»EPA
United States
Environmental Protection
Agency
Office of Research and
Development
Washington, DC 20460
                                 EPA/540/R-05/004
                                 March 2005
           Innovative Technology
           Verification Report
           Technologies for Monitoring
           and Measurement of Dioxin
           and Dioxin-like Compounds
           in Soil and Sediment
           CAPE Technologies LLC
           DF1 Dioxin/Furan Immunoassay Kit
           PCB TEQ Immunoassay Kit

-------

-------
                                EPA/540/R-05/004
                                March 2005
   Innovative Technology
     Verification  Report

     CAPE Technologies LLC
DF1 Dioxin/Furan Immunoassay Kit
    PCB TEQ Immunoassay Kit
                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's 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 CAPE Technologies DFl Dioxin/Furan and PCB
TEQ Immunoassay kits. 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 each 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 CAPE  Technologies DFl Dioxin/Furan and PCB TEQ Immunoassay kits are immunoassay techniques that report
the total toxicity equivalents (TEQ) of dioxin/furans and polychlorinated biphenyls (PCBs), respectively. As part of
the performance evaluation, the technology results were compared to TEQ results generated by a reference laboratory,
AXYS Analytical  Services, using EPA Methods 1613B and 1668A, which involve the use of HRMS.  It should be
noted that the results generated by the CAPE Technologies kits may not  directly correlate to HRMS TEQ in all cases
because it is known that the congener responses and cross-reactivities of the kits are not identical to the toxicity
equivalency factors that are used to convert congener HRMS concentration values to TEQ. The effect of
cross-reactivities may contribute  to this technology's reporting results that are biased high or low compared to HRMS
TEQ results.  Therefore, these kits should not be viewed as producing an equivalent measurement value to HRMS
TEQ, but as a screening value to approximate HRMS TEQ. As described in CAPE Technologies literature, the best
results for immunoassay screening are obtained on a single site basis. The ideal approach involves partially
characterizing a site by HRMS, using those results to develop a site specific immunoassay calibration, and refining
that calibration overtime, based on an ongoing stream of confirmatory HRMS samples. This approach was not
evaluated during this demonstration; samples from multiple sites were pooled and a single calibration was used.

A summary of the performance of the CAPE Technologies DFl Dioxin/Furan and PCB TEQ Immunoassay kits is as
follows: The CAPE Technologies kits generally reported data higher than the certified PE and reference laboratory
values. The technology's estimated method detection limit [12 to 35 picogram per gram (pg/g)] was higher than what
was reported by the developer (1 pg/g TEQ). The CAPE Technologies TEQD/F results that were generated in the
laboratory and in the field for replicate samples were statistically different for 19% of the samples, and of these
samples, CAPE Technologies laboratory results were more comparable to the reference laboratory results. No
significant effect was observed for the reproducibility of CAPE Technologies results by matrix type (soil vs. sediment
vs. extract)  or by sample type (PE vs. environmental vs extract). A slight effect was observed for total TEQ values by
PAH concentration, but the effect was not statistically significant for TEQD/F or TEQPCB.  The technology had a rate of
false negative results of 3 to 5% around 20 pg/g TEQ, with false positive rates ranging from 11 to 14%. However,
CAPE Technologies's false positive and false negative rates around 50 pg/g were generally lower for all three TEQ
types, ranging from 4 to 10%. These data suggest the CAPE Technologies kits 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 both the cost ($59,234 vs. $398,029) and the time (three weeks vs.
eight months) to analyze the 209 demonstration samples were significantly less than those of the reference laboratory.
                                                    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	 5
         1.2.3   Overview of Field Demonstration	 5

2   Description of CAPE Technologies DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits	 6
    2.1   Company History  	 6
    2.2   Product History	 6
    2.3   Technology Description  	 7
    2.4   Developer Contact Information 	 9

3   Demonstration and Environmental Site Descriptions  	 12
    3.1   Demonstration Site Description and Selection Process	 12
    3.2   Description of Sampling Locations 	 13
         3.2.1   Soil Sampling Locations	 13
         3.2.2   Sediment Sampling Sites	 15

4   Demonstration Approach 	 17
    4.1   Demonstration Objectives  	 17
    4.2   Toxicity Equivalents	 17
    4.3   Overview of Demonstration Samples	 19
         4.3.1   PE Samples  	 19
         4.3.2   Environmental Samples	 23
         4.3.3   Extracts  	 26
    4.4   Sample Handling	 26
    4.5   Pre-Demonstration Study	 27
    4.6   Execution of Field Demonstration 	 28

-------
                                       Contents (continued)

                                                                                                   Page

    4.7   Assessment of Primary and Secondary Objectives  	 28
         4.7.1   Primary Objective PI: Accuracy	 28
         4.7.2   Primary Objective P2: Precision	 28
         4.7.3   Primary Objective P3: Comparability	 29
         4.7.4   Primary Objective P4: Method Detection Limit	 30
         4.7.5   Primary Objective P5: False Positive/False Negative Results  	 30
         4.7.6   Primary Objective P6: Matrix Effects	 30
         4.7.7   Primary Objective P7: Technology Costs	 31
         4.7.8   Secondary Objective SI: Skills Level of Operator	 31
         4.7.9   Secondary Objective S2: Health and Safety Aspects 	 31
         4.7.10  Secondary Objective S3: Portability	 31
         4.7.11  Secondary Objective S4: Sample Throughput 	 31

5   Confirmatory Process  	 32
    5.1   Traditional Methods for Measurement of Dioxin and Dioxin-Like
         Compounds in Soil and Sediment  	 32
         5.1.1   Fiigh-Resolution Mass Spectrometry  	 32
         5.1.2   Low-Resolution Mass Spectrometry	 32
         5.1.3   PCB Methods	 32
         5.1.4   Reference Method Selection	 33
    5.2   Characterization of Environmental Samples 	 33
         5.2.1   Dioxins and Furans 	 33
         5.2.2   PCBs  	 34
         5.2.3   PAHs  	 34
    5.3   Reference Laboratory Selection 	 34
    5.4   Reference Laboratory Sample Preparation and Analytical Methods	 35
         5.4.1   Dioxin/Furan Analysis	 35
         5.4.2   PCB Analysis	 35
         5.4.3   TEQ Calculations 	 35

6   Assessment of Reference Method Data Quality  	 37
    6.1   QA Audits	 37
    6.2   QC Results 	 38
         6.2.1   Holding Times and Storage Conditions  	 38
         6.2.2   Chain of Custody	 38
         6.2.3   Standard Concentrations	 38
         6.2.4   Initial and Continuing Calibration  	 38
         6.2.5   Column Performance and Instrument Resolution 	 39
         6.2.6   Method Blanks	 39
         6.2.7   Internal Standard Recovery	 39
         6.2.8   Laboratory Control Spikes	 39
         6.2.9   Sample Batch Duplicates  	 39
    6.3   Evaluation of Primary Objective PI: Accuracy	 39
    6.4   Evaluation of Primary Objective P2: Precision	 40
    6.5   Comparability to Characterization Data	 42
    6.6   Performance Summary	 43
                                                   VI

-------
                                       Contents (continued)

                                                                                                  Page

7   Performance of CAPE Technologies DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits	 44
    7.1   Evaluation of DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits Performance	 44
         7.1.1   Evaluation of Primary Objective PI: Accuracy  	 44
         7.1.2   Evaluation of Primary Objective P2: Precision  	 44
         7.1.3   Evaluation of Primary Objective P3: Comparability 	 47
         7.1.4   Evaluation of Primary Objective P4: Estimated Method Detection Limit	 48
         7.1.5   Evaluation of Primary Objective P5: False Positive/False Negative Results	 49
         7.1.6   Evaluation of Primary Objective P6: Matrix Effects 	 50
         7.1.7   Evaluation of Primary Objective P7: Technology Costs  	 50
    7.2   Observer Report: Evaluation of Secondary Objectives 	 50
         7.2.1   Evaluation of Secondary Objective SI: Skill Level of Operator 	 54
         7.2.2   Evaluation of Secondary Objective S2: Health and Safety Aspects	 54
         7.2.3   Evaluation of Secondary Objective S3: Portability 	 54
         7.2.4   Evaluation of Secondary Objective S4: Throughput 	 55
         7.2.5   Miscellaneous Observer Notes  	 55

8   Economic Analysis	 56
    8.1   Issues and Assumptions  	 56
         8.1.1   Capital Equipment Cost 	 56
         8.1.2   Cost of Supplies	 56
         8.1.3   Support Equipment Cost	 57
         8.1.4   Labor Cost	 57
         8.1.5   Investigation-Derived Waste Disposal Cost	 57
         8.1.6   Costs Not Included  	 57
    8.2   DF1 Dioxin/Furan and PCB TEQ Immunoassay Kit Costs	 58
         8.2.1   Capital Equipment Cost 	 58
         8.2.2   Cost of Supplies	 58
         8.2.3   Support Equipment Cost	 60
         8.2.4   Labor Cost	 60
         8.2.5   Investigation-Derived Waste Disposal Cost	 60
         8.2.6   Summary of DF1 Dioxin/Furan and PCB Immunoassay Kit Costs	 60
    8.3   Reference Method Costs	 60
    8.4   Comparison of Economic Analysis Results	 61

9   Technology Performance Summary	 62

10  References	 65

Appendix A     SITE Monitoring and Measurement Technology Program Verification Statement	A-l
Appendix B      Supplemental Information Supplied by the Developer 	 B-l
Appendix C      Reference Laboratory Method Blank and Duplicate Results Summary	 C-l
Appendix D     Summary of Developer and Reference Laboratory Data  	D-l
                                                  vn

-------
                                       Contents (continued)

                                                                                                  Page

                                                Figures

Figure 1-1. Representative dioxin, furan, and polychlorinated biphenyl structure	 3
Figure 2-1. CAPE Technologies DF1 Dioxin/Furan Immunoassay kit	 7
Figure 2-2. CAPE Technologies DF1 Immunoassay kit in operation during the field demonstration	 9
Figure 6-1. Comparison of reference laboratory and characterization D/F data for environmental samples	42


                                                Tables

2-1.    Cross-Reactivity of the DF1 Immunoassay Kit	 10
2-2.    Cross-Reactivity of the PCB TEQ Immunoassay Kit 	 11
3-1.    Summary of Environmental Sampling Locations  	 14
4-1.    World Health Organization Toxicity Equivalency Factor Values	 18
4-2.    Distribution of Samples forthe Evaluation of Performance Objectives	 20
4-3.    Number and Type of Samples Analyzed in the Demonstration  	 20
4-4    Summary of Performance Evaluation Samples	 21
4-5.    Characterization and Homogenization Analysis Results for Environmental Samples	 25
4-6    Distribution of Extract Samples	 26
5-1.    Calibration Range of HRMS Dioxin/Furan Method	 32
5-2.    Calibration Range of LRMS Dioxin/Furan Method  	 32
6-1.    Objective PI Accuracy - Percent Recovery	 40
6-2.    Evaluation of Interferences  	 40
6-3a.   Objective P2 Precision - Relative Standard  Deviation	 41
6-3b.   Objective P2 Precision - Relative Standard  Deviation (By Sample Type)	 42
6-4.    Reference Method Performance Summary - Primary Objectives	 43
7-1.    Objective PI Accuracy - Percent Recovery	 45
7-2a.   Objective P2 - RSD as a Description of Precision by Sample 	 45
7-2b.   Objective P2 - RSD as a Description of Precision by Sample Type	 46
7-3    Objective P3 - Comparability Summary Statistics of RPD	 47
7-4.    Objective P3 - Comparability Using Interval Assessment 	 48
7-5.    Objective P3 - Comparability for Blank Samples	 48
7-6a.   Objective P4 - Estimated MDL for TEQD/F and TEQPCB	 49
7-6b.   Objective P4 - Estimated MDL for Total TEQ	 49
7-7.    Objective P5 - False Positive/False Negative Results	 49
7-8    Objective P6 - Matrix Effects Using Descriptive Statistics and ANOVA Results Comparing
       TEQD/F Replicate Analysis Conducted During the Field Demonstration and in the Laboratory	 51
7-9    Objective P6-Matrix Effects Using RSD as a Description of Precision by Soil, Sediment, and Extract ... 52
7-10   Objective P6 - Matrix Effects Using RSD as a Description of Precision by PAH Concentration Levels
       (Environmental Samples Only) 	 53
7-11   Objective P6 - Matrix Effects of Known Interferences Using PE Materials  	 53
8-1.    Cost Summary  	 59
8-2.    Reference Method Cost Summary	 61
9-1.    CAPE Technologies LLC DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits Performance Summary -
       Primary Objectives	 63
9-2.    CAPE Technologies LLC DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits Performance Summary -
       Secondary Objectives	 64
                                                  Vlll

-------
                          Abbreviations, Acronyms, and Symbols
Ah
ANOVA
ASTM
ATSDR
CIL
cm
CoA
COC
CRM
DER
D/F
DNR
D/QAPP
EIA
ELC
ELISA
EMDL
EMPC
EPA
ERA
FDA
g
GC
HPLC/GPC
HRP
HRGC
HRMS
i.d.
IDW
ITVR
kg
L
LRMS
aryl hydrocarbon
analysis of variance
American Society for Testing and Materials
Agency for Toxic Substances and Disease Registry
Cambridge Isotope Laboratories
centimeter
Certificate of Analysis
chain of custody
certified reference material
data evaluation report
dioxin/furan
Department of Natural Resources
demonstration and quality assurance project plan
enzyme immunoassay
Environmental Learning Center
enzyme-linked immunosorbent assay
estimated method detection limit
estimated maximum possible concentration
Environmental Protection Agency
Environmental Resource Associates
Food and Drug Administration
gram
gas chromatography
high-performance liquid chromatography/gel permeation chromatography
horseradish peroxidase
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
mg
mL
mm
MDL
MMT
MS
NERL
ng
NIST
NOAA
OD
ORD
PAH
PCB
PCDD/F
PCP
PE
Pg
ppb
ppm
ppt
psi
QA/QC
RM
RPD
RSD
SDL
SIM
SITE
SOP
SRM
TCDD
TEF
microliter
micrometer
meter
Michigan Department of Environmental Quality
milligram
milliliter
millimeter
method detection limit
Monitoring and Measurement Technology
mass spectrometry
National Exposure Research Laboratory
nanogram
National Institute for Standards and Technology
National Oceanic and Atmospheric Administration
optical density
Office of Research and Development
polynuclear aromatic hydrocarbons
polychlorinated biphenyl
polychlorinated dibenzo-p-dioxin/dibenzofuran
pentachlorophenol
performance evaluation
picogram
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

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

TEG                 tetraethylene glycol
TEQ                 toxicity equivalent
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 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.
                                                  xn

-------
                                              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); the CAPE Technologies LLC
DF1 Dioxin/Furan and polychlorinated biphenyl (PCB)
toxicity equivalent (TEQ) Immunoassay kits (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) demon-
stration 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
developers, 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-/>-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 environ-
ment. 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) The 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.
Battelle'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
environmental 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
responsibilities 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
distinguishing 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 characterization 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 CAPE
Technologies 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
performance evaluation of the CAPE Technologies LLC
DF1 Dioxin/Furan and  PCB TEQ Immunoassay kits is
presented in this ITVR. Separate ITVRs have been
published for the other four participating technologies.

-------
                                             Chapter 2
                    Description of CAPE Technologies DF1 Dioxin/Furan
                                and PCB TEQ Immunoassay Kits
This technology description is based on information
provided by CAPE Technologies 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 DF1 Dioxin/Furan
Immunoassay Kit from CAPE Technologies is an
enzyme immunoassay (EIA) test kit containing a
polyclonal antibody specific for PCDD/Fs. The
company's PCB TEQ Immunoassay Kit from CAPE
Technologies is an EIA test kit containing a polyclonal
antibody specific for dioxin-like PCBs. Both semi-
quantitative screening and quantitative analysis are
possible with these kits, but this evaluation focused only
on quantitative analysis. Samples can be prepared for
analysis by EIA using a variety of methods.  Extracts of
soil, sediment, food, water, fly ash, stack gas, tissue, or
other samples that have been prepared by conventional
extraction  methods can be exchanged to a water-miscible
solvent system for analysis using the CAPE
Technologies immunoassay kits. More commonly,
immunoassay specific sample  preparation methods are
used to reduce the time, effort, and cost of sample
preparation.  Design and operation of the two kits are
nearly identical except for the  combination of antibody
and enzyme conjugate that is responsible for the
specificity of each kit.  One sample preparation method
can be used for both kits, providing  separate
dioxin/furan and PCB fractions. These fractions can be
analyzed by the respective kits, giving separate TEQ
results for  both dioxin/furan and PCB.

2.1    Company History
CAPE Technologies was founded in 1996 by
Robert Carlson and Robert Harrison to develop and
market immunoassay test kits and support technology
for analysis of dioxins and related compounds. Its
headquarters are in South Portland, Maine. Primary
products are immunoassay kits and sample preparation
kits for analysis of dioxin and related compounds;
analytical services are also offered.

The principals of CAPE Technologies have more than
40 years combined experience in the design, develop-
ment, validation, marketing, and technical support of
immunoassays for environmental analysis, including five
EPA 4000 series methods. In 2000, CAPE Technologies
was selected by EPA Region 1 as an Environmental
Technology Innovator of the Year.

