«EPA
           United States       Office of Research and    EPA/540/R-05/001
           Environmental Protection   Development       March 2005
           Agency         Washington, DC 20460
           Innovative Technology
           Verification Report
           Technologies for Monitoring
           and Measurement of Dioxin
           and Dioxin-like Compounds
           in Soil and Sediment

           Xenobiotic Detection Systems, Inc.
           CALUX® by XDS

-------

-------
                                  EPA/540/R-05/001
                                  March 2005
   Innovative Technology
     Verification  Report

Xenobiotic Detection Systems, Inc.
         CALUX® by XDS
                 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-00-185. 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 Xenobiotic Detection Systems, Inc., CALUX® by
XDS. Four other technologies were evaluated as part of this demonstration, and separate reports have been prepared
for each technology. The objectives of the demonstration included evaluating the technology's accuracy, precision,
sensitivity, sample throughput, tendency for matrix effects, and cost.  The test also included an assessment of how
well the technology's results compared to those generated by established laboratory methods using high-resolution
mass spectrometry (HRMS). The demonstration objectives were accomplished by evaluating the results  generated by
the technology from 209 soil, sediment, and extract samples. The test samples included performance evaluation (PE)
samples (i.e., contaminant concentrations were certified or the samples were spiked with known contaminants) and
environmental samples collected from 10 different sampling locations.

The Xenobiotic Detection Systems, Inc., CALUX® by XDS is an aryl hydrocarbon-receptor bioassay that individually
reports the total toxicity equivalents (TEQ) of dioxins/furans and polychlorinated biphenyls (PCBs) in the sample. 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. When comparing the CALUX® by
XDS results with HRMS TEQ results from the certified samples and  the reference methods, the reader should keep in
mind the limitations of the TEQ approach, noting that it is possible that Ah-receptor binding compounds that are
being measured during the CALUX® by XDS analysis are not all accounted for in the reference laboratory TEQ result
and that the World Health Organization toxicity equivalency factors used to generate the reference laboratory TEQs
may differ from the assay Ah-receptor binding affinity for certain analytes. Therefore, the technology should not be
viewed as producing an equivalent measurement value to HRMS TEQ values for all samples. Since the technology
measures an actual biological response, it is possible that the technology may give a better representation of the true
toxicity from a risk assessment standpoint.

The CALUX® by XDS generally reported data higher than the certified PE and  reference laboratory values for TEQD/F
and total TEQ, but were generally lower than the certified PE and reference laboratory values for TEQPCB. The
technology's estimated method detection limit was similar to what was reported by the developer (0.53 to 0.63 pg/g
TEQD/F).  No statistically significant matrix effects were observed by matrix type (soil vs.  sediment vs. extract) or
polynuclear aromatic hydrocarbon concentration. Twenty-one percent of the CALUX® by XDS results from replicate
sample sets that were analyzed in the laboratory and in the field showed a significant statistical difference, and only
total TEQ value showed a statistically significant effect due to sample type (performance evaluation vs. environmental
vs. extract). The technology had a fairly high rate of false positive and false negative results around 1 picogram/gram
(pg/g) TEQPCB (15% and 23%, respectively), but it had significantly fewer false positives  and false negatives for total
TEQ (4% and 1%, respectively) and TEQD/F (6% and 0%, respectively). When comparing XDS's results to the
reference laboratory for samples above and below 50 pg/g TEQ, all of the false  positive and false negative rates for all
TEQ types were less than  10%. These data suggest that the XDS technology could be an effective tool to screen for
samples above or below 1  pg/g TEQ for TEQD/F and total TEQ, and that it could be effective for all three types of
TEQ values to determine results above or below 50 pg/g TEQ, particularly considering that both the cost ($89,564 vs.
$398,029) and the time (six weeks vs. eight months) to  analyze the 209 demonstration samples were significantly less
than that 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 Xenobiotic Detection Systems, Inc., CALUX® by XDS	 6
    2.1   Company History  	 6
    2.2   Product History	 7
    2.3   Technology Description 	 7
    2.4   Developer Contact Information  	 8

3   Demonstration and Environmental Site Descriptions  	 9
    3.1   Demonstration Site Description and Selection Process	 9
    3.2   Description of Sampling Locations  	 10
         3.2.1   Soil Sampling Locations 	 10
         3.2.2   Sediment Sampling Sites	 12

4   Demonstration Approach 	 14
    4.1   Demonstration Objectives  	 14
    4.2   Toxicity Equivalents	 14
    4.3   Overview of Demonstration Samples	 16
         4.3.1   PE Samples  	 16
         4.3.2   Environmental Samples	 19
         4.3.3   Extracts  	 21
    4.4   Sample Handling	 21
    4.5   Pre-Demonstration Study	 23
    4.6   Execution of Field Demonstration  	 24
    4.7   Assessment of Primary and  Secondary Objectives  	 24
         4.7.1   Primary Objective PI: Accuracy	 25
         4.7.2   Primary Objective P2: Precision	 25
         4.7.3   Primary Objective P3: Comparability	 25
         4.7.4   Primary Objective P4: Estimated Method Detection Limit 	 26
         4.7.5   Primary Objective P5: False Positive/False Negative Results  	 26

-------
                                       Contents (continued)

                                                                                                  Page

         4.7.6   Primary Objective P6: Matrix Effects	 26
         4.7.7   Primary Objective P7: Technology Costs	 27
         4.7.8   Secondary Objective SI: Skill Level of Operator	 27
         4.7.9   Secondary Objective S2: Health and Safety Aspects 	 27
         4.7.10  Secondary Objective S3: Portability	 27
         4.7.11  Secondary Objective S4: Sample Throughput  	 27

5   Confirmatory Process	 27
    5.1   Traditional Methods for Measurement of Dioxin and Dioxin-Like
         Compounds in Soil and Sediment  	 28
         5.1.1   High-Resolution Mass Spectrometry  	 28
         5.1.2   Low-Resolution Mass Spectrometry	 28
         5.1.3   PCB Methods	 29
         5.1.4   Reference Method Selection	 29
    5.2   Characterization of Environmental Samples 	 29
         5.2.1   Dioxins and Furans  	 29
         5.2.2   PCBs  	 30
         5.2.3   PAHs  	 30
    5.3   Reference Laboratory Selection  	 30
    5.4   Reference Laboratory Sample Preparation and Analytical Methods	 31
         5.4.1   Dioxin/Furan Analysis	 31
         5.4.2   PCB Analysis	 31
         5.4.3   TEQ Calculations 	 31

6   Assessment of Reference Method Data Quality  	 33
    6.1   QA Audits	 33
    6.2   QC Results  	 34
         6.2.1   Holding Times and Storage Conditions  	 34
         6.2.2   Chain of Custody	 34
         6.2.3   Standard Concentrations	 34
         6.2.4   Initial and Continuing Calibration  	 34
         6.2.5   Column Performance and Instrument Resolution  	 35
         6.2.6   Method Blanks	 35
         6.2.7   Internal Standard Recovery	 35
         6.2.8   Laboratory Control Spikes	 35
         6.2.9   Sample Batch Duplicates  	 35
    6.3   Evaluation of Primary Objective PI: Accuracy	 35
    6.4   Evaluation of Primary Objective P2: Precision  	 36
    6.5   Comparability to Characterization Data	 38
    6.6   Performance Summary	 39

7   Performance of Xenobiotic Detection Systems, Inc.,  CALUX® by XDS  	 40
    7.1   Evaluation of CALUX®byXDS Performance	 40
         7.1.1   Evaluation of Primary Objective PI: Accuracy  	 40
         7.1.2   Evaluation of Primary Objective P2: Precision  	 40
         7.1.3   Evaluation of Primary Objective P3: Comparability 	 42
         7.1.4   Evaluation of Primary Objective P4: Estimated Method Detection Limit	 43
         7.1.5   Evaluation of Primary Objective P5: False Positive/False Negative Results	 44
                                                   VI

-------
                                       Contents (continued)

                                                                                                 Page

         7.1.6   Evaluation of Primary Objective P6: Matrix Effects 	  45
         7.1.7   Evaluation of Primary Objective P7: Technology Costs  	  48
    7.2   Observer Report: Evaluation of Secondary Objectives 	  48
         7.2.1   Evaluation of Secondary Objective SI: Skill Level of Operator  	  49
         7.2.2   Evaluation of Secondary Objective S2: Health and Safety Aspects	  49
         7.2.3   Evaluation of Secondary Objective S3: Portability 	  50
         7.2.4   Evaluation of Secondary Objective S4: Throughput 	  50
         7.2.5   Miscellaneous Observer Notes 	  51

8   Economic Analysis	  52
    8.1   Issues and Assumptions  	  52
         8.1.1   Capital Equipment Cost 	  52
         8.1.2   Cost of Supplies	  52
         8.1.3   Support Equipment Cost	  53
         8.1.4   Labor Cost	  53
         8.1.5   Investigation-Derived Waste Disposal Cost	  53
         8.1.6   Costs Not Included  	  53
    8.2   CALUX®byXDS Costs	  54
         8.2.1   Capital Equipment Cost 	  54
         8.2.2   Cost of Supplies	  54
         8.2.3   Support Equipment Cost	  54
         8.2.4   Labor Cost	  57
         8.2.5   Investigation-Derived Waste Disposal Cost	  57
         8.2.6   Summary of CALUX® by XDS Costs	  57
    8.3   Reference Method Costs	  57
    8.4   Comparison of Economic Analysis Results	  58

9   Technology Performance Summary	  59

10  References	  62

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

Appendix B  Supplemental Information Provided 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

1-1    Representative dioxin, furan, and polychlorinated biphenyl structure	  3
2-1    XDS patented sample processing procedure	  8
2-2    Luminescence produced when CALUX® by XDS cells are exposed to dioxin and dioxin-like chemicals	8
2-3    XDS processing samples during the field demonstration	  8
6-1    Comparison of reference laboratory and characterization D/F data for environmental samples	38


                                                Tables

3-1    Summary of Environmental Sampling Locations  	 11
4-1    World Health Organization Toxicity Equivalency Factor Values	 15
4-2    Distribution of Samples for the Evaluation of Performance Objectives	 16
4-3    Number and Type of Samples Analyzed in the Demonstration  	 17
4-4    Summary of Performance Evaluation Samples	 18
4-5    Characterization and Homogenization Analysis Results for Environmental Samples	 22
4-6    Distribution of Extract Samples	 23
5-1    Calibration Range of HRMS Dioxin/Furan Method	 28
5-2    Calibration Range of LRMS Dioxin/Furan Method 	 28
6-1    Objective PI Accuracy - Percent Recovery	 36
6-2    Evaluation of Interferences 	 36
6-3a   Objective P2 Precision - Relative Standard Deviation	 37
6-3b   Objective P2 Precision - Relative Standard Deviation (By Sample Type)	 38
6-4    Reference Method Performance Summary - Primary Objectives	 39
7-1    Objective PI Accuracy - Percent Recovery	 41
7-2a   Objective P2 Precision - Relative Standard Deviation	 41
7-2b   Objective P2 Precision - Relative Standard Deviation (By Sample Type)	 42
7-3    Objective P3 Comparability - RPD Summary Statistics	 43
7-4    Objective P3 - Comparability Using An Interval Assessment 	 44
7-5    Objective P3 - Comparability for Blank Samples	 44
7-6    Objective P4 - Estimated Method Detection Limit	 44
7-7    Objective P5 - False Positive/False Negative Results	 45
7-8    Objective P6 - Matrix Effects Using Descriptive Statistics and ANOVA Results Comparing Replicate
       Analysis Conducted During Field Demonstration and in the Laboratory	 46
7-9    Objective P6-Matrix Effects Using RSD as a Description of Precision by Soil, Sediment, and Extract  ... 48
7-10   Objective P6 - Matrix Effects Using RSD as a Description of Precision by PAH Concentration Levels
       (Environmental Samples Only) 	 48
7-11   Objective P6 - Matrix Effects of Known Interferences Using PE Materials 	 48
8-1    Cost Summary 	 55
8-2    Reference Method Cost Summary	 58
9-1    CALUX® by XDS System Performance Summary - Primary Objectives	 60
9-2    CALUX® by XDS System Performance Summary - Secondary Objectives	 61
                                                  Vlll

-------
                          Abbreviations, Acronyms, and Symbols
Ah
ANOVA
ATSDR
CALUX
CIL
CoA
COC
CRM
DER
D/F
DIPS
DMSO
DNR
D/QAPP
ELC
EMDL
EMPC
EPA
ERA
FDA
g
GC
HPLC/GPC
HRGC
HRMS
i.d.
IDW
ITVR
kg
L
LRMS
|im
m
aryl hydrocarbon
analysis of variance
Agency for Toxic Substances and Disease Registry
Chemical-Activated LUciferase expression
Cambridge Isotope Laboratories
Certificate of Analysis
chain of custody
certified reference material
data evaluation report
dioxin/furan
Dioxin/Furan and PCB-Specific
dimethyl sulfoxide
Department of Natural Resources
demonstration and quality assurance project plan
Environmental Learning Center
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
high-resolution capillary gas chromatography
high-resolution mass spectrometry
internal diameter
investigation-derived waste
innovative technology verification report
kilogram
liter
low-resolution mass spectrometry
micrometer
meter
                                                IX

-------
                    Abbreviations, Acronyms, and Symbols (Continued)
MDEQ
MDL
mg
mL
mm
MMT
MS
NERL
ng
NIST
NOAA
ORD
PAH
PCB
PCDD/F
PCDH
PCP
PE
Pg
PHDH
ppb
ppm
ppt
psi
QA/QC
RM
RPD
RSD
SDL
SIM
SITE
SOP
SRM
TCDD
TEF
TEQ
TEQD/F
Michigan Department of Environmental Quality
method detection limit
milligram
milliliter
millimeter
Monitoring and Measurement Technology
mass spectrometry
National Exposure Research Laboratory
nanogram
National Institute for Standards and Technology
National Oceanic and Atmospheric Administration
Office of Research and Development
polynuclear aromatic hydrocarbons
polychlorinated biphenyl
polychlorinated dibenzo-p-dioxin/dibenzofuran
polychlorinated diaromatic hydrocarbon
pentachlorophenol
performance evaluation
picogram
polyhalogenated diaromatic hydrocarbon
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-p-dioxin
toxicity equivalency factor
toxicity equivalent
total toxicity equivalents of dioxins/furans
total toxicity equivalents of World Health Organization dioxin-like polychlorinated
biphenyls

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

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
X-CARB            proprietary carbon matrix developed by Xenobiotic Detection Systems, Inc.
XDS                Xenobiotic Detection Systems, Inc.
                                               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,
Sue Kaelber-Matlock, and Al Taylor 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 demon-
stration. We appreciate the support of the Dioxin SITE Demonstration Panel for their technical input to the
demonstration/quality assurance project plan.  In particular, we recognize Andy Beliveau, Nardina Turner,
Greg Rudloff, Allen Debus, Craig Smith, David Williams, Dwain Winters, Jon Josephs, Bob Mouringhan, Terry
Smith, and Joe Ferrario of the U.S. EPA. Thanks also go to EPA Region 2, EPA Region 3, EPA Region 4, EPA
Region 5, EPA Region 7, and the MDEQ for collecting and supplying environmental samples for inclusion in the
demonstration. Andy Beliveau, Allen Debus, and Nardina Turner served as EPA reviewers of this report. Michael
Jury (MDEQ), Sue Kaelber-Matlock (MDEQ), Jim Sanborn (California-EPA), and Jeffrey Archer (U.S. Food and
Drug Administration) served as additional reviewers of this report. Computer Sciences Corporation provided a
technical editing review of the report. This report was prepared for the EPA by Battelle. Special acknowledgment is
given to Amy Dindal, who was the Battelle Project Manager, and to Josh Finegold, Nicole Iroz-Elardo, Mark Misita,
Tim Pivetz, Mary Schrock, Rachel Sell, Bea Weaver, and Zack Willenberg for their contributions to the preparation
of this report.
                                                  Xll

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

This innovative technology verification report (ITVR)
describes the SITE MMT Program and the scope of this
demonstration (Chapter 1); the Xenobiotic Detection
Systems, Inc. (XDS),  CALUX® (Chemical-Activated
LUciferase expression) by XDS (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 characterization and remediation
       efforts than conventional laboratory
       technologies.

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

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

-------
            that promote the SITE Program and
            participating technologies. It also supports
            the technologies by offering technical
            assistance, training, and workshops.

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

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

•      is applicable to a variety of environmentally
       contaminated sites,

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

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

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

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

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

The second step of the technology verification process is
to plan and implement a demonstration that will generate
representative, high-quality data to assist potential users
in selecting a technology. Demonstration planning
activities include a pre-demonstration sampling and
analysis investigation that assesses existing conditions at
the proposed demonstration site or sites. The objectives
of the pre-demonstration investigation are to (1) confirm
available information on applicable physical, chemical,
and biological characteristics of contaminated  media at
the sites to justify selection of site areas for the
demonstration;  (2) provide the technology developers
with an opportunity to evaluate the areas, analyze
representative samples, and identify logistical require-
ments; (3) assess the overall logistical and quality
assurance requirements for conducting the demon-
stration; 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
          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.

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

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

1.2.1   Organization of Demonstration
The key organizations and personnel involved in the
demonstration, including the roles and responsibilities of
each, are fully described in the D/QAPP.(2) EPA/NERL
had overall responsibility for this project. The EPA
reviewed and concurred with all project deliverables
including the D/QAPP and the ITVRs, provided
oversight during the demonstration, and participated in
the Visitor's Day. Battelle served as the verification
testing organization for EPA/NERL. Battelle's
responsibilities included developing and implementing
all elements of the D/QAPP; scheduling and coordinat-
ing 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 technologies' ITVRs. 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 XDS
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 (ELC) 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 CALUX® by XDS is
presented in this ITVR. Separate ITVRs have been
published for the other four participating technologies.

-------
                                              Chapter 2
            Description of Xenobiotic Detection Systems, Inc., CALUX® by XDS
This technology description is based on information
provided by XDS 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.
CALUX® by XDS technology is based on a reporter
gene system using a genetically engineered cell line
capable of detecting all of the WHO-recognized dioxins,
furans, and  PCBs. Giving results for dioxins/furans and
PCBs separately or together, as well as being available
as a screening and/or quantitative analysis, CALUX® by
XDS is used to analyze soil, sediment, fly ash, stack gas
emissions, food, feed, blood, and water suspected of
being contaminated with dioxins/furans and PCBs.

