
ERA TCDD 30 *
ERA PAH *
ERA PCB 100*
ERA PCB 10000*
ERA Aroclor *
ERA Blank *
All Performance Samples
% Recovery
1574
10
253
299
80
NA
NA
NA
NA
NA
NA
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
5
10
1574
253
443
 NA = not applicable; insufficient data were reported to determine R
 or the sample was not spiked with those analytes
 * No D/Fs were in the sample or the concentrations of D/Fs were
   outside of Wako's reporting range
                                                   42

-------
ERA PAH contained only spiked PCBs or PAHs, so
these samples should not have been detections for D/F.
The lack of reported data for ERA TCDD 10 (certified
TEQD/F = 11 pg/g) and ERA TCDD 30 (certified TEQD/F
= 33 pg/g) indicates that these samples contained D/F
concentrations that were below the detection capabilities
of the kit. All four of the ERA TCDD 10 replicates were
reported accurately by Wako as < reporting limits.

7.1.2   Evaluation of Primary Objective P2:
        Precision
A summary of the Dioxin ELISA Kit Wako (for
environmental) RSD values is presented in Tables 7-2a
and 7-2b. The description of how RSD values were
calculated is presented in Section 4.7.2. The RSD values
are presented in Table 7-2a for each sample where Wako
reported values for three or more of the replicate
samples. A summary by sample type is presented in
Table 7-2b, along with the minimum RSD value, the
maximum RSD value, the median RSD value, and the
mean RSD value. Low RSD values (< 20 %) indicate
high precision. In terms of sample type, the Dioxin
ELISA Kit Wako (for environmental) values had the
most precise data for the extract results, with a mean
RSD value of 26%, but there were only two RSD values
averaged in the extract value while the environmental
and PE sample sets had considerably more values to
average (16 and 6, respectively). Overall RSD values
ranged from 11% to 145%, with a mean RSD of 62%.

7.1.3 Evaluation of Primary Objective P3:
      Comparability
The description of the statistical analyses used in the
comparability evaluations are described in Section 4.7.3.
The comparability of the Wako and reference laboratory
data was assessed by calculating RPD for the TEQD/F
values, as presented in Table 7-3. The summary statistics
presented in Table 7-3 provide an overall assessment of
the RPD values. The Wako values agreed best with the
reference laboratory D/F measurements for extract
samples, with a median RPD value of-1%. The median
RPD values for the environmental, PE, and overall
TEQD/F values were 52%, -37%, and 34%, respectively,
with minimum and maximum overall values around
minus 200% and positive 200%, respectively.  RPD
values between positive and negative 25% indicate good
agreement between the reference laboratory and
developer values.  Of the TEQD/F samples, 19 (18%) of
the samples had RPD values between positive and
negative 25%.

The agreement when sorting the developer and reference
laboratory results for TEQD/F into four intervals
(< 50 pg/g, 50 to 500 pg/g, 500 to 5,000 pg/g, and
> 5,000 pg/g) are described in Table 7-4. The agreement
between the developer and reference laboratory data was
62%. Note that, as described in Section 4.7.3, results that
were reported as semiquantitative results were counted
as in agreement if the reference laboratory data was
within that interval. For example, a reference laboratory
result reported as 3,400 pg/g TEQD/F was counted as in
agreement with a Wako result reported as > 2,967 pg/g
2,3,7,8 TCDD EQ, even though the absolute quantitative
values would be in different ranges.

Interval reporting addresses the question of whether a
value reported by the technology would result in the
same decision of what to do next with the sample if it
was analyzed by the reference method. This interval
assessment table indicates that 38% of the time, the
Wako result would have indicated a different interval
(and therefore a different decision to be made about the
sample) than if it was analyzed by the reference
laboratory, based on the concentration ranges chosen for
these intervals.

The ERA blank samples contained levels of D/Fs that
were below the reporting limits of the developer
technologies (see Table 4-4 certified value: 0.046 pg/g
TEQD/F). Wako reported concentrations were compared
with the reference laboratory reported data for these
samples in Table 7-5. Wako reported one of the eight
values as a detection (114.00 pg/g) and seven results
were reported as nondetects. As such, 88% agreed with
the reference laboratory results. It should be noted that
the reference laboratory data presented in Table 7-5 were
calculated with nondetect values assigned a zero
concentration.  When applying the TEQ calculation
method of assigning nondetects with a concentration of
one-half the SDL, the reference data increased, but the
conclusions regarding agreement with the developer data
remain the same.
                                                   43

-------
Table 7-2a. Objective P2 Precision - Relative Standard Deviation (All Samples)
Sample Type
Environmental
Extracts
Sample ID
Brunswick #1
Brunswick #2
Brunswick #3
Midland #1
Midland #2
Midland #3
Midland #4 *
NC PCB Site #1 *
NC PCB Site #2 *
NC PCB Site #3 *
Newark Bay #1 *
Newark Bay #2
Newark Bay #3
Newark Bay #4 *
RantanBay#l *
Raritan Bay #2 *
Raritan Bay #3 *
Saginaw River #1
Saginaw River #2
Saginaw River #3
Solutia#l
Solutia #2
Solutia #3
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #3
Titta. River Sed #1 *
Titta. River Sed #2
Titta. River Sed #3 *
WinonaPost#l
Winona Post #2
Winona Post #3
Envir. Extract #1
Envir. Extract #2
Spike #1 *
Spike #2
Spike #3 *
Relative Standard Deviation (%)a
NA
NA
74
60
35
65
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
88
34
43
48
60
108
NA
16
74
NA
105
NA
13
60
70
41
NA
NA
11
NA
                                                  44

-------
Sample Type
PE
Sample ID
Cambridge 5 183 *
Cambridge 5184
ERA Aroclor *
ERA Blank *
ERA PAH*
ERAPCB 100*
ERAPCB 10000*
ERATCDD 10*
ERA TCDD 30 *
LCG CRM-529
NIST 1944
Wellington WMS-01
Relative Standard Deviation (%)a
47
145
NA
NA
NA
NA
82
NA
NA
64
54
100
 NA = not applicable (i.e., one or more of the replicates were reported as a nondetect value).
 * No D/Fs were in the sample or the concentrations of D/Fs were outside of Wako's reporting range.
 a Three or four replicate results were used to calculate the RSD values.
Table 7-2b. Objective P2 Precision - Relative Standard Deviation (By Sample Type)
Sample
Type
Environ.
Extract
PE
Overall
Relative Standard Deviation (%)
No.
16
2
6
24
MIN
13
11
47
11
MAX
108
41
145
145
MED
60
26
73
60
MEAN
60
26
82
62
Table 7-3. Objective P3 - Comparability Summary Statistics of RPD
Sample Type
Environmental
Extract
PE
Overall
TEQn/F RPD (%)
N
76
10
20
106
MIN
-195
-198
-185
-198
MAX
198
98
180
198
MEDIAN
52
-1
-37
34
Table 7-4. Objective P3 - Comparability Using Interval Assessment
Agreement
Number Agree
% Agree
Number Disagree
% Disagree
TEQD/F
127
62
78
38
                                                      45

-------
Table 7-5. Objective P3 - Comparability for Blank Samples
Replicate
1
2
3
4
5
6
7
8
% agree
Wako
(pg/g 2,3,7,8-TCDD EQ)
<17.80
< 34.30
< 17.80
< 34.30
< 26.30
114.00
< 45.30
< 26.30
RefLab"
(pg/g TEQD/F)
J0.0942 b
J0.0728
J0.237
JO. 307
JO. 113
J0.0524
J0.211
J0.0692
Agree?
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
88% (7 of 8)
a All nondetect and EMPC values were assigned a zero concentration for the reference laboratory TEQ calculation.
b J flag was applied to any reported value between the SDL and the lowest level calibration.
7.1.4  Evaluation of Primary Objective P4:
       Estimated Method Detection Limit
It should be noted that these detection limit calculations
did not strictly follow the definition as presented in the
Code of Federal Regulations (i.e., t-value with 6 degrees
of freedom). PE samples in the precisely appropriate
range for evaluation of this technology's detection limit
were not available. Since detections were not reported
for all seven replicate samples, the degrees of freedom
and statistical power of the analysis were reduced
accordingly. The only approach that led to the use of the
definitional calculation with 6 degrees of freedom
required special treatment of the nondetect values (i.e.,
assigning values that were one-half or equal to the
nondetect value). However, these calculations are
provided as estimated method detection limits (EMDLs)
to give the reader a sense of the detection capabilities of
the technology.

The EMDL of the Dioxin ELISA Kit Wako  (for
environmental) was determined by assessing the values
that Wako reported for three PE samples:
Wellington WMS-01, Cambridge 5183, and Extract
Spike # 1. The data from the Wellington sample were
evaluated, but determined to have a D/F concentration
(62 pg/g TEQD/F) that was higher than appropriate for
evaluation of EMDL; they were not included.  As shown
in Table 7-6, because some of the results for the samples
were nondetects, the TEQD/F EMDL was calculated in
three ways: by setting nondetect values to zero, by
setting nondetect values to half of the reporting limit
value, and by setting nondetect values to the reporting
limit value itself. For the Extract Spike # 1 samples,
Wako reported only two samples as actual values (as
opposed to nondetects); therefore, an EMDL could not
be calculated for those samples by setting nondetect
values to zero. One replicate for Cambridge 5183 (Wako
115) was reported as > 3,167 pg/g 2,3,7,8-TCDD EQ;
that value was excluded as an outlier. The MDLs for the
Cambridge 5183 and Extract Spike #1 samples ranged
from 83 to 201 pg/g 2,3,7,8-TCDD EQ. The detection
limit reported by Wako in the demonstration plan was
20 pg/g 2,3,7,8-TCDD EQ. PE samples with TEQ
concentrations in the precisely appropriate range for
evaluation of this technology's detection limit were not
available, so these calculated values should  be
considered a rough estimate.

 7.1.5 Evaluation of Primary Objective P5: False
       Positive/False Negative Results
The description of false positive/false negative
calculations  is presented in Section 4.7.5. The summary
of false positive/false negative results is presented in
Table 7-7. In this evaluation, Wako results,  reported in
pg/g 2,3,7,8-TCDD EQ, were compared to the reference
laboratory's  results reported in pg/g TEQD/F. Wako
reported 20 false positive (10%) and 26 false negative
(13%) results, relative to the reference laboratory's
reporting of samples above and below 20 pg/g TEQD/F.
For samples  around 50 pg/g TEQD/F, Wako's rate of false
positives was the same (10%); but, there were fewer
false negatives. As described in Section 7.1.4, the
EMDLs were estimated to be 83 to 201 pg/g 2,3,7,8-
TCDD EQ; but, because the concentration of PE samples
                                                    46

-------
Table 7-6.  Objective P4 - Estimated Method Detection Limit
Statistic
Degrees of Freedom
SD(pg/g 2,3,7,8-
TCDD EQ)
EMDL (pg/g
2,3,7,8-TCDD EQ)
Cambridge 5183a
Nondetect
values set to
zero
2
29
201
Nondetect values
set to 1A value
5
30
102
Nondetect
values set to
reported value
5
25
84
Extract Spike #1
Nondetect values
set to zero
NA
Nondetect
values set to Vz
value
6
32
102
Nondetect
values set to
reported value
6
26
83
aExcludes Wako 115, which was reported as > 3,167 pg/g.
NA = not available (insufficient data to calculate).
Table 7-7.  Objective P5 - False Positive/False
Negative Results
Rate
False Positive
False Negative
TEQ:
20 pg/g
10%
(20outof207)a
13%
(26 out of 207)
)/F
50 pg/g
10%
(21 of 207)
8%
(16 of 207)
 a 207 sample results were evaluated instead of 209 due to two
   reference laboratory data points which were discarded due to
   sample preparation errors. See Section 6.4.

were not the most appropriate for evaluation of this
technology's MDL, the calculated EMDLs should be
considered rough estimates. Higher EMDLs are the
result of less precise data at low concentration levels.
The false positive and false negative results were
calculated using the entire range of concentrations and
demonstrate that Wako's results agreed with laboratory
results, relative to the screening levels, for the majority
(-80%) of the samples. As such, this evaluation suggests
that the Wako kit could be an effective screening tool for
determining sample results above and below 20 pg/g
TEQD/F and even more effective as a screen for samples
above and below 50 pg/g TEQD/F.