2.2    Product History
The CAPE Technologies DF1 Dioxin/Furan
Immunoassay Kit was first developed in 1996.
Optimization and validation of the immunoassay as a
TEQ predictor were pursued over the next two years in
collaboration with several established dioxin laboratories
around the world. During the same time period, the first
immunoassay-specific sample processing methods were
developed.  Commercial sales of the DPI kit began in
late 1998.  Concurrent refinement of sample preparation
methods resulted  in a simple extraction and one step
oxidative cleanup for high pg/g levels in soil. This
combination of sample preparation method and DF1
immunoassay was applied to rapid soil screening using
field samples from two well known U.S. dioxin sites.
The resulting data were reviewed by the U.S. EPA,
leading to the acceptance in June 2001 of SW-846
Method 4025 based on the DPI kit. During this
validation process a more rigorous cleanup method was
developed for soils, based on portions of the SW-846
Method 8290 cleanup. This method, when used with the
DF1 kit, is referred to as modified Method 4025, or
4025m. Method 4025m allows for low pg/g analysis in
solid samples using a 5-g sample and easily removes
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                   6

-------
high levels of aliphatic oils which occur commonly in
dioxin contaminated soils.  Commercial sale of this kit
based sample preparation system began in 2002. The
CAPE Technologies portion of the current study used
this system exclusively.

An early PCB TEQ kit was developed and partially
validated by CAPE Technologies before 1998, but was
not released to market because of unacceptably high
cross-reactivity for PCB 77. After various studies
suggested changes in congener recognition profiles, the
CAPE Technologies PCB TEQ kit was developed in
1998 based on a new antibody with improved
specificity. In 2000, the sample cleanup of Method
4025m was modified to provide a separate PCB fraction
for immunoassay analysis.  The resulting single cleanup
and fractionation were used by CAPE Technologies in
the current study.  In 2002, a validation study was started
for the purpose of obtaining EPA acceptance of this
method as SW-846 Method 4026.  Partly because of the
delay in EPA's dioxin reassessment, the validation study
was put on indefinite hold. Commercial sales of the
PCB TEQ Kit began in 2004 and are  expected to spur
reopening of the Method 4026 validation study.

2.3     Technology Description
The DF1 Dioxin/Furan Immunoassay Kit (Figure 2-1)
and the PCB TEQ Immunoassay Kit are nearly identical
in design and operation.  They differ primarily in the
antibody and competitor-horseradish  peroxidase (HRP)
conjugate used, and in the specificity resulting from
these specially developed reagents.  Both kits are
designed to provide results as TEQ concentrations by
responding to the toxic dioxin/furan or PCB congeners
in approximate correlation with their toxic equivalency
factors (TEFs).  Both tests recognize multiple congeners,
preferentially targeting congeners with high TEF values,
i.e., those with the highest toxicity relative to
2,3,7,8-TCDD.  The specificity of the dioxin/furan test is
predominantly for dioxins and furans that contain 3 to 6
chlorines, with a strong preference for the 2,3,7,8
chlorinated congeners. This specificity roughly parallels
the TEF values of the individual dioxin and furan
congeners.  The specificity of the  PCB TEQ test is
predominantly for non-ortho and mono-ortho chlorinated
congeners, with a strong preference  for PCBs 126 and
169. This specificity roughly parallels the TEF values of
the individual PCB congeners.  Both tests have only
minimal recognition of the target compounds of the
other test.

The immunoassay specific sample preparation begins
with an organic solvent extraction. The extracts are then
processed through an immunoassay specific cleanup.  In
the case of this evaluation, the cleanup combines two
familiar parts of the Method 8290 cleanup, but in a way
that allows for rapid batch processing using inexpensive
disposable columns and no specialized equipment.
Since the cleanup is performed in solvents incompatible
with the immunoassays, a solvent exchange is required
Figure 2-1.  CAPE Technologies DF1 Dioxin/Furan Immunoassay kit.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
after the cleanup.  Dioxins, ftirans, and dioxin-like PCBs
have very low volatility and are retained during this
solvent exchange in a small volume of a keeper solution
(Triton X-100 detergent in tetraethylene glycol [TEG])
after evaporation of the original solvent. Methanol is
added to dilute this solution, and the
methanol-TEG-Triton mixture is added  directly to the
immunoassay tubes. During the first immunoassay
incubation, analyte molecules are specifically bound by
the analyte-specific antibodies, which have been
immobilized on the immunoassay tube surface. After
washing away the unbound material, the bound analyte
molecules remain, and a competitor-HRP conjugate is
added. Bound analyte molecules occupy the binding
sites of the antibodies in proportion to the dioxin/furan
or dioxin-like PCB content of the sample, reducing the
binding of the competitor-HRP conjugate. After an
incubation period, unbound conjugate is removed, and
the test tubes are washed thoroughly. The incubation
period can be 2 to 24 hours; for convenience, during the
field demonstration, the samples were incubated
overnight (-12 hours). The amount of conjugate bound
by the anti-analyte antibody is  inversely related to the
amount of analyte originally present in the sample.
Finally, a solution of chromogenic HRP substrate and
hydrogen peroxide is added to the test tubes.  Color
development is directly proportional to enzyme
concentration and inversely related to the dioxin/furan or
dioxin-like PCB concentration in the original sample.
The test tubes are analyzed using a tube reader or
spectrophotometer to measure  the optical density (OD).
The OD values of unknown samples are compared to the
OD values of standards to determine the level of
dioxin/furan or dioxin-like PCB in the samples.

The final measured EIA response is the sum of the
individual congener responses. Both the dioxin/furan  kit
and the PCB TEQ kit correlate with TEQ because the
cross-reaction profile of each kit roughly correlates with
the TEF values of its respective target congeners.

Accuracy may vary solely because of the variability of
congener composition.  To maximize accuracy, the
variability of congener composition in the target sample
population should be known. The best performance is
achieved  when all samples are from a single group that
share as many properties as possible (common source of
contamination, similar congener composition, similar
sample matrix, etc.).

The limit of detection for the CAPE Technologies
Dioxin/Furan Immunoassay Kit is approximately 4 pg of
2,3,7,8-TCDD, equivalent to 4 pg of dioxin/furan TEQ.
The limit of detection for CAPE Technologies' PCB
TEQ Immunoassay Kit is approximately 10 pg of PCB
126, equivalent to 1 pg of PCB TEQ. These detection
limits make both tests sufficiently sensitive for analysis
at levels below 10 pg/g TEQ using a 5-g sample. Less
sensitive performance is possible by decreasing the
amount of sample extract added to the cleanup
procedure.

Regardless of sample load, the manufacturer's
recommendations for extract cleanup must be followed
closely in order to obtain acceptable results. Raw
pg/tube results must be converted to raw pg/g in the
original sample by use of the proper dilution and volume
factors.  For accurate absolute quantitation, raw pg/g
results must be adjusted by a calibration adjustment
factor.  This factor is empirically determined by the user
based on a variety of QA samples. Calibration adjust-
ment factors can be estimated before analysis, but they
are best refined on an ongoing basis by use of
appropriate QA samples  (see Appendix B).

During the demonstration, the dilution protocol used was
designed to provide approximate quantitation of samples
that were high relative to the primary target level, while
using a minimum of resources (i.e., the residue of the
sample already processed and analyzed). The protocol
was not designed for maximum accuracy and may
indeed have problems related only to  potential
overloading of cleanup columns. Most applications of
the kits would not require a more refined result than this,
but if such a result were required, the first result would
be used to select a lower sample load and another
(smaller) aliquot would be processed.

Matrix detection limits will vary according to matrix,
sample size, and dilution factor. A single experienced
analyst can process approximately 20 samples per day
using the procedure evaluated in this  study.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
The following kits are available from CAPE
Technologies (all DF1 kits have parallel PCB TEQ kits):
    DF1-ST-A, a small starter package containing two
    DF1-12 kits (40 antibody-coated tubes and matching
    liquid reagents), one Grip-Rack, and one set of
    dioxin standards, plus two check samples of dioxin
    in toluene made by Wellington Labs.

    DF1-ST-B, a large starter package containing one
    DF1-60 kit (100 antibody-coated tubes and matching
    liquid reagents), one Grip-Rack, and one set of
    dioxin standards, plus two check samples of dioxin
    in toluene made by Wellington Labs.

After the purchase of one starter package, subsequent
purchases are either the  DF1-12 or the DF1-60. These
kits do not include dioxin standards and check samples,
which must be ordered separately. The DF1-12 kit for
screening analysis of 12 samples includes 20 antibody-
coated tubes and matching liquid reagents. The DF1-60
kit for screening analysis of 60 samples includes 100
antibody-coated tubes and matching liquid reagents.

Tables 2-1 and 2-2 describe the cross-reactivity of the
DPI and PCB TEQ immunoassay kits, respectively.

This is the method that CAPE Technologies
implemented during the field demonstration. A photo of
the technology in operation during the demonstration is
presented in Figure 2-2. CAPE Technologies provided
supplemental information about the performance of their
technology during the demonstration and it is presented
in Appendix B.

2.4    Developer Contact Information
Additional information about the DF1 and PCB TEQ
Immunoassay kits can be obtained by contacting:
       CAPE Technologies LLC
       Bob Harrison
       3 Adams Street
       South Portland, Maine 04106-1604
       Telephone: (207) 741-2995
       E-mail: cape-tech@ceemaine.org
       Web site: www.cape-tech.com
     Figure 2-2. CAPE Technologies DF1
     Immunoassay kit in operation during the
     field demonstration.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                    9

-------
              Table 2-1. Cross-Reactivity of the DF1 Immunoassay Kit
                 Toxic Dioxin Congeners
                 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
                 OCDD

                 Toxic Furan Congeners
                 2,3,7,8-TCDF
                 1,2,3,7,8-PeCDF
                 2,3,4,7,8-PeCDF
                 1,2,3,4,7,8-HxCDF
                 1,2,3,6,7,8-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
                 OCDF

                 Other PCDD/F Congeners
                 2,3-dichlorodibenzo-p-dioxin
                 2,7-dichlorodibenzo-p-dioxin
                 2,3-dichlorodibenzofuran
                 2,7-dichlorodibenzofuran
                 2,3,7-trichlorodibenzo-p-dioxin
                 2,3,8-trichlorodibenzofuran
                 1,2,3,4-TCDD
                 1,2,3,4-TCDF
                 1,3,6,8-TCDD
                 1,3,6,8-TCDF

                 Polychlorinated Biphenyls
                 3,3',4,4' (PCB 77)
                 3,3',4,4',5 (PCB 126)
                 2,2',4,4',5 (PCB 153)
                 3,3',4,4',5,5' (PCB 169)
                 Aroclor 1254
% Crossreactivity3
         100
         105
           1.6
           7.9
          39
           0.7
          0.001
          20
           4.6
          17
           0.4
           1.0
           3.3
           4.9
           0.02
           0.9
          0.001
           0.13
           0.003
           0.02
          O.002
          24
           0.26
          0.001
          O.001
           0.05
           0.007
           0.4
           0.5
          0.1
          O.I
          O.I
                1 Response curves were prepared for each congener as noted. The
                 percent cross-reactivity = (((2,3,7,8-TCDD 150) ^ (congener 150))
                 x 100). Values are typically based on two to four independent
                 curves, each containing at least four concentrations.
Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                           10

-------
Table 2-2. Cross-Reactivity of the PCB TEQ Immunoassay Kit
Category
Non-Ortho



Mono -Ortho







Di-Ortho

Aroclor 1254
PCB No.
44
49
52
66
70
82
84
85
87
92
95
97
PCB No.
99
101
110
128
132
138
141
149
153
158
163
168
PCB No. Chlorination Pattern
77
81
126
169
105
114
118
123
156
157
167
189
170
180
3,4/3',4'
3,4,5/4'
3,4,5 /3',4'
3,4,5 /3',4',5'
2,3,4 /3',4'
2,3,4,5/4'
2,4,5 /3',4'
3,4,5 /2',4'
2,3,4,5 /3',4'
2,3,4 /3',4',5'
2,4,5 /3',4',5'
2,3,4,5 /3',4',5'
2,3,4,5 /3',4',5'
2,3,4,5 /2',4',5'
TEFa % Cross-Reactivity"
0.0001
0.0001
0.1
0.01
0.0001
0.0005
0.0001
0.0001
0.0005
0.0005
0.00001
0.0001
o.ooo r
0.00001 c
0.90
0.54
100
232
0.017
0.0063
0.0064
0.11
0.43
1.1
0.93
9.2
0.0083
0.0023
Common Congeners (no assigned TEF values)


























Chlorination Pattern
2,3/2',5'
2,4/2',5'
2,5/2',5'
2,4/3',4'
2,5/3',4'
2,3,4 /2',3'
2,3,6 /2',3'
2,3,4 /2',4'
2,3,4 /2',5'
2,3,5 /2',5'
2,3,6 /2',5'
2,4,5 /2',3'
Chlorination Pattern
2,4,5 /2',4'
2,4,5 /2',5'
2,3,6 /3',4'
2,3,4 /2',3',4'
2,3,4 /2',3',6'
2,3,4 /2',4',5'
2,3,4,5 /2',5'
2,3,6 /2',4',5'
2,4,5 /2',4',5'
2,3,4,6 /3',4'
2,3,5,6 /3',4'
2,4,6 /3',4',5'
% Cross-Reactivity
0.0002
0.0002
0.0002
0.0058
0.013
0.0009
0.0002
0.0005
0.0009
0.0005
0.0005
0.0005
% Cross-Reactivity
0.0002
0.0002
0.0005
0.0035
0.0005
0.0002
0.0005
0.0018
0.0023
0.0005
0.0021
0.0028


























1 TEF values are from Van den Berg et al.<5)
b Response curves were prepared for each congener as noted. The percent cross-reactivity = (((congener 150) ^ (PCB 126 159)) x 100). Values
  are typically based on two to four independent curves, each containing at least four concentrations.
c No TEF assigned by WHO.
               Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                            11

-------
                                              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,
                                                    12

-------
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
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,
pentachlorophenol (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 contamina-
tion, 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 part-per-
trillion (ppt) 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 ppt range. The final two samples
were collected from Department of Natural Resources
                                                   13

-------
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
o
J
4
3
3
4
3
o
3
o
3
3
32
 (DNR)-owned property in Saginaw, 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 milligrams/kilogram (mg/kg) for
pentachlorophenol 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 dichlorophenoxyacetic acid) and 2,4,5-T
(2,4,5 trichlorophenoxyacetic acid). Manufacture of the
chemical herbicide began at the site in 1948 and ceased
                                                    14

-------
in 1969. The source of the 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
information 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.
                                                   15

-------
3.2.2.3 Saginaw River
Saginaw River 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
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.
                                                   16

-------
                                               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:

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 of the
participating technologies 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.

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 TEF, according to the equation:
                  TEQ = Cc * TEF
                                                    17

-------
where Cc is the concentration of the congener. The TEF
(see Table 4-1) provides an equivalency factor for each
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 esti-
mate 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 report(5) 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
Table 4-1. World Health Organization Toxicity Equivalency Factor Values
Compound'3'
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

OCDD
Dioxin-like PCBs
Coplanar
3,3',4,4'-TCB (PCB 77)
3,4,4',5-TCB(PCB81)
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
,2,3,4,7,8-HxCDF
,2,3,7,8,9-HxCDF
,2,3,6,7,8-HxCDF
2,3,4,6,7,8-HxCDF
,2,3,4,6,7,8-HpCDF
,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 1 14)
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
                                                     18

-------
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,
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.(6) 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 demon-
stration, 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).
                                                    19

-------
Table 4-2. Distribution of Samples for the Evaluation of Performance Objectives
Performance Objective
P 1 : Accuracy
P2: Precision
P3: Comparability
P4: EMDL
P5: False positive/negative results
P6: Matrix effects
P7: Cost
S 1 : 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 technology
No. of Samples
58
128
23
209
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 in this section.

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 was 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
                                                   20

-------
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
CRM 529
WMS-01
RM5184
SRM 1944
custom
custom
custom
custom
custom
custom
056
(lot 56011)
Certified Concentration
TEQD/F
(Pg/g)
3.9
6583
62
171
251
11
33
NS
NS
NS
11
0.046
TEQPCB
(Pg/g)
5.0
424C
10.5
941
41°
NSf
NS
NS
11
1121
3,760C
0.01
PAH
(mg/kg)
0.18
NAd
NA
27
2.4e
0.33
<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.)
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
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
kgs+ 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 (|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
                                                     21

-------
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-01 is a
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
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
semivolatile 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
                                                   22

-------
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 milliliters (mL). For each PE sample, the blank soil
matrix was weighed into a two-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 samples since the data were not based
on analytically derived results. Further confirmation of
the concentrations 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 one-
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 demon-
stration 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 two hours. Approximately
500 g of material were put in a blender and blended for
two minutes. The blender sides were scraped with a
spatula, and the sample was blended for a second two-
  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.
                                                    23

-------
minute period. The sample was sieved [USA Standard
testing, No. 10, 2.00-millimeter (mm) 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 was
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.

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) dis-
carding 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.
                                                   24

-------
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 RSD for 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
roa
9, 10, 11
4
5
5
2,8
1,3
8
4
4
6,7
77%
72S
1 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.
                                                          25

-------
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 repre-
sented a 10-g sediment sample extraction and were
reported in pg/mL, which was calculated by the
following equation:

  pg/mL = Wgsamples)x(lOgaHquot) x (3Q Dp)
           (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 3Ox dilution factor is included. The
extracts were not processed through any cleanup steps,
but they 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.