2.1    Company History
XDS was started in 1995 by Drs. George C. Clark and
Michael S. Denison to develop biologically based
methods for analysis of toxic compounds that are
harmful to animals and humans. The primary
headquarters of the company are located in the city of
Durham, on the edge of North Carolina's Research
Triangle Park.

The CALUX® by XDS technology was first used
commercially in 1996 to test milk. Its effectiveness
became known internationally throughout the scientific
community after its much-publicized successes in the
United States. The Hiyoshi Corporation of Japan became
the first licensee of XDS technology in 2000. Years of
extensive burning of refuse that would normally go into
landfills in Japan has resulted in extensive low-level
dioxin contamination. The CALUX® by XDS
technology provided Hiyoshi a cost-effective method for
extensive screening of large areas of land.
In August of 2001, the Food and Drug Administration
(FDA) Center for Veterinary Medicine and the FDA
Office of Regulatory Affairs, Arkansas Regional
Laboratory, signed a licensing agreement to use the
CALUX® by XDS bioassay for investigation as a new
technology in the detection of dioxin-like compounds.

XDS was selected by the Belgium government in
September of 2000 to help protect the country's residents
and food supply from chemical contamination. The
Scientific Institute of Public Health of Belgium signed a
five-year licensing agreement after XDS won a
Belgium-sponsored competition that included
technology entries from six other companies. Also in
2002, BELTEST (the Belgium Government's
accreditation service) certified the XDS-patented
bioassay technology as a valid and accurate method for
screening detection of chlorinated dioxins and PCBs. As
a result of this certification, XDS's patented technology
is an accepted method throughout the European Union
for screening dioxins and PCBs in foodstuffs. In
December 2003, Prince Agri Products, Inc.  of Quincy,
Illinois,  selected and recommended XDS to its raw
material suppliers as a preferred dioxin analysis
laboratory.  Prince Agri Products, a leader in the trace
mineral  industry, manufactures and processes more trace
mineral  supplements than any other supplier for the
animal feed industry.

Currently, XDS is preparing to market an endocrine
disrupter detection bioassay. This is a cell-based
transcriptional method to evaluate the endocrine
disrupter activity of chemicals for the estrogen receptor.
XDS has termed this test method the LUMI-CELL™ ER
bioassay and has developed a standardized test
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

                                                    6

-------
procedure in a stably transfected recombinant cell line
that is sensitive, robust, and reproducible in detecting
estrogen-active chemicals.

The association of exposure to endocrine (hormone)
disrupter chemicals (EDCs) and adverse health effects in
human and wildlife populations has led to worldwide
concern.  Some of the health effects that have led to this
concern include global increases in testicular cancer,
regional declines in sperm counts, altered sex ratios in
wildlife populations, increases in the incidence of breast
cancer and endometriosis, and accelerated puberty in
females that are expected to result from exposure to
chemicals that adversely affect steroid hormone action.

The LUMI-CELL™ ER bioassay is an extremely rapid
in vitro method that can evaluate the estrogenic activity
of chemicals within two days.  The method also provides
relative activity of a chemical to the standard
beta-estradiol and provides dose response activity of the
chemical. The standardized protocol developed allows
for a very robust system with low variability and high
sensitivity. The cost of the LUMI-CELL™ ER bioassay
is a few hundred dollars per chemical, which is
substantially less than any animal base method. The
LUMI-CELL™ ER bioassay is a transcriptionally based
assay capable of testing for antagonistic responses of
EDCs, which is not possible using other binding assays.

2.2    Product History
In 1998, XDS was awarded a patent (U.S. patent number
5,854,010) for its proprietary CALUX® by XDS assay
for dioxin-like chemicals.  XDS genetically engineered
mammalian cell lines to contain the gene for luciferase,
an enzyme fireflies use to produce light.  In the patented
CALUX® by XDS process, firefly luciferase is produced
when dioxin-like chemicals are present. The amount of
light produced is directly related to the amount of
dioxin-like chemicals.  The process detects dioxin at
levels below one part per trillion, and costs 40% to 70%
less than traditional high-resolution GC/HRMS.

In April 2004, XDS was awarded a second U.S. patent
(U.S. patent number 6,720,431 B2), further improving
the CALUX® by XDS bioassay. This certification was
regarding a method for separating the polyhalogenated
diaromatic hydrocarbon (PHDH) toxicity equivalents
(TEQs) of the PCDD/F subgroup from the TEQs of the
PCB compounds and reporting these results separately.
This new method is a major step forward in toxin
detection as it allows for multiple analysis results from
one PHDH laboratory sample. This saves time and is
extremely cost efficient for both research and general
public applications.  The new process also provides a
method for eliminating compounds that are not of the
PHDH chemical group. This process provides nearly
identical savings to the first patented process.

Further development of the CALUX® by XDS
technology was  supported by Small Business Innovation
Research grants (1R43 ES08327-01 and 2R44
ES08372-02) from the National Institute of
Environmental Health  Sciences in Research Triangle
Park, North Carolina, one  of the National Institutes of
Health.

2.3    Technology Description
XDS has patented (U.S. patent number 5,854,010) a
genetically engineered cell line that contains the firefly
luciferase gene under transactivational control of the Ah
receptor. This cell line can be used for the detection and
quantification of the Ah-receptor agonists, the target
receptor of dioxins, furans, and PCBs.  The XDS term for
the in vitro assay is the CALUX® by XDS assay.  The
most widely studied compounds that activate this system
are the polychlorinated diaromatic hydrocarbons
(PCDH), such as 2,3,7,8-TCDD. Many PCDH com-
pounds are quantified relative to TCDD, since this is one
of the most potent activators of Ah-receptor mediated
gene transcription. These relative quantifications are
known as TEQs, and the results from the CALUX® by
XDS assay provide a measure of TEQs in a sample. By
using patented cleanup methods developed by XDS, it is
possible to  separate PCBs from dioxins/dibenzofurans
and to determine what portion of the total TEQ in a
sample is due to each of these classes of compounds.
XDS has termed this procedure the Dioxin/Furan and
PCB-Specific (DIPS) or DIPS-CALUX® by XDS
bioassay.

Prices start at $200 for a dioxin screening (single)
analysis and $250 for a dioxin and PCB analysis, with
analysis provided as a fee  for service at the XDS
laboratories. Field analysis is available with 96-well
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
plates being shipped to the site for analytical procedures
to be performed by trained personnel. Costs per 96-well
plates are approximately $2,400, with each plate capable
of analyzing up to 40 samples along with standard
curves and quality control standards. Rental of
equipment and proprietary software to perform the
CALUX® by XDS is also available.

 The CALUX® by XDS bioassay for dioxin-like
chemicals uses a patented sample processing procedure
(U.S. patent number 6,720,431) that allows separation of
coplanar PCBs and PCDDs/PCDFs so that estimates of
TEQ can be made for each chemical class.  This allows
reporting of TEQ estimates for chlorinated dioxins/
furans and for the PCBs. The samples are extracted
using a modification of the EPA SW-846 Method 8290
extraction method. Briefly, the dried samples are ground,
and 1-g aliquots are placed in solvent-cleaned glass vials
with polytetrafluoroethylene-lined caps. The sample is
extracted with a 20% solution of methanol in toluene
and then twice with toluene. During each extraction step,
the samples are sonicated  in  an ultrasonic water bath.
The three extracts from each sample are filtered, pooled,
and concentrated by vacuum centrifugation. The sample
extract is suspended in hexane and rapidly processed
through a patented two-column chromatographic
procedure to produce two  extracts, one containing
chlorinated dioxins/furans and one containing PCBs (see
Figure 2-1). The extracts are exchanged into dimethyl
sulfoxide (DMSO) and used to dose the genetically
engineered  cells in the CALUX® assay by XDS to
provide TEQ estimates for PCBs and PCDD/PCDFs.
Prior to dosing the cells, the  sample extracts in DMSO
are suspended in cell culture medium. This medium is
then used to expose monolayers of the H1L1 cell line
grown in 96-well culture plates (see Figure 2-2). In
addition to the samples, a  standard curve of
        Figure 2-2. Luminescence
        produced when CALUX* by XDS
        cells are exposed to dioxin and
        dioxin-like chemicals.

2,3,7,8-TCDD is assayed [250, 125, 62.5, 31.25, 15.63,
7.81, 3.91, 1.95, 0.98, 0.49, and 0.24 parts pertrillion
(ppt) TCDD]. The plates are incubated for a time to
produce optimal expression of the luciferase activity in a
humidified CO2 incubator. Following incubation, the
medium is removed and the cells are examined
microscopically for viability. The induction of luciferase
activity is quantified using the
luciferase assay kit from
Promega.
This is the developer method
that was implemented during
the field demonstration. A
photo of the technology in
operation during the
demonstration is presented in
Figure 2-3. XDS provided
supplemental information
about the performance of their
technology during the
demonstration and it is
presented in Appendix B.
Figure 2-3. XDS
processing samples
during the field
demonstration.
       Figure 2-1. XDS patented sample
       processing procedure.
2.4    Developer Contact Information
Additional information about this technology can be
obtained by contacting:

       Xenobiotic Detection Systems, Inc.
       Dr. John Gordon
       1601E. Geer Street, Suite S
       Durham, North Carolina 27704
       Telephone: (919) 688-4804
       E-mail: johngordon@dioxins.com
       Website: www.dioxins.com
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

-------
                                              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 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
Tittabawassee River. The flood plain is not heavily laden
with PCBs; however, low levels of PCBs have been
detected in some areas. Soil samples taken from areas
outside the flood plain were at typical background
levels. The source of the contamination was speculated
to be attributed to legacy contamination from chemical
manufacturing.

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

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

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

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

The demonstration was conducted at the Green Point
ELC over a 10-day period from April 26 to May 5, 2004.
All technologies were operated inside trailers equipped
with fume hoods  or inside mobile laboratories. As such,
the ambient weather conditions during the demonstration
had little impact on the operation of the technologies,

-------
since all of the work spaces were climate-controlled with
heat and air conditioning. The outdoor weather
conditions were generally cool and rainy, but the
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, pentachloro-
phenol (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 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 (DNR)-owned
property in Saginaw, which was formerly a farming area
                                                   10

-------
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
J
o
J
3
32
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 200I/early 2002.  This material was
placed into an approximate 2!/2-acre treatment cell
located on facility property. During 2002/2003, material
at the treatment cell was treated through addition of
amendments (high-ammonia fertilizer and manure) and
tilling. Final concentrations achieved in the treatment
cell averaged 26 milligram (mg)/kilogram (kg) for PCP
and from 8,000 to 10,000 for pg/g dioxin equivalents.
Samples obtained for this study from this site were
obtained from the treatment cell after these
concentrations had been achieved.

3.2.1.5 Solutia
The chemical production facility at the Solutia site in
Nitro, West Virginia, is located along the eastern bank of
the Kanawha River, in Putnam County, West Virginia.
The site has been used for chemical production since the
early 1910s. The initial production facility was
developed by the U.S. government for the production of
military munitions during the World War I era between
1918 and 1921. The facility was then purchased by a
small private chemical company, which began manu-
facturing chloride, phosphate, and phenol compounds at
the site. A major chemical manufacturer purchased the
facility in  1929 from Rubber Services Company. The
company continued to expand operations and accelerated
its growth in the 1940s. A variety of raw materials has
been used at the facility over the years, including
inorganic compounds, organic solvents, and other
organic compounds, including Agent Orange. Agent
Orange is a mixture of chemicals containing equal
amounts of two herbicides: 2,4-D (2,4
dichlorophenoxyacetic acid) and 2,4,5-T (2,4,5
trichlorophenoxyacetic acid). Manufacture of this
chemical herbicide began at the site in 1948 and ceased
in 1969. The source of the dioxin contamination in the
                                                    11

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

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

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

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

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

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

3.2.2.3  Saginaw River
Saginaw River samples were collected at six locations.
The first sampling location was in the Saginaw River
just downstream of Green Point Island. Samples were
collected near the middle of the river in about 21 feet of
water. The sample was granular with some organic
material. The estimated TEQ concentration was 100 ppt.
Another Saginaw River sample was taken upstream of
Genesee Bridge on the right side of the river. The sample
was a brown fine sand from about 15 feet of water. The
                                                   12

-------
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 (D/F) 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.
                                                    13

-------
                                               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 Ah-receptor-inducing compounds have been
characterized in the test samples, calibration of
technologies results to HRMS results on site-by-site
basis was not evaluated, etc.). However, the
demonstration  as outlined below was agreed upon by the
Dioxin SITE Demonstration Panel members to provide a
reasonable evaluation of the technologies.

4.2   Toxicity Equivalents
For risk assessment purposes, estimates of the toxicity of
samples that contain a mixture of dioxin, furan, and PCB
congeners are often expressed as TEQs. TEQs are
calculated by multiplying the concentration of each
congener with a toxicity equivalency factor (TEF),
according to the equation:
                                                    14

-------
                  TEQ = Cc * TEF

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 deter-
mined by the World Health Organization (WHO) for
mammalian species.(5) The total TEQ from dioxin and
furans (TEQD/F) in a sample is calculated by adding up all
of the TEQ values from the individual dioxin and furan
congeners. The total  TEQ contribution from PCBs
(referred to as TEQPCB) is calculated by summing up the
individual PCB TEQ  values. The total TEQ in a sample is
the  sum of the  TEQD/F and TEQPCB values. TEQ
concentrations for soils and sediments are typically
reported in pg/g, which is equivalent to ppt.

Concentrations of dioxins, furans, and PCBs, represented
as total TEQ concentration, provide a quantitative
estimate of toxicity for all congeners expressed as if the
mixture were a TEQ mass of 2,3,7,8-TCDD only. While
the TEQ concept provides a way to estimate potential
health or ecological effects, the limitations of this
approach should be understood. The WHO report noted
that the TEF indicates an order of magnitude estimate of
the toxicity of a compound relative to 2,3,7,8-TCDD.(5)
Therefore, the accuracy of the TEF factors could be
affected by differences in species, in the functional
responses elicited by the compounds, and in additive and
nonadditive effects when the congeners are present in
complex mixtures. The WHO report5 concluded,
however, that it is unlikely that a significant error would
be observed due to these differences. The larger impact to
the TEF concept is the presence of Ah-receptor binding
compounds, such as PAHs (including naphthalenes,
anthracenes, and fluorenes) and brominated and
chloro/bromo-substituted analogues of PCDD/Fs that
have not been assigned TEF values but which may
contribute to the total TEQ.  This potentially can result in
an underestimation of TEQs in environmental samples
using the TEF approach.(5)
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
555555 f
OCDD
Dioxin-like PCBs
Coplanar
3,3',4,4'-TCB (PCB 77)
3,4,4',5-TCB (PCB 81)
3,3',4,4',5-PeCB (PCB 126)
3,3',4,4',5,5'-HxCB (PCB 169)




WHO TEF

1
1

0.1
0.1
0.1

0.01

0.0001


0.0001
0.0001
0.1
0.01




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

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

0.1
0.05
0.5
0.1
0.1
0.1
0.1
0.01
0.01
0.0001


0.0001
0.0005
0.0001
0.0001
0.0005
0.0005
0.00001
0.0001
    T = Tetra, Pe = Penta, Hx = Hexa, Hp = Hepta, O = Octa, CDD = chlorinated dibenzo-/>-dioxin, CDF = chlorinated
    dibenzofuran, CB = chlorinated biphenyl
                                                    15

-------
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 utilized to
obtain PE samples for use in this demonstration, and
Table 4-4 summarizes the PE samples that were
included. PE samples consisted of three types of
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
SI : Skill level of operator
S2: Health and safety
S3: Portability
S4: Sample throughput
Sample Type Used in Evaluation
PE
PE, environmental, extracts
PE, environmental, extracts
PE, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
PE, environmental, extracts
                                                   16

-------
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
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
nondetectable or were considerably lower than the
detection capabilities of developer technologies. The PE
samples were selected based on availability and on the
correlation of the PE composition as it related to the
environmental samples that were chosen for the
demonstration (e.g., the PE sample had a similar
congener pattern to one or more of the environmental
sites).

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

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 2 hours at 121°C and 15 pounds
per square inch (psi). Analytical results indicated that the
soil had low levels of D/F and PCBs.
RM 5184 is a heavily contaminated soil sample with
relatively high levels of D/F and PCBs.  According to
the Certificate of Analysis (CoA), approximately 75 kg
of contaminated sediment were obtained from an EPA
Superfund site in Massachusetts that was known to
contain considerable contamination from PCBs and
other chemical pollutants. The sediment was sieved to
achieve uniform particle size and homogenized to within
5% using a disodium fluorescein indicator. Samples
were then sterilized three times for 2 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 LGCPromochem
Certified reference material (CRM) 529 was obtained
from LGC Promochem. The following description is
taken from the reference material report that
accompanied  CRM 529.  The soil for CRM 529 was
collected in Europe from a  site where chloro-organic and
other compounds had been in large-scale production for
several decades, but where  production had ceased more
than five years before sampling. The site had been
contaminated during long-term production of
trichlorophenoxyacetic acid. An area of sandy soil was
excavated to a depth of several meters. Several hundred
kilograms of this mixed soil were air-dried at about 15°C
for three months. After removal of stones and other
foreign matter by sieving, the remaining material was
                                                    17

-------
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
6,583
62
171
251
11
33
NS
NS
NS
11
0.046
TEQPCB
(Pg/g)
5.0
424C
10.5
941
4P
NSf
NS
NS
11
1,121
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.)
sterilized in air at 120°C for 2 hours, thoroughly mixed,
and ground in an Alpine air jet mill to a particle size of
<63 micrometers (j-im). The material was homogenized
once more in a Turbula mixer and packaged in 50-g
quantities. The final mean moisture content at the time
of bottling was found to be 1.5%. According to the CoA,
certified values are provided for five dioxin congeners,
seven furan congeners, three chlorobenzene compounds,
and three chlorophenol compounds.  No PCBs were
reported with certified values on the  CoA, so the mean
concentration determined by the reference laboratory
was used as the certified value.