7.1.6  Evaluation of Primary Objective P6:
       Matrix Effects
Six types of potential matrix effects were investigated:
(1) measurement location (field vs. laboratory
measurements), (2) matrix type (soil vs. sediment vs.
extract), (3) sample type (PE vs. environmental vs.
extract), (4) PAH concentration, (5) environmental site,
and (6) known interferences. A summary of the matrix
effects is provided in the bullets below, followed by a
detailed discussion:

    Measurement location: not evaluated (all samples
    analyzed on-site)
•   Matrix type: none
•   Sample type: none
•   PAH concentration: none
•   Environmental site: none
•   Known interferences: slight

In Table 7-8, precision summary values are presented by
matrix type.  A one-way ANOVA model was used to test
the effect of soil vs. sediment vs. extract on RSD. These
tests showed no significant effect on RSD. In Table 7-9,
precision summary values are presented by PAH
concentrations for environmental samples only. A one-
way ANOVA model was used to test the effect of PAH
concentration on RSD. These tests showed no effect.
The summary of RSD values segregated by sample type
is presented in Table 7-2b.  A one-way ANOVA model
was used to test the effect of sample type (PE vs.
environmental vs. extract) on RSD. These tests showed
no significant effect on RSD. The average RSD for the
extracts (26%) was considerably less than that of the
environmental samples (60%) and PE  samples (82%),
but there were only two RSD values averaged in the
extract value while the environmental  and PE sample
sets had considerably more values to average (16 and 6,
respectively). Based on the comparability results (RPD
values), Wako's results were not more or less
comparable for one particular environmental site,
suggesting that matrix effects were not dependent on
environmental site.
                                                    47

-------
Table 7-8.  Objective P6 - Matrix Effects Using RSD as a Description of Precision by Matrix Type
Matrix Type
Soil
Sediment
Extract
Overall
RSD for TEQD/F (%)
N
15
7
2
24
MIN
13
34
11
11
MAX
145
105
41
145
MED
63
71
26
62
MEAN
60
74
26
60
Table 7-9. Objective P6 - Matrix Effects Using RSD as a Description of Precision by PAH Concentration Levels
(Environmental Samples Only)
PAH Concentration Level (ng/g)
> 100,000
10,000-100,000
1,000-10,000
< 1,000
Overall (Environmental Samples Only)
RSD for TEQn/F (%)
N
1
3
7
5
16
MIN
74
13
34
16
13
MAX
74
70
108
105
108
MED
74
48
64
57
60
MEAN
74
60
60
48
60
The effect of known interferences was also assessed by
evaluating the results of PE materials that contained one
type of contaminant (PCBs or PAHs) but was not spiked
with D/Fs. Table 7-10 summarizes the detections for D/F
reported by Wako for these PE samples. For the ERA
PAH sample that contained only spiked PAHs, Wako
reported three values as nondetects and one as 38.5 pg/g
2,3,7,8-TCDD EQ.  The PCB-only spiked samples were
reported with two detections (mean 132 pg/g 2,3,7,8-
TCDD EQ) for the lower PCB spike and all four
replicates reported as D/F detects (mean 76 pg/g 2,3,7,8-
TCDD EQ) for the higher PCB spike.

7.1.7  Evaluation of Primary Objective P7:
       Technology Costs
Evaluation of this objective is fully described in
Chapter 8, Economic Analysis.

7.2    Observer  Report: Evaluation of
       Secondary Objectives
Wako's dioxin ELISA screening kit is an anti-dioxin
monoclonal antibody in an ELISA format. The following
procedural steps were observed: ASE of a 10-gram
sample aliquot (this procedure was omitted for samples
received as extracts), evaporation by nitrogen to dryness,
multilayer silica cleanup, concentration by large
TurboVap, phthalocyanine immobilized silica cleanup,
solvent exchange to 50 microliters of dimethyl sulfoxide
(DMSO) using nitrogen evaporation, addition of buffer
and antibody, incubation, addition of conjugate, addition
of reaction mixture to microplate wells, incubation, plate
washing, addition of color-developing solution and stop
solution, and analysis with a microplate reader at
450 nanometers. Each plate contained a seven-point
curve and  four QC samples prepared in the same manner
as the sample extracts. Each sample extract was
processed  in duplicate during the ELISA process. A total
of 40 samples could be analyzed with a 96-well plate.
Analysis was accomplished in a few seconds with the
operating software generating the curve and sample
results almost simultaneously. The software marked
samples that were out of range and these samples were
reanalyzed with a dilution of another 10-|iL aliquot.

The demonstration plan only stated the procedural steps
for the ELISA kit and did not address the extraction and
sample extract cleanup necessary for this technology.
Also, the operating procedure listed in the demonstration
plan did not match the operating procedure that was
observed.  The kit's instructions did match the
procedures observed including an extraction
                                                   48

-------
Table 7-10. Objective P6 - Matrix Effects Using
PE Materials





PE Sample
ERA PAH
ERAPCB 100
ERAPCB 10,000


%
Recovery
for Spiked
Analytes a
NA
NA
NA
Mean TEQ (pg/g
2,3,7,8-TCDD EQ)
Reported by Wako
for Analytes that
were not Spiked in
the PE Sample
38.5b
132 c
76
a NA = not applicable; percent recovery value could not be
  calculated.
b Three replicates were reported as nondetects.
0 Two replicates were reported as nondetects.

and cleanup flow diagram, as well as detailed
instructions for kit operation.

 7.2.1  Evaluation of Secondary Objective SI:
       Skill Level of Operator
From observation, a fully trained environmental
chemistry technician would be the minimum
qualifications of a person capable of meeting the kit's
extraction and cleanup needs. Wako's ELISA kit only
would require a user skilled in pipetting and computing
to generate accurate data. Wako prefers only
experienced personnel use the kit. The operators of the
technology at the demonstration were highly trained
laboratory staff. Each process in the system was
operated by personnel highly experienced with that
process. For example, all ASE extractions observed
were conducted by an ASE specialist. The ELISA
operation was also completed by someone with
significant ELISA experience. If one person was going
to perform all steps of the kit from sample extraction to
analysis, that person would have to be highly trained in
all areas of the sample process system.  A person would
also have to be very conscious of contamination
reduction in such areas as glassware washing and
equipment cleaning, considering several components of
the system are non-disposable and must be reused. The
personnel in the following tabulation were observed
processing demonstration samples for Wako:
Operator
Name
Tomohiro
Itoh
Sheldon
Henderson

Kimihiko
Sano
Nobukazu
Miyamoto
Minoru
Imokawa
Masako
Hayakawa
Job Title
Researcher

ASE Product
Marketing
Specialist
Experimental
Assistant
Researcher

Researcher

Experimental
Assistant
Education Level
Bachelor of
Engineering
Master of
Business
Administration
Master of
Pharmacy
Doctor of
Agriculture
Bachelor of
Agriculture
Bachelor of
Pharmacy
Years of
Experience
1 year

1 5 years


half year

3 years

3 years

half year

The provided instructions were in Japanese but were
translated for the observer by Wako. Once translated, the
instructions were very easy to follow. One step, which
might be better explained in the instructions, was the
time intervals between each sample receiving some
component of the reaction solutions. For example, the
analysts very carefully added a solution to a sample,
waited 10 seconds, and then added the same solution to
the next sample in line. A better explanation of this
practice and why it is necessary would be beneficial. The
sample process could be stopped, well sealed, and stored
at room temperature after extraction, after any
concentration, or after any column cleanup without any
adverse effects. In addition, the ELISA process contains
an 18- to 20-hour refrigerated incubation period.
According to the developers, once the extracts are in
DMSO, it is stable for 3 weeks. The recommended
extraction and extract cleanup requires large amounts of
hazardous solvents, large laboratory equipment,
significant amounts of dedicated bench-top space, and
fume hoods. Although the actual kit is simplistic, the
observed sample processing prior to kit use is very
complex. Wako only supplies the kit and does not
process samples for users.
7 7
   '.2  Evaluation of Secondary Objective S2:
       Health and Safety Aspects
The sample process generated a large amount of
hazardous waste including: flammable solvents, spent
reactants (corrosive impregnated silica, copper, sodium
                                                    49

-------
sulfate), spent samples, and glass (pipettes, disposable
cleanup columns, filters). The process also generated
small amounts of hazardous aqueous waste. A
considerable amount of lab trash was also generated
which included used personal protective equipment,
absorbent paper, and aluminum foil. A complete
inventory of the waste generated was performed after the
demonstration for the processing of 209 samples by
Wako and the following was recorded. None of the
containers was verified as full. Also, some items left as
waste could have been reused, but were not because of
the cost shipping them back to Japan.

(1) Three boxes  of used silica cleanup columns
(2) One 5-gallon container with broken glass and used
    phthalocyanine immobilized silica columns
(3) One large box with almost empty solvent bottles
    (some bottles contained approximately 100 mL of
    solvent)
(4) Four large boxes of used ASE collection vials
(5) Three gallons of spent organic solvent contained in a
    large container
(6) Eleven  5-gallon containers with spent organic
    solvents
(7) One 5-gallon container with used ASE filters
(8) Nine filled broken glass boxes that contained glass
    waste

The reader should be advised that, although no
difficulties were  encountered during this project,
difficulties could arise with disposal of dioxin-
contaminated waste.

7.2.3  Evaluation of Secondary Objective S3:
       Portability
The observed Dioxin ELISA Kit Wako (for
environmental) required an exacting sample extraction
and cleanup, similar to what is required for traditional
HRMS dioxin analysis. The extraction and sample
clean-up required one mobile lab fully outfitted with
fume hoods, water, nitrogen and/or purified compressed
air, and significant electrical power (-15 kVA power for
the mobile laboratory). The ELISA procedure required
less infrastructure, requiring only a refrigerator, watch,
pipettes, small bench space, microplate reader with
computer, and a controlled environment. Wako decided
to segregate the ELISA operations in another trailer,
which also accommodated material storage. This set up
took six Wako personnel approximately one and a half
days to become operational. Although the developer
used both a mobile lab and a trailer, the observer felt that
one mobile lab would be sufficient to house the sample
processing system if adequate storage cabinets were
provided. To accommodate the guidance given in the kit
instructions, one would have to bring a fully functional
environmental lab to the field. Once these
accommodations are met, the turnaround time of
samples would possibly be faster in the field than a
remote lab simply because no sample shipment is
needed. During the demonstration, Wako used four
nitrogen cylinders during sample preparation.

7.2.4   Evaluation of Secondary Objective S4:
        Throughput
All 209 samples were processed by Wako in the  field by
the eighth day. To accomplish this, two Dionex ASE
extractors were employed along with three various
nitrogen evaporators, fume hood, sink, refrigerator,
computer, microplate reader, at least 15 feet of bench
space for column cleanups, compressed filtered air
and/or compressed nitrogen, and hazardous and
nonhazardous material storage areas.  This was all
distributed between one mobile lab and one trailer. Five
personnel were needed to operate the process at this
speed. Wako lost 6 to 8 hours of sample processing due
to meetings, Visitor's Day, and nitrogen supply line
problems (leaks).