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.
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
#2
Environmental #7,
Sample #1
Spike #la
Spike #2a
Spike #3a
Sample Description
Soxhlet extraction in toluene; no
cleanup
Soxhlet extraction in toluene; no
cleanup
0.5 pg/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
*  Prepared in toluene.
b  Seven replicates were analyzed for EMDL evaluation.
c  This extract was spiked with PCBs only but a low-level (approximately 0.3 pg/mL) 2,3,7,8-TCDD contamination was confirmed by the
  reference laboratory.
                                                    26

-------
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. CAPE Technologies
elected to have soil and sediment samples identified. 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.
CAPE Technologies analyzed the samples in the
prescribed order. The extracts were the first 23 samples
in the analysis order. The randomization was generated
so that, to the extent possible, an equal split of the
sample replicates were analyzed in the field and in the
laboratory. For example, when four replicates of a
particular sample were included in the suite of
demonstration samples, two replicates were analyzed
among the first half of the samples and two replicates
were among the second half of the samples. In the field,
the samples were only analyzed by CAPE Technologies
for total TEQD/F. A 40-mL fraction of each D/F extract
that was generated in the field during the demonstration
was archived for analysis in the developer's laboratories
using the PCB TEQ Immunoassay Kit.

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 CAPE  Technologies 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,  CAPE Technologies was sent six soil/
sediment samples with the corresponding D/F, PCB, and
PAH characterization data to perform a self-evaluation of
their kits. In Phase 2, seven additional soil/sediment
samples and two extracts were sent to CAPE
Technologies  for blind evaluation. AXYS analyzed all 15
pre-demonstration samples blindly. The CAPE
Technologies  pre-demonstration results were paired with
the AXYS results and returned to CAPE Technologies so
they could use the HRMS pre-demonstration sample data
to refine the performance of their kits 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 CAPE Technologies was a viable
candidate to continue in the demonstration process.
                                                   27

-------
4.6    Execution of Field Demonstration
CAPE Technologies arrived on-site on Sunday, April
25, and spent several hours that day setting up its
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 that 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.

CAPE Technologies received its first batch of samples
by midmorning on April 26. CAPE Technologies
completed analysis of 95 samples for D/F only in 5
working days (on April 30). 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.
The remaining analyses (95 samples for TEQPCB and
114 samples for both TEQD/F and TEQPCB) were
completed by CAPE Technologies in their laboratories
and reported on August 27. CAPE Technologies
reported that it took them two weeks of analytical time
to  complete the 114 sample analyses in their
laboratories. CAPE Technologies was also offered the
opportunity to reanalyze any samples before reporting
final results. CAPE Technologies reanalyzed and reported
new results for two samples that were analyzed for D/Fs
in the field.

4.7    Assessment of Primary and Secondary
       Objectives
The purpose of this section is to describe how the
CAPE Technologies reported its results TEQD/F, TEQPCB
and total TEQ (all in pg/g). The CAPE Technologies
results were compared to the certified values and
reference laboratory results for TEQD/F, TEQPCB, and total
TEQ. The reference laboratory total TEQ values were
calculated by summing the TEQD/F and TEQPCB data.
Total TEQs value could not be calculated for two
reference laboratory samples that were excluded due to
sample preparation issues  (see  Section 6.4).

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. Percent  recovery values relative
to the certified or spiked concentrations were calculated.
To evaluate accuracy, the average  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=
where C is the mean concentration value calculated from
the technology replicate measurements (reported in pg/g
TEQ) and CR is the certified value (in pg/g TEQ).
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.
                                                   28

-------
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:
where SD is the standard deviation and C is the
average measurement. Both values are in pg/g TEQ.

The equation used to calculate RSD between replicate
measurements was:
                       SD
                  KSD=
            100%
RSD was calculated if detectable concentrations were
reported for at least three replicates. The mean, median,
minimum, and maximum RSD values, in percent, 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.

4.7.3  Primary ObjectiveP3: 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 CAPE Technologies
were compared to the corresponding reference
laboratory results by calculating a relative percent
difference (RPD). The equation for RPD, reported in
percent, is as follows:
        RPD =
   (MR-MD)
average (MR, MD)
xlOO%
where MR is the reference laboratory measurement (in
pg/g TEQ) and MD is the developer measurement (in pg/g
TEQ). Nondetects were not included in this evaluation.
The CAPE Technologies results were compared to the
reference laboratory for TEQD/F, TEQPCB and total TEQ.
For PE samples, TEQD/F and TEQPCB RPD calculations
were only performed for the analyte classes that the PE
sample contained. For example, PE sample #6 was only
spiked with 2,3,7,8-TCDD. Consequently, RPD
calculations were only performed for TEQD/F and not
TEQPCB or total TEQ.

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
CAPE Technologies 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 developer and
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.

The accuracy  of reporting blank samples was assessed.
The blanks included eight  replicate samples that
contained levels of D/Fs and PCBs 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%.
                                                   29

-------
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, CAPE
Technologies 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.  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 is shown in the following equation:
             EMDL= t.
,(SD)
where  t(n_, ,.^=0^ = Student's t-value appropriate for a
99 percent confidence level and a standard deviation
estimate with n-1 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 CAPE Technologies kits 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 above
and below 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 CAPE
Technologies were considered false positive. Conversely,
those samples that were reported as < 20 (or 50) pg/g
TEQ by CAPE Technologies, but reported as > 20 (or 50)
pg/g TEQ 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), sample type
(i.e., PE, environmental, and extract), varying levels of
PAHs, environmental site, and 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 for the 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
                                                    30

-------
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. To evaluate the effect of
measurement location, ANOVA tests were performed
for sample results within a replicate set that were
generated both in the laboratory and in the field. For
these analyses, p-values < 0.05 indicated statistically
different results between the laboratory and field
measurements and therefore a significant effect of the
measurement location on reported results. The
percentage of replicate sets having p-values < 0.05 was
reported.

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.

4.7.8  Secondary Objective SI: Skills Level of
       Operator
Based on observations during the field demonstration,
the type of background and training required to
properly operate the DF1 Dioxin/Furan Immunoassay
Kit 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 DF1 Dioxin/Furan
Immunoassay Kit 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 CAPE Technologies (e.g.,
a trailer and fume hood), 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 CAPE Technologies
worked in the field was documented using attendance log
sheets where CAPE Technologies 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.
                                                   31

-------
                                             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
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
                                                   32

-------
dioxin-like PCBs and can be used to determine all 209
PCB congeners. Not all PCBs are determined indi-
vidually with this method because some are determined
as sets of coeluting congeners. The calibration range for
PCBs based on atypical 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
Aroclors1 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
demonstration, and it was selected as the dioxin/furan
reference method. Reference data of equal quality
needed to be generated to determine the PCB contribu-
tion 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 demon-
stration. 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, spiked with 13C12-labeled
internal standards, and extracted with methylene chloride
using accelerated solvent extraction 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 chroma-
tography 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 (iL. 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 high-resolution gas
chromatography/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 non2,3,7,8-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. PCDD/F
data were reported as both concentration (pg/g dry) and
TEQs (pg TEQ/g dry).
                                                   33

-------
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
166 8A 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 gas
chromatography 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 (iL and 1 mL, depending on the anticipated
concentration of PCBs in the sample, as reported by the
sample providers. PAHs were separated by capillary gas
chromatography on a DB-5, 60-m column and were
identified and quantified using electron impact mass
spectrometry. 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.
                                                   34

-------
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 HRMS
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 FiRMS coupled to an FiRGC 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 1668 A 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)
                                                   35

-------
Detection limits were reported as sample-specific
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) were
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.
                                                   36

-------
                                               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
demonstration 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 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.
In addition, the audit confirmed that analytical
                                                    37

-------
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.
Checklists 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 meeting 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 login 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 166 8A acceptance  limits, the data were
accepted.
                                                    38

-------
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
performance evaluation 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 demonstra-
tion, 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 6-3. The R values 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
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
                                                   39

-------
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

Table 6-1. Objective PI Accuracy - Percent Recovery
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
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-0 1
Cambridge 5184
NIST 1944
ERATCDD10
ERA TCDD 30
ERA PAH
ERA PCB 100
ERA PCB 10000
ERA Aroclor
ERA Blank
All Performance Evaluation Samples
% Recovery
TEQprB
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.
Table 6-2.  Evaluation of Interferences
PE Material with Known Interference
ERA PAH
ERA PCB 100
ERA PCB 10000
ERA TCDD 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)
                                                   40

-------
Table 6-3a. Objective P2 Precision - Relative Standard Deviation
Sample Type
Environmental
Extract
Performance
Evaluation
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 Bav #1
Newark Bav #2
Newark Bay #3
Newark Bav #4
RaritanBav#l
Raritan Bay #2
Raritan Bay #3
S aginaw River #1
S aginaw 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
En vir 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
J
5
4
10
4
77
21
21
25
7
2
6
1
6
3
3
8
7
60
36
4
11
7
9
12
19
14
13
13
4
9
71
83
119
1
4
7
3
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
3
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.
                                                       41

-------
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
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
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.
                 16000
                                                    y = 0.8595x + 41.181
                                                         R2 = 0.9899
                                        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.
                                                   42

-------
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.
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 (%)
MeanRSD (%)
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).
                                                    43

-------
                                             Chapter 7
                Performance of CAPE Technologies DFl Dioxin/Furan and
                                  PCB TEQ Immunoassay Kits
7.1    Evaluation of DFl Dioxin/Furan and PCB
       TEQ Immunoassay Kits Performance
The CAPE Technologies DFl Dioxin/Furan and PCB
TEQ Immunoassay kits are immunoassay techniques
that report TEQ of dioxin/furans and PCBs, respectively.
It should be noted that the results generated by the
CAPE Technologies kits may not directly correlate to
HRMS TEQ in all cases because it is known that the
congener responses  and cross-reactivities of the kits are
not identical to the TEFs that are used to convert
congener HRMS concentration values to TEQ. The
effect of cross-reactivities may contribute to this
technology's reporting results that are biased high or low
compared to HRMS TEQ results.  Therefore, these kits
should not be viewed as producing an equivalent
measurement value to HRMS TEQ but as a screening
value to approximate HRMS TEQ. As described in
Appendix B and CAPE Technologies literature, the best
results for immunoassay screening are obtained on a
single site basis. The ideal approach involves partially
characterizing a site by HRMS, using those results to
develop a site  specific immunoassay calibration, and
refining that calibration over time, based on an ongoing
stream of confirmatory HRMS samples. This approach
was not evaluated during this demonstration; samples
from multiple  sites were pooled and a single calibration
was used.

The following sections describe the performance of the
DFl Dioxin/Furan Immunoassay Kit and the PCB TEQ
Immunoassay Kit, 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 percent recovery (R) values for the
CAPE Technologies D/F and PCB kits is presented in
Table 7-1. The description of how Rvalues were
calculated is presented in Section 4.7.1. The Rvalues
are presented for TEQPCB TEQD/F and total TEQ values.
The minimum R value, the maximum R value, median R
value, and the mean R value are presented for each set of
TEQ results. The mean Rvalues for the TEQPCB TEQD/F,
and total TEQ results were 195%, 236%, and 160%,
respectively. The Rvalues presented in Table 7-1
indicate that the CAPE Technologies kits generally
reported data that were biased high relative to the
certified values of the PE samples, although the bias for
one PE sample (NIST 1944) was consistently low (R
values between 30% and 49%) for all TEQ values.

7.1.2  Evaluation of Primary Objective P2:
       Precision
Summaries of the RSD values for the CAPE
Technologies D/F Immunoassay Kit and PCB TEQ
Immunoassay Kit are 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 for TEQPCB,TEQD/F, and total TEQ in
Table 7-2a, and 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 for each set of TEQ results and sample
types. Low RSD values (< 20%) would indicate high
precision. In terms of sample type, the CAPE Tech-
nologies D/F and PCB kits had the most precise data for
the PE TEQD/F results, with a mean  RSD value of 64%.
In terms of TEQ values, the CAPE Technologies kits
had the most precise data for the TEQD/F values with an
                                                  44

-------
Table 7-1. Objective PI Accuracy - Percent Recovery
PE Sample ID
1
2
o
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
ERATCDD10
ERA TCDD 30
ERA PAH
ERAPCB 100
ERAPCB 10000
ERA Aroclor
ERA Blank
All Performance Evaluation Samples
% Recovery
TEQPrR
171
26
302
131
49
NA
NA
NA
523
257
99
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
26
523
151
195
TEQn;F
436
NA
182
68
30
182
157
NA
NA
NA
595
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
7
30
595
182
236
Total TEQ
296
105
199
121
33
268
158
NA
NA
NA
101
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
33
296
139
160
 NA = not applicable.
Table 7-2a. Objective P2 - RSD as a Description of Precision by Sample
Sample
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
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
Relative Standard Deviation (%)a
TEQprR
117
141
38
163
187
136
172
93
82
145
77
96
62
120
181
79
65
46
145
199
132
162
96
95
177
TEQn;F
71
51
187
96
23
36
46
92
23
34
29
21
62
46
40
17
59
89
70
165
65
58
27
56
87
Total TEQ
68
49
174
95
31
41
46
42
17
52
19
19
61
68
43
20
53
87
72
128
63
64
27
57
88
                                                  45

-------
Sample
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
ERAPCB 100
ERAPCB 10000
ERATCDD 10
ERA TCDD 30
LCG CRM-529
NIST 1944
Wellington WMS - 01
Relative Standard Deviation (%)a
TEQprR
86
135
200
188
125
88
102
50
141
153
99
118
74
118
173
116
110
117
93
172
200
99
53
198
TEQn;F
NA
166
86
104
91
116
93
155
40
86
136
0
29
99
98
68
NA
NA
NA
55
51
NA
66
48
Total TEQ
67
128
85
119
91
114
86
154
51
99
98
117
18
105
169
61
NA
NA
NA
90
50
81
54
77
NA = not applicable (i.e., one or more of the replicates were reported as a nondetect value).
a Three or four replicate results were used to calculate the RSD values.
Table 7-2b.  Objective P2 - RSD as a Description of Precision by Sample Type
Sample
Type
Env
Ex
PE
All
RSD (%) for TEQpr
N
32
5
12
49
MIN
38
50
53
38
MAX
200
153
200
200
MED
122
118
117
118
R
MEAN
123
112
127
123
RSD (%) for TEQn;F
N
31
5
8
44
MIN
17
0
29
0
MAX
187
155
99
187
MED
62
86
60
63
MEAN
71
83
64
71
RSD (%) for Total TEQ
N
32
5
9
46
MIN
17
51
18
17
MAX
174
154
169
174
MED
63
99
77
67
MEAN
68
104
78
74
                                                        46

-------
overall RSD of 71%. Overall RSD values ranged from
0% to 200%. Note that the vendor reported several
TEQPCB results as "0" pg/g TEQ (see Appendix D), so
zero was used in the calculation of RSD. This is different
than the treatment of nondetects (reported as
"< reporting limits"), which were not included in the
analysis.

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 CAPE Technologies and
reference laboratory data was assessed by calculating
RPD values for TEQPCB, TEQD/F, and total TEQ, as
presented in Table 7-3. The summary statistics presented
in Table 7-3 provide an overall assessment of the RPD
values that is reported by TEQ value and sample type.
The CAPE Technologies values agreed best with the
reference laboratory D/F measurements for extract
samples, with a median RPD values of-4%. The median
RPD value for TEQPCB, TEQD/F, and total TEQ were
-13%, -26%, and -5%, respectively, with minimum and
maximum 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
TEQPCB, TEQD/F, and total TEQ values, 22 (12%), 28
(17%), and 21  (13%) of the samples, respectively, had
RPD values between positive and negative 25%. This
evaluation indicates that the CAPE Technologies results
were generally higher than the reference laboratory (as
evidenced by all median values being negative).

Comparability was also assessed using the interval
approach discussed in Section 4.7.3. The agreement
when sorting the developer and reference laboratory
results for TEQPCB, TEQD/F, and total TEQ data into four
intervals < 50 pg/g, 50 to 500 pg/g, 500 to 5,000 pg/g,
and > 5,000 pg/g) is described in Table 7-4. The
agreement between the developer and reference
laboratory was 82% for TEQPCB, 71% for TEQD/F, and
64% for total TEQ. Interval reporting addresses the
question 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
from 18 to 36% of the time, the CAPE Technologies
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, depending on which TEQ value was being
determined and the concentrations chosen for the
intervals.