4.3.1.3 Wellington
PE sample WMS-01 was obtained from TerraChem, the
U.S. distributor for Wellington, an Ontario-based
company. As described in the CoA, WMS-0 lisa
homogeneous lake sediment that was naturally
contaminated (and not fortified). The crude, untreated
sediment used to prepare WMS-01 was collected from
Lake Ontario. The sediment obtained was subsequently
air-dried; crushed to break up agglomerates; air-dried
again, and then sieved, milled, and re-sieved (100%
< 75 |im). The sediment was then subsampled into 25-g
aliquots. The demonstration samples for only the
Wellington PE samples were 25 g rather than 50 g based
on the packaging size that was 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
                                                     18

-------
epoxy-coated modified Van Veen-type grab sampler
designed to sample the sediment to a depth of
10 centimeters.  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% sandy 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, and
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
materials required for that PE sample. Typically,
additional solvent was added to this concentrate to yield
sufficient volume of solution, appropriate for the mass of
soil to be spiked. Based on a soil mass of 1,600 g, the
volume of spike concentrate was approximately 10 to
30 milliliter (mL). For each PE sample, the blank soil
matrix was weighed into a 2-liter (L) wide mouth glass
jar, the spike concentrate was distributed onto the soil,
and the soil was allowed to air-dry for 30 to 60 minutes.
The PE samples were then capped and mixed in a rotary
tumbler for 30 minutes. Each PE sample was certified as
the concentration of target analytes present in the blank
matrix, plus the amount added during manufacture,
based on volumetric and gravimetric measurements.
CoAs were provided by ERA for all six ERA-provided
PE samples. The certified values provided by ERA were
different from the commercially available certified
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.
                                                    19

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

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

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

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

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

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 indistinguishable. As
such, the soil and sediment samples were jointly referred
                                                       1 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.
                                                    20

-------
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 homoge-
nization criteria and was considered  for inclusion in the
demonstration. If either of these criteria was not met,
options for the sample included (a) discarding it and not
considering it for use in the demonstration, (b) reanalyz-
ing it to determine if the data outside the
homogenization criteria were due to  analytical issues, or
(c) rehomogenizing and reanalyzing  it. Of these options,
(a) and (b) were utilized, but (c) was not because an
adequate  number of environmental samples were
selected using criteria (a) and (b). The average D/F
concentration and RSDs for the homogenization
analyses of environmental samples are shown in
Table 4-5. The composition of two particular Saginaw
River samples was of interest for inclusion in the
demonstration because of their concentration and unique
congener pattern, but the homogenization criteria were
slightly exceeded (i.e., 28% and 34% RSD for Saginaw
River Sample #2 and Saginaw River Sample #3,
respectively). Since multiple replicates of every sample
were analyzed, those samples were included in the study
because of their unique nature, but are flagged as slightly
exceeding the homogenization criteria. A correlation of
environmental samples to PE samples, similar to that
presented in Table 4-4, is presented in Table 4-5.

4.3.3  Extracts
A summary of the extract samples is provided in
Table 4-6. The purpose of the extract samples was to
evaluate detection and measurement performance
independent of the sample extraction method. As shown
in Table 4-6, two environmental samples (both sedi-
ments) 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)
           (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 30x 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 environ-
mental 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
                                                   21

-------
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
o
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%
128
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.
the environmental and PE samples visually
indistinguishable.

The samples were randomized in two ways. First, the
order in which the filled jars were distributed was
randomized. All jars had two labels. The label on the
top of the jar was the analysis order and contained
sample numbers 1 through 209. A second label placed
on the side of the jar contained a coded identifier
including a series of 10 numbers coded to include the
                                                      22

-------
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 #1"
Spike #2a
Spike #3a
Sample Description
Soxhlet extraction in toluene; no
cleanup
Soxhlet extraction in toluene; no
cleanup
0.5pg/mL
2,3,7,8-TCDD
100 pg/mL 2,3,7,8-TCDD
1,000 pg/mL each WHO PCB
(TEQ-11)
10,000 pg/mL each WHO PCBC
(TEQ~ 1,000)
Total number of extracts
No. of Replicates per Sample
4
4
7b
4
4
23
1  Prepared in toluene.
b  Seven replicates were analyzed for EMDL evaluation.
c  This extract was spiked with only PCBs, but a low-level (approximately 0.3 pg/mL) 2,3,7,8-TCDD contamination was confirmed by the
  reference laboratory.
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. XDS elected not to
have any of this information. 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.
XDS analyzed the samples in the order received. The
extracts were the first 23 samples in the XDS analysis
order.

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 character-
ization 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 XDS 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
                                                    23

-------
included a representative of each environmental site
analyzed during the demonstration.

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

4.6    Execution of Field Demonstration
XDS arrived on-site on Sunday, April 25, and spent
several hours setting up its mobile laboratory. 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 all 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.(2) 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.

XDS received its first batch of samples by midmorning
on April 26. XDS completed analysis of 43 samples
(23 extracts and 20 soil/sediment samples) 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. XDS also
encountered  some equipment failures that were not the
fault of the developer that impeded progress. These are
described in  detail in Section 7.2.  The remaining
166 samples were completed by XDS in its laboratories.
These samples were  shipped to XDS on May 3 and
received at XDS on May 4.  The remaining 166 samples
analyzed in the XDS laboratories were reported on June
16. XDS reports that typical (nonexpedited) turn around
times for sample analyses in their laboratory is 21 to 30
days. Once the complete data set was submitted, XDS
was offered the opportunity to reanalyze any samples
before reporting final results, but it declined this offer
and elected to not re-run any of the samples.

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

XDS reported its results in TEQD/F and TEQPCB (both in
pg/g). The XDS results were compared to the certified
values and reference laboratory results for TEQD/F,
TEQPCB, and total TEQ. For the developer data, total
TEQ values were calculated by summing TEQD/F and
TEQPCB data. If one of the values was reported as a
nondetect  (i.e., "< reporting limits") or was not reported
(i.e., "NA"),  a value of zero was used. In the case where
one of the  values was reported as, "> reporting limit",
the reporting limit value was used. If both values were,
"< reporting  limits", "> reporting limits", and/or "NA", a
total TEQ  value could not be calculated. For the refer-
ence laboratory data, total TEQ values  were calculated
for all samples except for two which were excluded due
to sample preparation issues (see Section 6.4).
                                                   24

-------
4.7.1  Primary Objective PI: Accuracy
The determination of accuracy was based on agreement
with certified or spiked levels of PE samples. PE sam-
ples 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 sam-
ples to calculate percent recovery. The equation used
was:
                R= C/CflxlOO%

where C is the mean concentration value  calculated
from the technology replicate measurements (in pg/g
TEQ) and CR is the certified value (in pg/g TEQ). Non-
detects and values reported as "> (value)"  were not
included in the accuracy assessment. Mean concentration
values were determined when at least three of four repli-
cates were reported as actual values [i.e., were not re-
ported as, "< (value)" or "> (value)"]. The mean, me-
dian, 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, envi-
ronmental, and extract samples) were analyzed in at least
quadruplicate. Seven replicates of three different sam-
ples were analyzed to evaluate EMDLs.

Precision was evaluated at both low and high concentra-
tion 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:
                  r  i    »           11/2
                    I   X—\ /	   	\ 9
            SD =
where SD is the standard deviation and  C  is the average
measurement. Both are reported in pg/g TEQ.

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

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

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

                                  Technology results reported by XDS were compared to
                                  the corresponding reference laboratory results by calcu-
                                  lating a relative percent difference (RPD). The equation
                                  for RPD, reported in percent, is as follows:
                                                      K-MD)
                                             RPD =
                                                   average (M R , M D )
                                                                x 100%
x 100%
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 evalua-
tion. Because the CALUX® by XDS reported both
TEQD/F and TEQPCB values, the XDS results were com-
pared to the reference laboratory TEQD/F and TEQPCB as
well as the total TEQ values. 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 per-
formed for TEQD/F and not TEQPCB or total TEQ.

The absolute value of the difference between the refer-
ence 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 val-
ues) 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
                                                   25

-------
RPD in calculation of comparability between the XDS
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 concentra-
tion 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 con-
tained levels of D/Fs and PCBs that were below the
reporting limits of the developer technology but con-
tained 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 de-
veloper technology reporting limit, this result was con-
sidered accurately reported by the developer. The accu-
racy of the blank samples was reported in terms of %
agreement. Ideal % agreement values would be 100%.

4.7.4  Primary Objective P4: Estimated Method
       Detection Limit
The method detection limit (MDL) calculation procedure
described in the demonstration plan was 40 CFR
Part 136, Appendix B, Revision 1.11. This procedure is
based on an assumption that the replicates  are
homogeneous enough to allow proper measurement of
the analytical precision and that the concentration is in
the appropriate range for evaluation of the technology's
sensitivity. For this evaluation, XDS 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,
                        n-i,i-«>=o.99)
(SD)
where  t(n_1:1_m=A99) = Student's t-value appropriate for a
99 percent confidence level and a standard deviation
estimate with n-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 CALUX® by XDS 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 1 pg/g TEQ and above and below 50 pg/g TEQ.
As such, the samples that were reported as < 1 (or 50)
pg/g TEQ by the reference  laboratory but > 1 (or 50)
pg/g TEQ by XDS were considered false positive.
Conversely, those samples  that were reported as < 1 (or
50) pg/g TEQ by XDS, but reported as > 1 (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.
                                                    26

-------
Precision (RSD) data were summarized by soil,
sediment, and extract (matrix type); by environmental,
PE, and extract (sample type); and by PAH concentra-
tion. 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 environ-
mental 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 concen-
tration. For the environmental site evaluation, the
comparability (RPD) values from each of the
10 environmental sites were compared to see if the
developer results were more or less comparable to the
reference laboratory for a particular site. For known
interferences, the developer's reported results for PE
samples were summarized for samples where the PE
samples did not contain the target analyte (e.g., did the
developer report D/F detections  for a sample only spiked
with PCBs).

This objective also evaluated whether performance was
affected by measurement location (i.e., in-field versus
laboratory conducted measurements), although this is
not a traditional matrix effect. 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 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: Skill Level of
       Operator
Based on observations during the field demonstration,
the type of background and training required to properly
operate the CALUX® by XDS 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 CALUX® by XDS
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
XDS (e.g., a mobile laboratory) and an assessment of
whether the infrastructure was adequate (or more than
adequate) for the technology's operation. Limitations of
operating the technology in the field are also discussed.

4.7.11 Secondary Objective S4: Sample
       Throughput
Sample throughput was measured based on the observer
notes, which focused on the time-limiting steps of the
procedures, as well as the documentation of sample
custody. The number of hours XDS worked in the field
was documented using attendance log sheets where XDS
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.
                                                   27

-------
                                             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 quality assurance/quality control
(QA/QC) incorporated into the analyses, and the
reporting requirements.

5.1.1  High-Resolution Mass Spectrometry
EPA Method 1613B(3) and SW846 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 atypical 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
SW846 Method
8290
2-400 pg/g
5-1, 000 pg/g
10-2,000 pg/g
5.1.2  Low-Resolution Mass Spectrometry
SW846 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
SW846 Method 8280
1,000-20,000 pg/g
2,500-50,000 pg/g
5,000-1 00,000 pg/g
                                                   28

-------
5.1.3  PCB Methods
There are more options for analysis of dioxin-like
compounds such as PCBs. EPA Method 1668A(4) is for
low- and trace-level analysis of PCBs. It involves matrix-
specific extraction, analyte-specific cleanup, and HRGC
/HRMS analysis. This method provides very accurate
determination of the WHO-designated dioxin-like PCBs
and can be used to determine all 209 PCB congeners. Not
all PCBs are determined individually with this method
because some are determined as sets of coeluting
congeners. The calibration range for PCBs based on a
typical 10-g sample and 20-(iL final sample volume is
from 0.4 to 4,000 pg/g. PCBs also can be determined as
specific congeners by GC/LRMS or as 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 demon-
stration, and it was selected as the D/F reference method.
Reference data of equal quality needed to be generated to
determine the PCB contribution to the TEQ, since risk
assessment is often based on TEQ values that are not
class-specific. As such, the complementary HRMS
method for PCB TEQ determinations, Method 1668A,(4)
was selected as the reference method for PCBs. Total
TEQD/F concentrations were generated by Method 1613B,
and total TEQPCB concentrations were generated by
Method 1668A.  These data were summed to derive a
total TEQ value  for each sample.

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

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

Each extract was analyzed by HRGC/HRMS in the
selected ion monitoring (SIM) mode at a resolution of
10,000 or greater. A DB-5 column was used for analysis
of the seventeen 2,3,7,8-PCDD/F congeners. The
instrument was calibrated for PCDD/F at levels specified
in Method 1613B with one additional calibration
standard at concentrations equivalent to one-half the level
of Method 1613B's lowest calibration point. Using a
DBS column, 2,3,7,8-TCDF is not separated from other
non-2,3,7,8-TCDF isomers. However, since the primary
objective was to determine adequacy of homogenization
and not congener quantification, it was determined that
sufficient information on precision could be obtained
                                                   29

-------
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).

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-
FiPLC extract was concentrated and fortified with
recovery internal standards. Extracts were concentrated
to a final volume between 500 microliters 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 verifica-
tions) 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 MS. Extracts were
analyzed in the SIM mode to achieve the lowest possible
detection limits.

5.3    Reference Laboratory Selection
Based on a preliminary evaluation of performance and
credibility, 10 laboratories were contacted and were sent
a questionnaire geared toward understanding the
capabilities of the laboratories, their experience with
analyzing dioxin samples for EPA, and their ability to
meet the needs of this demonstration. Two laboratories
were selected for the next phase of the selection process
and were sent three blind audit samples. Each laboratory
went through a daylong audit that included a technical
systems audit and a quality  systems audit. At each
laboratory, the audit consisted of a short opening
conference; a full day of observation of laboratory
procedures, records, interviews with laboratory staff; and
a brief closing meeting. Auditors submitted followup
questions to each laboratory to address gaps in the
observations.
                                                   30

-------
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 meter (m), 0.25-mm internal
diameter (i.d.), 0.1-|im film thickness]. A second column,
DB-225 (30 m, 0.25-mm i.d., 0.15-|im film thickness),
was used for confirmation of 2,3,7,8-TCDF
identification.  Samples that were known to contain
extremely high levels of PCDD/F were extracted without
the addition of the surrogate standard, split, then spiked
with the isotopically labeled surrogate standard prior to
cleanup. This approach allowed extraction of the method
specified 10-g sample volume, and subsequent sufficient
dilution that high level analytes were brought within the
instrument calibrated linear range. While this approach
induces some uncertainty because the actual  recovery of
analytes from the extraction process is unknown, it was
decided by the demonstration panel that in general
analyte recovery through the extraction procedures are
known to be quite good and that the uncertainty
introduced by this approach would be less than the
uncertainty introduced by other approaches such as
extracting a significantly smaller sample size.

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

5.4.3  TEQ Calculations
For the reference laboratory data, D/F and PCB congener
concentrations were converted to TEQ and subsequently
summed to determine total TEQ, using the TEFs
established by WHO in 1998  (see Table 4-l).(5)
Detection limits were reported as sample-specific
                                                   31

-------
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 calcula-
tion 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, TEQ
values calculated by option #1 were used in comparison
with the developer technologies (as presented in
Appendix D). 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.
                                                   32

-------
                                              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 standard
operating procedures (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
instruments and equipment were maintained and
                                                    33

-------
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 D/F 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. The minimum signal-to-noise criteria for
analytes in the calibration verification solution were
always met.
                                                    34

-------
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 inter-
ferences.  As the cleanup standard is not used for
quantification of native analytes, these data were
accepted.  Any samples affected by internal standard
recoveries outside of the D/QAPP and outside of the
EPA method criteria were evaluated for possible impact
on total TEQ and for comparability with replicates
processed during the program before being accepted.

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

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

6.3    Evaluation of Primary Objective PI:
       Accuracy
Accuracy was assessed through the analysis of PE
samples consisting of certified standard reference
materials, certified spikes, and certified blanks. A
summary of reference method percent recovery (R)
values is presented in Table 6-1. The Rvalues are
presented for TEQPCB, TEQD/F, and total TEQ.  The
minimum, maximum, mean, and median R values are
presented for each set of TEQ results. The reference
method values were in best agreement with the certified
values for the TEQPCB results, with a mean R value of
96%. The mean R values for TEQD/F and total  TEQ were
125% and 94%, respectively. The mean and median R
values for the TEQPCB and total TEQ were identical. The
mean and median R values for TEQD/F were not similar
and were largely influenced by the TEQD/F recovery for
ERA Aroclor of 324%. The ERA Aroclor-certified
TEQD/F values were based on TCDD and TCDF only,
whereas the reference method TEQD/F values were based
on contributions from all 2,3,7,8-substituted D/F
analytes. The Rvalues presented in Table 6-2  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
                                                   35

-------
Table 6-1.  Objective PI Accuracy - Percent Recovery
PE Sample
ID
1
2
3
4
5
6
7
8
9
10
11
12
PE Sample
Description
Cambridge 5 183
LCG CRM-529
Wellington WMS-01
Cambridge 5 184
NIST 1944
ERATCDD 10
ERA TCDD 30
ERA PAH
ERAPCB 100
ERAPCB 10000
ERA Aroclor
ERA Blank
All Performance Evaluation Samples
% Recovery
TEQpr«
81
100
93
120
102
NA
NA
NA
96
95
82
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
81
120
96
96
TEQ™
111
106
106
106
91
79
77
NA
NA
NA
324
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
77
324
106
125
Total TEQ
94
106
105
118
93
79
77
NA
95
95
83
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
10
77
118
94
94
 NA = not applicable; insufficient data were reported to determine R or the sample was not spiked with those analytes.
Table 6-2.  Evaluation of Interferences
PE Material with Known Interference
ERA PAH
ERAPCB 100
ERAPCB 10000
ERATCDD 10
ERA TCDD 30
Mean TEQ (pg/g)
0.195(D/F+PCB)
0.073 (D/F)
0.220 (D/F)
0.025 (PCB)
0.036 (PCB)
reference method TEQs is listed in Table 6-3. D/F and
PCB TEQs were not affected by PAH as evidenced
through the analysis of ERA PAH standard reference
material. D/F and PCB TEQs were not affected by each
other as evidenced by spikes that contained only one set
of analytes having negligible influence on the TEQ of
the other analyte set.