The developer stated it could process 40 samples in three
days using the process that was observed in the field.
This process involved 3.5 personnel for sample
extraction and preparation and 1.5 personnel for sample
analysis. The process is not amenable to an
inexperienced user. One person had multiple roles,
including performing the phthalocyanine immobilized
silica cleanup, solvent exchange into DMSO, and
software operator of the microplate reader. As part of the
observation, 23 samples received as extracts were
processed in just over 24 hours. It was common during
the week of observation to see simultaneously 40
samples being extracted on an ASE; 40 samples  being
put through a multilayer silica column; and another 24
samples being put through phthalocyanine immobilized
silica columns. The process did seem to bottleneck at
the DMSO solvent exchange step early in the
demonstration, but near the end of the first week the
                                                    50

-------
DMSO solvent exchange operator was able to keep up
with the sample flow. Based on observation, regardless
of how small a sample batch size is, no data could be
generated in less than two days (two 12-hours shifts)
mainly because of the minimal 18-hour incubation
period. The developer's estimate that 40 samples could
be completed in three days seems reasonable. In fact,
with the system that was set up in the field, 40 samples
went through extraction one day, sample preparation the
second day and analysis the third day. Meaning, if the
flow of samples was maintained, eventually 40 samples
could be finished every day.

7.2.5  Miscellaneous  Observer Notes
According to the developer, the kit is usually sold to
experienced clients who  require very little support or
training. Wako provides training classes and phone
support that easily meets the needs of their experienced
clients. The kit is primarily sold in Japan, but Wako
would provide a training video if kits are sold in the
United States. Wako currently has several sales  offices
in the United States.

The developer stated that the kit is readily available and
is supplied with the following materials:

(1)    One 96-well plate with the wells coated with
      secondary antibody
(2)    Positive Control tube
(3)    PC solutions
(4)    Buffer B solution
(5)    Primary antibody solution
(6)    Purified water
(7)    Buffer A
(8)    POD-conjugate solution
(9)    Parafilm
(10)   Reaction mixture
(11)   Wash solution
(12)   Color developing solution
(13)  Citrate buffer
(14)  Stop solution
(15)  Dilution tubes
(16)  Instructions

All materials needed for the suggested extraction and
cleanup would have to be supplied by the user. The
extraction and cleanup materials are, however, listed in
the kit instructions. Also, the following materials for the
ELISA process would have to be supplied by the user
and, unless noted, are listed in the "materials needed"
section of the kit instructions:

(1)   Refrigerator
(2)   Sample tubes and rack
(3)   Micropipettes
(4)   Methanol
(5)   Acetone
(6)   Stirring rods
(7)   Paper towels*
(8)   Scotch tape*
(9)   Aluminum foil*
(10)  Microplate reader
(11)  Watch*
* not listed on the materials needed in the instructions

Each microplate submitted for analysis contains a
seven-point curve and three QC samples (samples with
known amount of dioxin).  Each sample, including the
three QC samples, is run in duplicate. These are
considered requirements and are outlined in the
instructions. The observer felt that it would be
beneficial to confirm results by traditional HRMS
analysis at or near any action levels required by a
program using this technology, but this was not required
by Wako for operation of the technology. No
extractions and/or cleanup QC (such as method blanks,
sample extraction duplicates, etc.) was mentioned by the
developers or the kit's instructions.
                                                     51

-------
                                              Chapter 8
                                          Economic Analysis
During the demonstration, the Wako kit and the
reference laboratory analytical methods were each used
to perform more than 200 sample analyses, including
samples with a variety of distinguishing characteristics
such as high levels of polychlorinated biphenyls and
PAHs. Collectively, the samples provided different
levels and types of contamination necessary to properly
evaluate the technologies and to perform a compre-
hensive economic analysis of each technology. The
purpose of the economic analysis was to estimate the
total cost of generating results by using the Wako kit and
then comparing this cost to the reference method. This
cost estimate also is provided so that potential users can
understand the costs involved with using this
technology.

This chapter provides information on the issues and
assumptions involved in the economic analysis (Section
8.1), discusses the costs associated with using the Dioxin
ELISA Kit Wako (for environmental) (Section 8.2),
discusses the costs associated with using the reference
method (Section 8.3), and presents a comparison of the
economic analysis results for the kit and the reference
laboratory (Section 8.4).

8.1    Issues and Assumptions
Several factors affect sample measurement costs.
Wherever possible in this chapter, these factors are
identified in  such a way that decision-makers can
independently complete a project-specific economic
analysis. The following five cost categories were
included in the economic analysis for the demonstration:
capital equipment, supplies, support equipment, labor,
and investigation-derived waste (IDW) disposal. The
issues and assumptions associated with these categories
and the costs not included in the analysis are briefly
discussed below. The issues and assumptions discussed
below only apply to the Dioxin ELISA Kit Wako (for
environmental) unless otherwise stated.
8.1.1  Capital Equipment Cost
The capital equipment cost was the cost associated with
the purchase of the Dioxin ELISA Kit Wako (for
environmental). Components of the kit are presented in
detail in Chapter 2 and 7. The cost information was
obtained from a standard price list provided by Wako.

8.1.2  Cost of Supplies
The cost of supplies was estimated based on the supplies
required to analyze all demonstration samples using the
Dioxin ELISA Kit Wako (for environmental) that were
not included in the capital equipment cost category.
Examples of such  supplies include filters, cleanup
columns, gas cylinders,  solvents, and distilled water. The
supplies that Wako used during the demonstration fall
into two general categories: consumable (or expendable)
and reusable. Examples of expendable supplies utilized
by Wako during the demonstration include hexane,
acetone, fluorobenzene, toluene, distilled water, nitrogen
cylinders, and Eppendorf pipettes. Examples of reusable
supplies include a microplate reader, accelerated solvent
extractor, and concentrators. It should be noted that this
type of equipment may or may not be already owned by
a potential Dioxin ELISA Kit Wako (for environmental)
user; however, for this economic analysis, an assumption
was made that the user does not possess these items.

The purchase price of these supplies was either obtained
from a standard price list provided by Wako or it was
estimated based on price quotes from independent
sources.

8.1.3  Support Equipment Cost
This section details the equipment used at the demon-
stration such as the mobile laboratory, construction
trailer, fume hood, and laptop computer required by the
technology. Costs  for these items will be reported per
actual costs for the demonstration.
                                                    52

-------
8.1.4  Labor Cost
The labor cost was estimated based on the time required
for work space setup, sample preparation, sample
analysis, and reporting. For the demonstration, devel-
opers reported results by submitting a chain-of-custody
(COC)/results form. The measurement of the time
required for Wako to complete all 209 sample analyses
during the field demonstration (522 labor-hours) was
estimated by the sign-in log sheets that recorded the time
the Wako operators were on-site. Time was removed for
site-specific training activities and Visitors Day. Time
estimates were rounded to the nearest hour.

During the demonstration, the skill level required for the
operators to complete analyses and report results was
evaluated. As stated in Section 7.2.1, based on the field
observations, a fully trained environmental chemistry
technician would be the minimum qualifications of a
person capable of meeting the kit's extraction and
cleanup needs. Wako's ELISA kit only would require a
user skilled in pipetting and computing to generate
accurate data. Four technicians are needed for pretreat-
ment, while two technicians are needed for ELISA.
Furthermore, specific technical training would be
necessary to use the accelerated solvent extractor. This
information was corroborated by Wako.

Education levels of the actual field operators are
included in Section 7.2.1. For the economic analysis,
costs were estimated using both actual and projected
necessary skill levels for operators.

8.1.5  Investigation-Derived Waste Disposal Cost
During the demonstration, Wako was provided with
5-gallon containers for collecting wastes generated
during the demonstration. Sample by-products such as
used samples, aqueous and solvent-based effluents
generated from analytical processes, used glassware, and
personal protective equipment were disposed of in the
containers. The total cost to dispose of these wastes
generated during the demonstration is included in the
economic analysis. Items such as coffee cups, food
waste, and office waste were  disposed of in regular
public refuse containers and were not included as IDW
and therefore not discussed in this economic  analysis.
8.1.6   Costs Not Included
Items whose costs were not included in the economic
analysis are identified below along with a rationale for
the exclusion of each.

Electricity.  During the demonstration, some of the
capital equipment was operated using AC power. The
costs associated with providing the power supply were
not included in the economic analysis as it is difficult to
estimate the electricity used solely by the Wako
technology. The total cost for electricity usage over the
10-day demonstration was $288. With seven mobile
labs/trailers and miscellaneous equipment being operated
continuously during the course of the demonstration, the
cost of Wako electricity usage would be no more than
$82. There was significantly more cost (approximately
$13,000) to install an electrical board and additional
power at the demonstration site, but this was a function
of the demonstration site and not the responsibility of the
individual developers, so this cost was not included in
the economic analysis.

Oversight of Demonstration Activities. A typical user
of the Dioxin ELISA Kit Wako (for environmental)
would not be required to pay for customer oversight of
sample analysis. The EPA, the MDEQ, and Battelle
representatives were present during the field demon-
stration, but costs for oversight were not included in the
economic analysis because these activities were project-
specific. For these same reasons, cost for auditing
activities (i.e., technical systems audits at the reference
laboratory and during the field demonstration) were also
not included.

Travel and Per Diem for Operators. Operators may be
available locally. Because the availability of operators is
primarily a function of the location of the project site,
travel and per diem costs for operators were not included
in the economic analysis.

Sample Collection and Management. Costs for sample
collection and management activities, including sample
homogenization and labeling, were not included in the
economic analysis because these activities were project-
specific and were not dependent upon the selected
reference method or developer technology. Additionally,
sample shipping, COC activities, preservation of
samples, and distribution of samples were specific
                                                    53

-------
requirements of this project that applied to all developer
technologies and may vary from site to site. None of
these costs were included in the economic analysis.

Shipping. Costs for (1) shipping equipment and supplies
to the demonstration site and (2) sample coolers to the
reference laboratory were not included in the economic
analysis because such costs vary depending on the
shipping distance and the service used (for example, a
courier or overnight shipping versus economy shipping).

Items Costing Less Than $10. The cost of inexpensive
items was not included in the economic analysis when
the estimated cost was less than $10. Items where it is
estimated that the cost was less than $10 included:
- Distilled water
- Personal protective equipment
- Waste containers
- Lab stools.

8.2     Dioxin ELISA Kit Costs
This section presents information on the individual costs
of capital equipment, supplies, support equipment, labor,
and IDW disposal for the Dioxin ELISA Kit Wako (for
environmental), as well as a summary of these costs.
Additionally, Table 8-1 summarizes the kit costs. As
described in Section 4.6, Wako analyzed all  209 samples
during the field demonstration and zero samples in its
laboratory (total 209 demonstration samples). It is
important to note that costs estimated in this section are
based on actual costs to analyze the samples during the
field demonstration. Cost estimates for analyzing the
entire set of 209 demonstration samples were then
determined based on the field demonstration costs, even
though a subset of 24 samples was reanalyzed in Wako's
laboratories. Because the number of samples analyzed
in the field is equal to the number of samples in the
entire demonstration set for Wako, itemized costs for the
field demonstration samples and the entire set of
demonstration samples will be identical.

8.2.1 Capital Equipment Cost
The capital equipment cost was the cost associated with
the purchase of the technology in order to perform
sample preparation and analysis. The Dioxin ELISA Kit
Wako (for environmental) can be purchased for $898.
One kit contains enough supplies for 42 samples
(duplicate analyses per sample). In the demonstration,
eight wells were used per sample, so only 10 samples
were analyzed per plate. Because the kit is consumable,
Wako does not rent the kit. During the field
demonstration, Wako utilized 21 ELISA kits for
approximately nine days to analyze 209 samples.

8.2.2  Cost of Supplies
The supplies that Wako used during the demonstration
fall into two general categories: expendable  or reusable.
Table 8-1 lists all the expendable and reusable supplies
that Wako used during the demonstration and the
corresponding costs. Expendable supplies are ones that
are consumed during the preparation or analysis.
Reusable costs are items that must be used during the
analysis but ones that can be repeatedly reused. The
estimated life of reusable supplies could not be assessed
during this economic analysis.

The total cost of the supplies employed by Wako during
the demonstration was $81,958. Supplies have to be
purchased from a retail vendor of laboratory supplies.
Reusable items listed in Table 8-1 can be substituted
with other models that operate under the  same
specifications, thereby modifying the cost of supplies to
the potential kit user. Costs to rent or lease rather than
purchase resusable items were not provided  by Wako.