The ERA blank samples contained levels of D/Fs and
PCBs that were below the reporting limits of the
developer technologies (see Table 4-4 certified values:
0.046 pg/g TEQD/F and 0.01 pg/g TEQPCB). The CAPE
Technologies reported concentrations were compared
with the reference laboratory reported data for these
samples in Table 7-5. CAPE Technologies reported four
of the eight TEQPCB values as detections (ranging from
2 to 4 pg/g), so the four results that were reported as
nondetects (i.e., zero) agreed with the reference
laboratory results. For TEQD/F, five of the results were
reported as detections (13 to 46 pg/g), so three of eight
results agreed with the reference  laboratory's reporting
of blank samples. For total TEQ  values, two of the
CAPE Technologies results were reported as nondetects
and 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 remained the same.
Table 7-3.  Objective P3 - Comparability Summary Statistics of RPD
Sample
Type
Env
Ex
PE
All
TEQ
N
128
16
38
182
MIN
-200
-198
-179
-200
rR RPD (%)
MAX
200
189
189
200
MEDIAN
-10
-115
15
-13
TEQn;F RPD (%)
N
120
16
31
167
MIN
-199
-189
-129
-199
MAX
198
179
169
198
MEDIAN
-26
-4
-30
-26
Total TEQ RPD (%)
N
122
12
25
159
MIN
-199
-134
-138
-199
MAX
198
145
171
198
MEDIAN
-8
18
-5
-5
                                                    47

-------
Table 7-4.  Objective P3 - Comparability Using Interval Assessment
Agreement
Number Agree
% Agree
Number
Disagree
% Disagree
TEQPCB
172
82
37
18
TEQn/F
148
71
60
29
Total TEQ
132
64
75
36
Table 7-5. Objective P3 - Comparability for Blank Samples
Rep
1
2
3
4
5
6
7
8
% agree
TEQPCB
CAPE
Technologies
(Pg/g)
0
0
2
2
0
4
3
0
RefLab"
(Pg/g)
J0.0243"
0.00385
0.00277
J0.042
J0.0229
J0.0191
J0.0325
J0.0225
Agree?
Yes
Yes
No
No
Yes
No
No
Yes
50% (4 of 8)
TEQD/F
CAPE
Technologies
(Pg/g)
<14
15
<50
17
46
13
<11
13
RefLab"
(Pg/g)
0.0942
0.0728
0.237
0.307
0.113
0.0524
0.211
0.0692
Agree?
Yes
No
Yes
No
No
No
Yes
No
38% (3 of 8)
Total TEQ
CAPE
Technologies
(Pg/g)
14
15
<52
19
46
17
<14
13
RefLab"
(Pg/g)
JO. 12
J0.08
J0.24
J0.35
JO. 14
JO. 07
J0.24
J0.09
Agree?
No
No
Yes
No
No
No
Yes
No
25% (2 of 8)
1 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 calculations did not strictly
follow the definition in the Code of Federal Regulations
(i.e., t value with 6 degrees of freedom). 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 EMDLs of the CAPE Technologies D/F and PCB
kits were determined for TEQD/F, TEQPCB, and total TEQ
values using the Extract Spike #1  and Cambridge 5183
sample results in Tables 7-6a and b. Since 2,3,7,8-TCDD
was the only congener in Extract Spike #1, only an
EMDL for TEQD/F could be determined. As shown in
Tables 7-6a and 7-6b, because some of the results were
nondetects, the EMDLs were calculated in three ways
for nondetect values: 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. While the number of degrees of
freedom was reduced from 6 to 2 or 3  for the Cambridge
5183 and Extract Spike #1 samples because of the
nondetect values, the EMDLs for all calculations were in
the range of 12 to 35 pg/g TEQD/F. The detection limit
reported by CAPE Technologies in the demonstration
plan was 1 pg/g TEQ.  The EMDLs determined using
the Wellington samples were significantly higher (200 to
300 pg/g TEQ) than for the other two samples. Since the
D/F and PCB concentrations were much higher than the
projected reporting limit, the Wellington samples did not
seem appropriate for calculating EMDLs, so the MDLs
for this sample were not included in the calculations.
                                                    48

-------
Table 7-6a.  Objective P4 - Estimated MDL for TEQD/F and TEQP
Statistic
Degrees of Freedom
Standard Deviation
(PR/R)
EMDL (pg/g)
TEQD/F
Cambridge 5183
Nondetect
values set
to zero
2
5
35
Nondetect
values set to
Vz value
6
6
20
Nondetect
set to
reported
value
6
4
12
Extract Spike #1
Nondetect
values set
to zero
3
7
33
Nondetect
values set
to Vz value
6
6
20
Nondetect
set to
reported
value
6
6
18
TEQPrR
Cambridge
5183
6
6
20
Table 7-6b. Objective P4 - Estimated MDL for Total TEQ
Statistic
Degrees of Freedom
Standard Deviation (pg/g, Total TEQ)
EMDL (pg/g, Total TEQ)
Cambridge 5183
Nondetect values
set to zero
2
4.73
33
Nondetect
values set to Vz
value
6
9.83
30
Nondetect
values set to
reported value
6
8.52
25
7.1.5  Evaluation of Primary Objective P5: False
       Positive/False Negative Results
The summary of false positive/false negative results is
presented in Table 7-7. CAPE Technologies reported
14% false positive and 5% false negative results, relative
to the reference laboratory's reporting of samples above
and below 20 pg/g TEQ, for the TEQPCB results. For
TEQD/F, the percentage of false positive/negative results
were slightly less (11% and 4%, respectively) than for
TEQPCB. CAPE Technologies reported 14% false
positives and 3% false negatives around 20  pg/g for total
TEQ. CAPE Technologies's false positive and false
negative rates around 50 pg/g were generally lower for
all three TEQ types, ranging from 4% to 10%.

These data suggest the CAPE Technologies kits as
processed in the demonstration 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 sample above and below 50 pg/g TEQ. CAPE
Technologies notes that the user can optimize
performance of the kit at other desired screening levels
by changing the size of sample extracted so that the
desired screening level concentration falls at the
optimum point on the kit response curve.
Table 7-7.  Objective P5 - False Positive/False Negative Results
Rate
False Positive
False Negative
TEQPCB
20 pg/g
14%
(29 of 209)
5%
(11 of 209)
50 pg/g
8%
(17 of 209)
4%
(9 of 209)
TEQD/F
20 pg/g
11%
(22 of 207)
4%
(9 of 207)
50 pg/g
8%
(17 of 207)
8%
(16 of 207)
Total TEQ
20 pg/g
14%
(28 of 207)
3%
(6 of 207)
50 pg/g
10%
(20 of 207)
5%
(11 of 207)
                                                   49

-------
7.1.6  Evaluation of Primary Objective P6: Matrix
       Effects
Six types of potential matrix effects were investigated:
(1) measurement analysis location (field vs. laboratory),
(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:  19% statistically different
•   Matrix type: none
•   Sample type: none
•   PAH concentration: slight effect on total TEQ
•   Environmental site: none
•   Known interferences: slight

A one-way ANOVA was performed on samples that had
at least one  detected replicate  analyzed in the field and in
the laboratory to determine if performance was affected
by the samples being analyzed in the field. A p-value
less than  0.05 in Table 7-8 indicates that the mean of
samples analyzed in the field was significantly different
from the mean of those analyzed in the laboratory. Only
TEQD/F values were included in this evaluation because
all TEQPCB and total TEQ values were generated by
CAPE Technologies in its  laboratories. Seven of 37 sets
of TEQD/F values (19% overall)  showed statistically
significant location effects, and of these samples, CAPE
Technologies generally reported the laboratory result
more comparably to the reference laboratory result. In
Table 7-9, 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 for TEQPCB,
TEQD/F, or total TEQ. In Table  7-10, 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 a slight effect  (p = 0.0327) on total
TEQ, but the effects for TEQD/F(p = 0.0526) or TEQPCB
(p = 0.0771) were not statistically significant.  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 for TEQPCB  TEQD/F or total
TEQ. Based on the comparability results (RPD), CAPE
Technologies's results were not more or less comparable
for one particular environmental site, suggesting that
matrix effects were not dependent on environmental site.

The effect of known interferences was also assessed by
evaluating the results of PE materials that contained one
type of contaminant (D/F, PCBs,  or PAHs) but not
another. Table 7-11 summarizes  the detection of
analytes not spiked in the PE samples along with the
percent recovery values (from Table 7-1) for the spiked
analytes. For the ERA PAH sample that contained no
spike D/Fs or PCBs, CAPE Technologies reported a
mean total TEQ value of 17 pg/g. The PCB-only spiked
samples were reported with only  one D/F detection.
CAPE Technologies reported only one sample as a slight
PCB detection for the ERA TCDD 30 D/F-only spiked
PE samples, but three TEQPCB detections (mean = 9.5
pg/g TEQPCB) for the ERA TCDD 10 D/F only spiked PE
sample.

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
The secondary objectives described in this section were
assessed only for the DF1 kit because the PCB TEQ
Immunoassay Kit was not deployed during the field
demonstration. Because of the similar principles and
procedures for the two kits, it is likely that similar
conclusions could be drawn, but this was not confirmed
by observation.

The technology used by CAPE Technologies at the
demonstration was composed of two kits, the SP-3
sample preparation kit and the DPI dioxin/furan
immunoassay. All steps of these procedures were
observed during the field demonstration. The sample
preparation consisted of first adding sodium sulfate to the
sample to make it free flowing, and then extracting the
samples for 2 to 4 hours using a 1:1 mixture of hexane-
to-acetone. After extraction, samples were centrifuged at
1,000 x g or less for 10 to 15 minutes. A portion of the
supernatant was removed and evaporated onto 0.25 mL
of tetradecane. Hexane was added to dilute the residue.
The acid silica-activated carbon coupled column system
was prewashed with hexane.
                                                    50

-------
Table 7-8. Objective P6 - Matrix Effects Using Descriptive Statistics and ANOVA Results Comparing TEQD/F
Replicate Analysis Conducted During the Field Demonstration and in the Laboratory
Sample Type
Environmental
Sample
Brunswick #1
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
RaritanBay #1
Raritan Bay #2
Raritan Bay #3
S aginaw River #1
Saginaw River #2
Saginaw River #3
Solutia #1
Solutia #2
Solutia #3
Titta. River Soil #1
Location
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
TEQ
N
1
3
2
2
2
2
1
3
1
3
1
2
2
2
2
2
1
2
1
2
1
2
1
3
1
3
2
2
2
2
1
2
2
2
1
3
2
2
2
2
2
2
2
2
2
2
Mean (SD)
(Pg/g)
59.0
137.3 (89.9)
5,025.0 (6696.3)
111.5(38.9)
471.0(482.2)
240.5 (259.5)
64.0
59.7(17.1)
119.0
215.0(60.3)
37.0
28.5(19.1)
13,800.0(3535.5)
1,949.0(2320.7)
25,550.0 (1484.9)
27,350.0(10111.6)
33700.0
53,700.0(15980.6)
31.0
21.5(6.4)
75.0
59.0(14.1)
11.0
57.0(19.9)
63.0
42.0 (23.6)
58.5(23.3)
69.5(36.1)
42.5 (6.4)
44.0(11.3)
56.0
49.5(43.1)
1,114.0(503.5)
239.5 (256.7)
1690.0
1,373.0(1237.6)
1,485.5(1760.0)
86.5 (50.2)
109.0(41.0)
37.5 (0.7)
712.5(221.3)
2,030.0 (353.6)
2,840.0(1343.5)
3,230.0(127.3)
118.0(97.6)
286.5 (16.3)
n/F
p-Value Comparing
Field to Laboratory
0.5292
0.4084
0.6121
0.8466
0.3017
0.7780
0.0582
0.8265
0.4931
0.4374
0.5252
0.1836
0.5212
0.7519
0.8852
0.9221
0.1601
0.8450
0.3779
0.1326
0.0466a
0.7224
0.1376
                                                 51

-------
Sample Type

PE
Sample
Titta. River Soil #2
Titta. River Sed #1
Titta. River Sed #2
Titta. River Sed #3
WinonaPost#l
Winona Post #2
Winona Post #3
Cambridge 5 183
Cambridge 5 184
ERA Aroclor
ERATCDD10
ERA TCDD 30
NIST 1944
Wellington WMS -01
Location
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
TEQ
N
2
2
2
2
2
2
1
3
2
1
2
2
2
2
1
2
1
3
1
3
2
2
1
2
2
2
4
3
Mean (SD)
(Pg/g)
1,251.5(68.6)
174.5 (24.7)
355.0(468.1)
38.0(31.1)
190.5 (95.5)
46.0 (28.3)
184.0
35.7(16.7)
1,945.0(742.5)
68.0
3,970.0(1018.2)
51.0(12.7)
987.0(748.1)
319.5(326.0)
17.0
17.0(7.1)
280.0
62.7 (47.8)
161.0
33.7(12.7)
15.0(9.9)
25.0(12.7)
82.0
36.5 (0.7)
119.0(8.5)
34.0(18.4)
121.3 (62.6)
101.7(51.1)
n/F
p-Value Comparing
Field to Laboratory
0.0023
0.4401
0.1765
0.0166
0.2872
0.0321
0.3669
1.0000
0.0589
0.0129
0.4730
0.0121
0.0272
0.6782
  Bold signifies in-field measurement statistically different from the laboratory measurement at the p<0.05 significance level.
Table 7-9. Objective P6 - Matrix Effects Using RSD as a Description of Precision by Soil, Sediment, and Extract
Matrix
Type
Soil
Sediment
Extract
Overall
RSD for TEQprR (%)
N
26
18
5
49
MIN
74
38
50
38
MAX
200
200
153
200
MED
118
119
118
118
MEAN
127
119
112
123
RSD for TEQn;F (%)
N
21
18
5
44
MIN
23
17
0
0
MAX
116
187
155
187
MED
58
64
86
63
MEAN
64
77
83
71
RSD for Total TEQ (%)
N
23
18
5
46
MIN
17
19
51
17
MAX
169
174
154
174
MED
63
68
99
67
MEAN
68
74
104
74
                                                        52

-------
Table 7-10. 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 TEQPCB (%)
N
3
4
16
9
32
MIN
38
88
46
86
38
MAX
145
125
187
200
200
MED
82
110
108
172
122
MEAN
88
108
115
154
123
RSD for TEQD/F (%)
N
3
4
16
8
31
MIN
23
71
17
46
17
MAX
187
116
96
166
187
MED
34
92
49
87
62
MEAN
81
93
51
97
71
RSD for Total TEQ (%)
N
3
4
16
9
32
MIN
17
68
19
46
17
MAX
174
114
95
128
174
MED
52
89
46
85
63
MEAN
81
90
49
87
68
Table 7-11. Objective P6 - Matrix Effects of Known Interferences Using PE Materials
PE Sample
ERA PAH
ERAPCB 100
ERAPCB 10000
ERATCDD10
ERA TCDD 30
% Recovery for Spiked
Analytes a
NAb
523% (PCB)
257% (PCB)
182% (D/F)
157(D/F)
Mean TEQ (pg/g)
Reported by CAPE Technologies for Analytes
that were not Spiked in the PE Sample
17 (total)
23 (D/F)C
all nondetects for D/F
9.5 (PCB)
1 (PCB)C
1 Percent recovery values taken from Table 7-1.
b NA = not applicable because R value could not be calculated.
c Three replicates were reported as nondetects.
At this point a slight change was made in the method
described in the demonstration plan by including a
pretreatment for samples that had significant color. For
such samples, a small amount of bulk fine acid silica was
added to the sample tube, mixed briefly, and the sample
loaded onto the acid silica column as a slurry of acid
silica in hexane.  Samples with no pretreatment were
loaded as hexane solutions. The columns were washed
with hexane. The carbon column was removed, washed
with a small amount of hexane, then eluted with 1:1
toluene/hexane to give the PCB fraction, which was
stored for later analysis.  The column was flipped and the
dioxin/furan fraction eluted with toluene. A keeper
solution was added to the eluted dioxin/furan fraction
(TEG-methanol-Triton X-100) and the solvent was
evaporated and centrifuged at 1-2000 x g to concentrate
all of the keeper solution at the bottom of the tube. The
samples were then diluted with methanol and continued
forward with the analysis.

The immunoassay was performed by first rinsing the
tubes with American Society for Testing and
Materials(ASTM)-grade distilled water. The rinse was
dumped and the tubes were inverted and tapped to
remove the excess water. After the rinse, 500 |iL of
reagent grade water was added along with 50 |iL of the
Triton X-100 in methanol. The tubes were then mixed for
10 seconds. To this,  10 |iL of control, standard, or
sample was added and the tubes were mixed. Tubes were
then covered and allowed to incubate. Product literature
indicates that incubation should occur for two hours
(acceptable results) to 12 to 24 hours (best results). In the
demonstration, the tubes were incubated overnight. After
incubation, the tube contents were dumped and the tubes
were washed four times with ASTM-grade distilled
water. Competitor conjugate (HRP conjugate, 500 |iL)
was added and allowed to incubate for 15 minutes. The
tubes were then emptied and washed an additional four
times. A 500-|iL aliquot of the substrate solution was
added to each tube and allowed to incubate for an
additional 30 minutes. After incubation, 500 |iL of stop
solution was added and the samples were analyzed with a
differential photometer.
                                                    53

-------
7.2.1  Evaluation of Secondary Objective SI: Skill
       Level of Operator
In the field demonstration, samples were processed with
the CAPE Technologies kit by Dr. Bob Harrison,who has
a Ph.D. in environmental toxicology, 23 years experience
with environmental immunochemistry, and eight years
with this specific technology. The developer
recommends that users have at least a bachelor's degree
in the sciences, and experience with both dioxin cleanup
methods and enzyme-linked immunosorbent assay
(ELISA) would be helpful.  From observation, the
education  level may not be strictly necessary, but the
experience with dioxin cleanup methods seemed very
important  in the use of the kit.

The instructions contained with the kits are detailed, but
at times they seemed difficult to understand. The kits are
supplied with basic instructions,  procedural notes, and
application notes. The addition of these notes, which may
have conflicting instructions, seemed somewhat
confusing  and did not always seem clear as to what
procedural changes should be followed without first
seeking input from the developer. However, once the
appropriate sample processing procedures are established
and the supplies are in place, the equipment contained in
the kit seemed straightforward and easy to use.  All
cleanup columns are premade and prepacked, and the
assay itself is uncomplicated. CAPE Technologies
provides an Excel file to aid in data analysis. This
spreadsheet automatically processes the raw data into a
useful form after data entry by the user. Accurate sample
weights are necessary for later calculations. Accurate
volumes are especially important when using the
immunoassay tubes, since variations in volume can cause
differences in results. The kit requires some safety
precautions when dealing with the solvents. Because the
fine acid silica for cleanup columns is prepacked, mask
precautions are not necessary.