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

A summary of the  reference method replicate RSD
values is presented in Tables 6-3a and 6-3b. The RSD
values are presented for TEQPCB, TEQD/F, and total TEQ
in Table 6-3a, and a summary by sample type is
presented in Table 6-3b, along with the minimum R
value, the maximum R value, and the mean R value for
each set of TEQ results and sample types. In terms of
sample type, the reference method had the most precise
                                                   36

-------
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 Bav #3
S aginaw River #1
S aginaw River #2
Saginaw River #3
Solutia#l
Solutia #2
Solutia #3
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #3
Titta. River Sed #1
Titta. River Sed #2
Titta. River Sed #3
WinonaPost#l
Winona Post #2
Winona Post #3
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
ERATCDD10
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
o
J
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.
                                                       37

-------
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
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.
      re
      Q
      re O
      ~" a
                                                 y = 0.8595x + 41.181
                                     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.
                                                   38

-------
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).
                                                    39

-------
                                  Chapter 7
Performance of Xenobiotic Detection  Systems, Inc., CALUX® by XDS
7.1    Evaluation of CALUX® by XDS
       Performance
The Xenobiotic Detection Systems, Inc. CALUX® by
XDS is an aryl hydrocarbon-receptor bioassay that
individually reports the TEQ of D/Fs and PCBs in the
sample.  When comparing the CALUX® by XDS results
with HRMS TEQ results from the certified samples and
the reference methods, the reader should keep in mind
the limitations of the TEQ approach described in Section
4.2. Note that it is possible that Ah-receptor binding
compounds that are being measured during the XDS
analysis are not all accounted for in the reference
laboratory TEQ result and that the 1998 WHO TEFs
used to generate the reference laboratory TEQs may
differ from the assay Ah-receptor binding affinity for
certain analytes.  Therefore, the technology should not
be viewed as producing an equivalent measurement
value to HRMS TEQ values for all samples. Since the
technology measures an actual biological response, it is
possible that the technology may give a better
representation of the true toxicity from a risk
assessment standpoint.

The following sections describe the performance of
CALUX® by XDS, 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 CALUX® by XDS percent recovery
(R) values is presented in Table 7-1. The description of
how R values were calculated is presented in Section
4.7.1. 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 CALUX® by XDS values were in best
                                          agreement with the certified values for the total TEQ
                                          results, with a mean R value of 217%. The mean R value
                                          for the TEQPCB and TEQD/F results were 548% and
                                          514%, respectively. The Rvalues presented in Table 7-1
                                          indicate that the CALUX® by XDS generally reported
                                          TEQD/F and total TEQ data that were biased high relative
                                          to the certified values of the PE samples, and TEQPCB
                                          data that were generally biased low.  Exceptions to this
                                          were the total TEQ R values for the PCB-only spiked PE
                                          samples and the Aroclor-spiked PE sample which also
                                          contained a low-level D/F spike. As shown in
                                          Appendix D, the PCB results  reported by XDS for these
                                          samples were considerably lower than the certified
                                          values, causing the total TEQ results to also be low.

                                          7.1.2  Evaluation of Primary Objective P2:
                                                 Precision
                                          A summary of the CALUX® by XDS RSD values is
                                          presented in Tables 7-2a and 7-2b. The description of
                                          how RSD values were calculated is presented in Section
                                          4.7.2. The RSD values  are presented 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 R value, the maximum R value, and the mean
                                          R value for each set of TEQ results and sample types.
                                          Low RSD values (< 20 %) indicate high precision. In
                                          terms of sample type, the CALUX® by XDS values had
                                          the most precise data for the PE TEQD/F results, with a
                                          mean RSD value of 34%. In terms of TEQ values, the
                                          CALUX® by XDS values had the most precise data for
                                          the TEQD/F values, with an overall RSD of 41%. Overall
                                          RSD values ranged from 2% to 199%.
                                       40

-------
Table 7-1.  Objective PI Accuracy - Percent Recovery
PE Sample
ID
1
2
3
4
5
6
7
8
9
10
11
12
PE Sample Description
Cambridge 5 1 83
LCG CRM-529
Wellington WMS-01
Cambridge 5 184
NIST 1944
ERATCDD 10
ERA TCDD 30
ERA PAH
ERAPCB 100
ERAPCB 10000
ERA Aroclor
ERA Blank
All Performance Evaluation Samples
% Recovery
TEQPrR
1,487
38
1,736
3
12
NA
NA
NA
NA
o
3
NA
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
6
3
1,736
25
548
TEQn;F
614
239
332
538
282
148
120
NA
NA
NA
1,842
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
8
120
1,842
307
514
Total TEQ
868
226
392
85
243
160
121
NA
45
15
17
NA
NUMBER
MIN
MAX
MEDIAN
MEAN
10
15
868
141
217
 NA = not applicable; insufficient data were reported to determine R or the sample was not spiked with those analytes
 1 Three or four replicate results were used to calculate the RSD values.
Table 7-2a. Objective P2 Precision - Relative Standard Deviation
Sample Type
Environmental
Sample ID
Brunswick #1
Brunswick #2
Brunswick #3
Midland #1
Midland #2
Midland #3
Midland #4
NC Site #1
NC Site #2
NC 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#l
Solutia #2
Solutia #3
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #3
Titta. River Sed #1
Relative Standard Deviation (% RSD)a
TEQprR
NA
NA
83
96
NA
113
NA
51
31
32
NA
NA
NA
NA
NA
NA
NA
151
146
NA
NA
165
189
84
194
46
NA
TEQn;F
82
34
52
20
28
31
23
11
62
23
32
62
37
50
18
21
23
23
32
2
24
84
56
46
16
124
85
Total TEQ
83
43
52
24
28
32
24
20
58
21
31
61
37
50
18
13
23
22
32
3
24
83
64
45
25
124
85
                                                         41

-------
Sample Type

Extracts
Performance
Evaluation
Sample ID
Brunswick #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
ERATCDD10
ERA TCDD 30
LCG CRM-529
NIST 1944
Wellington WMS-01
Relative Standard Deviation (% RSD)a
TEQprR
NA
NA
NA
NA
114
NA
64
NA
NA
NA
NA
199
162
NA
99
NA
NA
31
NA
NA
26
67
163
TEQn;F
82
57
42
42
81
76
94
9
35
18
35
24
20
31
NA
NA
76
98
31
16
3
21
23
Total TEQ
83
57
42
42
82
76
92
9
60
14
69
165
22
125
117
140
143
76
20
18
3
22
74
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 Precision - Relative Standard Deviation (By Sample Type)
Sample
Type
Env
Ex
PE
All
Relative Standard Deviation (% RSD)
TEQprR
No.
14
1
7
22
MIN
31
64
26
26
MAX
194
64
199
199

MEAN
107
64
107
105
MED
104
64
99
97
TEQ™
No.
32
5
10
47
MIN
2
9
o
3
2
MAX
124
94
98
124
MEAN
44
38
34
41
MED
35
35
23
32
Total TEQ
No.
32
5
12
49
MIN
o
J
9
o
J
3
MAX
124
92
165
165
MEAN
44
49
77
53
MED
39
60
75
42
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. In Table 7-3, the comparability of the XDS and
reference laboratory data was assessed by calculating
RPD values for TEQPCB, TEQD/F, and total TEQ is
summarized. Table 7-3 provides an overall assessment
of the RPD values that is reported by TEQ value and
sample type. The XDS values agreed best with the
reference laboratory D/F measurements for extract
samples, with a median RPD value of-8%. The
median RPD values for TEQPCB, TEQD/F, and total
TEQ were -17%, -102%, and -92%, with minimum
and maximum values around -200% and +200%,
respectively. This evaluation indicates that the XDS
results were generally higher than the reference
laboratory (as evidenced by all median values being
negative) and that the TEQPCB results were reported
most consistently with the reference laboratory results.
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, five (5%), seventeen
(9%), and nineteen (11%) of the samples, respectively,
had RPD values between positive and negative 25%.

Comparability was also assessed using the interval
approach discussed in Section 4.7.3.  The agreement
                                                   42

-------
Table 7-3. Objective P3 Comparability - RPD Summary Statistics
Sample Type
Environmental
Extract
PE
Overall
TEQPrRRPD(%)
N
71
9
25
105
MIN
-200
-190
-195
-200
MAX
175
166
200
200
MEDIAN
-91
49
115
-17
TEQn;F RPD (%)
N
127
16
37
180
MIN
-198
-136
-160
-198
MAX
196
74
-21
196
MEDIAN
-105
-8
-107
-102
TOTAL TEQ RPD (%)
N
127
12
29
168
MIN
-188
-136
-191
-191
MAX
186
98
183
186
MEDIAN
-95
-98
-85
-92
when sorting the developer and reference laboratory
results for TEQPCB, TEQD/F, and total TEQ data into four
intervals (< 50 pg/g, 50-500 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, 69% for TEQD/F, and 72% 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 31% of the
time, the XDS analysis would have resulted in a
different decision about the  sample than if it was
analyzed by the reference laboratory, based on the TEQs
determined for this demonstration and the concentrations
chosen for the interval.

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 XDS-
reported concentrations were compared with the
reference laboratory reported data for these samples in
Table 7-5. XDS reported 6 of the 8 TEQPCB values as
detections (ranging from 0.88 to 13.72 pg/g), so only
two results were reported as nondetects and agreed with
the reference laboratory results. For TEQD/F, only two of
the results were reported as detections (0.75 pg/g and
13.74 pg/g), so six of eight results agreed with the
reference laboratory's reporting of blank samples. It
should be  noted that the reference laboratory data
presented  in Table 7-5 were calculated with nondetect
values assigned a zero concentration. When applying
the TEQ calculation method of assigning nondetects
with a concentration of one-half the SDL, the reference
data increased, but the conclusions regarding agreement
with the developer data remain the same.
7.1.4  Evaluation of Primary Objective P4:
       Estimated Method Detection Limit
It should be noted that these detection limit calculations
did not strictly follow the definition as presented in the
Code of Federal Regulations (i.e., t-value with 6 degrees
of freedom). 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  non-detect values (i.e.,
assigning values that were one-half or equal to the
nondetect value). However, these calculations are
provided as EMDLs to give the reader a sense of the
detection capabilities of the technology.

The EMDL of the CALUX® by XDS was determined
using Extract Spike # 1. Seven samples were prepared in
toluene spiked with 0.5 pg/mL of 2,3,7,8-TCDD only.
Two other PE samples, Cambridge 5183 and Wellington
WMS-01, were included in the demonstration in
replicates of seven so that these samples could
potentially be used for the EMDL calculation. These
samples were not included in the EMDL evaluation,
since the D/F and PCB levels were considerably higher
than the detection capabilities of the CALUX® by XDS.
Since 2,3,7,8-TCDD was the only congener spiked in
Extract Spike #1, only an EMDL for TEQD/F could be
determined. As shown in Table 7-6, because some of the
results for the samples were nondetects, the TEQD/F
EMDL was calculated in three ways: by setting
nondetect values to zero, by setting nondetect values to
half of the reporting limit value, and by setting nondetect
values to the reporting limit value itself. For the seven
Extract Spike #1 samples, XDS reported three as
                                                    43

-------
Table 7-4. Objective P3 - Comparability Using An Interval Assessment
Agreement
Number Agree
% Agree
Number Disagree
% Disagree
TEQPCB
160
82
35
18
TEQn/F
142
69
65
31
Total TEQ
146
72
57
28
Table 7-5. Objective P3 - Comparability for Blank Samples
Rep
1
2
o
J
4
5
6
7
8
% agreement
TEQPCB
XDS
(Pg/g)
13.72
0.88
1.02
6.87
ND<1.26
1.91
ND 0.50
5.67
Ref Lab"
(Pg/g)
J0.0243b
0.00385
0.00277
J0.042
J0.0229
J0.0191
J0.0325
J0.0225
Agree?
No
No
No
No
Yes
No
Yes
No
25%
(2 of 8)
TEQD/F
XDS
(Pg/g)
ND O.45
ND 0.23
ND O.45
ND O.45
ND O.45
ND 0.45
0.75
13.74
Ref Lab"
(Pg/g)
J0.0942
J0.0728
J0.237
JO. 307
JO. 113
J0.0524
J0.211
J0.0692
Agree?
Yes
Yes
Yes
Yes
Yes
Yes
No
No
75%
(6 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.
ND = nondetect.
Table 7-6. Objective P4 - Estimated Method Detection Limit
Statistic
Degrees of Freedom
SD (pg/g TEQ™)
EMDL (pg/g TEQ™)
Extract Spike #1
Nondetect values set to zero
3
0.136
0.62
Nondetect values set to 1A value
6
0.198
0.63
Nondetect values set to
reported value
6
0.170
0.53
nondetects (<0.13 pg/g TEQ). While the number of
degrees of freedom ranged from 3 to 6 because of the
nondetect values, the EMDLs for all three calculations
were very similar (0.62 pg/g TEQ, 0.63 pg/g TEQ, and
0.53 pg/g TEQ). The detection limit reported by XDS in
the demonstration plan was 0.3 pg/g TEQ.
7.1.5  Evaluation of Primary Objective P5: False
       Positive/False Negative Results
The description of false positive/false negative
calculations is presented in Section 4.7.5. The summary
of false positive/false negative results is presented in
Table 7-7.
                                                    44

-------
Table 7-7.  Objective P5 - False Positive/False Negative Results
Rate
False Positive
False Negative
TEQPCB
lpg/g
15%
(29 of 194)
23%
(45 of 194)
50 pg/g
9%
(18 of 194)
6%
(11 of 194)
TEQD/F
lpg/g
6%
(12 of 207)
0%
(Oof 207)
50 pg/g
10%
(20 of 207)
0.5%
(1 of 207)
Total TEQ
lpg/g
4%
(8 of 207)
1%
(2 of 207)
50 pg/g
6%
(12 of 207)
0%
(Oof 207)
The technology had a fairly high rate of false positive
and false negative results around 1 pg/g TEQPCB (15%
and 23%, respectively), but it had significantly fewer
false positives and false negatives for total TEQ (4%
and 1%, respectively) and TEQD/F (6% and 0%,
respectively). When the XDS results were compared to
the reference laboratory for values around 50 pg/g
TEQ, the false positive and false negative rates for all
TEQ types were 10% or below.

These data suggest that the XDS technology could be
an effective tool to screen samples as being above or
below 1 pg/g TEQ for TEQD/F and total TEQ, and that
it could be effective for all three types of TEQ values
to determine results above or below 50 pg/g TEQ.

7.1.6  Evaluation of Primary Objective P6:
       Matrix Effects
Six types of potential matrix effects were investigated:
(1) sample analysis location (field vs. laboratory), (2)
matrix type (soil vs. sediment vs. extract), (3) PAH
concentration, (4) sample type (PE vs. environmental
vs. extract) (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: 21% statistically different
•    Matrix type: none
    Sample type: slight for total TEQ
•    PAH concentration: none
•    Environmental site: none
•    Known inteferences: 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. Three of six TEQPCB measurements,
one of 15 TEQD/F measurements, and four of 18 total
TEQ measurements showed statistically significant
location effects. The majority (50%) of the TEQPCB
values showed a significant difference, but only six
sets had data that could be evaluated.  Overall, 21% of
the samples tested showed a statistically significant
difference by sample analysis location, and of these
samples, generally XDS reported the laboratory result
more comparably to the reference laboratory result
location.  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 no significant effect on RSD for TEQPCB,
TEQD/F, or total TEQ. 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 or TEQD/F but it did show a slight effect
(p = 0.0471) for total TEQ. Based on the  compara-
bility results (RPD), XDS's results were not more or
less comparable for one particular environmental site,
suggesting that matrix effects were not dependent on
environmental sites.
                                                   45

-------
Table 7-8. Objective P6 - Matrix Effects Using Descriptive Statistics and ANOVA Results Comparing Replicate Analysis Conducted During Field
Demonstration and in the Laboratory
Sample Type
Environmental
Sample
Brunswick
#1
Midland #3
NCPCB
Site #2
Saginaw
River #1
Saginaw
River #3
Saginaw
River #4
Solutia#l
Titta. River
Soil #2
Titta. River
Sed#2
Winona
Post #3
Location
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
TEQPCB
N
0
2
1
2
1
o
5
i
3
0
1
0
0
1
1
1
o
5
i
i
i
i
Mean (SD)
(Pg/g)
NAa
19.9(24.9)
2.6
7.7 (8.6)
> 39302.4
82,944.0 (6,460.9)
117.4
8.8 (4.2)
NA
1.6
NA
NA
2.4
3.1
819.4
6.3 (3.8)
1.9
6.3
172.7
99.2
p-Value
Comparing
Field to
Laboratory
b
0.7119
--
0.0020C
--
--
--
0.0000
--
--
TEQD/F
N
1
3
1
3
1
o
3
1
3
1
o
3
1
3
1
3
1
o
3
1
o
j
1
3
Mean (SD)
(Pg/g)
678.4
852.2 (805.2)
569.0
664.4 (235.9)
> 47,853. 3
73,5906.8
(5,7349.2)
2,340.0
3,406.8 (605.7)
551.3
569.6(12.7)
83.4
39.4(15.5)
293.2
193.1 (16.4)
1668.3
1,695.8(338.2)
303.4
419.9(263.6)
15502.3
73,561.7
(42,016.4)
p-Value
Comparing
Field to
Laboratory
0.8689
0.7595
--
0.2667
0.3371
0.1333
0.0341
0.9502
0.7388
0.3540
Total TEQ
N
1
3
1
3
1
o
J
1
3
1
o
J
1
3
1
3
1
o
J
1
o
J
1
3
Mean (SD)
(Pg/g)
678.4
865.5(826.1)
571.5
669.5 (243.3)
8,7155.7
818,850.8
(5,7033.9)
2,457.4
3,415.6(607.4)
551.3
570.1 (13.1)
83.4
39.4(15.5)
295.6
194.2(15.0)
2,487.6
1,702.0(339.1)
305.3
422.0 (264.4)
15,675.0
73,594.8 (42073.6)
p-Value
Comparing
Field to
Laboratory
0.8626
0.7605
0.0080
0.3051
0.3384
0.1333
0.0279
0.1827
0.7391
0.3555
                                                                46

-------
Sample Type
PE
Sample
Cambridge
5183
Cambridge
5184
ERA
Aroclor
ERA Blank
ERA PAH
ERA TCDD
30
LCG CRM-
529
Wellington
WMS-01
Location
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
field
lab
TEQPCB
N
1
4
0
o
5
i
i
i
5
0
1
1
0
0
3
1
2
Mean (SD)
(Pg/g)
3.2
92.5 (164.7)
NA
28.3 (45.9)
1,690.2
110.7
5.7
4.9(5.5)
NA
27.3
1.9
NA
NA
163.1 (42.1)
524.5
11.2(2.8)
p-Value
Comparing
Field to
Laboratory
0.6612
--
--
0.9023
--
--
--
0.0043
TEQD/F
N
1
6
1
o
3
2
2
1
1
1
1
1
o
5
i
3
1
6
Mean (SD)
(Pg/g)
28.5
23.2(5.9)
807.5
956.1 (211.5)
168.9(2.1)
236.4 (86.7)
13.7
0.8
2.9
1.2
45.7
37.6(6.1)
1,5684.8
1,5713.6(644.9)
228.6
202.3 (50.0)
p-Value
Comparing
Field to
Laboratory
0.4423
0.6047
0.3857
--
--
0.3704
0.9727
0.6475
Total TEQ
N
1
6
1
•"»
5
2
2
1
6
1
2
1
o
5
i
3
1
6
Mean (SD)
(Pg/g)
31.7
84.8(138.0)
807.5
984.3 (229.3)
1,014.0(1,197.3)
291.8(165.0)
19.4
4.2 (5.2)
2.9
14.3(18.5)
47.5
37.6(6.1)
1,5684.8
1,5876.7 (616.4)
753.1
206.1 (45.9)
p-Value
Comparing
Field to
Laboratory
0.7360
0.5730
0.4870
0.0427
0.7050
0.2943
0.8128
0.0001
1 NA = not available; data reported as < or > (value).
b p-Value could not be determined because either the field or lab value was NA.
c Bold indicates field measurement statistically different from the laboratory measurement at the p<0.05 significance level.
                                                                                    47

-------
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 TEQPCB (%)
N
16
5
1
22
MIN
26
67
64
26
MAX
199
163
64
199
MED
97
146
64
97
MEAN
102
122
64
105
RSD for TEQD/F (%)
N
24
18
5
47
MIN
3
2
9
2
MAX
124
85
94
124
MED
31
33
35
32
MEAN
44
39
38
41
RSD for Total TEQ (%)
N
26
18
5
49
MIN
3
3
9
3
MAX
165
85
92
165
MED
43
39
60
42
MEAN
61
41
49
53
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
o
J
1
7
o
J
14
MIN
31
114
51
46
31
MAX
83
114
189
194
194
MED
32
114
146
84
104
MEAN
49
114
130
108
107
RSD for TEQD/F (%)
N
3
4
16
9
32
MIN
23
42
11
2
2
MAX
62
82
84
124
124
MED
52
79
31
42
35
MEAN
46
70
35
47
44
RSD for Total TEQ (%)
N
3
4
16
9
32
MIN
21
42
13
o
J
3
MAX
58
83
83
124
124
MED
52
79
31
42
39
MEAN
44
71
36
47
44
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, XDS
reported a mean total TEQ value of 10.5 pg/g. The
PCB-only spiked samples were reported with D/F
concentrations that were 10% of the PCB certified
concentration. XDS reported only one sample as a
slight PCB detection for the D/F-only spiked PE
samples.