8.2.3  Support Equipment Cost
Wako analyzed demonstration samples in a 32-foot
mobile lab equipped with two fume hoods and one
32-foot construction trailer equipment with a
refrigerator.  The rental  cost for the mobile lab for use
during sample extraction and sample analysis was
$3,500, while the rental cost for the construction trailer
was $1,919. The minimum rental rate for both the
mobile lab and construction trailer was one month.
Wako only used the mobile laboratory and construction
trailer for two weeks.  Since weekly or daily rental rates
for the mobile lab were not an option, the entire cost is
reported. A laptop computer is a necessity for the
efficient operation of this technology. This is a one-time
purchase that is reusable.
                                                    54

-------
Table 8-1. Cost Summary
Item

Capital Equipment
Purchase of Dioxin ELISA Kit
Supplies
Expendable
Nitrogen Cylinder
Cylinder Regulator
Presep® Multilayer Silica Gel (5 pcs.)
Presep® Phthalocyanine Immobilized
Silica Gel (10 pcs.)
Copper, Reduced, Granular (100 g)
Toluene (3 -liter)
Acetone (3 -liter)
Hexane (3 -liter)
Fluorobenzene (10 g)
DMSO (500 mL)
Eppendorf Pipette and Tips
Reusable
Accelerated Solvent Extractor
Turbo Vap-LV Concentrator
Turbo Vap-500 Concentrator
Pressurized Gas Blowing Concentrator
Low Temperature Circulator
Compressor
Microplate Reader
Refrigerator
Support Equipment
Mobile Laboratory
Construction Trailer
Laptop Computer
Labor
Operator
row Disposal*
Total Cost
Quantity Used
During

21


5
2
1

10
1
1
1
1
1
1
1

1
1
1
1
1
2
1
1

1
1
1

522
1

Field Demo

Kits


unit
unit
unit

unit
unit
unit
unit
unit
unit
unit
unit

unit
unit
unit
unit
unit
unit
unit
unit

unit
unit
unit

labor hours
unit

Unit Cost
($)•

898


31
182
165

330
91
76
63
61
26
27
1,000

44,000
8,000
8,000
100
5,000
5,000
9,000
500

3,500
1,919
1,000

80C
1,168

Cost ($)
209 samples

18,858 b


155
364
165

330
91
76
63
61
26
27
1,000

44,000
8,000
8,000
100
5,000
5,000
9,000
500

3,500
1,919
1,000

41,891
1,168
$150,294
   Itemized costs were rounded to the nearest $1.
   Wako used 21 kits during the demonstration because they analyzed more sample replicates than are normally
   performed.  One kit can be used for 42 samples when duplicate sample analyses are performed, so the cost of the
   kits to perform 209 sample analyses would have been greatly reduced ($4,490) if duplicates were performed.
   Labor rate for field technicians to operate technology rather than research scientists was $50.75 an hour, $26,492
   for 209 samples.
   Further discussion about waste generated during demonstration can be found in Chapter 7.
                                                               55

-------
8.2.4  Labor Cost
As described in Section 8.1.4, 522 labor-hours were
spent in the field to analyze 209 samples. An hourly rate
of $32.10 was used for a research scientist performing
sample extractions and sample analysis, and a
multiplication factor of 2.5 was applied to labor costs in
order to account for overhead costs.(9) Based on this
hourly rate and multiplication factor, a labor rate of
$41,891 was determined for the analysis of the 209
samples during the  field demonstration.

Based on observation, it is anticipated that lower-cost
field technicians, with proper training and skill levels,
could have analyzed the samples in a similar amount of
time. As such, the labor rate for the analysis  of 209
samples during the  field demonstration could have been
as low as $26,492 (hourly rate of $20.30 with 2.5
multiplication factor for 522 labor-hours).

8.2.5  Investigation-Derived Waste Disposal Cost
As discussed in Chapter 7, Wako was provided with
5-gallon containers for collecting wastes generated
during the demonstration.  Chapter 7 discusses the type
and amount of waste generated by the technology during
the field demonstration in  more detail.

During the demonstration, Wako analyzed 209 samples.
The total cost to dispose of the waste generated for these
samples was $1,168.

8.2.6  Summary ofDioxin ELISA Kit Costs
The total cost for performing dioxin analyses using the
Dioxin ELISA Kit Wako (for environmental) during the
field demonstration was $150,294. The dioxin analyses
were performed for 58 soil and sediment PE samples,
128 soil and sediment environmental samples, and 23
extracts. When Wako performed  multiple dilutions or
reanalyses for a sample, these were not included  in the
number of samples  analyzed.

The total cost of $150,294 for analyzing the
demonstration samples under the Wako kit included
$18,858 for capital  equipment; $81,958 for supplies;
$6,419 for support equipment; $41,891 for labor; and
$1,168 for IDW disposal. Of these five costs, the largest
cost was for the supplies (55 percent of the total cost).
8.3    Reference Method Costs
This section presents the costs associated with the
reference method used to analyze the 209 demonstration
samples for dioxin. Typical costs of these analyses can
range from $800 to $1,100 per sample, depending on the
method selected, the level of quality assurance/quality
control incorporated into the analyses, and reporting
requirements. The reference laboratory utilized EPA
Method 1613B for all soil and sediment samples for
comparison with the Dioxin ELISA Kit Wako (for
environmental). The reference method costs were
calculated using information from the reference
laboratory invoices.

To allow an accurate comparison of the Dioxin ELISA
Kit Wako (for environmental) and reference method
costs, the reference method costs were estimated for the
same number and type of samples as was analyzed by
Wako. For example, although the reference laboratory
analyzed soil and sediment samples for coplanar PCBs,
the associated sample analytical costs were not included
in the reference method costs because Wako did not
analyze samples for PCBs during the demonstration.

Table 8-2 summarizes the projected and actual reference
method costs. At the start of the demonstration, the
reference laboratory's projected cost per sample was
$785 for dioxin/furan analysis. This cost covered the
preparation and analysis of the demonstration samples,
required method QC samples, electronic data
deliverable, and the data package for each. The actual
cost  for the 209 demonstration analyses was $213,580
for D/F. This was higher than the projected ($162,915)
due to reanalysis, re-extractions, dilutions, and
additional cleanups that were above the 30% repeats
allowable by the original quote.  The turnaround time  by
the reference laboratory for reporting all 209 samples
was approximately eight months (171 business days).
The quoted turnaround time was three months.

8.4    Comparison of Economic Analysis
       Results
The total costs for the Dioxin ELISA kit Wako (for
environmental) ($150,294) and the reference method
($213,580) are listed in Tables 8-1 and 8-2, respectively.
The total cost for the use of the Wako kit was $63,286
less than the reference method. Furthermore, the Wako
                                                    56

-------
Table 8-2. Reference Method Cost Summary
Analyses Performed
Dioxin/Furans, EPA Method
1613B, GC/HRMS
Total Cost
Number of Samples
Analyzed
23 extracts
186 soil/sediment
209 samples
Cost per
sample
Quotation ($)
735
785

Itemized Cost ($)
Quotation3
16,905
146,010
162,915
Actual
213,580
 a Price includes up to 30% of samples requiring additional work of some kind (dilutions or extra cleanup). Greater
   than that would require additional work with further charges associated to them ($150 to $180 per sample per
   procedure).
analyses were completed on-site in nine days during the
field demonstration where the reference analyses took
eight months. Use of the kit in the field will likely
produce additional cost savings because the results will
be available within a few hours of sample collection;
therefore, critical decisions regarding sampling and
analysis can be made in the field, resulting in a more
complete data set. Additional possible advantages to
using field technologies include reduction of multiple
crew and equipment mobilization-demobilization cycles
to a single cycle, dramatically increased spatial
resolution mapping for higher statistical confidence,
leading to reduced insurance costs and reduced disposal
costs, and compression of total project time to reduce
administrative overhead. However, these savings cannot
be accurately estimated and thus were not included in the
economic analysis.  Project-specific costs associated
with the use of the technology, such as the cost for
HRMS confirmation analyses and training costs to be
proficient in the use of the technology, were also not
accounted for in this analysis.

The Dioxin ELISA Kit Wako (for environmental) is a
screening method only that reports 2,3,7,8-TCDD EQ
(total D/F concentration in the sample), unlike the
reference method which reports concentrations for
individual congeners. Although the kit's analytical
results did not have the same level of detail as the
reference method analytical results (or comparable
QA/QC data), the kit provided dioxin/furan analytical
results on site in a nine-day at significant cost and time
savings compared to the reference laboratory.
                                                      57

-------
                                             Chapter 9
                               Technology Performance Summary
The purpose of this chapter is to provide a performance
summary of the Wako Pure Chemical Industries, Ltd.
Dioxin ELISA Kit Wako (for environmental) by
summarizing the evaluation of the primary and
secondary objectives of this demonstration in Tables 9-1
and 9-2, respectively. Detailed information about these
evaluations, including a complete evaluation of the
reference laboratory data, can be found in previous
sections of this report.

The data generated and evaluated during this
demonstration showed that the Wako kit in many cases
did not directly correlate with HRMS TEQD/F values. It
did show that the kit could be an effective screening tool
for determining sample results above and below 20 pg/g
TEQD/F and even more effective as a screen for samples
above and below 50 pg/g TEQD/F, particularly
considering that both the cost ($150,294 vs. $213,580)
and the time (nine days on-site vs. eight months) to
analyze the 209 demonstration samples were
significantly less than the reference laboratory. Because
the Wako kit is not expected to directly correlate to
HRMS TEQD/F in all cases, the technology should not be
viewed as producing an equivalent measurement value to
HRMS TEQD/F values. It has been suggested that
correlation between the Wako and HRMS TEQ could be
improved by first characterizing a site and calibrating the
Wako results to HRMS results. Subsequent analysis
using the Wako kit for samples obtained from this  site
may then show better correlation with the HRMS TEQ
result. This approach was not evaluated during this
demonstration.
                                                   58

-------
Table 9-1. Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit Wako (for environmental) Performance
Summary - Primary Objectives
Objective
P 1 : Accuracy
P2: Precision
P3: Comparability
P4: Estimated method
detection limit
P5: False Positive/False
Negative Rate
P6: Matrix Effects
P7: Cost
Statistic
Number of data points
Median Recovery (%)
Mean Recovery (%)
Number of data points
Median RSD (%)
MeanRSD (%)
Number of data points
Median RPD (%)
Interval agreement (%)
Blank agreement (%)
EMDL (pg/g 2,3,7,8-TCDD EQ)
False positive rate at 20 pg/g TEQ (%)
False positive rate at 50 pg/g TEQ (%)
False negative rate at 20 pg/g TEQ (%)
False negative rate at 50 pg/g TEQ (%)
Performance
5
253
443
24
60
62
106
34
62
88
83-201
10
10
13
8
• Measurement location: not evaluated (all samples analyzed on-site)
• Matrix type: none
• Sample type: none
• PAH concentration: none
• Environmental site: none
• Known interferences: slight
Cost for Wako to deploy at the field demonstration site and analyze all 209 samples: $150,294.
                                                59

-------
Table 9-2.  Wako Pure Chemical Industries, Ltd. Dioxin ELISA Kit Wako (for environmental) Performance
Summary - Secondary Objectives
           Objective
                                    Performance
 SI: Skill level of Operator
Based on observation during the field demonstration, a fully trained environmental chemistry
technician would be the minimum qualifications of a person capable of meeting the kit's
extraction and cleanup needs. The Dioxin ELISA Kit Wako (for environmental) only would
require a user skilled in pipetting and computing to generate accurate data. Wako prefers
only experienced personnel use the kit.
 S2: Health and Safely Aspects
The sample process generated a large amount of hazardous waste including: flammable
solvents, spent reactants (corrosive impregnated silica, copper, sodium sulfate), spent
samples, and glass (pipettes, disposable cleanup columns, filters). The process also generated
small amounts of hazardous aqueous waste. A considerable amount of lab trash was also
generated that included used personal protective equipment, absorbent paper, and aluminum
foil.  A fume hood is necessary for the operation of this technology.
 S3 Portability
The observed Dioxin ELISA kit provided by Wako required an exacting sample extraction
and cleanup, similar to what is required for traditional HRMS dioxin analysis.  The extraction
and sample cleanup required one mobile lab fully outfitted with fume hoods, water, nitrogen
and/or purified compressed air, and significant electrical power (~ 15 kVA power for the
mobile laboratory).  The ELISA procedure required less infrastructure, requiring only a
refrigerator, watch, pipettes, small bench space, microplate reader with computer, and a
controlled environment.
 S4: Sample Throughput
During the field demonstration, all 209 demonstration samples were processed by Wako,
equating to a sample throughput rate of 23 samples per day. This was accomplished in about
9 full working days (522 labor-hours), with six people performing various aspects of the
sample preparation and analytical procedure.	
                                                          60

-------
                                            Chapter 10
                                            References
1.  EPA. 2001. Database of Sources of Environmental
   Release ofDioxin-like Compounds in the United
   States, EPA/600/C-01/012, March.