7.2.2  Evaluation of Secondary Objective S2:
       Health and Safety Aspects
It can be expected that around 60 mL of hexane will be
used for each sample. Not all of this becomes waste, but
a portion does. The disposable columns as well as the
disposable glassware  create solid waste. The solvent
waste itself is not inherently more hazardous than would
be expected, unless the samples are very contaminated.
A complete inventory of the waste generated was
performed after the demonstration for processing of
95 samples by CAPE Technologies and the following
was recorded. None of the containers was verified as full.
Note that this summary does not include the samples that
were analyzed in the CAPE Technologies laboratories:

(1)  One 1-gallon container filled with aqueous waste
(2)  One used broken glass container
(3)  One box of 40 mL vials containing sample/solvent
    mixture.

Based on observation, the amount of waste recorded
seemed low in comparison to what should have been
generated, but portions of the hexane rinses of the
columns were lost to volatilization and 6 mL of the
column eluate was saved for later PCB analysis. 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
A trailer was used by CAPE Technologies during the
demonstration. The need for power, nitrogen tanks, a
fume hood, shakers, vacuums, and other equipment
requires a trailer at the very least for successful operation
of this technology. While the  assay itself is very easy to
use in the field, the extraction and cleanup methods
require level work space and could create some difficulty
in the field unless there was a mobile lab or trailer
available. The cleanup method requires a large amount of
space within a hood.  The space issue can be mitigated by
the use of racks that have been designed by CAPE
Technologies and are available for purchase. During the
demonstration, the individual using the technology was
able to easily overcome the hood space constraints and
work quite efficiently. The photometer recommended by
CAPE Technologies is also very easy to use and is
designed with field portability in mind. In all, while this
technology does require a mobile lab or trailer, it did not
seem difficult to use in the field. The trailer used for this
demonstration took approximately half of a day to set up
before samples could be processed. Differences in
reported results due to measurement location (in field vs.
laboratory) are described in Section 7.1.6.
                                                    54

-------
7.2.4  Evaluation of Secondary Objective S4:
       Throughput
During the demonstration, 95 samples were processed by
CAPE Technologies, including sample extraction, extract
cleanup, dioxin/furan immunoassay, and capture of the
PCB fraction for later analysis. This was done entirely by
one person over the full five-day field period, with a total
work time of approximately 55 hours. The developer
stated that, with one experienced user, 20 to 25 samples
that are dry and relatively clean could be finished per
day. The developer gives a range of acceptable times for
the first sample incubation of both immunoassays, from
2 hours to overnight (12 to 24 hours). The shorter time
gives a slight reduction in sensitivity (less than twofold),
and the longer time eliminates the possibility of variable
incubation time effects due to the time required for
sample addition. Otherwise the two options provide
equivalent results and the choice between the two is
made based on ease of use and turnaround time
requirements. Sample turnaround time is affected by this
protocol choice, but overall sample throughput is not.
During the entire project, except for the final day of the
field phase, the longer incubation time was used because
of flexibility and ease of use. On that last day, samples
extracted in the morning were processed to completion
that day.

The number of samples processed in a day could be
higher with two experienced users, but would be far less
for novice users. The first data would be available after
about 8 hours using the shorter incubation time, but not
until the next morning using the longer incubation time.
The developer reports that shorter run times have been
used in certain cases, but at the cost of decreased
sensitivity. For the fastest results, it is possible to
compress the entire immunoassay portion of the
procedure into 1.5 hours. This could give first results in
less than 8 hours. Smaller batch sizes will also give faster
turnaround, but throughput would decrease because of
the lesser efficiency of small batches. Based on
observation, it seemed far more efficient to extract a
large number of samples first, and then allow a large
batch to incubate overnight, with the immunoassay
completed the morning of the second day.

Because the  sample preparation method provides both
dioxin/furan and PCB fractions from a single procedure,
the additional time required for completion of the PCB
test for a set  of samples would be limited to the time
needed for solvent exchange, immunoassay set up, and
later immunoassay manipulations (the developer
estimates an extra 2 hours per day per batch of
20 samples). If the dioxin/furan and PCB tests for a
single set of samples were run concurrently and
incubated overnight, both dioxin/furan and PCB results
would be available by the middle of the next morning.

7.2.5  Miscellaneous Observer Notes
CAPE Technologies is a U.S.-based company. The
developer offers a training class that lasts a minimum of
1 !/2 days that can be held either at the developer's
laboratory or at the user's site. Training is completely
flexible and tailored to the needs of the customer. CAPE
Technologies also offers extensive phone support for
customers and encourages users to discuss their projects
so that CAPE Technologies can help guide the user's
choices of sample preparation and assay procedures.

The materials that come with the DPI kit are the coated
antibody tubes (amount varies by kit purchased), the
competitor conjugate, the HRP substrate, stop solution, a
vial of the Triton X-100, a test tube rack, and a bag of
uncoated tubes. The kit will also include a set of
standards and controls. For the analysis alone, the user
must supply methanol, a nitrogen evaporator, glass tubes,
a differential photometer (or any photometer capable of
reading at 450 nm), a timer, marking pens, a tray for
waste disposal, and reagent grade water. The extraction
and clean-up kit comes with disposable carbon
minicolumns, disposable acid silica columns, containers
for sample extraction, and steel bearings used for mixing.
The user must supply anhydrous sodium sulfate, hexane,
acetone, toluene, a balance, an orbital platform shaker, a
centrifuge, fume hood, small vacuum pump (optional),
luer ports, computer with Microsoft Excel 97 or higher,
glass pipettes, glass vials, a repeater pipettor, a variable
volume pipettor, and catch basins for column waste.

The developer assumes that there will be approximately
40% QC samples processed per kit.  The developer leaves
QC choices to the individual user, but it would
recommend method blanks, method spikes, duplicates,
and evaporation controls (solvent spikes that undergo the
evaporation step to determine loss).  All of the  above QC
samples were used during the demonstration, along with
matrix spikes and known samples (RMs or previously
analyzed samples). See Appendix B for developer FIRMS
calibration recommendations.
                                                    55

-------
                                              Chapter 8
                                          Economic Analysis
During the demonstration, the CAPE Technologies kits
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 comprehensive
economic analysis of each technology. The purpose of the
economic analysis was to estimate the total cost of
generating results by using the CAPE Technologies kits
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
DF1 Dioxin/Furan and PCB TEQ Immunoassay kits
(Section 8.2), discusses the costs associated with using the
reference methods (Section 8.3),  and presents a
comparison of the economic analysis results for the CAPE
Technologies kits 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 DPI Dioxin/Furan and PCB TEQ
Immunoassay kits unless otherwise stated.

8.1.1 Capital Equipment Cost
The capital equipment cost was the cost associated with
the purchase of the CAPE Technologies kits.
Components of the kits are presented in detail in
Chapters 2 and 7. The purchase price information was
obtained from a standard price list provided by CAPE
Technologies.

8.1.2 Cost of Supplies
The cost of supplies was estimated based on the supplies
required to analyze all demonstration samples using the
DF1  Dioxin/Furan and PCB TEQ Immunoassay kits that
were not included in the capital equipment cost
category. Examples of such supplies include filters,
cleanup columns, gas cylinders, solvents, and distilled
water. Only one sample preparation kit is required for
assay using both the DPI and PCB TEQ kits. The
supplies that CAPE Technologies used during the
demonstration fall into two general categories:
consumable (or expendable) and reusable. Examples of
expendable supplies utilized by CAPE Technologies
during the demonstration include hexane, acetone,
distilled water, toluene, methanol, tetradecane, nitrogen
cylinders, sodium sulfate, Pasteur pipets, and glass
disposable extraction tubes. Examples of reusable
supplies include a top loading balance, orbital platform
shaker, tabletop centrifuge, hot plate, and sample
evaporation system. It should  be noted that this type of
equipment may or may not be already owned by a
potential immunoassay kit user; however, for this
economic analysis, an assumption was made that the
user does not possess these items.
                                                    56

-------
The purchase price of these supplies was either obtained
from a standard price list provided by CAPE
Technologies, 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
demonstration such as the construction trailer, fume hood,
and laptop computer required by the technology. Costs for
these items will be reported per actual costs for the
demonstration.

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, developers reported
results by submitting a chain-of-custody (COC)/results
form. The measurement of the time required for CAPE
Technologies to complete 95 sample analyses during the
field demonstration (55 labor-hours) was estimated by the
sign-in log sheets that recorded the time the CAPE
Technologies operator was on-site. Time was removed for
site-specific training activities and Visitor's Day. Time
estimates were rounded to the nearest hour.

During the demonstration, the skill level required for the
operator to complete analyses and report results was
evaluated. As stated in Section 7.2.1, based on the field
observations, the education level may not be strictly
necessary, but the experience with dioxin cleanup
methods seemed very important in the use of the CAPE
Technologies kits. A technician with at least a bachelor's
degree in the sciences, and experience with both dioxin
cleanup methods and ELISA would be helpful.
Nonscientists with significant analytical experience
should be able to perform the method without undue
difficulty. Method-specific training could be obtained in
as little as two days. This information was corroborated by
CAPE Technologies.

The education level of the actual field operator includes a
Ph.D. degree for the primary operator. For the economic
analysis, costs were estimated using both actual and
projected necessary skill levels for the operator.

8.1.5  Investigation-Derived Waste Disposal Cost
During the demonstration, CAPE Technologies 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 CAPE
Technologies. 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 CAPE Technologies
electricity usage would be no more than $41. 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 CAPE Technologies kits would not be required to
pay for customer oversight of sample analysis. The
EPA, the MDEQ, and Battelle representatives were
present during  the field demonstration, 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
                                                    57

-------
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 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     DF1 Dioxin/Furan and PCB TEQ
        Immunoassay Kit Costs
This section presents information on the individual costs
of capital equipment, supplies, support equipment, labor,
and IDW disposal for the DF1 Dioxin/Furan
Immunoassay kit as well as a summary of these costs.
Additionally, Table 8-1 summarizes the DF1 Dioxin/
Furan Immunoassay kit and PCB TEQ Immunoassay kit
costs. As described in Section 4.6, CAPE Technologies
analyzed 95 samples during the field demonstration using
the DF1 kit, 114 samples in their laboratory using the DF1
kit, and all 209 demonstration samples using the PCB
TEQ kit in its laboratory. It is important to note that costs
estimated in this section are based on actual costs to
analyze the 95 samples for D/Fs during the field demon-
stration.  Cost estimates for analyzing the entire set of
209 demonstration samples for both D/F and PCBs were
then determined based on the field demonstration costs.

8.2.1   Capital Equipment Cost
The capital equipment cost was the cost associated with
the purchase of the technology to perform sample
preparation and analysis. The DF1 Dioxin/Furan
Immunoassay kit can be purchased from CAPE
Technologies for approximately $60 per sample. Two
sizes of immunoassay kits are offered and contain
enough supplies for 12 or 60 samples to be analyzed
respectively. In conjunction with the immunoassay kits,
CAPE Technologies offers sample preparation kits for
extraction and cleanup of samples in preparation of
immunoassay analysis. The sample preparation kit can
be purchased from CAPE Technologies for
approximately $15 per sample. Sample preparation kits
are sold for 12 or 60 samples corresponding to the
immunoassay kit sizes. All CAPE Technologies kits
assume that 40% of the resources in each kit will be
used for various quality control samples, including
standards, reference samples, replicates, and matrix
blanks. Because the kit is consumable, CAPE
Technologies does not rent the immunoassay kits.
During the demonstration, CAPE Technologies utilized
two DF1 Dioxin/Furan Immunoassay kits (assuming the
larger sized kit) over five days to analyze 95 samples
and a total of four DF1 kits for the entire 209 samples.
CAPE Technologies also used four PCB TEQ kits for
the 209 samples.

8.2.2   Cost  of Supplies
The supplies that CAPE Technologies used during the
demonstration fall into two general categories:
expendable or reusable. Table 8-1 lists all the
expendable and reusable supplies that CAPE
Technologies used during the demonstration and their
corresponding costs. The cost of each item was rounded
to the  nearest $ 1. 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 CAPE
Technologies during the demonstration was $7,143.
Supplies have to be purchased from a retail vendor of
                                                    58

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

Capital Equipment
Purchase of DF1 Dioxin/Furan Immunoassay
Kit
Purchase of PCB TEQ Immunoassay Kit
Purchase of Sample Preparation Kit
Supplies
Expendable13
Hexane (4-L bottle)
Acetone (1-L bottle)
Toluene (1-L bottle, 99.9%)
Nitrogen Cylinder
Cylinder Regulator
Methanol( 1-L bottle)
Tetradecane (25 mL)
Glass Tubes (16x125 mm; case of 1,000)
Sodium Sulfate (anhydrous, granular, 500 g)
Pasteur Pipets (package of 250; 2-mL size)
Reusable
Top Loading 0.1 -g Balance
Orbital Platform Shaker
Tabletop Centrifuge
Sample Evaporation System
Hot Plate
Photometer"1
Positive Displacement Pipettor
Repeater Pipettor
Support Equipment
Construction Trailer
Fume Hood
Laptop Computer
Labor

Operator
IDW Disposal6
Total Cost
During Field



2
2
2


2
1
2
1
1
1
1
1
1
1

1
1
1
1
1
1
1
1

1
1
1


55
1

Demo


kits
kits
kits


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
($)
\WJ


3,600
3,600
900


65
27
28
31
182
30
8
84
40
10

550
1,521
1,613
750
135
1,200
100
425

1,919
1,100
1,000


80C
133

Itemized
95
samples


7,200
7,200
1,800


130
27
28
31
182
30
8
84
40
10

550
1,521
1,613
750
135
1,200
100
425

1,919
1,100
1,000


4,414
133
31,630
Cost3 ($)
209
samples


15,840
15,840
3,960


286
59
62
68
182
30
18
84
40
20

550
1,521
1,613
750
135
1,200
100
425

1,919
1,100
1,000


12,139
293
59,234
    a Itemized costs were rounded to the nearest $1.
    b All reagents are HPLC grade, unless otherwise noted.
    0 Labor rate for field technicians to operate technology rather than research scientists was $50.75 an
     hour, $2,791 for 95 samples, and $7,675 for 209 samples.
    dA photometer can be rented from CAPE Technologies for $120 per month.
    e Further discussion about waste generated during demonstration can be found in Chapter 7.
                                                         59

-------
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.

8.2.3  Support Equipment Cost
CAPE Technologies analyzed demonstration samples in
a 32-foot construction trailer equipped with a fume hood.
As determined by the observers, a construction trailer
with fume hood would be necessary for operation of this
technology. The rental cost for the construction trailer for
use during sample extraction and sample analysis was
$1,919. The minimum rental rate for the construction
trailer was one month. CAPE Technologies only used the
construction trailer for five days. Since weekly or daily
rental rates for the construction trailer were not an option,
the entire cost is reported. The fume hood rental and
installation was $1,100.

A laptop computer is a necessary for the efficient
operation of this technology. This  is a one-time purchase
that is reusable.

8.2.4  Labor Cost
As described in Section 8.1.4, 55 labor-hours were spent
in the field to analyze 95 samples for D/F only. 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
$4,414 was determined for the analysis of the 95 samples
during the field demonstration. It was estimated that the
labor cost for the total 209 samples (for both D/F and
PCBs) was $12,139.

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
95 samples during the field demonstration could have
been as low as $2,791 (hourly rate of $20.30 with 2.5
multiplication factor for 55 labor-hours), and $7,675 for
all 209 demonstration samples (for both D/F and PCBs).

8.2.5  Investigation-Derived Waste Disposal Cost
As discussed in Chapter 7,  CAPE Technologies 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, CAPE Technologies analyzed
95 samples. The total cost to dispose of the waste
generated for these samples was $133. The cost to
dispose of waste  for all 209 samples is estimated at $293.

8.2.6  Summary ofDFl Dioxin/Furan andPCB
       TEQ Immunoassay Kit Costs
The total cost for performing dioxin and PCB analyses
using the DF1 Dioxin/Furan Immunoassay Kit and PCB
TEQ Immunoassay Kit was $59,234. The analyses were
performed for 58 soil and sediment PE samples, 128 soil
and sediment environmental samples, and 23 extracts.
When CAPE Technologies performed multiple dilutions
or reanalyses for  a sample, these were not included in the
number of samples analyzed.

The total cost of $59,234 for analyzing the 209 demon-
stration samples using  the DF1 Dioxin/Furan and PCB
TEQ Immunoassay kits included $35,640 for capital
equipment; $7,143 for supplies; $4,019 for support
equipment; $12,139 for labor; and $293 for IDW
disposal. Of these five  costs, the largest cost was for the
purchase of the kits (60% 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 and dioxin-like PCBs. 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 dioxin/furan
analysis and EPA Method 166 8A for coplanar PCB
analysis for all soil and sediment samples for comparison
with the CAPE Technologies kits. The reference method
costs were calculated using cost information from the
reference laboratory invoices.