7.1.7  Evaluation of Primary Objective P7:
        Technology Costs
Evaluation of this objective is fully described in
Chapter 8, Economic Analysis.
Table 7-11. Objective P6 - Matrix Effects of
Known Interferences Using PE Materials
PE Sample
ERA PAH
ERA PCB 100
ERA PCB
10,000
ERATCDD10
ERA TCDD 30
% Recovery
for Spiked
Analytes a
NAb
NA
3% (PCB)
148% (D/F)
120% (D/F)
Mean TEQ (pg/g)
Reported by XDS
for Analytes that
were not Spiked in
the PE Sample
10.5 (total)
1.1 (D/F)
125 (D/F)
5.35 (PCB)C
1.86(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.
7.2     Observer Report: Evaluation of
        Secondary Objectives
The CALUX® by XDS technology is based on a
genetically engineered cell line containing the firefly
luciferase gene under transactivational control of the
                                                 48

-------
Ah receptor. This cell line is used to detect and
quantify Ah-receptor agonists in a sample extract.
Increasing Ah receptor activity in a sample extract
will cause increasing expression of firefly luciferase,
which is detected as light emission from the activated
cells.  XDS has developed and patented proprietary
cleanup procedures to separate PCBs from PCDD/Fs
in a sample extract prior to analysis and can, there-
fore, give results for PCBs and PCDD/Fs separately or
combined.  This technology may be used to screen
samples or to provide a quantitative analysis.
Currently, samples may be  sent to XDS for analysis or
the technology may be licensed.  Steps observed
during the demonstration included transferring extract
samples, extracting soil samples, processing extracts
through cleanup columns, dosing cells, and final
read-out of results. Samples were prepared as out-
lined in the demonstration plan with the exception that
2 g of each solid sample were extracted instead of 1 g.
7.2.1   Evaluation of Secondary Objective SI:
        Skill Level of Operator
In the field demonstration, this technology was
operated solely by Dr. John Gordon. Dr. Gordon is
the research director at XDS and has a Ph.D. in
biochemical genetics with over seven years of
experience in biochemistry, cell culture, molecular
biology, and chemistry. A second person (a
nonscientist) was available to assist as necessary, but
Dr. Gordon ran the technology during the field
demonstration independently.

The developer states that good organic/analytical
laboratory skills and cell culture experience would be
useful for successful operation of the technology.
Based on observation, the extract cleanup is similar to
that used for HRMS sample preparation. Good skills
in processing  cleanup columns would be important for
accurate and precise measurements and how  rapidly
the samples could be processed. Experience  with cell
culture would also be useful.  Overall, a good
technician or  entry-level chemist could operate this
technology once trained.

Instructions are provided in the form of SOPs once
the technology has been licensed. Based on a quick
look through the SOPs available at the demonstration,
the instructions appeared to be detailed and thorough.
Comprehensive training is included with licensing the
technology and this would greatly assist the user.
This technology has several steps where attention to
detail is critical to obtain acceptable sample results.
This includes careful processing of samples through
the cleanup procedures, pipetting small volumes, and
accurately weighing out samples. All standards can
be kept at room temperature for a period of one year.
After one year, the standards should be remade to
ensure confidence. All reagents are valid for one year
and should be kept in proper storage (i.e., solvents
should be kept in a standard reinforced-metal solvent
cabinet at ambient room temperature).  Preliminary
range finding of the  sample extract uses half or less of
the extract volume, so there is plenty of extract
available to reprocess an analysis without having to
re-extract a second sample.  This technology can be
stopped at several places without adversely affecting
sample results, including after extraction, after
cleanup, and after solvent evaporation (using a
vacuum centrifuge). This technology does require a
fair amount of standard laboratory equipment such as
an ultrasonic water bath, a vacuum centrifuge, a
humidified CO2 incubator, and a luminometer that
could be difficult to  troubleshoot in the field if
problems occurred.  However, a stocked mobile lab
would make it convenient to have spare equipment
and parts available, and staff would be properly
trained in troubleshooting these instruments if any
problems were encountered while in the field. This
developer will also analyze samples for customers as a
fee for service at the developer's location.

7.2.2   Evaluation of Secondary Objective S2:
        Health and Safety Aspects
Wastes generated with this technology include vials,
spent solvent and spent sample from extraction;
disposable cleanup columns and solvents from the
cleanup steps; and test tubes, solvents, pipette tips,
and 96-well plates from the assay.  A complete
inventory of the waste generated was performed after
the demonstration for processing 43 samples by XDS,
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
XDS laboratories.

(1)  One 5-gallon  container marked "low
     concentration" containing 58 used acid silica
     columns, used X-CARB columns, 27 columns,
     400 pipette tips, bench paper, and 20 tubes.
                                                  49

-------
(2) One 5-gallon container marked "high
    concentration" containing 27 caps from soil jars,
    23 ampoules from the extract samples, 23 extract
    tubes, 46 cleanup tubes, 100 pipette tips, and
    bench paper.

(3) One 5-gallon container with bench paper and 400
    culture tubes.

(4) One 5-gallon container with 600 pipette tips,
    bench paper, and sixteen 96-well plates.

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
As observed, this technology required a fume hood
(especially for processing the cleanup columns) and
several standard bio-analytical laboratory pieces of
equipment such as an ultrasonic water bath, vacuum
centrifuge, humidified CO2 incubator, and a
luminometer. Therefore, a trailer with a fume hood
would be the minimum required for successful field
operation. During the course of the demonstration,
the developer also tested a portable airtight chamber
that could be used in lieu of a humidified CO2
incubator. The developer intends for such innovations
as the airtight chamber to enhance the field portability
of the technology. According to the developer, XDS
is working toward increased field portability and is
considering equipping its own mobile lab for
responding to field requests.

Setup for this demonstration took approximately
8 hours. This included adjusting to a last-minute
equipment change by the vendor, who supplied the
incubator to  Battelle and the developer performing
initial maintenance on the instrument essential for its
basic operation. The developer believes that having its
own mobile lab in the field would greatly reduce setup
time, perhaps 1 to 2 hours. With a well-equipped
trailer, samples could be processed as efficiently in
the field as in the laboratory.  For the demonstration,
the space constraints of the 28-foot mobile laboratory
provided to the developer, including placement of the
bench-top double-cabinet incubator on the floor, made
processing in the field more cumbersome. In
addition, stacked cleanup columns were awkward to
process in the short hood height of the mobile lab's
fume hood, but the developer made modifications to
make the process more manageable.

Differences in reported results due to measurement
location (in field vs. laboratory) are described in
Section 7.1.6.

7.2.4   Evaluation of Secondary Objective S4:
        Throughput
XDS processed 43 of the 209 demonstration samples
in the field. For the demonstration samples, XDS
analyzed the samples once (referred to as "XDS
Screen") and reported the results. For greater
accuracy, XDS recommends triplicate comprehensive
sample analyses so that the average and standard
deviation can be reported for the results. These
samples were processed by one person and were
completed in five days.  Approximately 8 hours were
lost due to startup meetings and participation in
Visitor's Day.  Another approximately 8 hours were
lost due to a blown hose on the provided CO2
incubator and an additional malfunction of the
instrument. These were failures of the incubator and
were out of the developer's control.  In view of the
condition and failures of the incubator, the developer
proceeded more cautiously with dosing the cells, so
the final incubations were not completed in batches as
large as would have been performed had the incubator
not malfunctioned.

The XDS process takes 2 days, with samples being
extracted and put through cleanup the first day,
incubated overnight, and then results read the second
day.  The developer felt that one person could process
approximately 75 samples during a two-day period
and that a two-person team could process even more.
Capacity to analyze samples is initially limited by
laboratory space and available equipment rather than
staffing.  A large number of samples can incubate
overnight and the read-out of results is relatively
quick.  The earliest results that would be available
from this technology is 36 to 48 hours.  The developer
states that for samples submitted to XDS for analysis,
the standard turnaround time is 30 days; however,
preliminary results can be available as quickly as
36 to 48 hours. Based on observation during the
demonstration, the target of 75 samples in two days
by a single person seems ambitious strictly based on
                                                  50

-------
the time to weigh  samples, extract, and clean the
extracts; however, some limitations during the
demonstration such as the space restrictions
(double-cabinet bench-top instrument placed on the
floor) and malfunctions of equipment provided to the
developer hindered the production process so that
optimal production was not observed. In a
malfunction-free environment where the operator had
conditions set up to ensure optimal production, one
person may have been able to process more samples
with greater ease.  The observer felt that in spite of the
limitations that occurred during the demonstration,
two people could accomplish 75 samples in two days.
The XDS technology is not sold as a kit but rather as a
licensed technology or as a fee for service at the XDS
laboratory.  The technology is based on using 96-well
plates. In the range finding portion of the testing,
typically 6 to  12 samples can be analyzed depending
on what is known about the sample. After this step,
40 samples, a standard curve, and quality control
samples can be analyzed per plate, with each plate
being read every half-hour.

7.2.5   Miscellaneous Observer Notes
XDS is a U.S. company.  Upon licensing the
technology, the user is supplied with a complete set of
SOPs, full training, and XDS validation of the lab.
Samples may also be  submitted to XDS for analysis.
Phone support is available for both customers who
send samples to XDS for analysis and for those
licensed to use the technology.

After being licensed, the user would be provided the
cells and an initial quantity of the  XDS-patented
X-CARB used for cleanup columns.  Periodically,
licensees would need to purchase additional X-CARB
from the developer.

Other materials and equipment that the user would
need include glass vials with polytetrafluoroethylene-
lined caps, 18-mm glass tubes, methanol, toluene, an
ultrasonic water bath, a filter, a vacuum centrifuge,
hexane, DMSO, the cell culture medium, 96-well
culture plates, 2,3,7,8-TCDD standard, a humidified
CO2 incubator, a microscope, a Promega luciferase
assay, 5-mL glass disposable pipettes, test tubes, glass
columns for the column cleanup, micro-pipettes, a
balance, a luminometer, an automated plate shaker,
and software for data reduction.  In the past, the
developer has assisted licensees with acquiring these
items.

XDS recommends the following QC with each plate:
three blanks, one recovery spike (spiked blank), one
matrix spike, one PCDD/F QC standard, one PCB QC
standard, four DMSO blanks, and one media blank.
Standard sample range-finding analysis is six
dilutions of each extract. The dilutions increase
accuracy and minimize the need to repeat analyses to
generate results within calibration. The number of
dilutions varies depending upon what is  known of the
specific sample (i.e., if the sample is considered
low-level, fewer dilutions are needed.) XDS does not
recommend a specific frequency of HRMS confirma-
tion of results (they offer European Union regulations
as a guide), but it defers this decision to  client
preference. In general, XDS stated that it would not
be as necessary to confirm very high level or very
low-level results, but that results near an action level
or threshold level for the matrix might benefit from
independent confirmation.
                                                  51

-------
                                           Chapter 8
                                       Economic Analysis
During the demonstration, the CALUX® by XDS
assay 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
PCBs 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
CALUX® by XDS assay 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 CALUX® by XDS assay (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 CALUX® by
XDS assay 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 demon-
stration: 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 CALUX® by XDS assay unless otherwise
stated.
8.1.1   Capital Equipment Cost
The capital equipment cost was the cost associated
with the purchase of the CALUX® by XDS assay.
Components of the CALUX® by XDS assay are
presented in detail in Chapters 2 and 7. XDS offers a
licensing agreement option for potential CALUX®
users. Licenses are renewable five-year agreements
and include support from XDS in the form of training
of client staff, providing laboratory equipment,
proprietary software, and laboratory validation. Price
information was obtained from a standard price list
provided by XDS.

8.1.2   Cost of Supplies
The cost of supplies was estimated based on the
supplies required to analyze all demonstration samples
using the CALUX® by XDS assay that were not
included in the capital equipment cost category.
Examples of such supplies include filters, cleanup
columns, gas cylinders, solvents, and distilled water.
The supplies that XDS used during the demonstration
fall into two general categories: consumable (or
expendable) and reusable. Examples of expendable
supplies utilized by XDS during the demonstration
include hexane, toluene, methanol, silica gel, culture
flasks, carbon dioxide cylinders, and plastic pipettes.
Examples of reusable supplies include a cell culture
incubator, low-speed centrifuge, centrifuge
concentrator, and a luminometer. It should be noted
that this type of equipment may or may not be already
owned by a potential CALUX® by XDS assay user;
however, for this economic analysis, an assumption
was made that the user does not possess these items.

The purchase price of these supplies was either
obtained from a standard price list provided by XDS,
or it was estimated based on price  quotes from
independent sources.
                                                 52

-------
XDS is the sole provider of X-CARB (an expendable
supply). Recommendations as to where to obtain all
other items can be provided by XDS.

8.1.3    Support Equipment Cost
This section details the equipment used at the
demonstration such as the mobile laboratory, fume
hood, and laptop computer required by the
technology. Costs for these items will be reported per
actual costs for the demonstration.
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.
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
COC/results form. The measurement of the time
required for XDS to complete 43 sample analyses in
the field (42 labor-hours) was estimated by the sign-in
log sheets that recorded the time the XDS operator
was on-site. Time was removed for site-specific
training activities and Visitor's Day. Additionally,
8 hours  was subtracted from the total time XDS spent
in the field to account for problems with the CO2
incubator. Time estimates were rounded to the nearest
hour.

During the demonstration, the skill level  required for
the operators to complete analyses and report results
was evaluated. As stated in Section 7.2.1, based on the
field observations, a good technician or entry-level
chemist could operate this technology once trained,
and a single operator could successfully perform the
assay. This information was corroborated by XDS.

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

8.1.5   Investigation-Derived Waste Disposal
        Cost
During the demonstration, XDS 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 PPE were disposed of in the containers. The total
cost to dispose of these wastes generated during the
Electricity.  During the demonstration, some of the
capital equipment was operated using AC power. The
costs associated with providing the power supply were
not included in the economic analysis as it is difficult
to estimate the electricity used solely by the XDS
technology. The total cost for electricity usage over
the 10-day demonstration was $288. With seven
mobile labs/trailers and miscellaneous equipment
being operated continuously during the course of the
demonstration, the cost of XDS 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 CALUX® by XDS assay 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
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
                                                  53

-------
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 (PPE)
- Waste containers
- Lab stools.

8.2     CALUX9 by XDS Costs
This section presents information on the individual
costs of capital equipment, supplies, support
equipment, labor, and IDW disposal for the CALUX®
by XDS assay as well as a summary of these costs.
Additionally,  Table 8-1 summarizes the CALUX® by
XDS costs. As described in Section 4.6, XDS
analyzed 43 samples during the field demonstration
and 166 samples in its  laboratory (total 209
demonstration samples). It is important to note that
costs estimated in this section are based on actual
costs to analyze the 43  samples during the field
demonstration. Cost estimates for analyzing the entire
set of 209  demonstration samples were then
determined based on the field demonstration costs.

This cost is based on the presumption that the
technology would be licensed and used by the user.
XDS also offers an analytical service. The cost for
XDS to analyze 209  samples is $250 per sample, for a
total of $52,250, and does not reflect the usual XDS-
provided discounts for this number of samples.

8.2.1   Capital Equipment Cost
The capital equipment cost was the cost associated
with the purchase of the technology in order to
perform sample preparation and analysis. The
CALUX® by XDS assay can be licensed from XDS
for $2,400. During the field demonstration, XDS
utilized the CALUX® by XDS assay for five days to
analyze 43 samples. Because the components of the
assay itself are consumable, XDS does not rent the
CALUX®; however, the rental of equipment to
perform the CALUX® assay is available from XDS.

8.2.2   Cost of Supplies
The supplies that XDS used during the demonstration
fall into two general categories: expendable or
reusable. Table 8-1 lists all the expendable and
reusable supplies that XDS used during the
demonstration and the 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 XDS
during the demonstration was $40,662. Supplies have
to be purchased from  a retail vendor of laboratory
supplies. Reusable items listed in Table 8-1 can be
substituted with other models that operate under the
same specifications, thereby modifying the cost of
supplies to the potential user.