2.  EPA. 2004."Technologies for the Monitoring and
   Measurement of Dioxin and Dioxin-like Compounds
   in Soil and Sediment," Demonstration and Quality
   Assurance Project Plan, U.S. EPA/600/R-04/036,
   April.

3.  EPA Method 1613B. 1994. Dioxins, Tetra- thru
   Octa-(CDDs) and Furans (CDFs), EPA/82 l/B-94-
   005, 40 Code of Federal Regulations Part 136,
   Appendix A, October.

4.  EPA Method 1668A. 1999. Chlorinated biphenyl
   congeners by HRGC/HRMS, EPA/821/R-00-002,
   December.

5.  van den Berg, M., Birnbaum, L., Bosveld, A. T. C.,
   Brunstrom, B., Cook, P., Feeley, M., Giesy, J. P.,
   Hanberg, A., Hasagawa, R., Kennedy, S. W.,
   Kubiak, T., Larsen, J. C., van Leeuwen, F. X. R.,
   Liem, A. K. D., Nolt, C., Peterson, R. E., Poellinger,
   L., Safe, S., Schrenk, D., Tillitt, D., Tysklind, M.,
   Younes, M., Waern, F., and Zacharewski, T. 1998.
   Toxic equivalency factors (TEFs) for PCBs, PCDDs,
   PCDFs for humans and wildlife. Environmental
   Health Perspectives 106: 775-792.

6.  De Rosa, Christopher T., et al. 1997. Dioxin and
   dioxin-like compounds in soil, Part 1: ATSDR
   Interim Policy Guideline. Toxicology and Industrial
   Health, Vol. 13, No. 6, 1997. pp. 759-768.

7.  NOAA. 1998.  Sampling and analytical methods of
   the national status and trends program mussel watch
   project:  1993-1996 update. NOAA Technical
   Memorandum NOSORCA 130.  Silver Spring,
   Maryland.

8.  EPA SW-846 Method 8290. 1994. Polychlorinated
   dibenzodioxins (PCDDs) and polychlorinated
   dibenzofurans (PCDFs) by high-resolution gas
   chromatography/high-resolution mass spectrometry
   (HRGC/HRMS), September.

9.  U.S. Bureau of Labor Statistics, National
   Compensation Survey. Accessed on 7/26/04.
   Available at:
   http://data.bls.gov/labjava/outside jsp?survey=nc.
                                                  61

-------

-------
                   Appendix A
SITE Monitoring and Measurement Technology Program
               Verification Statement

-------

-------
          United States Environmental Protection Agency
                              Office of Research and Development
                                   Washington, DC 20460
        SITE Monitoring and Measurement Technology Program
                              Verification Statement
        TECHNOLOGY TYPE:  Enzyme-Linked Immunosorbent Assay
        APPLICATION:
MEASUREMENT OF DIOXIN AND DIOXIN-LIKE
COMPOUNDS
        TECHNOLOGY NAME: Dioxin ELISA Kit Wako (for environmental)
        COMPANY:
        ADDRESS:

        PHONE:
        WEB SITE:
        E-MAIL:

        U.S. SUBSIDIARY:
        ADDRESS:

        PHONE:
        WEB SITE:
        E-MAIL:
Wako Pure Chemical Industries, Ltd.
1-2 Doshomachi 3-Chome Chuo-ku
Osaka 540-8605 Japan
+81-6-6203-3841
http://wako-chem.co.jp
cservice@wako-chem.co.jp

Wako Chemicals USA, Inc.
1600 Bellwood Road
Richmond, Virginia 23237-1326 USA
(804) 271-7677
http://www.wakousa.com
emmy@wakousa.com
 VERIFICATION PROGRAM DESCRIPTION

 The U.S. Environmental Protection Agency (EPA) created the Superfund Innovative Technology Evaluation (SITE)
 Monitoring and Measurement Technology (MMT) Program to facilitate deployment of innovative technologies
 through performance verification and information dissemination. The goal of this program is to further environmental
 protection by substantially accelerating the acceptance and use of improved and cost-effective technologies. The
 program assists and informs those involved in designing, distributing, permitting, and purchasing environmental
 technologies. This document summarizes results of a demonstration of the Wako Pure Chemical Industries, Ltd.
 Dioxin Enzyme-Linked Immunosorbent Assay (ELISA) Kit Wako (for environmental).

PROGRAM OPERATION

Under the SITE MMT Program, with the full participation of the technology developers, the EPA evaluates and
documents the performance of innovative technologies by developing demonstration plans, conducting field tests,
collecting and analyzing demonstration data, and preparing reports. The technologies are evaluated under rigorous
quality assurance protocols to produce well-documented data of known quality. The EPA's National Exposure
                                            A-l

-------
Research Laboratory, which demonstrates field sampling, monitoring, and measurement technologies, selected Battelle
as the verification organization to assist in field testing technologies for measuring dioxin and dioxin-like compounds
in soil and sediment.

DEMONSTRATION DESCRIPTION

The demonstration of technologies for the measurement of dioxin and dioxin-like compounds was conducted at the
Green Point Environmental Learning Center in Saginaw, Michigan, from April 26 to May 5, 2004. The primary
objectives for the demonstration were as follows:
P1.  Determine the accuracy.
P2.  Determine the precision.
P3.  Determine the comparability of the technology to EPA standard methods.
P4.  Determine the estimated method detection limit (EMDL).
P5.  Determine the frequency of false positive and false negative results.
P6.  Evaluate the impact of matrix effects on technology performance.
P7.  Estimate costs associated with the operation of the technology.

The secondary objectives for the demonstration were as follows:
S1.  Assess the skills and training required to properly operate the technology.
S2.  Document health and safety aspects associated with the technology.
S3.  Evaluate the portability of the technology.
S4.  Determine the sample throughput.

A total of 209 samples was analyzed by each technology, including a mix of performance evaluation (PE) samples,
environmentally contaminated samples, and extracts. Wako analyzed all 209 samples on-site during the field
demonstration. The PE samples were used primarily to determine the accuracy of the technology and consisted of
purchased reference materials  with certified concentrations. The PE samples also were used to evaluate precision,
comparability, EMDL, false positive/negative results, and matrix effects. Dioxin-contaminated samples from Warren
County, North Carolina; the Saginaw River, Michigan; the Tittabawassee River, Michigan; Midland, Michigan;
Winona Post, Missouri; Nitro, West Virginia; Newark Bay, New Jersey; Raritan Bay, New Jersey; and Brunswick,
Georgia were used to evaluate precision, comparability, false positive/negative results, and matrix effects. Extracts
prepared in toluene were used to evaluate precision, EMDL, and matrix effects. All samples were used to evaluate
qualitative performance objectives such as technology cost, the required skill level of the operator, health and safety
aspects, portability, and sample throughput. AXYS Analytical Services (Sidney, British Columbia) was contracted to
perform the reference analyses by high-resolution mass spectrometry (HRMS) (EPA Method  1613B). The purpose of
the verification statement is to provide a summary of the demonstration and its results; detailed information is available
in Technologies for Monitoring and Measurement of Dioxin and Dioxin-like Compounds in Soil and Sediment— Wako
Pure Chemical Industries, Dioxin ELISA Kit Wako (for environmental) (EPA/540/R-05/002).

TECHNOLOGY DESCRIPTION

The technology description and operating procedure below are based on information provided by Wako Pure Chemical
Industries, Ltd. A monoclonal  antibody specific to dioxin is mixed with a sample solution or the positive control (PC)
provided with the Dioxin ELISA Kit Wako (for environmental). Peroxidase conjugated with a dioxin analog
(POD-conjugate) is then added, reacting with a primary antibody to dioxin in the sample. The mixture is added to a
microplate coated with a secondary antibody that captures the antibody-POD-conjugate and incubated at 2 to 8°C for
18 to 20 hours. After washing the resultant microplate with  a buffer, the antibody-POD-conjugate complex formed on
the plate is reacted with substrate for peroxidase. The reaction is stopped by adding stop solution, and the microplate
reader reads the signal. The Dioxin ELISA Kit Wako (for environmental) contains the secondary antibody microplate,
the PC, buffers A and B, the primary antibody, peroxidase conjugate (lyophilized), wash solution concentrate,
substrate, citrate buffer, stop solution, and a plate seal. The monoclonal antibodies used for the Dioxin ELISA Kit
Wako (for environmental) indicate cross-reactivity nearly equal to the positive control (2,7,8-trichlorodibenzo [1,4]
dioxin-1-yl) acrylic acid) and 2,3,7,8-tetrachlorodibenzo-/>-dioxin (TCDD). The technology results are reported in
picogram/gram (pg/g) 2,3,7,8-TCDD equivalents (EQ).
                                                    A-2

-------
VERIFICATION OF PERFORMANCE

The Wako kit is an immunoassay technology that reports total dioxin/furan concentration in the sample. It should be
noted that the results generated by this technology may not directly correlate to HRMS total toxicity equivalents of
dioxins/furans (TEQD/F) in all cases because it is known that the congener responses and cross-reactivity of the kit are
not identical to the toxicity equivalency factors that are used to convert congener HRMS concentration values to
TEQD/F. Therefore, this technology should not be viewed as producing an equivalent measurement value to HRMS
TEQD/F but as a screening value to approximate HRMS TEQD/F concentration. It has been suggested that correlation
between the Wako and HRMS TEQ could be improved by first characterizing a site and calibrating the Wako results to
HRMS results. Subsequent analysis using the Wako kit for samples obtained from this site may then show better
correlation with the HRMS TEQ result. This approach was not evaluated during this demonstration.

Accuracy: The determination of accuracy was based on the agreement of the Wako results with the certified or spiked
levels of the  PE samples that were obtained from commercial sources. Accuracy was assessed by percent recovery (R),
which is the  average of the replicate results from the Dioxin ELISA Kit Wako (for environmental)  divided by the
certified or spiked value of the PE sample, multiplied by 100%.  Ideal R values are near 100%. The overall R values
were 443% (mean), 253% (median), 10% (minimum), and 1,574% (maximum).

Precision: Replicates were incorporated for all samples (PE, environmental, and extracts) included in the 209 samples
analyzed in the demonstration. Replicates were incorporated for all samples (PE, environmental, and extracts) included
in the 209 samples analyzed in the demonstration. Three samples had seven replicates in the experimental design, one
sample had eight replicates, and all other samples had four replicates. Precision was determined by calculating the
standard deviation of the replicates, dividing by the average concentration of the replicates, and multiplying by 100%.
Ideal RSD values are less than 20%. The overall RSD values were 62% (mean),  60% (median), 11% (minimum), and
145% (maximum).