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 and $885 for PCB
analysis. This cost covered  the preparation and analysis
                                                    60

-------
Table 8-2. Reference Method Cost Summary
Analyses Performed
Dioxin/Furans, EPA Method
161 3B, GC/HRMS
WHO PCBs EPA Method 1668A,
GC/HRMS
1668 Optional Carbon Column
DB1
Total Cost
Number of
Samples
Analyzed
23 extracts
186
soil/sediment
23 extracts
186
soil/sediment
40
209 samples
Cost per sample
Quotation ($)
735
785
685
735
150

Itemized Cost ($)
Quotation3
16,905
146,010
15,755
136,710
6,000
321,380
Actual
213,580
184,449
398,029
" 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 with them ($150 to $180 per sample per procedure).
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 and
$184,449 for PCBs, and atotal of $398,029. This was
higher than the projected ($321,380) 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 DF1 Dioxin/Furan Immunoassay
kit and PCB TEQ Immunoassay Kit analyses ($59,234)
and the reference method ($398,029) are listed in Tables
8-1 and 8-2, respectively. The total cost for the CAPE
Technologies kits was $338,795 less than the reference
method. It should be noted that CAPE Technologies
analyzed 95 samples for D/F in five days on-site during
the field demonstration and it completed the remaining
samples off-site in its laboratory. CAPE Technologies
reported that the total analysis time to  analyze the
remaining 114 samples for D/F and all 209 samples for
PCBs was two weeks. For comparison, the reference
laboratory took 8 months to report all 209 samples.

In addition, use of the immunoassay kits 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 CAPE Technologies immunoassay kits are screening
methods that report TEQ values, unlike the reference
method, which reports concentrations for individual
congeners. Although the CAPE Technologies kit
analytical results did not have the same level of detail as
the reference method analytical results (or comparable
QA/QC data), the  DPI Dioxin/Furan and PCB TEQ
Immunoassay kits provided analytical results that could
be generated on site at significant cost and time savings
compared to the reference laboratory.
                                                     61

-------
                                              Chapter 9
                                Technology Performance Summary
The purpose of this chapter is to provide a performance
summary of the CAPE Technologies DF1 Dioxin/Furan
and PCB TEQ Immunoassay kits 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 CAPE Technologies kits
in many cases did not directly correlate with HRMS TEQ
values, but that the kits 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
sample above and below 50 pg/g TEQ, particularly
considering that both the cost ($59,234 vs. $398,029) and
the time (three weeks vs. eight months) to analyze the
209 demonstration samples were significantly less than
those of the reference laboratory. Because the CAPE
Technologies kits are not expected to directly correlate to
FiRMS TEQ in all cases, the technology should not be
viewed as producing an equivalent measurement value to
FiRMS TEQ but as a screening value to approximate
FiRMS TEQ. As described in CAPE Technologies
literature, the best results for immunoassay screening are
obtained on a single site basis.  The ideal approach
involves partially characterizing a site by FIRMS, using
those results to develop a site specific immunoassay
calibration, and refining that calibration over time, based
on an ongoing stream of confirmatory FIRMS samples.
This approach was not evaluated during this demon-
stration; samples from multiple sites were pooled and a
single calibration was used.
                                                    62

-------
Table 9-1.  CAPE Technologies LLC DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits Performance
Summary - Primary Objectives
Objective
P 1 : Accuracy
P2: Precision
P3: Comparability
P4: Estimated
Method Detection
Limit
P5: False
Positive/False
Negative Ratea
P6: Matrix Effects
P7: Cost
Performance
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)
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 (%)
TEQPCB
8
151
195
49
118
123
182
-13
82
50
20
14
8
5
4
TEQD/F
7
182
236
44
63
71
167
-26
71
38
12-35
11
8
4
8
Total TEQ
8
139
160
46
67
74
159
-5
64
25
25-33
14
10
o
J
5
• Measurement location: 1 9% statistically different
• Matrix type: none
• Sample type: none
• PAH concentration: slight effect on total TEQ
• Environmental site: none
• Known interferences: slight
Cost for the analysis of 95 samples for D/F only during field demonstration: $3 1 ,630
Cost for the analysis of all 209 samples for both D/F and PCBs: $59,234
a CAPE Technologies notes that the user can optimize performance of the kit at other desired screening levels by changing the size of sample
extracted so that the desired screening level concentration falls at the optimum point on the kit response curve.
                                                         63

-------
Table 9-2.  CAPE Technologies LLC DF1 Dioxin/Furan and PCB TEQ Immunoassay Kits Performance
Summary - Secondary Objectives3
           Objective
                                    Performance
 SI: Skill level of Operator
The developer recommends that users have at least a bachelor degree in the sciences, and
experience with both dioxin cleanup methods and ELISA would be helpful. From
observation, the education level may not be strictly necessary, but the experience with dioxin
clean-up methods seemed very important in the use of the kit. During the field demonstration,
95 samples were processed by CAPE Technologies equating to a sample throughput rate of
19 samples per day.  This was accomplished in about 5 full working days (55 labor-hours),
with a single operator performing all aspects of the technology operation.
 S2: Health and Safety Aspects
Approximately 60 mL of hexane will be used for each sample. Not all of this becomes solvent
waste, but a portion does. The disposable columns as well as the disposable glassware create
solid waste. The solvent waste itself is not inherently more hazardous than would be expected,
unless the samples are very contaminated. A fume hood is necessary for use during solvent
extraction.
 S3: Portability
This technology is readily deployable in a field or mobile environment. The need for power,
nitrogen tanks, a fume hood, shakers, vacuums, and other equipment requires a trailer at the
very least for successful operation of this technology. While the assay itself is very easy to use
in the field, the extraction and cleanup methods require level work space and could create
some difficulty in the field unless there was a mobile lab or trailer available.
 S4: Sample Throughput
During the field demonstration, 95 samples were processed by CAPE Technologies equating
to a sample throughput rate of 19 samples per day. This was accomplished in about five full
working days (55 labor-hours), with a single operator performing all aspects of the technology
operation. CAPE Technologies reports that the analysis time for the remaining 114 samples
for D/F and all 209 samples for PCBs was approximately two weeks in its laboratory.
a The secondary objectives were assessed only for the DF1 kit because the PCB TEQ Immunoassay Kit was not deployed during the field
  demonstration. Because of the similar principles and procedures for the two kits, it is likely that similar conclusions could be drawn, but this
  was not confirmed by observation.
                                                          64

-------
                                            Chapter 10
                                             References
1.   EPA. 2001. Database of Sources of Environmental
    Release of Dioxin-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/82l/B-94-
    005, 40 Code of Federal Regulations Part 136,
    Appendix A, October.

4.   EPA Method 1668A. 1999. Chlorinated biphenyl
    congeners byHRGC/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, 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 NOS ORCA 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
                                                  65

-------

-------
                   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: DF1 Dioxin/Furan Immunoassay Kit and PCB TEQ Immunoassay Kit

      COMPANY:            CAPE Technologies LLC
      ADDRESS:             3 Adams Street
                             South Portland, Maine 04106-1604
      PHONE:               (207) 741-2995

      WEB SITE:             http://www.cape-tech.com
      E-MAIL:               cape-tech@ceemaine.org
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 CAPE Technologies DF1 Dioxin/Furan and
polychlorinated biphenyl (PCB) toxicity equivalent (TEQ) immunoassay kits.

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
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.

                                             A-l

-------
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. CAPE Technologies analyzed 95 samples for D/F only during the
field demonstration; the remaining 114 samples for D/F and all 209 samples for PCBs were analyzed in its laboratory.
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; 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 (FiRMS) (EPA Method 1613B and EPA Method 1668A). 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 ofDioxin and Dioxin-like Compounds in Soil and Sediment—CAPE
Technologies DPI and TEQ PCB Immunoassay kits (EPA/540/R-05/004).

TECHNOLOGY DESCRIPTION

The technology description and operating procedure below are based on information provided by CAPE  Technologies.
The DPI Dioxin/Furan Immunoassay Kit and the PCB TEQ Immunoassay Kit are nearly identical in design and
operation. They differ primarily in the antibody and competitor-horseradish peroxidase (HRP) conjugate used, and in
the specificity resulting from these specially developed reagents.  Both kits are designed to provide results as TEQ
concentrations by responding to the toxic dioxin/furan or PCB congeners in approximate correlation with their toxic
equivalency factors (TEFs).  Both tests recognize multiple congeners, preferentially targeting congeners with high TEF
values, i.e., those with the highest toxicity relative to  2,3,7,8-tetrachlorodibenzo-/?-dioxin (TCDD).  The specificity of
the dioxin/furan test is predominantly for dioxins and furans that contain from 3 to 6 chlorines, with a strong
preference for the 2,3,7,8 chlorinated congeners. This specificity roughly parallels the TEF values of the individual
dioxin and furan congeners. The specificity of the  PCB TEQ test is predominantly for non-ortho and mono-ortho
chlorinated congeners, with a strong preference for PCBs 126 and 169. This specificity roughly parallels the TEF
values of the individual PCB congeners. Both tests have only minimal recognition of the target compounds of the
other test. The immunoassay specific sample preparation begins with an organic solvent extraction. The extracts are
then processed through an immunoassay specific cleanup.  In the case of this evaluation, the cleanup combines two
familiar parts of the EPA Method  8290 cleanup, but in a way that allows for rapid batch processing using inexpensive
disposable columns and no specialized equipment. Since the cleanup is performed in solvents incompatible with the
immunoassays, a solvent exchange is required after the cleanup. Dioxins, furans, and dioxin-like PCBs have very low
volatility and are retained during this solvent exchange in a small volume of a keeper solution (Triton X-100 detergent
in tetraethylene glycol [TEG]) after evaporation of the original solvent. Methanol is added to dilute this solution, and
the methanol-TEG-Triton X mixture is added directly to the immunoassay tubes. During the first immunoassay
incubation, analyte molecules are  specifically bound by the analyte-specific antibodies, which have been immobilized
on the immunoassay tube surface. After washing away the unbound material, the bound analyte molecules remain, and
a competitor-HRP conjugate is added. Bound analyte molecules occupy the binding sites of the antibodies in
proportion to the dioxin/furan or dioxin-like PCB content of the sample, reducing the binding of the competitor-HRP

                                                   A-2

-------
conjugate. After a short incubation, unbound conjugate is removed, and the test tubes are washed thoroughly. Finally,
a solution of chromogenic HRP substrate and hydrogen peroxide is added to the test tubes. Color development is
directly proportional to enzyme concentration and inversely related to the dioxin/furan or dioxin-like PCB concentra-
tion in the original sample. The test tubes are analyzed using a tube reader or spectrophotometer to measure the optical
density (OD). The OD values of unknown samples are compared to the OD values of standards to determine the level
of dioxin/furan or dioxin-like PCB in the samples.

VERIFICATION OF PERFORMANCE

The CAPE Technologies kits are immunoassay technologies that report total dioxin/furan TEQ (TEQD/F) and total
coplanar PCB TEQ (TEQPCB) in the sample in picogram/gram (pg/g). It should be noted that the results generated by
the CAPE Technologies kits may not directly correlate to HRMS TEQ in all cases because it is known that the
congener responses and cross-reactivities of the kits are not identical to the TEFs that are used to convert congener
FIRMS concentration values to TEQ. Therefore, these kits should not be viewed as producing an equivalent
measurement value to HRMS TEQ but as a screening value to approximate HRMS TEQ. As described in CAPE
Technologies literature, the best results for immunoassay screening are obtained on a single  site basis. The ideal
approach involves partially characterizing a site by HRMS, using those results to develop a site specific immunoassay
calibration, and refining that calibration over time, based on an ongoing stream of confirmatory HRMS samples. This
approach was not evaluated during this demonstration; samples from multiple sites were pooled and a single
calibration was used.

Accuracy: The determination of accuracy was based on the agreement of the CAPE Technologies results with the
certified 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 kits divided by the certified or spiked value of the
PE sample, multiplied by 100%. Ideal Rvalues are near 100%. The overall Rvalues were 195% (mean),  151%
(median), 26% (minimum), and 523% (maximum) for TEQPCB; 236% (mean), 182% (median), 30% (minimum), and
595% (maximum) for TEQD/F, and 160% (mean), 139% (median), 33% (minimum), and 296% (maximum) for total
TEQ.

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 123% (mean), 118% (median), 38% (minimum),
and 200% (maximum) for TEQPCB 71% (mean), 63% (median), 0% (minimum), and 187% (maximum) for TEQD/F;
and 74% (mean), 67% (median), 17% (minimum), and 174% (maximum) for total TEQ.

Comparability: The CAPE Technologies DF1 and PCB TEQ kit results were compared to EPA Method 1613B and
EPA Method 1668A results for TEQPCB, TEQD/F, and total TEQ. 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 -13%
(median), -200% (minimum), and 200% (maximum) for TEQPCB; -26% (median), -199% (minimum), and 198%
(maximum) for TEQD/F; and -5% (median), -199% (minimum), and 198% (maximum) for total TEQ.  The CAPE
Technologies results were also  compared to the reference laboratory results using an interval approach to determine if
the CAPE Technologies 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 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  82% for TEQPCB71% for TEQD/F and 64% for total TEQ.

Estimated method detection limit: EMDL was calculated for the for the technology generally according to the
procedure described in 40 CRF Part 136, Appendix B, Revision 1.11. Lower EMDL values indicate better sensitivity.
The calculated EMDLs ranged from 12 to 35 pg/g TEQ, depending on whether nondetect values were assigned values


                                                   A-3

-------
of zero, one-half the reporting limit value, or the reporting limit value itself. The detection limit reported by CAPE
Technologies in the demonstration plan was 1 pg/g TEQ.

False positive/negative results: Samples that were reported as less than a specified level by the reference laboratory
but above that level by CAPE Technologies were considered false positive. Conversely, those samples that were
reported as less than a specified level by CAPE Technologies but reported as greater than the specified level by the
reference laboratory were considered false negatives. Ideal false positive and negative percentages were zero. The
CAPE Technologies kits had a false positive rate of 14% and a false negative rate of 5% for TEQPCB; 11% and 6%,
respectively, for TEQD/F; and 14% and 3%, respectively, for total TEQ for reporting data above and below 20 pg/g TEQ
relative to the reference laboratory data. CAPE Technologies's false positive and false negative rates around 50 pg/g
were generally lower for all three TEQ types, ranging from 4 to 10%. These data suggest the CAPE Technologies kits
as processed in the demonstration 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 sample above and below 50 pg/g TEQ. CAPE Technologies notes
that the user can optimize performance of the kit at other desired screening levels by changing the size of sample
extracted so that the desired screening level concentration falls at the optimum point on the kit response curve.

Matrix effects: The likelihood of matrix-dependent effects on performance was investigated by evaluating results in a
variety of ways. The CAPE Technologies TEQD/F results that were generated in the laboratory and in the field for
replicate samples were statistically different for 19% of the samples, and of these samples, CAPE Technologies results
were most comparable to the reference laboratory results. No  significant effect was observed for the reproducibility of
CAPE Technologies results by matrix type (soil, sediment, and extract) or by sample type (PE vs. environmental vs.
extract). A slight effect was observed for total TEQ values by PAH concentration, but the effect was not statistically
significant for TEQD/F or TEQPCB. PE samples spiked for a particular contaminant (e.g., D/Fs) were sometimes reported
as detections for other analytes that were not spiked in the sample (e.g.,  PCBs). The CAPE Technologies results were
not more or less comparable to the reference laboratory results based on environmental site.

Cost: The full cost of using the CAPE Technologies kits  was documented and compared to the cost of the reference
analyses. As demonstrated, the total cost for the CAPE Technologies kits  to analyze all 209 samples was $59,234. The
cost for the reference laboratory to analyze all 209 samples by Method 1613B and Method 1668A was $398,029. The
total cost for the CAPE Technologies kits was $338,795 less than the reference method.

Skills and training required: The developer recommends that users have at least a bachelor degree in the sciences,
and experience with both dioxin cleanup methods and enzyme-linked immunosorbent assay would be helpful.  From
observation, the education level may not be strictly necessary, but the experience with dioxin cleanup methods seemed
very important in the use of the kit.

Health and safety aspects: Approximately 60 mL of hexane  will be used for each sample. Not all of this becomes
solvent waste, but a portion does. The disposable columns as  well as the disposable glassware create solid waste. The
solvent waste itself is not inherently more hazardous than would be expected, unless the samples are very
contaminated. A fume hood is necessary for use  during solvent extraction.

Portability: This technology is readily deployable in a field or mobile environment. The need for power, nitrogen
tanks, a fume hood, shakers, vacuums, and other equipment requires a trailer at the very least for successful operation
of this technology. While the assay itself is very easy to use in the field,  the extraction and cleanup methods require
level work space and could create some difficulty in the field  unless there was a mobile lab or trailer available.

Sample throughput: During the field demonstration, 95  samples were processed by CAPE Technologies equating to a
sample throughput rate of 19 samples per day. This was accomplished in about five full working days (55 labor-hours),
with a single operator performing all aspects of the technology operation.  CAPE Technologies reported that the
analytical time to complete the remaining 114 samples for D/F and all 209 sample analyses for PCBs was two weeks.
For comparison, the reference laboratory took eight months to report all 209 samples.
     NOTICE: Verifications are based on an evaluation of technology performance under specific, predetermined
     criteria and the appropriate quality assurance procedures. 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.

-------

-------
                                             CAPE Comments

General Comments
CAPE Technologies wishes to express its gratitude to both EPA andBattelle for the skillful and professional manner in
which this studywas designed and conducted. This task was obviously massive and probably seemed thankless.
However, we understand the value of the effort that was made, and we sincerely appreciate the consideration
andflexibility shown during all phases of the study.