8.2.3   Support Equipment Cost
XDS analyzed demonstration samples in a 24-foot
mobile lab equipped with a fume hood.  The rental
cost for the mobile lab for use during sample
extraction  and sample analysis was $2,750. The
minimum rental rate for the mobile lab was 1 month.
XDS only used the mobile laboratory for five days.
Since weekly or daily rental rates for the mobile lab
were not an option, the entire cost is reported. As
determined by the  observers, a construction trailer
with a fume hood could have been sufficient for
                                                 54

-------
Table 8-1. Cost Summary
Quantity Used
During Field
Item Demo
Capital equipment
Licensing Agreement to use CALUX® 1
Supplies
Expendable
5-3/4" Pasteur Pipettes 1
Aluminum Foil 1
Bench-top Paper, 2 rolls of 20" x 300' 1
16 x 125 mm Tubes 1
50-mL glass centrifuge Tubes 1
25-mL Drying Tubes 1
10-mm Drying Tubes 1
Glass Rods 1
Pipet Tips (P200) 1
PipetTips(PlO) 1
Pipet Tips (PI 000) 1
Scintillation Vials 1
Scintillation Vial Caps 1
Silica Gel 1
Sulfunc Acid (2. 5-L bottle) 1
Hexane (4-L bottle) 1
Toluene(4-L bottle) 1
Methanol (4-L bottle) 1
Ethyl Acetate (4-L bottle) 1
Acetone (4-L bottle) 1
Celite (500 grams) 1
Sodium Sulfate (1,000 grams) 1
CarbonMatnx (X-CARB) 1
4-mL Teflon Vial 1
13 x 100 Test Tubes 1
DMSO(lOOmL) 1
Pipet Bulbs, 2-mL Capacity (pack of 72) 1
Tridecane (25 mL) 1
Glasswool 1
9" Pasteur Pipettes 1
15-mL Plastic Centrifuge Tubes, Sterile 1
50 ml Plastic Centrifuge Tubes 1
Phosphate Buffered Saline (3,000 mL) 1
RPMI Medium (3 ,000 mL) 1
Trypsm (600 mL) 1
Pen/Strep Solution (600 mL) 1
Fetal Serum (500 mL) 1
Lysis Solution (150 mL) 1
Substrate Solution (10 mL) 1
75 centimeter2 Tissue Culture Flasks 1
96-Well Plates 1
Backing Tape 1
Ethanol 1
Latex Gloves 1
Pipet Tips, Stenle (P200) 1
unit


unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
unit
Unit Cost ($)
2,400


84
5
126
57
58
82
72
130
50
49
36
110
187
150
54
22
36
44
79
55
106
21
Proprietary
33
29
25
39
8
50
688
163
207
42
51
77
48
104
30
39
209
226
43
62
85
30
Itemized
43
Samples
2,400


84
5
126
57
58
82
72
130
50
49
36
110
187
150
54
22
36
44
79
55
106
21
Proprietary
33
29
25
39
8
50
688
163
207
42
51
77
48
104
30
39
209
226
43
62
85
30
Cost3 ($)
209
Samples
2,400


84
5
126
57
58
82
72
130
50
49
36
110
187
150
54
22
36
44
79
55
106
21
Proprietary
33
29
25
39
8
50
688
163
207
42
51
77
48
104
30
39
209
226
43
62
85
30
                                               55

-------


Item
2-mL Sterile Pipettes, Plastic (500/case)
10-mL Sterile Pipettes, Plastic (200/case)
1.0-mL Multipipettor Syringes (100/case)
10.0-mL Multipipettor Syringes
(100/case)
Sodium Hydroxide
175 centimer2 Tissue Culture Flasks
75-mL Culture Flasks
Cryogenic 2mL Tubes
CO2 Gas Cylinder
CO2 Cylinder Regulator
Reusable
Cell Culture Incubator
Centrifuge (Low-Speed, Table Top)
Microscope, Inverted
Microscope
Hemocytometer, Cell Counter
Shaker for 96-Well Plates
Balance
Centrifuge Concentrator
Sonicating Water Bath
Luminometer
Support Equipment
Mobile Laboratory
Laptop Computer
Labor
Operator
IDW Disposal0
Total Cost if performed all 209 in field
Total Cost as performed (43 samples
in field and 166 in XDS laboratory)
Total Cost if Performed by XDS in its
laboratory
Quantity Used
During Field
Demo
1 unit
1 unit
1 unit

1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit

1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit
1 unit

1 unit
1 unit

42 labor hours
1 unit







Unit Cost ($)
108
62
101

101
39
176
195
37
14
265

4,197
915
400
750
105
790
2,500
3,500
506
22,000

2,750
1,000

80b
292





Itemized
43
Samples
108
62
101

101
39
176
195
37
14
265

4,197
915
400
750
105
790
2,500
3,500
506
22,000

2,750
1,000

3,360
292
$48,064

$48,064

$10,750
Cost3 ($)
209
Samples
108
62
101

101
39
176
195
37
14
265

4,197
915
400
750
105
790
2,500
3,500
506
22,000

2,750
1,000

16,331
1,419
$62,162

$89,564

$52,250
a Itemized costs were rounded to the nearest $1.
b Labor rate for field technicians to operate technology rather than research scientists
  was $50.75 an hour, $2,132 for 43 samples and $10,360 for 209 samples.
0 Further discussion about waste generated during demonstration can be found in Chapter 7.
                                                       56

-------
operation of this technology in the field. Use of a
construction trailer with a fume hood would have
been more cost efficient, lowering the support
equipment cost by at least $ 1,000.

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, 42 labor-hours were
spent in the field to analyze 43 samples. An hourly
rate of $32.10 was used for a research scientist
performing sample extractions and sample analysis,
and a multiplication factor of 2.5 was applied to labor
costs in order to account for overhead costs.(9) Based
on this hourly rate and multiplication factor, a labor
rate of $3,360 was determined for the analysis of the
43 samples during the field demonstration. It was
estimated that the labor cost for the total 209 samples
was $16,331.

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
43 samples during the field demonstration could have
been as low as $2,132 (hourly rate of $20.30 with
2.5 multiplication factor for 42 labor-hours), and
$10,360 for all 209 demonstration samples.

8.2.5  Investigation-Derived Waste Disposal
       Cost
As discussed in Chapter 7,  XDS 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, XDS analyzed 43 samples.
The total cost to dispose of the waste generated for
these samples was $292. The cost to dispose of waste
for all 209 samples is estimated at $1,419.

8.2.6  Summary of CALUX9 by XDS Costs
The total cost for performing dioxin and PCB
analyses using the CALUX® by XDS assay in the
field for all 209 samples was $62,162. The dioxin and
PCB analyses were performed for 58 soil and
sediment PE samples, 128 soil and sediment
environmental samples, and 23 extracts. When XDS
performed multiple dilutions for a sample, these were
not included in the number of samples analyzed. The
cost to have XDS analyze the 209 samples as an
analytical service would have been $52,250. The cost
to analyze the samples as it was performed
(43 samples in the field and 166 samples in the XDS
laboratories) was $89,564.

The total cost of $62,162 for analyzing the demon-
stration samples under the CALUX® by XDS licensing
option included $2,400  for capital equipment
(licensing agreement); $40,662 for supplies; $3,750
for support equipment; $16,331 for labor; and $1,419
for IDW disposal. Of these five costs, the largest cost
was for the supplies (65% of the total cost).

8.3     Reference Method Costs
This section presents the costs associated with the
reference method used to analyze the 209 demonstra-
tion 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
D/F analysis  and EPA Method 166 8A for coplanar
PCB analysis for all soil and sediment samples for
comparison with the CALUX®. 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 D/F analysis and $885
for PCB analysis. This cost covered the preparation
and analysis of the demonstration samples, required
method QC samples, electronic data deliverable, and
the data package for each. The actual cost for the 209
demonstration analyses  was $213,580 for D/F 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
                                                 57

-------
months (171 business days). The quoted turnaround
time was three months.

8.4      Comparison of Economic Analysis
         Results
The total costs for the CALUX® by XDS ($89,564)
and the reference method ($398,029) are listed in
Tables 8-1 and 8-2, respectively. The total cost for the
CALUX® by XDS was $308,465 less than the
reference method. It should be noted that XDS
analyzed 43 samples in five days on-site during the
demonstration and completed the remaining
166 samples in its laboratory within six weeks of the
demonstration. XDS reports a typical (non-expedited)
turnaround time of 21 to 30 days for sample analyses
in their laboratory. The demonstration analyses took
slightly longer than normal due to the volume of
samples and other sample analyses already in their
queue. For comparison, the reference laboratory took
8 months to report all 209 samples.

Use of the CALUX® by XDS assay 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.

CALUX® by XDS is a method that reports both
TEQD/F and TEQPCB. The reference method reports
these TEQ values as well as concentrations for
individual congeners. Although the CALUX® by XDS
analytical results did not have the same level of detail
as the reference method analytical results (or
comparable QA/QC data), the CALUX® by XDS
assay provided D/F and coplanar PCB analytical
results on-site at significant cost and time savings
compared to the reference laboratory.
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
1 Price includes up to 30% of samples requiring additional work of some kind (dilutions or extra cleanup). Greater than that
  would require additional work with further charges associated to them ($150 to $180 per sample per procedure).
                                                  58

-------
                                           Chapter 9
                             Technology Performance Summary
The purpose of this chapter is to provide a
performance summary of the CALUX® by XDS 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.

When comparing the CALUX® by XDS results with
HRMS TEQ results from the certified samples and the
reference methods, the  reader should keep in mind the
limitations of the  TEQ approach described in
Section 4.2. Note that it is possible that Ah-receptor
binding compounds that are being measured during
the CALUX® by XDS analysis are not all accounted
for in the reference laboratory TEQ result and that the
1998 WHO TEFs used  to generate the reference
laboratory TEQs may differ from the assay
Ah-receptor binding affinity for certain analytes. The
data generated and evaluated during this demonstra-
tion showed that the XDS technology was not directly
comparable to the HRMS TEQ values in many cases.
Since the technology measures an actual biological
response, it is possible that the technology may give a
better representation of the true toxicity from a risk
assessment standpoint.  However, it showed it could
be an effective tool to screen for samples above or
below 1 pg/g TEQ for TEQD/F and total TEQ and
above or below 50 pg/g TEQ for TEQPCB, TEQD/F, and
total TEQ, particularly considering that both the cost
($89,564 vs. $398,029) and the time (six weeks vs.
eight months) to analyze the 209 demonstration
samples were significantly less than that of the
reference laboratory.
                                                59

-------
Table 9-1.  CALUX® by XDS System Performance Summary - Primary Objectives
Objective
P 1 : Accuracy
P2: Precision
P3:
Comparability
P4: Estimated
Method
Detection Limit
P5: False
Positive/False
Negative Rate
P6: Matrix
Effects
P7: Cost
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 1 pg/g
TEQ (%)
False positive rate at 50 pg/g
TEQ (%)
False negative rate at 1 pg/g
TEQ (%)
False negative rate at 50 pg/g
TEQ (%)
TEQPCB
6
25
548
22
97
105
105
-17
82
25
not determined
15%
9%
23%
6%
TEQD/F
8
307
514
47
32
41
180
-102
69
75
0.53 to 0.63
6%
10%
0%
0.5%
Total TEQ
10
141
217
49
42
53
168
-92
72
not determined
not determined
4%
6%
1%
0%
• Measurement location: 21% statistically different
• Matrix type: none
• Sample type: slight for total TEQ
• PAH concentration: none
• Environmental site: none
• Known interferences: slight
As demonstrated, total cost was $89,564 (43 samples during field demonstration: $48,064 and
166 samples analyzed by XDS in its laboratories: $41,500).
Projected if all 209 demonstration samples were analyzed in field: $62,162.
Projected if all 209 demonstration samples were analyzed in XDS laboratory: $52,250.
                                            60

-------
Table 9-2.  CALUX® by XDS System Performance Summary - Secondary Objectives
         Objective
                                  Performance
 SI: Skill level of Operator
Good skills in processing cleanup columns would be important for accurate and precise
measurements and how rapidly the samples could be processed. Experience with cell
culture would also be useful.  Overall, a good technician or entry-level chemist could
operate this technology once trained.
 S2: Health and Safety
 Aspects
Wastes generated with this technology include vials, spent solvent, and spent sample
from extraction; disposable cleanup columns and solvents from the cleanup steps; and
test tubes, solvents, pipette tips, and 96-well plates from the assay. Cost for waste
disposal of 209 samples was estimated at $1,419. A fume hood is necessary for the
operation of this technology.
 S3: Portability
In addition to a fume hood, this technology required several standard bio-analytical
laboratory pieces of equipment such as an ultrasonic water bath, vacuum centrifuge,
humidified CO2 incubator, and a luminometer.  Therefore, a trailer with a fume hood
would be the minimum required for successful field operation. XDS is working toward
increased field portability and  is considering equipping its own mobile lab for responding
to field requests.
 S4: Sample Throughput
During the field demonstration, 43 samples were processed by XDS, equating to a
sample throughput rate of 9 samples per day. This was accomplished in about five full
working days (42 labor-hours), with one person exclusively performing the work. (See
Section 7.2.4 regarding nondeveloper-related instrumentation problems and throughput
delays.) XDS reported the remaining sample 166 results that were analyzed in their
laboratories in six weeks (normal, nonexpedited turnaround times are 21 to 30 days).
                                                       61

-------
                                          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-re solution
    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
                                                62

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

-------

-------
  UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
                            Office of Research and Development
                                 Washington, DC 20460
                                  xvEPA
     SITE Monitoring and Measurement Technology Program
                           Verification Statement
     TECHNOLOGY TYPE:    Aryl Hydrocarbon Receptor Bioassay

     APPLICATION:         MEASUREMENT OF DIOXIN AND DIOXIN-LIKE
                            COMPOUNDS

     TECHNOLOGY NAME:  CALUX® by XDS

     COMPANY:            Xenobiotic Detection Systems, Inc.
     ADDRESS:             1601 E. Geer Street, Suite S
                            Durham, North Carolina 27704
     PHONE:                (919) 688-4804

     WEB SITE:             www.dioxins.com
     E-MAIL:               johngordon@dioxins.com
VERIFICATION PROGRAM DESCRIPTION

The U.S. Environmental Protection Agency (EPA) created the Superfund Innovative Technology Evaluation
(SITE) Monitoring and Measurement Technology (MMT) Program to facilitate deployment of innovative
technologies through performance verification and information dissemination. The goal of this program is to
further environmental protection by substantially accelerating the acceptance and use of improved and
cost-effective technologies. The program assists and informs those involved in designing, distributing,
permitting, and purchasing environmental technologies. This document summarizes results of a demonstration of
Chemical-Activated LUciferase expression (CALUX®) by Xenobiotic Detection Systems, (XDS) Inc.

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.
                                         A-l

-------
DEMONSTRATION DESCRIPTION

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

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

A total of 209 samples was analyzed by each technology, including a mix of performance evaluation (PE)
samples, environmentally contaminated samples, and extracts. XDS analyzed 43 of these samples during the
field demonstration and 166 samples in their 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 (HRMS) (EPA Method 1613B and EPA Method 1668A) to
compare to the CALUX® by XDS assay. The purpose of the verification statement is to provide a summary of
the demonstration and  its results; detailed information is available in Technologies for Monitoring and
Measurement of Dioxin and Dioxin-like Compounds in Soil and Sediment—Xenobiotic Detection Systems
CALUX* by XDS (EPA/540/R-05/001).

TECHNOLOGY DESCRIPTION

The technology description and operating procedure below are based on information provided by XDS. XDS
has patented (U.S. patent number 5,854,010) a genetically engineered cell line that contains the firefly luciferase
gene under transactivational control of the aryl hydrocarbon (Ah) receptor. This cell line can be used for the
detection and quantification of the Ah receptor agonists, the target receptor of dioxins, furans, and
polychlorinated biphenyls (PCBs). The XDS term for the in vitro assay  is the CALUX® by XDS assay. The most
widely studied compounds that activate this system are the polychlorinated diaromatic hydrocarbons (PCDJrl),
such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). Many PCDH compounds are quantified relative to TCDD,
since this is one of the  most potent activators of Ah-receptor mediated gene transcription. These relative
quantifications are known as toxicity equivalents (TEQs), and the results from the CALUX® by XDS assay
provide a measure of TEQs in a sample. By using patented cleanup methods developed by XDS (U.S patent
number 6,720,432 B2), it is possible to separate PCBs from dioxins/dibenzofurans and to determine what portion
                                                 A-2

-------
of the total TEQ in a sample is due to each of these classes of compounds. XDS has termed this procedure the
Dioxin/Furan and PCB-Specific (DIPS) or DIPS-CALUX® by XDS bioassay. The TEQs were reported
individually for dioxins/furans and PCBs.

VERIFICATION OF PERFORMANCE

The CALUX® by XDS technology is an Ah-receptor bioassay that individually reports total dioxin/furan TEQ
(TEQD/F) and total PCB TEQ (TEQPCB) in picogram/gram (pg/g). When comparing the CALUX® by XDS results
with HRMS TEQ results from the certified samples and the reference methods, the reader should keep in mind
the limitations of the TEQ approach, noting that it is possible that Ah-receptor binding compounds that are being
measured during the CALUX® by XDS analysis are not all accounted for in the reference laboratory TEQ result
and that the World Health Organization toxicity equivalency factors used to generate the reference laboratory
TEQs may differ from the assay Ah-receptor binding affinity for certain analytes. Therefore, the technology
should not be viewed as producing an equivalent measurement value to HRMS TEQ values for all samples.
Since the technology measures an actual biological response, it is possible that the technology may give a better
representation of the true toxicity from a risk assessment standpoint.

Accuracy: The determination of accuracy was based on the agreement of the XDS results with the certified or
spiked levels of the PE samples that were obtained from commercial sources. Accuracy was assessed by percent
recovery (R), which is the average of the replicate results from CALUX® by XDS divided by the certified or
spiked value  of the PE sample, multiplied by 100%. Ideal R values are near 100%. The overall R values were
548% (mean), 25% (median), 3% (minimum), and 1,736% (maximum) for TEQPCB values; 514% (mean), 307%
(median), 120% (minimum), and 1,842% (maximum) for TEQD/F values; and 217% (mean), 141% (median),
15% (minimum), and 868% (maximum) for total TEQ values.