Comparability: The Dioxin ELISA Kit Wako (for environmental) results were compared to EPA Method 1613B
results for total TEQD/F. The results were compared by determining the relative percent difference  (RPD) by dividing
the difference of the two numbers by the average of the two numbers and multiplying by 100%. Ideal RPD values are
between positive and negative 25%. The overall RPD values were 34% (median), -198% (minimum), and 198%
(maximum).  The Wako results were also compared to the reference laboratory results using an interval approach to
determine if the Wako results and the reference laboratory results would place the samples in the same action-level
interval, thereby resulting in the same action-oriented decision. The developer and reference data were grouped  into
four toxicity equivalent (TEQ) concentration ranges. The ranges were < 50 pg/g, 50 to 500 pg/g, 500 to 5,000 pg/g,
and > 5,000 pg/g.  The intervals were determined based on current guidance for cleanup levels. The percentage  of
developer results that agreed with reference laboratory results for TEQD/F was 62%.

Estimated method detection limit: EMDL was calculated for the technology generally according  to the procedure
described in  40 CFR Part 136, Appendix B,  Revision 1.11. Lower EMDL values indicate better sensitivity. The
calculated EMDLs ranged from 83 to 201 pg/g 2,3,7,8-TCDD EQ values, depending on whether nondetect values were
assigned values of zero, one-half the reporting limit value, or the reporting limit value itself. The detection limit
reported by Wako in the demonstration plan was 20 pg/g 2,3,7,8-TCDD  EQ. PE samples with TEQ concentrations in
the precisely appropriate range for evaluation of this technology's detection limit were not available, so these
calculated values should be considered a rough estimate.

False positive/negative results: Samples that were reported as less than a specified level by the reference laboratory
but greater than the specified level by Wako were considered false positive. Conversely, those samples that were
reported as less than the specified level by Wako but reported as  greater than the specified level by the reference
laboratory were considered false negatives. Note that results that were reported as semiquantitative results were
counted as in agreement if the reference laboratory data was within that interval. For example, a reference laboratory
result reported as 3,400 pg/g TEQD/F was counted as in agreement with a Wako result reported as > 2,967 pg/g
2,3,7,8-TCDD EQ, even though the absolute quantitative values would be in different ranges. Ideal false positive and
false negative rates were zero. The kit had a false positive rate of 10% and a false negative rate of 13% around 20 pg/g
TEQ. Comparison of values around 50 pg/g  indicated the same false positive rate (10%), but less false negatives (8%).
                                                    A-3

-------
These data suggest the Wako kit could be an effective screening tool for determining sample results above and below
20 pg/g TEQ and even more effective as a screen for samples above and below 50 pg/g TEQ.

Matrix effects: The likelihood of matrix-dependent effects on performance was investigated by evaluating results in a
variety of ways. No significant effect was observed for the reproducibility of Wako results by matrix type (soil,
sediment, and extract), sample type (PE vs. environmental vs. extract), or by polynuclear aromatic hydrocarbon
concentration. The Wako results were not more or less comparable to the reference laboratory results based on
environmental site.

Cost: The full cost of the technology was documented and compared to the cost of the reference analyses. The total
cost for the Dioxin ELISA kit Wako (for environmental) to analyze all 209 samples was $150,294. The total cost for
the reference laboratory to analyze all 209  samples by EPA Method 1613B was $213,580. The total cost for the use of
the Wako kit was $63,286 less than the reference method.

Skills and training required: Based on observation during the field demonstration, a fully trained environmental
chemistry technician would be the minimum qualifications of a person capable of meeting the kit's extraction and
cleanup needs. Wako prefers only experienced personnel use the kit and to be skilled in pipetting and computing skills
to generate accurate data.

Health and safety aspects: The sample process generated a large amount of hazardous waste, including flammable
solvents, spent reactants (corrosive impregnated silica, copper, sodium sulfate), spent samples, and glass (pipettes,
disposable cleanup columns,  and filters). The process also generated small amounts of hazardous aqueous waste. A
considerable amount of lab trash was also generated, which included used personal protective equipment, absorbent
paper, and aluminum foil. A fume hood is  necessary for the operation of this technology.

Portability: The observed Dioxin ELISA kit provided by Wako required an exacting sample extraction and cleanup,
similar to what is required for traditional HRMS dioxin analysis. The extraction and sample cleanup required one
mobile lab fully outfitted with fume hoods, water, nitrogen and/or purified compressed air, and significant electrical
power (~ 15 kilowatt/ampere (kVA) power for the mobile laboratory). The  ELISA procedure required less
infrastructure, requiring only a refrigerator, watch, pipettes, small bench space, microplate reader with computer, and a
controlled environment.

Sample throughput: During the  field demonstration, all 209 demonstration samples were processed by Wako,
equating  to a sample throughput rate of 23 samples per day.  This was accomplished in about 9 full working days (522
labor-hours), with six people  performing various aspects  of the sample preparation and analytical procedure. The
reference analyses took 8 months to complete the same 209 sample analyses.
        NOTICE: Verifications are based on an evaluation of technology performance under specific, predetermined
        criteria and the appropriate quality assurance procedures. The EPA makes no expressed or implied warranties as
        to the performance of the technology and does not certify that a technology will always operate as verified. The
        end user is solely responsible for complying with any and all applicable federal, state, and local requirements.
                                                    A-4

-------
                                          Appendix B
                   Supplemental Information Supplied by the Developer


 The purpose of this section is for the developer to provide additional information about the technology. This can
include updates/changes/modifications planned for the technology or which have occurred since the technology was
       tested. The developers can also use this section to comment and expand on the findings of the report.
            Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------

-------
   \Afeko's comment toward EPA report

   1, Goal of Demonstration
 Wako Pure Chemical Ind., Ltd. should design a project of field determination of
 dioxins  as the  goal. \Nako  designed its  measurement program as  a
 screening method for this demonstration.

 1.1 .Reporting results
 1.1.1 The concentration of dioxins:
 (1)  If the sample concentration was 50 pg/g or less of dioxins, we decided
      to report semiquantitative results under the demonstration coordinator's advice.
 (2)  If the sample concentration was  between 50 to 1,000 pg/g of dioxins, we
      decided to determine the exact concentration of dioxins in the sample.
 (3)  If the sample concentration was  1,000 pg/g or more of dioxins, we
      decided to report semiquantitative  results under the demonstration coordinator's
      advice.

 1.1.2 The term of determination
 We planned the time frame for determination to be just in the demonstration period from
 April 26 to May 5 in Saginaw, Michigan.

 1.1.3 The working hour in the demonstration  period
 We made a timetable for each day and it was the working hours from 7:30 am to 7:30
 pm

 1.1.4 Determination of the number of sample
 We believed that  we should analyze  all samples (209 samples) under the term of
 demonstration.  And, we did not have enough  time to re-analyze on site  for  the
 demonstration period.

 1.1.5 Demonstration site
 We came from Japan to  participate in the demonstration at Saginaw, Michigan

 1.1.6 The properties of samples
 The samples distributed by EPA contained various concentrations of dioxins, furans
 and PCBs from extremely high to spiked. The samples included a variety of matrices such
Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                       B-l

-------
  as the polyaromatic compounds.  The samples were distributed blindly and randomly.

  2. Experimental design
  We made the experimental design to  meet these requirements from the demonstration
  as follows.
  2.1 The amount of sample
  We designed the method to take 10g of each sample from distributed materials on site.

  2.2 The frequency of dilution
  Our consideration was samples needed to divide with dilution at four stages.

  2.3 The range of fixed quantity
  Our kit has the range of fixed quantity is from about 20 to 2,000 pg/g.

  We did the same preprocessing "full clean up" as  HRGC/HRMS  because we had to
  treat  samples including  various  matrices. However we used  the multilayer
  silica-gel column for the preprocessing to simplify the process. Due to "full clean  up", waste
  (glassware and solvent, etc.) had increased and sample throughput decreased more than
  originally projected.  One of our main targets was to measure 40 to 50 samples a day.

  3. About the measurement result
  Our first priority of this demonstration was "In the period, all samples (209 samples) are
  measured  on the  environmental  site".  Therefore,  neither  the  minimum limit of
  determination nor the upper limit was measured again. We reported results as below
  the fixed quantity lower bound (<20pg/g) or above the fixed quantity upper bound (>2,000pg/g).
  We considered that  it was enough as the report of this demonstration whatever we were
  able to measure  20pg/g,  because  it  was a prevailing opinion that  50pg/g or less of
  dioxins value was  an unnecessary  concentration  for the  treatment. And  a  high
  concentration sample (more than >)  could be measured if diluted.  However, the
  percent recovery (R) and the   relative  percent difference (RPD)  values were  not
  calculated for the samples which our  reported as "< (value)" and "> (value)". They are
  reported as "NA" (see table 7-1,  7-2a and 7-3) because the equations that are used to
  calculate those values (see section 4.7 of the  report) require that an actual number be
  used. These values were used the  results as reported (in < and >) in other evaluations,
  such  as the interval  assessment. For these evaluations, our data could use as reported
  because of the nature of the assessment.
Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                       B-2

-------
  of  various  characteristics  in  the period  in  the  environmental  site,  (conventional
  alternative: Standard pretreatment procedure in Japan, "full clean up")
  However, we would like to say it is not likely that various matrices like these samples exist
  at one contaminated site. It seems that the extraction and preprocessing would be able
  to be simplified according to the situation if the sample characteristics are definable.
  We want to examine further better extraction and pretreatment procedure, and
  to find a more effective method.
Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                        B-3

-------

-------
                         Appendix C
Reference Laboratory Method Blank and Duplicate Results Summary

-------

-------
Table C-l. Summary of Method Blank Performance
Sample
Batch
Number
D/F
WG12107
D/F
WG12148
D/F
WG 12264
D/F
WG12534
D/F
WG 12641
D/F
WG12737
D/F
WG12804
D/F
WG13547
Criteria
Met
Y
N
N
N
N
N
N
N
Method
Blank TEQa
(Pg/g)
0.000812
0.133
0.0437
0.610
0.0475
0.348
0.0153
0.0553
Sample TEQ Range3 (pg/g)
26.1-74.1 (Newark Bay)
9.93-13.3 (RaritanBay)
13.5-50.4 (Newark Bay)
49.5-15,200 (Brunswick)
1.0-94.1 (Titta. River sediment)
0.237-6900 (PE)
25. 3-71 00 (PE)
3 1-269 (Midland)
72.8 (Brunswick)
123 (Titta. River sediment)
0.1 59-7690 (PE)
25.7-1 92 (Midland)
35.2- 1300 (Titta. River soil)
3. 89-188 (PE)
57.5- 3,000 (Nitro)
37.9 (North Carolina)
122 (Saginaw River)
26.4- 222 (Midland)
Comments

Many samples had concentrations >20x
blank. Few that didn't were not
significantly affected on a total TEQ basis.
Most samples had concentrations >20x
blank. Low level Tittabawassee River
sediment samples L6749-2 (Ref 48b), -9
(Ref 130), -10 (Ref 183), and -12 (Ref 207)
were evaluated based on their replication
within the demonstration analyses and
comparison to characterization results and
considered unaffected by method blank
exceedances. Low level PE samples
L6760-1 (Ref 25), -3 (Ref 28), and -4 (Ref
29) were D/F blanks with resulting TEQs
sufficiently low enough to still be
distinguished as blank samples.
Sample concentrations > 20x blank.
All but PE sample Ref 177 (0.159 TEQ)
had significantly higher total TEQ than
blank. Ref 177 was confirmed by running
in another batch and results, which agreed
within 18%. Additionally, Ref 177 was
compared to its replicates within the
program and considered acceptable.
Sample concentrations >20x blank.
A few analytes higher than criteria but no
significant contribution to total TEQ.
Several analytes exceeded criteria, but
blank total TEQ contribution to sample is
relatively small.
                                                C-l

-------
Sample
Batch
Number
D/F
WG13548






















D/F
WG13549

D/F
WG13551

D/F

WG13552
D/F
WG13984








D/F
WG14274









Criteria
Met











N













N


N


Y





N









N





Method
Blank TEQa
(Pg/g)