The purpose of this appendix is to provide CAPE Technologies, as a technology developer, the opportunity to comment
on the study and its results. Additionally, this is the place for information that the developer may deem to be relevant
to the technology and its use but outside the scope of the main part of the report.  Thus, the comments offered below
are a critical part of the study and should be read carefully by any potential user of this technology. Additionally, much
information beyond what is presented here can be found on the CAPE Technologies Web site (www.cape-tech.com).
This Web site covers topics from the principles of immunoassay technology to the practical issues of quality assurance
and data interpretation, and many important things in between. The CAPE site also includes links to EPA sites in two
key areas: (1) SW-846 solid waste methods, including Method 4025, which uses the CAPE Technologies DF1 kit; and
(2) the Technology Innovation Program(TIP), which seeks to educate the environmental community about new
technologies, such as field analytical methods, and how to use them effectively.

Any research on Method 4025 or on CAPE Technologies products in general must refer first to the company Web site,
since this is the single most important source of up-to-date information about the products themselves and the methods
for which they are used.

Comments on the Reported Economic Value of the Kits Evaluated vs. the True Value of Field Analytical
Methods
The primary conclusion regarding the two CAPE  Technologies immunoassay kits studied was expressed in several
places in the report. This evaluation project has demonstrated that CAPE Technologies kits "...could be an effective
screening tool..." at low- to mid-pg/g levels in soil and sediment.  Secondarily, the  report notes that"... the
DF-lDioxin/Furan and PCB TEQ Immunoassay Kits provided analytical results that could be generated on-site at
significant cost and time savings compared to the reference laboratory." In summary, the data on cost, turnaround time,
and the ability  of the method to work in the field demonstrate profound advantages  over conventional methods.

These attributes are certain to be attractive to many who simply see analternative analytical method. However, the
most significant advantage  to using CAPE Technologies kits is that they offer the site manager a completely new tool
to address a set of very old  and familiar problems. The advantages of Method  4025  for site assessment and remediation
are very similar to what was demonstrated by other EPA SW-846 4000 series  methods in the early 1990s (such as 4020
for total PCB).  These include (a) potential for reduction of multiple crew and equipment mobilization-demobilization
cycles to a single cycle, (b) dramatically increased spatial resolution mapping  for higher statistical confidence, leading
to reduced insurance costs and reduced disposal costs, (c) compression of total project time to reduce administrative
overhead, and (d) better defensibility of site actions for reduced legal costs.

Field analytical technologies, especially for dioxin and dioxin-likePCBs, represent a paradigm shift in site assessment
and remediation. Quite simply, the speed, simplicity, and low cost of the 4000 series methods provide site managers the
ability to generate  site characterization maps and to guide remediation with speed and statistical confidence that are
literally impossible with conventional lab-based methods.

As shown clearly by the EPA materials linked to from CAPE's site, the true cost savings from the use of field
analytical technologies come from these points, rather than just the six-fold cost differential between two "competing"
analytical methods. In fact, an important lesson to be learned from TIP's efforts is that field analytical methods and
lab-based methods are complementary rather than competing, and the value of using the two methods in concert is
much greater than the value of using either method alone.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                    B-l

-------
Technical Issues in the Study Design and Interpretation of Results-Pooled Calibration and FP/FN Rates
Readers of this report likely know that analysis of dioxin and dioxin-like PCBs is expensive, time-consuming,
technically difficult,and applied to an extremely wide variety of sample types and congener profiles. Because of these
factors, the design of this evaluation study necessarily represents a balancing act between comprehensiveness on the
one hand and cost control on the other. Some important ramifications of this compromise are discussed in more detail
below.

Users of field analytical methods such as the two immunoassay kits evaluated in this report will almost always be
seeking data from a single site or a group of related sites. In such cases, CAPE Technologies makes specific
recommendations on how to apply the kits, especially including quality assurance samples and quantitative data
interpretation. A key part of this recommendation is to use prior GC-MS data, if available, to perform an initial site-
specific calibration adjustment, then to generate an ongoing stream of confirmatory GC-MS samples to for continuing
support refinement of this calibration. A draft CAPE Technologies document that addresses these concerns, Technical
Note TN-004, is included as part of this appendix.

Virtually no one uses the Dioxin/Furan or PCB-TEQ immunoassays as theyare presented in the report: a single large
data set with many sample types from multiple unrelated locations, pooled and using a single calibration adjustment
from samples that may or may not be from the same site as the unknowns. Because of this and the blind nature of the
study (i.e., not knowing which samples came from one site), the blanket calibration adjustment applied is necessarily a
compromise.  It serves the nominal statistical purposes of the study, but in no way represents how individual kit users
apply the raw data to decision making in the field. Differences among sites may arise for many reasons and can
contribute to significant errors unless the differences are controlled.  Site-specific calibration does this.

An important ramification of this issue is that in a single site, application with site-specific calibration based on selected
GC-MS results, false negative and false positive rates at a predetermined level should decrease from the present study,
with its pooled and and nonspecific calibration.  Because of the study design, as dictated primarily by cost constraints,
retrospectively imposing such a calibration is impossible, as that would require a set of known samplesfor calibration of
each sample type/location subset.

An example of quantitative results from a blind, single-site study is given in CAPE Technologies Application Note
AN-008. The methodology is nearly identical to that used here, except that none of the samples were  analyzed in a
field lab situation.  The data summary page of AN-008 is included as part of this appendix.  The full document is
available on CAPE's Web site. An extreme case of the need for site-specific calibration can be found in Finland, where
the primary wood preservative (with use patterns parallel to pentachlorophenol in the US) contains  a dioxin/furan
congener profile unlike any other sample type. Because these samples get about 70% of their TEQ  from one
heptachlorodibenzofuran, which is poorly recognized by the DF1 kit, a calibration adjustment of approximately
sevenfold is required, due solely to the congenerprofile. However, because the DPI kit results still  correlate with  TEQ
and calibration is a separate issue, once the calibration adjustment is made, the adjusted results can be  used effectively
for screening at apreselected level. The use of the DF1 kit for this type of sample is heading into its second field season
in Finland.

Technical Issues in the Study Design and Interpretation of Results-Curve Slope, Dilutions, and High End
Quantitation
Both kits evaluated in this study are competitive immunoassays having anegative slope response curve, which is
intrinsically sigmoid in shape. Thus the theoretical precision and the working range trade off againsteach other. As the
curve slope increases, the precision improves, butthe working range shrinks.  As the curve slope decreases, the
workingrange increases, but the precision deteriorates. The Dioxin/Furan testand the PCB TEQ test have slightly
different curves, but both representreasonable compromises between working range and precision.

Because of the very wide range of concentrations in the study samples, an amount of sample was chosen that  would
give the best results for low concentration samples. Many of the study samples were therefore off the high end of the
standard curve in their first run and required dilution to give a quantitative result. There are two ways  of doing this;
and, because of the time and fiscal constraints of the study, the faster and less expensive of these methods was used.
The approach used is also the less accurate of the two, but the effect of this on decision making should be nearly nil.

             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                    B-2

-------
The reason is simple. Even though dilution and retesting of an off-scale high sample may give a result with a greater
proportional error than a sample not requiring dilution, the first result of off-scale high remains unchanged.

Proper use of the Dioxin/Furan and PCB-TEQ kits for a single decision level is simple. Choose an amount of sample
that would place your result in the steepest part of the curve if the sample were at the decision level. This approach is
required if the best possible false positive/false negative rates at that decision level are to be obtained.  Use of the kits
for a second decision level would require either the dilution protocol used in the study or, more accurately, repeat
cleanup of another aliquot of sample extract that would allow the analyst to also make the second decision in the
steepest part of the response curve.  This is generally not done because it increases the sample preparation resources
required.
              Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                      B-3

-------
CAPE Technologies                                                 DRAFT

High Performance Dioxin/Furan  Immunoassay  Kit


Technical Note TN-004

Quantitation, Calibration,  and Quality Assurance for Method 4025m

Quantitation:  Dioxin/furan  analysis by  US EPA  Method  4025m using  the  CAPE  Technologies DF1
Immunoassay Kit gives quantitative results which correlate with TEQ (per Application Note AN-008).  However,
just as with conventional  chemical analysis, proper calibration and quality assurance are required for maximum
reliability.

The DF1 immunoassay is inherently quantitative.  Each immunoassay run should include 2378-TCDD standards
to define a standard curve as described in Section D (Table 1) of the kit insert IN-DF1.  This  curve is applied to
unknowns using Calculation  Module C,  a special purpose  Microsoft Excel file  available from the  CAPE
Technologies  web site (www.cape-tech.com).   Module C  uses an  iterative  non-linear curve fitting procedure
based on the  same four  parameter equation which is the basis for  a variety of commercial  immunoassay data
analysis software.  Module C calculates the best fit standard curve and the  concentrations of unknowns  based
on that curve.   Background information and instructions are included  with Calculation Module C.

The process described above  produces raw quantitative results based on the standard curve,  which may or may
not be an  acceptable endpoint.  If the analyst's goal is  relative quantitation (i.e.  looking for hot spots- finding
deviations from a certain baseline  and  estimating their  concentration  relative to that baseline), then  no
calibration  adjustment is required.  However, if the goal is absolute quantitation (as for virtually all dioxin analysis
by GC-MS), then a calibration  adjustment must be applied to the raw quantitative results.  Calculation Module C
has this calibration  adjustment  calculation built in, but the  analyst must determine the  actual  calibration
adjustment factor (CAP) and provide the QA data supporting its use.

Calibration of other  4000  series  methods: In order to  articulate the rationale supporting  this calibration
adjustment, it  is helpful to first describe the  approach to  calibration for the other 4000  series immunoassay
methods approved by the US  EPA (www.epa.gov/epaoswer/hazwaste/test/4_series.htm). These  methods, such
as Method 4020 for PCBs, have a calibration  adjustment built into the method.  This adjustment is determined
by the kit manufacturer and is applied on the front end,  through the  use of  immunoassay calibrators instead of
standards. These calibrators are designed to let the analyst make  semi-quantitative decisions  at pre-selected
levels, such as 1, 5, 10, or 50 ug/g. Once the kit user compares the  sample  to a calibrator in  the  same run and
makes a decision, no further data interpretation is required.  The calibration rationale assumes that the  samples
to  be analyzed and the decision levels to be used are the same as those used for the validation study.

The actual concentrations of these calibrators may differ from the decision level  by a factor of two or more.  For
example, users of one of the Method 4020 PCS kits would make a decision on whether the sample PCS level is
less than 10 ug/g by comparing it to a  calibrator  in the same  run that  actually contains 5 ug/g  PCB.  This
difference  between decision level and actual  concentration used for the calibrator is determined by splitting
samples and analyzing by both the conventional  method and the immunoassay, in quantitative  mode and with
no adjustment of the data. The resulting quantitative relationship between the two data  sets is used to set the
calibrator level so that a minimum false  negative rate is  achieved  in the   semi-quantitative decision  making
process.

There are  several good reasons why these quantitative  results from the  two methods  might not follow a  1:1
relationship (regression line slope of 1), even if the correlation is excellent.  These include, but are not limited  to,
reduced efficiency of the rapid  extraction  method, effects of differences in congener profile between the PCB in
the sample and standard, and random variation.  The front end calibration  procedure  described above  allows
compensation for all such factors together, without explicitly determining their individual contributions.  The
calibration adjustment  described above is effectively the same as  obtaining unadjusted quantivative  results,
then multiplying them by a uniform adjustment  factor. The approach to calibration for Method 4025m  is similar
and accomplishes the same goal,  but with some very important differences.  The rationale for this approach is
described below.
TN-004            Quantitation, Calibration, and Quality Assurance for Method 4025m           10/7/04
     Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

                                                B-4

-------
Calibration rationale and procedure for Method 4025m:  The same factors noted above which can cause the
regression line slope to be less than 1 must also be dealt with when calibrating Method 4025m.  However,  there
are more potential factors because  of the increased complexity of the procedure  (e.g.  recovery through cleanup
and solvent exchange as well  as extraction) and because of the greater variability of the analyte composition
(congener profile)  among the  population of possible  samples.   For these  and other reasons,  the front end
calibration approach described above for other 4000  series immunoassays is not  viable for  Method 4025m.
Therefore Method 4025m analysis  uses standards  rather  than calibrators, and the analyst applies a back end
calibration adjustment to the raw  quantitative  results.

The calibration procedure supported by the above rationale is straightforward.  A  set of split samples is analyzed
by the reference method (GC-MS)  and also  by Method 4025m. The comparison data set will likely  have  some
deviation from the  ideal 1:1  relationship  noted above (regression line slope  other than 1).  A  new data set  of
adjusted 4025m results is created by multiplying each raw 4025m result by the CAP (starting at 1).  The CAP is
then changed and the  regression line slope is calculated for the  adjusted 4025m data.   The final CAP value is
that which gives a regression line slope of 1 for the adjusted 4025m data.  This CAP is then uniformly applied  to
all raw 4025m results.  Once a CAP is determined, it should be checked and  refined continuously using the
stream of GC-MS data from ongoing quality  assurance samples.  On a larger project, from 5 to  20 percent  of
samples screened by  Method 4025m should be split for conventional analysis.   These are the most important
quality assurance samples,  but are  by no means the only ones that should be run.

Notes on calibration quality:  For  best results, calibration adjustment should  be  done on a site specific basis if
possible.   Differences  in dioxin source,  sample  matrix,  and congener  profile will all  increase  the variability  of
quantitative results and decrease the probability of success.  The  effect of congener profile on calibration can
be estimated  in advance using Calculation  Module A.  More  samples will obviously give better results.   It is
theoretically possible to base a CAP on a single sample, but statistically unwise.  Likewise, it is statistically best
for the samples on which the CAP is based to cover as wide a concentration range as possible.

The closer the calibration samples are to the target sample population, the better the  calibration adjustment will
be.   It is  possible to use other reference samples for calibration, but the results will  not be as good as  when
using samples from the same set as the unknowns.  For  example,  calibration based solely on spiked samples
can be used,  but  is less than ideal, since  it will not  account for extraction differences between  spikes and
incurred  residues.   Likewise,  calibration based solely on unrelated  samples, such  as standard  reference
materials,  will not account for matrix differences between the reference sample and the unknown samples.
TN-004             Quantitation, Calibration, and Quality Assurance for Method 4025m           10/7/04
      Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                 B-5

-------
CAPE Technologies High Performance Dioxin/Furan Immunoassay Kit

Table 1.  Quality assurance data within modified EPA Method 4025

QA sample	n	mean±SD    units (comment)
Solvent exchange negative controls    24     1.9±1.0
Unspiked method blanks              22     2.9±1.7
                pg (lowest standard = 3.2 pg)
                pg (lowest standard = 3.2 pg)
Solvent exchange positive controls     22     102±20    % of nominal pg (generally 50 pg)
Spikes into method blank extracts      14      79±27     % of nominal pg/g (30 to 195 pg/g)
Spikes into sample extracts            32      67±28     % of nominal pg/g (30 to 390 pg/g)
Duplicate precision (two aliquots of one extract cleaned and analyzed in parallel)
                                    23      13±14     % cv (range 5 to 750 pg/g)
Figure 1. Correlation between modified EPA Method 4025 and EPA Method 8290

A set of 23 soil  samples from a sewage  treatment facility were  prepared and analyzed as described in
Sections F and G above.  Each EIA tube received  prepared extract equivalent to 500  to 640 mg of original
sample. Samples which gave off-scale high results were  diluted and the EIA was repeated so that each  EIA
tube received the equivalent  of 50 to 100 mg of prepared  sample.  Subsamples of each  sample were
analyzed separately by HRGC-HRMS.  The TEQ values were  calculated from TEF values and individual
congener concentrations  as measured  by Method  8290.   The  line represents  x = y.   The correlation
coefficient was 0.95. The  mean relative percent difference value for the samples plotted below was 24±15%
(±SD), with a maximum of 47%.