Precision: 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 105% (mean), 97% (median), 26%
(minimum), and 199% (maximum) for TEQPCB; 41% (mean), 32% (median), 2% (minimum), and 124%
(maximum) for TEQD/F; and 53% (mean), 42% (median), 3% (minimum), and 165% (maximum) for total TEQ.

Comparability: The XDS results were compared to EPA Method 1613B and 1668A results. 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 -17% (median), -200% (minimum), and 200% (maximum) for TEQPCB;
-102% (median), -198% (minimum), and 196% (maximum) for TEQD/F; and -92% (median), -191% (minimum),
and 186% (maximum) for total TEQ. The XDS results were also compared to the reference laboratory results
using an interval approach to determine if the XDS 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 was 82% for TEQPCB,
69% for TEQD/F and 72% for total TEQ.

Estimated method detection limit: EMDL was calculated generally according to the procedure described in 40
CFR Part 136, Appendix B, Revision 1.11. Lower EMDL values indicate better sensitivity. The calculated
EMDLs ranged from 0.53 to 0.63 pg/g TEQD/F, depending on whether nondetect values were assigned values of
zero, one-half the reporting limit value, or the reporting limit value itself. The detection limit reported by XDS in
the demonstration plan was 0.3 pg/g TEQD/F.
                                                A-3

-------
False positive/negative results: Samples that were reported as less than a specified level by the reference
laboratory but greater than the specified level by XDS were considered false positive. Conversely, those samples
that were reported as less than the specified level by XDS but reported as greater than the specified level by the
reference laboratory were considered false negatives. Ideal false positive and false negative rates were zero. The
technology had a fairly high rate of false positive and false negative results around 1 pg/g TEQPCB (15% and
23%, respectively), but it had significantly fewer false positives and false negatives for total TEQ (4% and 1%,
respectively) and TEQD/F (6% and 0%, respectively). When comparing XDS's results to the reference laboratory
for samples above and below 50 pg/g TEQ, all of the false positive and false negative rates for all TEQ types
were less than 10%. These data suggest that the XDS technology could be an effective tool to screen samples
above or below 1 pg/g TEQ for TEQD/F and total TEQ, and that it could be effective for all three types of TEQ
values to determine results above or below 50 pg/g TEQ.

Matrix effects: The likelihood of matrix-dependent effects on performance was investigated by evaluating
results in a variety of ways. The XDS results that were generated in the laboratory and in the field for replicate
samples were statistically different for 21% of the samples, and, in these cases, the XDS laboratory result was
generally more comparable to the reference laboratory. No significant effect was observed for the reproducibility
of XDS results by matrix type (soil, sediment, and extract) or by PAH concentration. A slight effect was
observed for total TEQ when comparing XDS's results by sample type (PE vs. environmental vs. extract), but
TEQD/F and TEQPCB showed no statistical difference. PE samples spiked for a particular contaminant (e.g., PCBs)
were sometimes reported as detections for other analytes that were not spiked in the sample (e.g., D/Fs). The
XDS results were not more or less comparable to the reference laboratory results based on environmental site.

Cost: The cost of the technology was documented and compared to the cost of the reference analyses. As
demonstrated, the total cost  for the CALUX® by XDS to analyze all 209 samples was $89,564. 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 CALUX® by XDS was $308,465 less than the reference method.

Skills and training required: Based on observation during the field demonstration, good skills in processing
cleanup columns would be important for accurate and precise measurements and how rapidly the samples could
be processed. Experience with cell culture would also be useful.  Overall, a good technician or entry-level
chemist could operate this technology once trained.

Health and safety aspects:  Wastes generated with this technology include vials, spent solvent, and spent sample
from extraction; disposable cleanup columns and solvents from the cleanup steps; and test tubes, solvents,
pipette tips, and 96-well plates from the assay.  A fume hood is necessary for the operation of this technology.

Portability: In addition to a fume hood, this technology required several standard bioanalytical laboratory pieces
of equipment such as an ultrasonic water bath, vacuum centrifuge, humidified CO2 incubator, and luminometer.
Therefore, a trailer with a fume hood would be the minimum required for successful field operation.

Sample throughput: During the field demonstration, 43 samples were processed by XDS, equating to a sample
throughput rate of 9 samples per day. This was accomplished in about 5 full working days (42 labor-hours), with
one person exclusively performing the work. (See Section 7.2.4 regarding nondeveloper-related instrumentation
problems and throughput delays.) XDS completed the remaining 166 samples in their laboratory within 6 weeks
of the demonstration. (Note  that typical, non-expedited turnaround times are 21 to 30 days in the XDS
laboratory.) For comparison, the reference laboratory took 8 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.

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

-------
               Xenobiotic Detection Systems Comments on EPA Site Program Data

Xenobiotic Detection Systems is pleased to have been invited and to have participated in this study. The EPA
and Battelle have run a rigorous cross validation study comparing our XDS CALUX screening estimates of
dioxin-like chemicals to high resolution gas chromatography/mass spectrometry analysis of chlorinated
dioxins/furans and PCBs.  This was a highly complex project demanding analytical precision over 7 logs of
concentration for these toxic chemicals, and was an exceedingly difficult accomplishment for any analytical
procedure.

XDS is proud of the characterization of the XDS CALUX technology provided in this report. It clearly
illustrates the value of our technology in examining toxicity issues in soil and sediment samples. The study also
demonstrates the technology is applicable to other matrices and situations.

       "These data suggest that the XDS technology could be an effective tool to screen samples as being above
       or below 1 pg/g TEQ or TEQ D/F and total TEQ, and that it could be effective for all three types of TEQ
       values to determine results above or below 50 pg/g."

       "The total cost for the CALUX by XDS to analyze all 209 samples was $89,564.  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 CALUX by XDS was $308,465 less than the reference method.

       "During the field demonstration, 43 samples were processed by XDS... .in about 5 full working days.
       XDS completed the remaining 166 samples in their laboratory within 6 weeks of the demonstration.  For
       comparison, the reference laboratory took 8 months to report all 209 samples."

XDS CALUX was designed to be a screening tool to evaluate contamination by these chemicals. The bioassay
was able to provide estimates of contamination particularly at the action levels of regulatory agencies for soil
samples and at significant cost savings.  The method detection limit was  determined to be between 0.53 pg to
0.63 pg TEQ/g sample and provides sufficient sensitivity for screening samples at 1 pg/g TEQ and 50 pg/g TEQ
for dioxins/furans yielding approximately 6 % false positives and 0% false negatives.

The screening mode of the XDS CALUX analysis used in this study entails extraction of the sample once,
processing with a single determination at a variety of dilutions of the extract to provide  a crude estimate of the
concentration of dioxin-like activity. This screening mode for the XDS-CALUX assay is not appropriate for
defining precision and accuracy with any confidence. However, our quantitative mode  of analysis is more
appropriate to provide this level of sample detail in an environment of high confidence and at detection levels
below one part per trillion.

Due to fiscal, time constraints, and the need to demonstrate the technology in the field in this study, our
participation was limited to providing the screening analyses of these EPA-Battelle samples.  A more appropriate
XDS CALUX analysis for defining precision and accuracy is provided by our quantitative mode of the assay
using triplicate analysis (preparing three individual extracts from a single sample and analyzing these
independently).  This comprehensive triplicate analysis allows for the determination of the relative standard
deviation (RSD) of the analyses and this figure is generally less than 20%.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

                                                 B-l

-------
Below is a chart (Figure 1) demonstrating the relationship between results on chlorinated dioxins/furan TEQ
provided by XDS CALUX analyses for the EPA-Battelle samples verses the reference GC/MS laboratory results
using a log-log plot.
        100000
         10000
          1000
    0)
    a:
    a O   100
    g£
    §§
    5  a    10
    "5
    o
            0.1

           0.01
y = 0.5641X08389
  R2 = 0.8448
               0.1
                  10
100
1000      10000    100000   1000000
                                         XDS CALUX D/F Results
                                               (pg/g TEQ)
                                                Figure 1
The data correlate well (R2= 0.8448).
These screening data points do not demonstrate a strict one to one correspondence.  This is expected since the
TEQ estimates by XDS CALUX receptor based technology provides an estimate of activation of the receptor by
the chemical extract. Many biological responses are logarithmically related to concentration.  The plotting of
this relationship generates one model that describes the relationship.

The deviation from a direct relationship occurs for a number of reasons including such factors as presence of
other halogenated dioxins and furans, differences in the REP values of XDS cells versus the TEF values used to
scale the GC/MS estimates of TEQ, and the kinetics of binding and activation of the receptor. Modeling the data
we can derive a formula to transform the CALUX data to provide a better estimate of the GC/MS data.

The formula for Figure 2 is y = 0.8389x - 0.2467, where  y = log GC/MS and x = log CALUX.
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                 B-2

-------
    3
    (A
    (1)
   o:
   LL ___
   S o
    CD Q.
   i£- • — •
    a>
   o
   o
5 -

4

3 -

2 -

1 -
                                          y=0.8389x-0.2487
                                              R2 = 0.8448
                                           XDS CALUX D/F Results
                                                 (pg/gTEQ)
                                                 Figure 2
Determining the appropriate model would improve the comparability of the GC/MS and XDS-CALUX estimates
of dioxin/furan and Total TEQ concentrations. The XDS CALUX method does underestimate the concentration
of PCBs. This is due to the differences in the relative response factors for these chemicals and the WHO TEF
values used to scale the GC/MS data.

A point to be noted in the execution of this study is that many of the developers requested or required
information about the contaminant levels of the provided samples.  Xenobiotic Detection Systems chose not to
accept any prior information on any of the sample materials and preferred to keep the study conditions
completely double blind and much closer to a real-world analysis scenario.

Savings

This report cited the large difference in the cost of the XDS CALUX analyses and the reference laboratory
analyses. Our clients are already aware of this and many use the XDS technology to screen their samples and
reduce their need for the more expensive GC/MS congener-specific analyses. The table (Figure 3) below
illustrates the savings of using XDS CALUX in conjunction with a GC/MS follow-up confirmation analysis.
This table uses the XDS CALUX costs ($89,564) for the analyses of the 209 SITE samples and the actual cost of
the  SITE reference laboratory analyses ($398,029).
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                 B-3

-------
209 Samples
No follow-up required
1% GC/MS Follow-up
2% GC/MS Follow-up
5% GC/MS Follow-up
10 % GC/MS Follow-up
20 % GC/MS Follow-up
30 % GC/MS Follow-up
40 % GC/MS Follow-up
50 % GC/MS Follow-up
60 % GC/MS Follow-up
70 % GC/MS Follow-up

100% GC/MS
CALUX by XDS
Screening Analysis
$89,564
$89,564
$89,564
$89,564
$89,564
$89,564
$89,564
$89,564
$89,564
$89,565
$89,566

No CALUX Analyses
GC/MS
Analys is
$0
$5,712
$9,520
$20,944
$39,984
$79,968
$119,952
$159,936
$199,920
$239,904
$279,888

$398,029
Screening plus
GC/MS
$89,564
$95,276
$99,084
$110,508
$129,548
$169,532
$209,516
$249,500
$289,484
$329,469
$369,454


Savings vs.
100% GC/MS
$308,372
$302,660
$298,852
$287,428
$268,388
$228,404
$188,420
$148,436
$108,452
$68,467
$28,482


                                              Figure 3
XDS Clients
Xenobiotic Detection Systems holds and regards the names of clients as highly confidential information.  We do
not release client names or provide any client information, without the client's consent.  We observe this same
confidentiality policy in regard to our CALUX by XDS licensees.
XDS has clients in the animal feed industry, environmental consulting and engineering companies, food
producers, leading colleges and universities, municipalities, incineration plants, manufacturing industries and
other categories.  Additional information, including research abstracts, is available on the XDS web site,
www.dioxins.com.

If you are interested in contacting current XDS CALUX clients or licensees, please contact Xenobiotic Detection
Systems.

We welcome opportunities to further explain and answer questions on our CALUX by XDS technology.  Please
contact us at info@dioxins.com or 1-888-DIOXINS (346-9467).
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.
                                                 B-4

-------
Summary:

Advantages of using CALUX by XDS:
209 SITE samples
Cost
Time
CALUX by XDS
$ 89,564
6 weeks*
Reference Laboratory
$398,029
8 months
*If necessary, the SITE samples could have been analyzed within a three- to four-week period. Our normal (non-expedited)
turn around time for analyses is 21 to 30 days.
Faster than GC/MS
        Results available in hours and days verses weeks

Less expensive than GC/MS
        Costs are in hundreds of dollars verses one thousand dollars or more

Flexible - as screening detection levels and threshold action levels can be client specific

Sensitive - detecting Dioxin/Furans below 1 ppt

Accurate - results are reproducible

Multiple samples can be processed during the same analysis procedure.

Already accepted in the European Union as a screening tool for foodstuffs

Can be rapidly set up in remote mobile facilities with minimal construction

Requires standard laboratory equipment, not excessive expensive instrumentation

Minimal laboratory staff required
             Information was provided by the developer and does not necessarily reflect the opinion of the EPA.

                                                   B-5

-------

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

-------

-------
Table C-l. Summary of Method Blank Performance

Sample
Batch
Number
D/F

WG12107
D/F
WG12148


D/F
WG 12264












D/F
WG12534
D/F
WG 12641






D/F
WG12737
D/F
WG12804
D/F
WG13547




Criteria
Met

Y


N








N







N




N




N

N


N


Method
Blank
TEQa
(Pg/g)

0 000812


0.133








0.0437







0.610




0.0475




0.348

0.0153


0.0553





Sample TEQ Range3 (pg/g)
26. 1-74. 1 (Newark Bay)

9.93-13.3(RantanBay)
13.5-50.4 (Newark Bay)
49.5-15,200 (Brunswick)


1.0-94.1 (Titta. River
sediment)
0.237-6,900 (PE)











25. 3-7,100 (PE)

3 1-269 (Midland)
72.8 (Brunswick)
123 (Titta. River sediment)
0.1 59-7,690 (PE)




25. 7-1 92 (Midland)
35.2- 1,300 (Titta. River soil)
3. 89-1 88 (PE)

57.5-3,000(Nitro)
37.9 (North Carolina)
122 (Saginaw River)
26.4-222 (Midland)



Comments



Many samples had concentrations >20x
blank. Few that didn't were not
significantly affected on a total TEQ
basis.
Most samples had concentrations >20x
blank. Low level Tittabawassee River
sediment samples L6749-2 (Ref 48b), -9
(Ref 130), -10 (Ref 183), and -12
(Ref 207) were evaluated based on their
replication within the demonstration
analyses and comparison to
characterization results and considered
unaffected by method blank
exceedances. Low level PE samples
L6760-1 (Ref 25), -3 (Ref 28), and -4
(Ref 29) were D/F blanks with resulting
TEQs sufficiently low enough to still be
distinguished as blank samples.
Sample concentrations > 20x blank.

All but PE sample Ref 177 (0. 159 TEQ)
had significantly higher total TEQ than
blank. Ref 177 was confirmed by
running in another batch and results,
which agreed within 18%. Additionally,
Ref 177 was compared to its replicates
within the program and considered
acceptable.
Sample concentrations >20x blank.

A few analytes higher than criteria but
no significant contribution to total TEQ.
Several analytes exceeded criteria, but
blank total TEQ contribution to sample is
relatively small.

                                              C-l

-------

Sample
Batch
Number
D/F
WG13548























D/F
WG13549

D/F
WG13551

D/F

WG13552
D/F
WG13984











Criteria
Met












N













N


N


Y






N





Method
Blank
TEQa
(Pg/g)












0.0114













0.0925


2.40


0.000969






0.0154








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)
1,1 00-1 0,800 (North
Carolina)
7,160-11,300 (Winona Post)
0.0386-9.28 (PE)

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











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



Blank contribution to total TEQ was
negligible except for PE samples L7179-
7 (Ref 94), -8 (Ref 96), -1 1 (Ref 108), -
12 (Ref 109), -17 (Ref 132), andL7182-
6 (Ref 150). 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.
C-2

-------
Sample
Batch
Number
D/F
WG14274
PCB
WG12108
PCB
WG12147
PCB
WG12265
PCB
WG12457
PCB
WG12687
PCB
WG12834
PCB
WG12835
PCB
WG12836
PCB
WG13008
PCB
WG13256
PCB
WG13257
Criteria
Met
N
N
Y
Y
N
N
N
N

N
N
Y
Y
Method
Blank
TEQa
(Pg/g)
0.0434
0.000137
0.000
0.0000584
0.000208
0.0183
0.000405
0.000125

0.0499
0.0221
0.000102
0.000251
Sample TEQ Range3 (pg/g)
2,800 (Nitro)
35.5-8,320 (North Carolina)
0.0530-5.93 (PE)
2.63-5. 19 (Newark Bay)
2.04-2.82 (RaritanBay)
1.21-5. 06 (Newark Bay)
0.104-0.330 (Brunswick)
0.132-0.369 (Brunswick)
0.034-0.649 (Titta. River
sediment)
0.00277-1, 030 (PE)
4.20-1, 020 (PE)
0.974-2.73 (Midland)
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)
1060-904,000 (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)
Comments
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 184). 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.
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 1 56 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.