0.0114













0.0925


2.40


0.000969





0.0154









0.0434







Sample TEQ Range3 (pg/g)
99.6-99.7 (Sagmaw River)
32.9-36.4 (North Carolina)
0.268-100 (Extracts)





















2,160-3,080 (Nitro)
146-1,320 (Saginaw River)
788-8,410 (North Carolina)
1,100-10,800 (North Carolina)
7,160-11,300 (WmonaPost)

0.0386-9.28 (PE)

25.8 (Midland)
0.524-24.8 (PE)
10.4(RantanBay)
53. 1-444 (Extracts)







2,800 (Nitro)
35.5-8,320 (North Carolina)
0.0530-5.93 (PE)









Comments
Several analytes exceeded criteria. In
general, the blank contribution to total TEQ
was negligible and in those cases results
were accepted. Several low-level extract
samples were evaluated as follows: Extract
Spike #1 samples L6754-4 (Ref 4), -8 (Ref
8), -10 (Ref 10), -14 (Ref 14), -19 (Ref 19),
-22 (Ref 22), and -23 (Ref 23) were known
TCDD spikes at 0.5 pg/mL. Results were
compared to the known spiked TEQ and
considered unaffected by blank
contribution to TEQ. Extract Spike #3
samples L6754-1 (Ref 1), -7 (Ref 7), -12
(Ref 12), and -15 (Ref 15) were PCB spikes
and not expected to contain D/F. These
spikes consistently contained a D/F TEQ of
~0.3. However, this came from a
consistent ~0.3 pg/mL of TCDD detected
in these extracts that was confirmed as a
low-level TCDD contamination by AXYS.
Since TCDD was not present in the lab
blank, these results were accepted as
unaffected by any blank contribution to
TEQ.
Many analytes exceeded limits, but the
blank contribution to total TEQ is small
relative to sample TEQs.
Many analytes exceeded limits, but the
blank contribution to total TEQ is small
relative to sample TEQs.



Blank contribution to total TEQ was
negligible except for PE samples L7 179-7
(Ref 94), -8 (Ref 96), -1 1 (Ref 108), -12
(Ref 109), -17 (Ref 132), and L7182-6 (Ref
1 50). All but L7 179-8 were certified
blanks. L7 179-8 was a PAH spike with no
D/F TEQ expected. The TEQs of these
samples were considered sufficiently low
enough to still be distinguished as blank
samples and were accepted.
Sample TEQs were large enough to be
unaffected by the blank TEQ except for
four PE samples L7 179-4 (Ref 85), -16
(Ref 124) and L7182-12 (Ref 169) and -14
(Ref 1 84). These PE samples were either
certified blanks or PCB spikes with no
expected D/F TEQ. Resulting TEQs for
these samples were considered low enough
to be distinguished as blank samples and
were accepted.
C-2

-------
Sample
Batch
Number
PCB
WG12108
PCB
WG12147
PCB
WG12265
PCB
WG12457
PCB
WG12687
PCB
WG12834
PCB
WG12835
PCB
WG12836
PCB
WG13008
PCB
WG13256
PCB
WG13257
PCB
WG13258
PCB
WG13554
PCB
WG14109
Criteria
Met
N
Y
Y
N
N
N
N

N
N
Y
Y
Y
N
N
Method
Blank TEQa
(Pg/g)
0.000137
0.000
0.0000584
0.000208
0.0183
0.000405
0.000125

0.0499
0.0221
0.000102
0.000251
0.000301
0.0000900
0.000288
Sample TEQ Range3 (pg/g)
2.63-5. 19 (Newark Bay)
2.04-2. 82 (Rantan Bay)
1.21-5. 06 (Newark Bay)
0.104-0.330 (Brunswick)
0.132-0.369 (Brunswick)
0.034-0.649 (Titta. River
sediment)
0.00277-1, 030 (PE)
4.20-1, 020 (PE)
0.974-2.73 (Midland)
10.3-1, 180 (PE)
0.0157-62.4 (Sagmaw River)
0.181-0.203 (Brunswick)
0.986-7.57 (Titta. River Soil)
0.822-2.06 (Wmona Post)
1,060-904,000 (North Carolina)

2.38-3. 15 (Midland)
1.03-8.37 (Titta. River soil)
4 1.0-1,1 40 (PE)
0.00385-0.051 (PE)
0.253-0.3 18 (Midland)
0.1 35-2. 08 (Extracts)
3.53-9.62 (PE)
1.14-1. 33 (Titta. River Soil)
0. 163-37. 0(Nitro)
29.8-73.6 (Saginaw River)
40.1-42.1 (PE)
0.000103-1,080 (Extracts)
435-1, 160 (PE)
0.388-0.452 (Nitro)
0.0467 (Saginaw River)
0.654-1. 87 (Wmona Post)
0.00300-0.0420 (PE)
Comments
PCB 77 slightly high, but all samples >20x
blank levels.


PCB 77 slightly high. Did not report any
samples where PCB 77 was <10x blank.
No significant effect on total TEQ.
PCB 77 and 156 high, but all samples >20x
blank levels.
PCB 77 slightly high. Does not affect total
TEQ.
PCB 77 slightly high. Sample TEQs much
greater than blank TEQ.
PCBs77, 123, 126, 156, 167, and 118 high,
but most samples significantly > 20x blank
levels.
PCBs 77 and 1 18 high, but all samples
>20x blank levels.



PCB 77 slightly high. Does not affect total
TEQ.
PCB 77 high. PE certified blanks Ref 85,
Ref 85 duplicate, and Ref 108 were the
only samples where PCB 77 was not >20x
blank. TEQs for these certified blanks
were considered low enough to be
distinguished as blank samples and were
accepted.
3 All nondetect and EMPC values were assigned a zero concentration for the TEQ calculation.
b "Ref XX" is a reference laboratory sample ID number.
                                                              C-3

-------
Table C-2. Sample Batch Duplicate Summary
Sample Batch
Number
D/FWG12107
D/FWG12148
D/F WG 12264
D/F WG12534
D/F WG 12641
D/F WG12737
D/F WG 12804
D/F WG 13 547
D/FWG13548
D/F WG13549
D/F WG13551
D/FWG13552
D/F WG 13 984
D/F WG 14274
PCBWG12108
Criteria
Met
N
Y
Y
Y
Y
Y
N
Y
Y
Y
Y
Y
Y
N
N
Duplicate RPDa (%)
23
2.1
1.2
5.7
4.6
14
none
16
5.9
3.6
0.0
20
(onU=l/2DLbasisb)
3.4
54
22
Comments
L6744-5, Ref 100 Newark Bay
Because this was above the 20% criteria, an additional
aliquot of this sample was prepared. Results for the
additional aliquot were within 1 1 % RPD from the original
results; therefore, this duplicate result was accepted.
L6744-9, Ref 122 Newark Bay
L6760-2, Ref 27 PE
L6760-14, RefSSPE
L6747-1, Ref 32 Midland
L6750-3, Ref 78 Tittabawassee River Soil
The duplicate processed with this batch was to be repeated
due to some analytes being <20x blank level. However, it
was reprocessed as a single sample and not a duplicate.
Samples in this set were accepted based on their agreement
with other replicates within the demonstration program.
L7 163-1, Ref 26 Nitro
L6751-14, Ref 83 North Carolina
L6751-7, Ref 135 North Carolina
L675 1 -1 , Ref 42 North Carolina
L7179-3, Ref 74 PE. Fails on a U=0 DL basis due to
presence of "K" flagged analytes in one replicate. When
compared on U-l/2 DL basis where "K" concentrations are
included in the TEQ calculation, the duplicate passed.
L7179-14, Ref 113PE
L7179-16, Ref 124PE
This was a PCB PE sample and contained only trace levels
of D/F. Replicate precision is affected because D/F content
is so low. This is not expected to indicate any problems
with precision within this sample set. Samples in this set
were accepted based on their agreement with other
replicates within the demonstration program.
L6744-2, Ref 49 Newark Bay
This result is only slightly above the acceptance criteria of
20%. The variability was influenced by 25% RPD for
PCB 126 (which has the highest TEF of the PCBs and,
therefore, a larger influence on total TEQ). The slight
exceedance in duplicate criteria was not considered to have
any significant impact on the data reported in this sample
batch. All samples in this set were also evaluated based on
their agreement with other replicates within the
demonstration program and deemed to be acceptable.
                                              C-4

-------
Sample Batch
Number
PCBWG12147
PCB WG12265
PCB WG12457
PCB WG12687
PCB WG12834
PCB WG12835
PCB WG12836
PCB WG1 3008
PCB WG13256
PCB WG13257
PCBWG13258
PCB WG13554
PCB WG14109
Criteria
Met
N
Y
N
Y
Y
N
Y
Y
Y
Y
Y
Y
N
Duplicate RPDa (%)
none
2.5
none
4.3
4.2
none
2.6
5.1
1.7
(on U=1/2DL basis)
15
19
12
85
(on U=1/2DL basis)
Comments
L6748-9, Ref 129 Brunswick
The duplicate sample for this batch required reprocessing.
When reprocessed, it was not prepared in duplicate.
Samples in this set were accepted based on the RPD of site
replicates that were processed within the batch (RPDs
<10%).
L6760-5, Ref35PE
L6760-16, Ref62PE
This duplicate set was to be repeated due to low internal
standard recovery. When repeated, it was not prepared in
duplicate. Data for this set was accepted because all
samples in the set were PE samples. These PE samples met
accuracy criteria and reproducibility criteria to other
replicates of the same PE material processed within the
demonstration.
L6762-12, Ref 169PE
L6750-8, Ref 164 Tittabawassee River Soil
Duplicate sample repeated in WG13258. Results reported
with that sample set. Three sets of sample replicates within
this batch were also compared and found to have <13.5%
RPD showing acceptable precision with this sample set.
L6751-6, Ref 126 North Carolina
L6750-6, Ref 121 Tittabawassee River Soil
L6761-3, Ref 74 PE. Fails on a U=0 DL basis due to
presence of "K" flagged analytes in one replicate. When
compared on U=l/2 DL basis where "K" concentrations are
included in the TEQ calculation, the duplicate passed.
L7 187-5, Ref 92 Tittabawassee River Soil
L6743-2, Ref36Nitro
L6762-l,Ref202PE
L7179-4, PE. Fails based on both U=0 and U=l/2 DL.
This was a blank PE sample and contained only trace levels
of PCBs. Replicate precision is affected because the PCB
content is so low. This is not expected to indicate any
problems with precision within this sample set. Samples in
this set were accepted based on their agreement with other
replicates within the demonstration program.
a Nondetects were assigned a concentration of zero unless otherwise noted and are referred to as U=0 DL values.
bU=l/2 DL indicates that non-detects were assigned a concentration equal to one-half the SDL and EMPC concentrations were assigned a value
   equal to the EMPC.
                                                               C-5

-------

-------
                  Appendix D
Summary of Developer and Reference Laboratory Data