                10000
                 1000
          pg/g by
          Method
          4025m
                  100
                   10






Correlation between EPA Method 8290
and EPA Method 4025 (modified)




















;





















/



















/














t


/


1 X = Y





















/




















/




















/


m















4

/



















z


•
















•
1*
1


















•
^
/

















J

?
•
1

















\
/




















/



















4
•


















p



















/
j;— •-
















/




















/




















JF




















































/

























/
























j
s

























                      10
  100                 1000
pg/g TEQ by Method 8290
                                                                               10000
www.cape-tech.com
                               AN-008
11/6/03    page 6
           Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                 B-6

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

-------

-------
Table C-1. 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. 3-7,100 (PE)
3 1-269 (Midland)
72.8 (Brunswick)
123 (Titta. River sediment)
0.1 59-7690 (PE)
25. 7-1 92 (Midland)
35.2-1,300 (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-1

-------

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 (Saginaw River)
32.9-36.4 (North Carolina)
0.268-100 (Extracts)




















2,1 60-3,080 (Nitro)
146-1,320 (Saginaw River)
788-8,410 (North Carolina)
1,100-10,800 (North Carolina)
7,160-11,300 (Winona Post)

0.0386-9.28 (PE)

25.8 (Midland)
0.524-24.8 (PE)
10.4(RaritanBay)
5 3. 1-444 (Extracts)








2800 (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 1 5) 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 L7 182-6 (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 (RaritanBay)
1.21-5. 06 (Newark Bay)
0.104-0.330 (Brunswick)
0.1 32-0. 369 (Brunswick)
0.034-0.649 (Titta. River
sediment)
0.00277-1030 (PE)
4.20-1, 020 (PE)
0.974-2.73 (Midland)
1 0.3-1,1 80 (PE)
0.0157-62.4 (Saginaw 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)
41. 0-1, 140 (PE)
0.00385-0.051 (PE)
0.253-0.318 (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)
40.1^2.1 (PE)
0.000103-1,080 (Extracts)
435-1, 160 (PE)
0.388-0.452 (Nitro)
0.0467 (Saginaw River)
0.654-1.87 (WinonaPost)
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.
1 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/FWG13551
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,Ref27PE
L6760-14, Ref 55PE
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-1/2 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
PCB WG12147
PCB WG12265
PCB WG12457
PCB WG12687
PCB WG12834
PCB WG12835
PCB WG12836
PCB WG1 3008
PCBWG13256
PCBWG13257
PCB WG13258
PCB WG13554
PCBWG14109
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= 1/2 DL basis)
15
19
12
85
(on U= 1/2 DL 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.
1 Nondetects were assigned a concentration of zero unless otherwise noted and are referred to as U=0 DL values.
b U=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. CAPE Technologies 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
Sample
Number
CAPE 105
CAPE 139
CAPE 130
CAPE 76
CAPE 158
CAPE 115
CAPE 136
CAPE 84
CAPE 195
CAPE 25
CAPE 83
CAPE 117
CAPE 49
CAPE 56
CAPE 172
CAPE 209
CAPE 197
CAPE 113
CAPE 164
CAPE 94
CAPE 173
CAPE 98
CAPE 27
CAPE 122
CAPE 166
CAPE 96
CAPE 71
CAPE 191
CAPE 38
CAPE 148
CAPE 170
CAPE 40
CAPE 31
CAPE 155
CAPE 121
Measurement
Location3
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Field
Laboratory
Field
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Field
Field
Laboratory
Laboratory
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
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
o
J
4
1
2
3
4
1
2
o
5
TEQPrR (pg/g)
Developer1"
12
2
4
0
2
10
1
0
153
77
212
194
1
3
24
0
0
19
1
0
33
1
0
11
8
0
1
0
6506
3714
22756
5217
7766
7564
1092
Reference
Laboratory0
0.314
0.342
0.369
0.313
0.127
0.128
0.132
0.123
0.19
0.181
0.203
0.182
2.59
2.73
2.5
2.53
2.7
2.81
2.48
3.15
2.28
2.17
2.23
2.38
0.253
0.318
0.974
0.263
53000
65300
80500
85100
311000
305000
210000
TEQ™ (pg/g)
Developer1"
81
90
241
59
24
79
116
79
84
9760
290
139
130
812
424
57
40
71
68
64
269
150
119
226
<13
42
37
15
16300
3590
308
11300
26600
20200
34500
Reference
Laboratory0
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
Total TEQ (pg/g)
Developer1"
93
92
245
59
26
89
117
79
237
9837
502
333
131
815
448
57
40
90
69
64
302
151
119
237
<21
42
38
15
22806
7304
23064
16517
34366
27764
35592
Reference
Laboratory011
67.51
71.94
62.07
68.11
49.63
72.93
56.13
60.52
12600.19
15200.18
13100.20
13600.18
224.59
243.73
271.50
270.53
210.70
181.81
199.48
195.15
187.28
176.17
178.23
163.38
25.95
26.72
31.97
26.06
53788.00
66400.00
81352.00
86006.00
314400.00
308300.00
213430.00
                                                               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
Sample
Number
CAPE 81
CAPE 154
CAPE 102
CAPE 41
CAPE 134
CAPE 36
CAPE 109
CAPE 137
CAPE 145
CAPE 99
CAPE 193
CAPE 204
CAPE 61
CAPE 138
CAPE 110
CAPE 66
CAPE 142
CAPE 82
CAPE 108
CAPE 175
CAPE 167
CAPE 34
CAPE 131
CAPE 55
CAPE 179
CAPE 128
CAPE 146
CAPE 73
CAPE 58
CAPE 107
CAPE 64
CAPE 162
CAPE 147
CAPE 89
CAPE 78
CAPE 123
CAPE 150
CAPE 29
Measurement
Location3
Field
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Field
Laboratory
Field
Laboratory
Laboratory
Field
Field
Laboratory
Laboratory
Field
Sample Description
NC PCB Site #2
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
Newark B ay #1
Newark B ay #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
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
Saginaw River #1
Saginaw River #1
S aginaw River #1
S aginaw River #1
Saginaw River #2
REP
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
o
J
4
1
2
3
4
1
2
o
J
4
1
2
3
4
1
2
3
4
1
2
3
4
1
TEQPrR (pg/g)
Developer1"
18250
2550
27180
3714
847
1
5
5
1
9
5
1
1
8
4
3
2
8
5
82
24
2
3
2
90
8
5
1
2
5
3
1
7
22
25
32
8
14
Reference
Laboratory0
361000
848000
618000
533000
904000
1.22
1.44
1.39
1.34
5.01
5.19
5.14
5.09
4.61
5.04
4.5
5.03
2.73
2.65
2.72
2.7
2.33
2.06
2.35
2.25
2.7
2.67
2.68
2.85
2.43
2.43
2.3
2.33
62.4
73.6
69.9
63.7
30.6
TEQ™ (pg/g)
Developer1"
24500
>8300
65000
33700
42400
31
17
26
<23
69
49
<11
75
80
46
11
45
63
22
68
36
42
95
75
44
52
36
38
47
19
56
80
<23
1470
758
421
58
1690
Reference
Laboratory0
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
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
Total TEQ (pg/g)
Developer1"
42750
>10850
92180
37414
43247
32
22
31
<24
78
54
<12
76
88
50
14
47
71
27
150
60
44
98
77
134
60
41
39
49
24
59
81
<30
1492
783
453
66
1704
Reference
Laboratory011
364490.00
856320.00
626410.00
542360.00
914800.00
24.22
15.44
15.89
14.84
55.61
52.59
79.24
55.49
43.51
49.94
44.70
46.93
36.33
28.75
30.32
29.50
12.53
12.36
12.75
13.65
16.00
15.77
15.48
15.85
12.83
13.53
12.90
12.26
1112.40
756.60
1139.90
757.70
1140.60
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
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Environmental
Sample
Number
CAPE 106
CAPE 178
CAPE 201
CAPE 68
CAPE 198
CAPE 119
CAPE 53
CAPE 153
CAPE 165
CAPE 59
CAPE 45
CAPE 177
CAPE 176
CAPE 50
CAPE 30
CAPE 77
CAPE 127
CAPE 192
CAPE 51
CAPE 174
CAPE 143
CAPE 87
CAPE 85
CAPE 163
CAPE 199
CAPE 39
CAPE 62
CAPE 161
CAPE 92
CAPE 135
CAPE 79
CAPE 182
CAPE 46
CAPE 156
CAPE 33
CAPE 63
CAPE 120
CAPE 88
Measurement
Location3
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Field
Laboratory
Laboratory
Field
Field
Field
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Field
Laboratory
Laboratory
Field
Field
Laboratory
Field
Laboratory
Field
Laboratory
Field
Laboratory
Field
Field
Laboratory
Field
Sample Description
Saginaw River #2
Saginaw River #2
Saginaw River #2
Saginaw River #3
Saginaw River #3
Saginaw River #3
Saginaw River #3
Solutia #1
Solutia #1
Solutia #1
Solutia #1
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 #1
Titta. River Sed #1
Titta. River Sed #1
Titta. River Sed #1
Titta. River Sed #2
Titta. River Sed #2
Titta. River Sed #2
REP
2
3
4
1
2
3
4
1
2
3
4
1
2
o
J
4
1
2
3
4
1
2
o
J
4
1
2
3
4
1
2
o
J
4
1
2
3
4
1
2
o
5
TEQPrR (pg/g)
Developer1"
26
189
10
6119
0
14
2
3
0
2
14
55
364
5
4
7
17
48
8
38
24
6
2
1
0
21
1
0
1
3
2
219
74
1
6
0
1
0
Reference
Laboratory0
31
26.7
29.8
0.0202
0.0164
0.0467
0.0157
0.452
0.163
0.388
0.391
17.6
18.8
19.2
18.5
29.7
36.9
37
31.5
7.32
8.26
7.57
8.37
0.986
1.2
1.03
1.06
1.26
1.16
1.54
1.33
0.0527
0.034
0.0407
0.0403
0.649
0.71
0.566
TEQ™ (pg/g)
Developer1"
1290
2650
179
241
51
122
2730
37
38
138
80
1780
2280
556
869
3790
3320
3140
1890
298
275
49
187
192
157
1300
1203
1220
>330
820
<280
60
686
16
24
123
66
258
Reference
Laboratory0
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
1
1.7
52.8
123
66.1
Total TEQ (pg/g)
Developer1"
1316
2839
189
6360
51
136
2732
40
38
140
94
1835
2644
561
873
3797
3337
3188
1898
336
299
55
189
193
157
1321
1204
1220
331
823
282
279
760
17
30
123
67
258
Reference
Laboratory011
984.00
1346.70
893.80
99.72
146.02
122.05
99.62
57.95
77.06
62.39
61.99
2107.60
1968.80
1879.20
2178.50
2839.70
2836.90
3037.00
3111.50
42.32
43.46
47.57
44.17
420.99
451.20
524.03
507.06
1051.26
677.16
1221.54
1301.33
1.10
1.14
1.04
1.74
53.45
123.71
66.67
D-3

-------
Sample Type
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
Sample
Number
CAPE 208
CAPE 43
CAPE 1 1 1
CAPE 203
CAPE 132
CAPE 65
CAPE 160
CAPE 74
CAPE 124
CAPE 159
CAPE 48
CAPE 189
CAPE 26
CAPE 32
CAPE 144
CAPE 194
CAPE 72
CAPE 3
CAPE 2
CAPE 9
CAPES
CAPE 11
CAPE 13
CAPES
CAPE 10
CAPE 17
CAPE 14
CAPE 12
CAPE 20
CAPE 7
CAPE 21
CAPE 4
CAPE 23
CAPE1
CAPE 18
CAPE 6
CAPE 22
CAPE 15
Measurement
Location3
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Field
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
Sample Description
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
REP
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
o
J
4
1
2
3
4
5
6
7
1
2
3
4
1
2
TEQPrR (pg/g)
Developer1"
0
65
2
0
1
2
1
1
10
14
23
66
12
10
73
42
2
6
6
2
9
2
28
3
3
<50
2
0
15
2
2
1
623
92
3
548
2174
42
Reference
Laboratory0
0.515
0.0719
0.0973
0.083
0.09
0.654
0.904
0.829
0.822
1.2
1.3
1.32
1.28
1.68
1.87
1.8
2.06
0.629
0.673
0.64
2.08
0.742
0.135
0.297
0.17
0.0638
0.00013
0.0001
0.0275
0.0562
0.00724
0.139
113
113
111
113
1060
1080
TEQ™ (pg/g)
Developer1"
26
184
26
26
55
1420
68
2470
>1800
42
3250
60
4690
458
550
89
1516
154
2240
134
168
31
67
33
66
<6
4
<3
19
3
8
<3
143
5
29
13
6
6
Reference
Laboratory0
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
Total TEQ (pg/g)
Developer1"
26
249
28
26
56
1422
69
2471
>1810
56
3273
126
4702
468
623
131
1518
160
2246
136
177
33
95
36
69
<56
6
<3
34
5
10
<4
766
97
32
561
2180
48
Reference
Laboratory011
94.62
13.07
11.30
12.78
13.89
7290.65
7370.90
7450.83
7160.82
9721.20
9771.30
9201.32
11301.28
10301.68
9771.87
9321.80
9872.06
175.63
444.67
176.64
441.08
56.04
53.44
53.40
53.77
0.57
0.51
0.54
0.55
0.64
0.58
0.66
204.60
204.80
200.10
213.00
1060.32
1080.35
D-4

-------
Sample Type
Extract
Extract
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
CAPE 19
CAPE 16
CAPE 126
CAPE 116
CAPE 112
CAPE 157
CAPE 205
CAPE 42
CAPE 152
CAPE 151
CAPE 101
CAPE 202
CAPE 69
CAPE 100
CAPE 171
CAPE 207
CAPE 57
CAPE 90
CAPE 86
CAPE 149
CAPE 104
CAPE 133
CAPE 103
CAPE 180
CAPE 184
CAPE 28
CAPE 125
CAPE 141
CAPE 60
CAPE 185
CAPE 168
CAPE 47
CAPE 75
CAPE 37
CAPE 67
CAPE 200
CAPE 169
CAPE 70
Measurement
Location3
Field
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Field
Field
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Laboratory
Laboratory
Field
Field
Field
Field
Laboratory
Laboratory
Field
Sample Description
Spike #3
Spike #3
Cambridge 5 183
Cambridge 5 183
Cambridge 5183
Cambridge 5183
Cambridge 5183
Cambridge 5183
Cambridge 5183
Cambridge 5 184
Cambridge 5 184
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
ERATCDD 10
REP
o
5
4
1
2
3
4
5
6
7
1
2
3
4
1
2
o
J
4
1
2
3
4
5
6
7
8
1
2
3
4
1
2
o
J
4
1
2
3
4
1
TEQPrR (pg/g)
Developer1"
3392
36
8
9
21
2
o
J
11
6
160
520
3372
874
588
258
13400
695
0
0
2
2
0
4
o
J
0
0
5
9
1
34
158
22
16
880
2634
1258
6748
1
Reference
Laboratory0
1060
990
3.81
4.33
4.2
4.24
4.25
3.86
3.53
1080
1120
1140
1160
1060
3690
3790
3800
0.0243
0.00385
0.00277
0.042
0.0229
0.0191
0.0325
0.0225
0.0254
0.00429
0.00423
0.026
10.6
11.1
10.6
9.95
1030
1030
1180
1020
0.0147
TEQ™ (pg/g)
Developer1"
6
6
22
12
<14
<13
<11
17
<13
44
117
27
280
45
36
20
161
<14
15
<50
17
46
13
<11
13
<14
14
<12
<13
<12
23
<12
<16
<12
<16
<11
<11
22
Reference
Laboratory0
0.363
0.268
4.78
4.08
4.06
3.56
3.89
5.93
3.89
187
188
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
NAe
0.053
0.127
0.204
0.507
0.105
0.0628
8.69
Total TEQ (pg/g)
Developer1"
3398
42
30
21
<35
<15
<14
28
<19
204
637
3399
1154
633
294
13420
856
14
15
<52
19
46
17
<14
13
<14
19
<21
14
<46
181
<34
32
<892
2650
<1269
<6760
23
Reference
Laboratory011
1060.36
990.27
8.59
8.41
8.26
7.80
8.14
9.79
7.42
1267.00
1308.00
1313.00
1340.00
1096.40
3722.90
3827.90
3835.50
0.12
0.08
0.24
0.35
0.14
0.07
0.24
0.09
0.18
0.15
0.17
0.27
10.64
NAe
10.65
10.08
1030.20
1030.51
1180.11
1020.06
8.70
D-5

-------
Sample Type
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
CAPE 24
CAPE 181
CAPE 183
CAPE 52
CAPE 206
CAPE 97
CAPE 140
CAPE 114
CAPE 91
CAPE 129
CAPE 187
CAPE 186
CAPE 54
CAPE 196
CAPE 93
CAPE 188
CAPE 190
CAPE 44
CAPE 118
CAPE 35
CAPE 95
CAPE 80
Measurement
Location3
Freld
Laboratory
Laboratory
Freld
Laboratory
Laboratory
Laboratory
Laboratory
Freld
Laboratory
Laboratory
Laboratory
Freld
Laboratory
Freld
Laboratory
Laboratory
Freld
Laboratory
Freld
Laboratory
Freld
Sample Description
ERATCDD 10
ERATCDD10
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
WellrngtonWMS-01
WellrngtonWMS-01
WellmgtonWMS-01
WellmgtonWMS-01
Wellington WMS -01
WellmgtonWMS-01
WellmgtonWMS-01
REP
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
TEQPrR (pg/g)
Developer1"
0
34
3
0
0
1
0
133
254
44
10
35
19
9
18
11
11
174
12
4
3
7
Reference
Laboratory0
0.0123
0.0299
0.045
0.0451
0.0153
0.0436
0.04
435
405
498
356
40.1
43.7
42.1
41
10.6
9.4
9.62
9.07
10.3
9.62
9.68
TEQ™ (pg/g)
Developer1"
8
34
16
82
<11
36
37
>350
>330
11600
9760
47
113
21
125
65
80
213
160
108
76
88
Reference
Laboratory0
9.28
8.44
8.2
27.4
25.3
24.8
23.9
NAe
6930
6900
7190
237
206
252
219
68
65.7
61.9
66.1
68
65.7
65.4
Total TEQ (pg/g)
Developer1"
8
68
19
82
<11
37
37
>483
584
11644
9770
82
132
30
143
76
91
387
172
112
79
95
Reference
Laboratory011
9.29
8.47
8.25
27.45
25.32
24.84
23.94
NAe
7335.00
7398.00
7546.00
277.10
249.70
294.10
260.00
78.60
75.10
71.52
75.17
78.30
75.32
75.08
All PCBs results were generated in the laboratory.
Data listed exactly as reported by the developer.
Qualifier flags (e.g., J and K flags) included in the raw data have been removed for the purposes of statistical analysis.
Data calculated by the developer by summing TEQPCB and TEQD/F.
Reference laboratory data was discarded due to laboratory sample preparation error.
                                                                             D-6

-------