C-3

-------
Sample
Batch
Number
PCB
WG13258
PCB
WG13554
PCB
WG14109
Criteria
Met
Y
N
N
Method
Blank
TEQa
(Pg/g)
0.000301
0.0000900
0.000288
Sample TEQ Range3 (pg/g)
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 (Wmona Post)
0.00300-0.0420 (PE)
Comments

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-4

-------
Table C-2. Sample Batch Duplicate Summary
Sample Batch
Number
D/FWG12107





D/FWG12148
D/F WG 12264
D/F WG12534
D/F WG 12641
D/F WG12737
D/F WG 12804





D/F WG 13 547
D/FWG13548
D/F WG13549
D/F WG13551
D/FWG13552




D/F WG 13 984
D/F WG 14274







Criteria
Met
N





Y
Y
Y
Y
Y
N





Y
Y
Y
Y
Y




Y
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








Comments
L6744-5, Ref 100 Newark Bay
Because this was above the 20% criteria, an additional
aliquot of this sample was prepared. Results for the
additional aliquot were within 1 1 % RPD from the
original results; therefore, this duplicate result was
accepted.
L6744-9, Ref 122 Newark Bay
L6760-2, Ref 27 PE
L6760-14,Ref55PE
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.
                                           C-5

-------
Sample Batch
Number
PCB WG12108
PCB WG12147
PCB WG12265
PCB WG12457
PCB WG12687
PCB WG12834
PCB WG12835
PCB WG12836
PCB WG1 3008
Criteria
Met
N
N
Y
N
Y
Y
N
Y
Y
Duplicate RPDa (%)
22
none
2.5
none
4.3
4.2
none
2.6
5.1
Comments
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.
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
C-6

-------
Sample Batch
Number
PCB WG13256




PCB WG13257
PCBWG13258
PCB WG13554
PCB WG14109







Criteria
Met
Y




Y
Y
Y
N








Duplicate RPDa (%)
1.7
(on U=1/2DL basis)



15
19
12
85
(on U=1/2DL basis)







Comments
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= 1II DL indicates that nondetects were assigned a concentration equal to one-half the SDL and EMPC concentrations were assigned a
value equal to the EMPC.
                                                            C-7

-------

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

-------

-------
Appendix D. XDS 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
Sample
Number
XDS 30
XDS 180
XDS 148
XDS 66
XDS 202
XDS 168
XDS 46
XDS 70
XDS 48
XDS 150
XDS 121
XDS 120
XDS 76
XDS 169
XDS 209
XDS 79
XDS 201
XDS 137
XDS 207
XDS 164
XDS 122
XDS 40
XDS 103
XDS 186
XDS 60
XDS 126
XDS 161
XDS 63
XDS 185
XDS 133
XDS 127
Measurement
Location
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
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
REP
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
o
5
TEQPrR (pg/g)
Developer3
NRd
37.50
NDO.88
(187.95)6
2.31
187.95
(ND<0.88)e
NDO.88
NDO.63
2.60
NDO.63
176.26
1710.37
2702.31
NDO.63
58.26
106.33
0.87
7.62
NDO.88
NDO.88
NDO.88
NDO.63
2.58
1.62
13.78
26.63
ND0.42
ND0.42
1.03
9611.90
13487.00
28334.32
Reference
Laboratory1"
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
TEQn
Developer3
678.4
1781.95
391.18
383.57
713.45
675.77
370.87
390.37
58756.79
282954.07
327462.3
341129.02
652.4
811.10
1018.37
734.47
837.69
456.94
744.96
935.82
558.19
568.95
500.23
934.75
35.43
43.23
37.58
57.74
32412.81
25719.60
26370.35
IF (Pg/g)
Reference
Laboratory1"
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
Total TEQ (pg/g) c
Developer
678.40
1819.45
391.18
385.88
901.40
675.77
370.87
392.97
58756.79
283130.33
329172.67
343831.33
652.40
869.36
1124.70
735.34
845.31
456.94
744.96
935.82
558.19
571.53
501.85
948.53
62.06
43.23
37.58
58.77
42024.71
39206.60
54704.67
Reference
Laboratory
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
                                                            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
Sample
Number
XDS90
XDS 160
XDS 172
XDS 181
XDS 35
XDS 128
XDS 187
XDS 197
XDS 117
XDS 88
XDS 141
XDS 112
XDS 154
XDS 93
XDS 132
XDS 77
XDS 50
XDS 191
XDS 75
XDS 182
XDS 64
XDS 29
XDS 146
XDS 156
XDS 178
XDS 193
XDS 176
XDS 131
XDS 56
XDS 53
XDS 174
XDS 205
XDS 80
XDS 167
XDS 67
Measurement
Location
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Sample Description
NC PCB Site #1
NC PCB Site #2
NC PCB Site #2
NC PCB Site #2
NC PCB Site #2
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
NC PCB Site #3
Newark B ay #1
Newark Bay #1
Newark B ay #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
Raritan Bay #1
Raritan Bay #1
Raritan Bay #1
Raritan B ay #1
Raritan Bay #2
Raritan Bay #2
Raritan Bay #2
Raritan Bay #2
Raritan Bay #3
Laboratory Raritan Bay #3
REP
4
1
2
o
J
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
TEQPrR (pg/g)
Developer3
30376.54
81581.94
77272.76
89977.32
>39302.35
156638.77
94523.48
95840.47
79055.94
NDO.50
NDO.42
2.98
NDO.42
1.31
NDO.42
NDO.50
6.39
NDO.42
NDO.50
NDO.42
ND0.5
NR
NDO.42
NDO.42
NDO.42
NDO.42
NDO.42
NDO.42
NDO.50
7.25
NDO.42
ND0.75
NEK1.90
NDO.42
NDO.50
Reference
Laboratory1"
85100
311000
305000
210000
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
TEQn
Developer3
28586.20
801924.49
698405.79
707390.20
>47853.33
631785.36
465880.06
416820.65
677660.44
17.43
33.76
28.80
40.35
92.25
256.22
83.38
98.99
48.12
59.28
40.83
90.68
83.4
54.83
39.47
23.79
23.42
29.64
19.04
23.75
22.54
29.05
31.29
38.20
25.92
23.74
IF (Pg/g)
Reference
Laboratory1"
906
3400
3300
3430
3490
8320
8410
9360
10800
23
14
14.5
13.5
50.6
47.4
74.1
50.4
38.9
44.9
40.2
41.9
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
Total TEQ (pg/g) c
Developer
58962.74
883506.43
775678.55
797367.52
87155.68
788424.13
560403.54
512661.12
756716.38
17.43
33.76
31.78
40.35
93.56
256.22
83.38
105.38
48.12
59.28
40.83
90.68
83.40
54.83
39.47
23.79
23.42
29.64
19.04
23.75
29.79
29.05
31.29
38.20
25.92
23.74
Reference
Laboratory
86006.00
314400.00
308300.00
213430.00
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
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
Sample
Number
XDS 165
XDS 190
XDS 82
XDS 55
XDS 44
XDS 39
XDS 124
XDS 84
XDS 159
XDS 104
XDS 25
XDS 188
XDS 95
XDS 118
XDS 109
XDS 91
XDS 42
XDS 52
XDS 86
XDS 199
XDS 115
XDS 136
XDS 98
XDS 177
XDS 200
XDS 142
XDS 119
XDS 69
XDS 140
XDS 71
XDS 34
XDS 143
XDS 107
XDS 173
XDS 183
Measurement
Location
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Sample Description
Raritan Bay #3
Raritan Bay #3
Saginaw River #1
Saginaw River #1
Saginaw River #1
Saginaw River #1
Saginaw River #2
Saginaw River #2
Saginaw River #2
Saginaw River #2
Saginaw River #3
Saginaw River #3
Saginaw River #3
Saginaw River #3
Solutia#l
Solutia#l
Solutia#l
Solutia#l
Solutia #2
Solutia #2
Solutia #2
Solutia #2
Solutia #3
Solutia #3
Solutia #3
Solutia #3
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #1
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #2
Titta. River Soil #3
REP
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
TEQPrR (pg/g)
Developer3
NDO.42
NDO.42
4.20
12.43
9.77
117.35
13.76
11.92
121.67
5.77
NR
1.55
ND0.50
ND0.50
ND0.63
ND0.63
2.44
3.11
2.57
4.24
1.86
56.53
4.21 (NR)e
1820.53
38.16
32.98
0.76
0.80
1.84
4.15
819.39
10.01
2.46
6.34
9.83
Reference
Laboratory1"
2.3
2.33
62.4
73.6
69.9
63.7
30.6
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
TEQn
Developer3
37.32
24.58
2940.06
3189.20
4091.26
2340
2729.40
5209.46
4096.12
2838.39
551.27
575.77
555.02
578.03
210.27
191.53
293.15
177.54
2310.65
2063.31
4730.74
23.96
2930.80
7621.37
2122.88
6480.03
225.88
132.25
88.89
97.65
1668.25
1587.89
1424.72
2074.71
3132.01
IF (Pg/g)
Reference
Laboratory1"
10.6
9.93
1050
683
1070
694
1110
953
1320
864
99.7
146
122
99.6
57.5
76.9
62
61.6
2090
1950
1860
2160
2810
2800
3000
3080
35
35.2
40
35.8
420
450
523
506
1050
Total TEQ (pg/g) c
Developer
37.32
24.58
2944.26
3201.63
4101.03
2457.35
2743.16
5221.38
4217.79
2844.16
551.27
577.32
555.02
578.03
210.27
191.53
295.59
180.65
2313.22
2067.55
4732.60
80.49
2935.01
9441.90
2161.04
6513.01
226.64
133.05
90.73
101.80
2487.64
1597.90
1427.18
2081.05
3141.84
Reference
Laboratory
12.90
12.26
1112.40
756.60
1139.90
757.70
1140.60
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
D-3

-------
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
Extract
Extract
Extract
Extract
Extract
Sample
Number
XDS74
XDS59
XDS 144
XDS 145
XDS 73
XDS 43
XDS 163
XDS 101
XDS 54
XDS 58
XDS 24
XDS 153
XDS 203
XDS 139
XDS 196
XDS 130
XDS 171
XDS 89
XDS 97
XDS 110
XDS 198
XDS 123
XDS 152
XDS 184
XDS 61
XDS 41
XDS 47
XDS 22
XDS 5
XDS 20
XDS 15
XDS 6
Measurement
Location
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Field
Field
Field
Field
Field
Sample Description
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
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
En vir Extract #1
Envir Extract #1
Envir Extract #1
Envir Extract #2
REP
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
o
5
4
1
TEQPrR (pg/g)
Developer3
NDO.50
23.44
13.10
NDO.42
NDO.50
0.60
NDO.42
NDO.50
6.27
NDO.50
1.91
NDO.42
NDO.42
NDO.42
NDO.42
186.90
51.43
NDO.63
NDO.63
2133.92
21.33
2149.30
12.94
99.17
NDO.63
172.69
NDO.63
9.18
26.11
9.84
NA
0.45
Reference
Laboratory1"
1.16
1.54
1.33
0.0527
0.034
0.0407
0.0403
0.649
0.71
0.566
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
TEQn
Developer3
1932.38
16722.28
(3353. 51)e
1650.17
6.90
1.88
4.42
1.66
131.24
480.42
647.96
303.39
32.37
24.56
49.49
19.60
30696.45
37880.91
34205.98
11048.42
(28400. 57)e
252424.73
61670.05
286476.09
34424.23
(43807. 75)e
122062.58
50372.99
15502.28
48249.56
489.67
523.29
516.89
2304.98
169.79
IF (Pg/g)
Reference
Laboratory1"
676
1220
1300
1.05
1.11
1
1.7
52.8
123
66.1
94.1
13
11.2
12.7
13.8
7290
7370
7450
7160
9720
9770
9200
11300
10300
9770
9320
9870
175
444
176
439
55.3
Total TEQ (pg/g) c
Developer
1932.38
16745.72
1663.27
6.90
1.88
4.42
1.66
131.24
486.69
647.96
305.30
32.37
24.56
49.49
19.60
30883.35
37932.34
34205.98
11048.42
254558.65
61691.38
288625.39
34437.17
122161.75
50372.99
15674.97
48249.56
498.85
549.40
526.73
2304.98
170.24
Reference
Laboratory
677.16
1221.54
1301.33
1.10
1.14
1.04
1.74
53.45
123.71
66.67
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
D-4

-------
Sample Type
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Extract
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
XDS 18
XDS9
XDS 16
XDS 19
XDS1
XDS 7
XDS 17
XDS 4
XDS 12
XDS 3
XDS 21
XDS 10
XDS 23
XDS 14
XDS 8
XDS 2
XDS 11
XDS 13
XDS 195
XDS 113
XDS 72
XDS 38
XDS 108
XDS 87
XDS 111
XDS 155
XDS 114
XDS 192
XDS31
XDS 37
XDS 138
XDS 28
XDS 189
XDS 106
Measurement
Location
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Field
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Field
Laboratory
Field
Laboratory
Laboratory
Sample Description
Envir Extract #2
Envir Extract #2
Envir Extract #2
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #1
Spike #2
Spike #2
Spike #2
Spike #2
Spike #3
Spike #3
Spike #3
Spike #3
Cambridge 5183
Cambridge 5183
Cambridge 5 183
Cambridge 5 183
Cambridge 5183
Cambridge 5183
Cambridge 5 183
Cambridge 5 1 84
Cambridge 5 1 84
Cambridge 5 1 84
Cambridge 5 184
ERA Aroclor
ERA Aroclor
ERA Aroclor
ERA Aroclor
ERA Blank
REP
2
o
J
4
1
2
3
4
5
6
7
1
2
3
4
1
2
3
4
1
2
3
4
5
6
7
1
2
3
4
1
2
3
4
1
TEQPrR (pg/g)
Developer3
1.20
NR
NR
O.07
0.09
0.09
0.07
0.1
NA
0.12
10.66
NR
12.2
NR
225.24
211.91
NR
NR
NDO.42
339.41
19.18
3.23
ND0.5
6.70
4.63
81.22
2.74
0.79
NR
1690.23
110.72
NR
NDO.42
(69.77)e
13.72
Reference
Laboratory1"
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
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
TEQn
Developer3
196.45
193.02
211.13
0.43
0.13
0.34
0.24
0.13
0.56
O.I 3
74.22
70.56
69.42
98.64
97.57
46.27
95.06
58.67
26.69
26.94
14.22
28.47
22.91
29.80
18.58
1033.74
716.75
1117.68
807.46
170.37
297.76
167.41
175.10
NDO.45
IF (Pg/g)
Reference
Laboratory1"
53.3
53.1
53.6
0.504
0.509
0.537
0.524
0.585
0.576
0.52
91.6
91.8
89.1
100
0.324
0.348
0.363
0.268
4.78
4.08
4.06
3.56
3.89
5.93
3.89
187
188
173
180
36.4
32.9
37.9
35.5
0.0942
Total TEQ (pg/g) c
Developer
197.65
193.02
211.13
0.43
NR
0.34
0.24
0.10
0.56
0.12
84.88
70.56
81.62
98.64
322.81
258.18
95.06
58.67
26.69
366.35
33.40
31.70
22.91
36.50
23.21
1114.96
719.49
1118.47
807.46
1860.60
408.48
167.41
175.10
13.72
Reference
Laboratory
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
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
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
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
XDS 116
XDS92
XDS 51
XDS 105
XDS 81
XDS 100
XDS 32
XDS 147
XDS 135
XDS 96
XDS 27
XDS 166
XDS 94
XDS 83
XDS 125
XDS 204
XDS 68
XDS 134
XDS 162
XDS 45
XDS 149
XDS 175
XDS 158
XDS 36
XDS 129
XDS 85
XDS 151
XDS 78
XDS 65
XDS 62
XDS 26
XDS 57
XDS 157
XDS 179
XDS 102
Measurement
Location
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Laboratory
Laboratory
Laboratory
Freld
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Freld
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Freld
Laboratory
Laboratory
Laboratory
Laboratory
Sample Description
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
ERATCDD10
ERATCDD 10
ERATCDD 10
ERA TCDD 30
ERA TCDD 30
ERA TCDD 30
ERA TCDD 30
LCG CRM-529
LCG CRM-529
LCG CRM-529
LCG CRM-529
NIST 1944
NIST 1944
NIST 1944
NIST 1944
REP
2
3
4
5
6
7
8
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
TEQPrR (pg/g)
Developer3
0.88
1.02
6.87
NEK1.26
1.91
ND0.50
5.67
ND0.42
NDO.42
27.34
NR
NDO.42
1.88
14.77
NDO.42
51.17
26.31
46.47
31.27
5.35
NDO.42
NDO.42
NDO.42
1.86
NDO.42
NDO.5
NDO.42
211.50
142.50
135.27
NR
8.27
NDO.88
4.01
2.01
Reference
Laboratory1"
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
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
TEQn
Developer3
NDO.23
NDO.45
NDO.45
NDO.45
NDO.45
0.75
13.74
NDO.45
1.16
NDO.45
2.92
0.46
NDO.45
0.75
1.99
252.75
206.93
19.60
21.62
10.78
22.81
14.67
16.83
45.66
44.46
32.77
35.42
15243.48
15448.41
16448.80
15684.8
885.45
591.64
776.07
573.89
IF (Pg/g)
Reference
Laboratory1"
0.0728
0.237
0.307
0.113
0.0524
0.211
0.0692
0.159
0.141
0.161
0.248
0.0386
NAf
0.053
0.127
0.204
0.507
0.105
0.0628
8.69
9.28
8.44
8.2
27.4
25.3
24.8
23.9
NAf
6930
6900
7190
237
206
252
219
Total TEQ (pg/g) c
Developer
0.88
1.02
6.87
NA
1.91
0.75
19.41
NR
1.16
27.34
2 92
0.46
1.88
15.52
1.99
303.92
233.24
66.07
52.89
16.13
22.81
14.67
16.83
47.52
44.46
32.77
35.42
15454.98
15590.91
16584.07
15684.80
893.72
591.64
780.08
575.90
Reference
Laboratory
0.08
0.24
0.35
0.14
0.07
0.24
0.09
0.18
0.15
0.17
0.27
10.64
NAf
10.65
10.08
1030.20
1030.51
1180.11
1020.06
8.70
9.29
8.47
8.25
27.45
25.32
24.84
23.94
NAf
7335.00
7398.00
7546.00
277.10
249.70
294.10
260.00
D-6

-------
Sample Type
Performance
Performance
Performance
Performance
Performance
Performance
Performance
Sample
Number
XDS 206
XDS 49
XDS 194
XDS 99
XDS 170
XDS 208
XDS 33
Measurement
Location
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Laboratory
Field
Sample Description
WellrngtonWMS-01
WellrngtonWMS-01
WellrngtonWMS-01
WellrngtonWMS-01
Wellington WMS-01
Wellington WMS-01
Wellmgton WMS-01
REP
1
2
3
4
5
6
7
TEQPrR (pg/g)
Developer3
NDO.75
9.21
NDO.42
13.22
NDO.42
NDO.75
524.54
Reference
Laboratory1"
10.6
9.4
9.62
9.07
10.3
9.62
9.68
TEQn
Developer3
201.61
177.64
203.56
138.57
290.81
201.83
228.59
IF (Pg/g)
Reference
Laboratory1"
68
65.7
61.9
66.1
68
65.7
65.4
Total TEQ (pg/g) c
Developer
201.61
186.85
203.56
151.79
290.81
201.83
753.13
Reference
Laboratory
78.60
75.10
71.52
75.17
78.30
75.32
75.08
a Data listed exactly as reported by the developer.
b Qualifier flags (e.g., J and K flags) included in the raw data have been removed for the purposes of statistical analysis.
0 Data calculated by summing TEQPCB and TEQD/F.
d NR = result not available.
s Revised result provided by XDS after demonstration period. Original result was used in the data analysis.
' Reference laboratory data was discarded due to laboratory sample preparation error.
                                                                             D-7

-------