-------

-------
Appendix D. Wako and Reference Laboratory One-to-One Matching
Sample Type
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Sample
Number
WAKO 176
WAKO 97
WAKO 141
WAKO 53
WAKO 208
WAKO 151
WAKO 33
WAKO 37
WAKO 194
WAKO 78
WAKO 185
WAKO 56
WAKO 207
WAKO 162
WAKO 82
WAKO 81
WAKO 147
WAKO 202
WAKO 26
WAKO 84
WAKO 72
WAKO 32
WAKO 187
WAKO 191
WAKO 159
WAKO 60
WAKO 132
WAKO 86
WAKO 51
WAKO 127
WAKO 117
WAKO 29
WAKO 160
WAKO 106
WAKO 199
WAKO 76
WAKO 130
WAKO 134
WAKO 41
WAKO 58
WAKO 182
WAKO 100
WAKO 183
WAKO 66
WAKO 68
WAKO 90
WAKO 124
WAKO 135
WAKO 171
WAKO 104
WAKO 116
WAKO 178
Measurement
Location
Field
Laboratory
Laboratory
Field
Field
Field
Field
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Field
Field
Field
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Sample Description
Brunswick #1
Brunswick #1
Brunswick #1
Brunswick #1
Brunswick #2
Brunswick #2
Brunswick #2
Brunswick #2
Brunswick #3
Brunswick #3
Brunswick #3
Brunswick #3
Midland #1
Midland #1
Midland #1
Midland #1
Midland #2
Midland #2
Midland #2
Midland #2
Midland #3
Midland #3
Midland #3
Midland #3
Midland #4
Midland #4
Midland #4
Midland #4
NC PCB Site #1
NC PCB Site #1
NC PCB Site #1
NC PCB Site #1
NC PCB Site #2
NC PCB Site #2
NC PCB Site #2
NC PCB Site #2
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
Newark Bay #1
Newark Bay #1
Newark Bay #1
Newark Bay #1
Newark Bay #2
Newark Bay #2
Newark Bay #2
Newark Bay #2
Newark Bay #3
Newark Bay #3
Newark Bay #3
Newark Bay #3
REP
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
Developer3
pg/g 2,3,7,8-
TCDD EQ
<17.8
415
159
<27.2
<34.3
<22.8
90.2
<45.3
99
250
64.5
65.2
238
64.5
110
290
132
59.3
117
155
78.8
323
126
133
<21.8
<28.5
<28.5
<51.0
>2800
1817
>3167
>4400
>2967
>3783
>3733
>4517
>3217
>4517
>3783
>2800
<28.2
<25.2
<28.2
28.5
72.7
<33.3
63.2
<28.5
<24.8
<25.2
>3167
<17.8
Reference
Laboratory1"
TEQn;F (pg/g)
67.2
71.6
61.7
67.8
49.5
72.8
56
60.4
12600
15200
13100
13600
222
241
269
268
208
179
197
192
185
174
176
161
25.7
26.4
31
25.8
788
1100
852
906
3400
3300
3430
3490
8320
8410
9360
10800
23
14
14.5
13.5
50.6
47.4
74.1
50.4
38.9
44.9
40.2
41.9
                                               D-l

-------
Sample Type
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Sample
Number
WAKO 131
WAKO 44
WAKO 52
WAKO 152
WAKO 156
WAKO 70
WAKO 149
WAKO 89
WAKO 95
WAKO 189
WAKO 105
WAKO 181
WAKO 73
WAKO 28
WAKO 125
WAKO 177
WAKO 54
WAKO 164
WAKO 197
WAKO 59
WAKO 80
WAKO 121
WAKO 205
WAKO 42
WAKO 123
WAKO 158
WAKO 113
WAKO 62
WAKO 128
WAKO 67
WAKO 118
WAKO 112
WAKO 34
WAKO 129
WAKO 150
WAKO 101
WAKO 98
WAKO 143
WAKO 55
WAKO 203
WAKO 27
WAKO 155
WAKO 36
WAKO 198
WAKO 200
WAKO 209
WAKO 50
WAKO 103
WAKO 85
WAKO 92
WAKO 161
WAKO 144
WAKO 24
WAKO 46
Measurement
Location
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Field
Field
Laboratory
Field
Laboratory
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Field
Field
Field
Field
Field
Field
Field
Sample Description
Newark Bay #4
Newark Bay #4
Newark Bay #4
Newark Bay #4
RaritanBay #1
RaritanBay #1
RaritanBay #1
RaritanBay #1
Raritan Bay #2
Raritan Bay #2
Raritan Bay #2
Raritan Bay #2
Raritan Bay #3
Raritan Bay #3
Raritan Bay #3
Raritan Bay #3
S aginaw River #1
Saginaw River #1
S aginaw River #1
Saginaw River #1
Saginaw River #2
Saginaw River #2
Saginaw River #2
Saginaw River #2
Saginaw River #3
Saginaw River #3
Saginaw River #3
Saginaw River #3
Solutia#l
Solutia#l
Solutia#l
Solutia#l
Solutia #2
Solutia #2
Solutia #2
Solutia #2
Solutia #3
Solutia #3
Solutia #3
Solutia #3
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #3
Titta. River Soil #3
Titta. River Soil #3
Titta. River Soil #3
Titta.. River Sed#l
Titta. River Sed #1
REP
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
Developer3
pg/g 2,3,7,8-
TCDD EQ
36.5
59.7
<37.5
<22.8
<21.8
<25.7
<22.8
<33.3
<33.3
<47.0
<25.2
<28.2
<25.7
<32.0
<24.8
<17.8
197
692
217
97.5
258
358
228
477
105.0
118
106.7
32.5
60
154
>3167
86.2
1252
410
432
1372
435
>3817
2367
362
1330
<21.8
<45.3
38.5
270
217
303
227
355
597
72.3
918
<32.0
86.5
Reference
Laboratory1"
TEQn;F (pg/g)
33.6
26.1
27.6
26.8
10.2
10.3
10.4
11.4
13.3
13.1
12.8
13
10.4
11.1
10.6
9.93
1050
683
1070
694
1110
953
1320
864
99.7
146
122
99.6
57.5
76.9
62
61.6
2090
1950
1860
2160
2810
2800
3000
3080
35
35.2
40
35.8
420
450
523
506
1050
676
1220
1300
1.05
1.11
D-2

-------
Sample Type
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
WAKO 188
WAKO 169
WAKO 102
WAKO 195
WAKO 153
WAKO 61
WAKO 154
WAKO 126
WAKO 65
WAKO 74
WAKO 30
WAKO 99
WAKO 140
WAKO 167
WAKO 71
WAKO 47
WAKO 122
WAKO 186
WAKO 91
WAKO 108
WAKO 136
WAKO 145
WAKO 19
WAKO 21
WAKO 12
WAKO 1 1
WAKO 5
WAKO 15
WAKO 22
WAKO 23
WAKO 16
WAKO 8
WAKO 9
WAKO 6
WAKO 2
WAKO 13
WAKO 1
WAKO 3
WAKO 17
WAKO 14
WAKO 20
WAKO 4
WAKO 10
WAKO 18
WAKO 7
WAKO 31
WAKO 175
WAKO 166
WAKO 87
WAKO 192
WAKO 115
WAKO 110
WAKO 49
WAKO 138
Measurement
Location
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Field
Field
Field
Field
Laboratory
Field
Field
Laboratory
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Field
Field
Field
Field
Laboratory
Field
Field
Field
Field
Field
Sample Description
Titta. River Sed #1
Titta. River Sed #1
Titta. River Sed #2
Titta. River Sed #2
Titta. River Sed #2
Titta. River Sed #2
Titta. River Sed #3
Titta. River Sed #3
Titta. River Sed #3
Titta. River Sed #3
WinonaPost#l
WinonaPost#l
WinonaPost#l
WinonaPost#l
Winona Post #2
Winona Post #2
Winona Post #2
Winona Post #2
Winona Post #3
Winona Post #3
Winona Post #3
Winona Post #3
Envir Extract #1
Envir Extract #1
Envir Extract #1
Envir Extract #1
Envir Extract #2
Envir Extract #2
Envir Extract #2
Envir Extract #2
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #2
Spike #2
Spike #2
Spike #2
Spike #3
Spike #3
Spike #3
Spike #3
Cambridge 5183
Cambridge 5 183
Cambridge 5183
Cambridge 5 183
Cambridge 5183
Cambridge 5 183
Cambridge 5183
Cambridge 5 1 84
Cambridge 5184
REP
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
1
2
Developer3
pg/g 2,3,7,8-
TCDD EQ
<47.0
50.2
55.2
302
<22.8
52.2
<21.8
39.8
34.7
<25.7
64.3
77.2
68.3
57
68.2
190
65.3
227
293
91.2
63.7
138
405
243
383
150
68.3
<24.5
<34.3
<32.0
<24.5
<24.5
<24.5
87.5
71.0
<24.5
<37.5
69.2
76.2
<24.5
60.5
61.5
<24.5
<29.7
<34.3
<46.2
<17.8
31.3
88.7
<47.0
>3167
64.3
1620
85.5
Reference
Laboratory1"
TEQn;F (pg/g)
1
1.7
52.8
123
66.1
94.1
13
11.2
12.7
13.8
7290
7370
7450
7160
9720
9770
9200
11300
10300
9770
9320
9870
175
444
176
439
55.3
53.3
53.1
53.6
0.504
0.509
0.537
0.524
0.585
0.576
0.52
91.6
91.8
89.1
100
0.324
0.348
0.363
0.268
4.78
4.08
4.06
3.56
3.89
5.93
3.89
187
188
D-3

-------
Sample Type
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
WAKO 63
WAKO 163
WAKO 179
WAKO 114
WAKO 109
WAKO 119
WAKO 173
WAKO 196
WAKO 172
WAKO 206
WAKO 35
WAKO 69
WAKO 43
WAKO 57
WAKO 96
WAKO 165
WAKO 88
WAKO 157
WAKO 204
WAKO 83
WAKO 48
WAKO 201
WAKO 107
WAKO 148
WAKO 39
WAKO 139
WAKO 93
WAKO 174
WAKO 94
WAKO 193
WAKO 40
WAKO 168
WAKO 75
WAKO 180
WAKO 133
WAKO 64
WAKO 170
WAKO 25
WAKO 38
WAKO 137
WAKO 45
WAKO 142
WAKO 1 1 1
WAKO 190
WAKO 120
WAKO 146
WAKO 77
WAKO 184
WAKO 79
Measurement
Location
Field
Field
Field
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Laboratory
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Freld
Laboratory
Freld
Freld
Freld
Freld
Sample Description
Cambridge 5 184
Cambridge 5 184
ERA Aroclor
ERA Aroclor
ERA Aroclor
ERA Aroclor
ERA Blank
ERA Blank
ERA Blank
ERA Blank
ERA Blank
ERA Blank
ERA Blank
ERA Blank
ERA PAH
ERA PAH
ERA PAH
ERA PAH
ERAPCB 100
ERAPCB 100
ERAPCB 100
ERAPCB 100
ERAPCB 10000
ERAPCB 10000
ERAPCB 10000
ERAPCB 10000
ERATCDD10
ERATCDD 10
ERATCDD 10
ERATCDD 10
ERA TCDD 30
ERA TCDD 30
ERA TCDD 30
ERA TCDD 30
LCG CRM-529
LCG CRM-529
LCG CRM-529
LCG CRM-529
NIST 1944
NIST 1944
NIST 1944
NIST 1944
Wellington WMS-01
WellrngtonWMS-01
Wellrngton WMS-01
Wellrngton WMS-01
Wellmgton WMS-01
Wellrngton WMS-01
Wellmgton WMS-01
REP
3
4
1
2
3
4
1
2
3
4
5
6
7
8
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
Developer3
pg/g 2,3,7,8-
TCDD EQ
217
122
<17.8
658
60.5
>3900
<17.8
<34.3
<17.8
<34.3
<26.3
114
<45.3
<26.3
<33.3
38.5
<33.3
<21.8
57.8
<35.2
207
<34.3
50.2
25.7
168
61.5
<33.3
<17.8
<33.3
<47.0
63.8
<24.8
<25.7
<17.8
457
367
1127
<32.0
332
75.8
223
168
109
<47.0
93.2
387
37
<28.2
<35.2
Reference
Laboratory1"
TEQn;F (pg/g)
173
180
36.4
32.9
37.9
35.5
0.0942
0.0728
0.237
0.307
0.113
0.0524
0.211
0.0692
0.159
0.141
0.161
0.248
0.0386
NAC
0.053
0.127
0.204
0.507
0.105
0.0628
8.69
9.28
8.44
8.2
27.4
25.3
24.8
23.9
NA(C)
6930
6900
7190
237
206
252
219
68
65.7
61.9
66.1
68
65.7
65.4
a Data listed exactly as reported by the developer.
b Qualifier flags (e.g., J and K flags) included in the raw data have been removed for the purposes of statistical analysis.
0 Reference laboratory data was discarded due to laboratory sample preparation error.
                                                            D-4

-------