Ł% United States
Environmental Protectio
^1 M^k. Agency
Office of Water
www.epa.gov April 2021
Report on the Multi-laboratory
Validation of Clean Water Act
Method 1628 for PCB Congeners
-------
U.S. Environmental Protection Agency
Office of Water (4303T)
Office of Science and Technology
Engineering and Analysis Division
1200 Pennsylvania Avenue, NW
Washington, DC 20460
EPA 820-R-21-003
-------
Disclaimer
This report was prepared by Adrian Hanley of the Engineering and Analysis Division within EPA's
Office of Water, with assistance from Harry McCarty and Mirna Alpizar of CSRA, a General Dynamics
Information Technology company. Mention of company names, trade names, or commercial products in
this report does not constitute endorsement or recommendation for use.
Questions or comments regarding this report should be addressed to:
Adrian Hanley
Engineering and Analysis Division (4303T)
Office of Science and Technology
USEPA Office of Water
William Jefferson Clinton Building West
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
Phone: 202-564-1564
E-mail: hanley.adrian@epa.gov
Method 1628 Multi-Laboratory Validation Study Report
l
April 2021
-------
Acknowledgements
EPA acknowledges the support of a number of organizations and individuals in the development and
validation of the draft PCB congener method, including the developers of the original procedure, the
members of EPA's workgroup, the organizations that provided the bulk samples of wastewater and
biosolids, the performance testing sample provider that prepared the individual samples, the vendors that
provided the custom standards used in the study, and EPA's support contractor staff who oversaw the
day-to-day operations during the study and assisted EPA in the preparation of this report. At a minimum,
that includes the following:
Coreen Hamilton
Sean Campbell
Louis Haviland
Steve Leahy
Jenny Pape
Vicki Reesor
Brian Watson
Charlie Appleby
Jason Collum
David Hair
Jim Keating
Danielle Kleinmaier
Geniece Lehmann
Michael Mahoney
Brett Moody
Lantis Osemwnegie
Santhini Ramasamy
Steve Reimer
Brian Shuhler
Scott Sivertsen
Brian D'Amico
Jan Goodwin
Adrian Hanley
Jan Matuszko
Michael Scozzafava
SGS-AXYS Analytical
SGS-AXYS Analytical
SGS-AXYS Analytical
SGS-AXYS Analytical
SGS-AXYS Analytical
SGS-AXYS Analytical
SGS-AXYS Analytical
EPA OLEM/OSRTI/TIFSD
EPA Region 4
EPA OW/OWM
EPA OW/OST/SHPD
EPA Region 5
EPAORD
EPA Region 3
EPA OLEM/OSRTI/TIFSD
EPAORD
EPA OW/OST/HECD
EPA Region 10
EPA Region 4
EPA Region 4
EPA OW/OST/EAD
EPA OW/OST/EAD
EPA OW/OST/EAD
EPA OW/OST/EAD
EPA OW/OST/EAD
Lemuel Walker EPA OW/OST/EAD
Robert Wood EPA OW/OST/EAD
Lynn Zipf EPA OW/OST/EAD
Colleen Chilson ERA, a Waters Company
Katherine Pauling ERA, a Waters Company
Anthony Ciacco ERA, a Waters Company
Don Shelly LGC Standards
Ben Priest Cambridge Isotope Laboratories
Kayla Meehan Cambridge Isotope Laboratories
Carrie Gamber Euro fins TestAmerica, Pittsburgh
Craig Forbes Hampton Roads Sanitation District
Blaine Hoffman Hampton Roads Sanitation District
Mike Martin Hampton Roads Sanitation District
Louis Logan Massachusetts Water Resources Authority
Steve Rhode Massachusetts Water Resources Authority
Greg Cavallo Delaware River Basin Commission
Ron MacGillvray Delaware River Basin Commission
Colin Conerly National Council for Air and Stream Improvement
Paul Wiegand National Council for Air and Stream Improvement
Mahesh Pujari Los Angeles City Sanitation
Mirna Alpizar General Dynamics Information Technology
Eric Boring General Dynamics Information Technology
Harry McCarty General Dynamics Information Technology
Kenneth Miller General Dynamics Information Technology
Method 1628 Multi-Laboratory Validation Study Report
li
April 2021
-------
Executive Summary
The goal of this project was to develop and validate a polychlorinated biphenyl (PCB) congener method
for use in Clean Water Act compliance monitoring. Currently, Method 608.3 is the commonly used EPA
method that is approved at 40 CFR Part 136 for compliance monitoring of PCBs, and it only measures
seven common Aroclor mixtures, not the total amount of PCB contamination, nor the specific PCB
congeners designated as toxic by the World Health Organization (WHO). To address these shortcomings
of Method 608.3, a new PCB congener method should meet the following criteria:
1. Identifies and quantifies PCB contamination using individual congeners, rather than attempting to
recognize and quantify the patterns generated from Aroclor mixtures.
2. Is more sensitive than the currently approved Method 608.3, but is not so sensitive that it is adversely
affected by typical laboratory background contamination.
3. Can be implemented at a typical mid-sized full-service environmental laboratory.
In response to this need, the US Environmental Protection Agency (EPA) Office of Water convened a
workgroup of EPA, laboratory, and utility staff, supported by contractors. After examining a suite of
candidate procedures, the workgroup prioritized an unpublished laboratory procedure and a specific set of
congeners for further development and validation efforts, designed the study described in this report, and
reviewed all the study products (see Section 1).
The newly validated Method 1628 detects all 209 PCB congeners, and quantifies them either directly or
indirectly. A total of 29 carbon-13 labeled congeners are used as isotope dilution quantification
standards. An additional 19 congeners are quantified by an extracted internal standard procedure, using
one of the isotope dilution standards. The remaining 144 congeners are quantified against a labeled
standard in the same homolog by assuming that it has a similar response (see Section 4). The method
requires the laboratory to run standards containing all 209 congeners to establish retention times and
method detection limits, but not during routine analysis. This approach strikes a balance between
enabling the laboratory to detect and quantify all 209 congeners, while not making the method too
arduous. Based on the results of matrix spike samples (see Section 8), method performance was similar
across all the congeners, regardless of the quantification approach.
The primary focus of this validation study was on wastewater compliance monitoring, but there is a need
for testing biosolids, soils/sediments, and fish tissue as well. Therefore, the multi-laboratory study tested:
Nine aqueous sample types, including wastewater effluents and influents collected from publicly
owned treatment work and indirect industrial dischargers
Three sediments collected from the Great Lakes region
Three biosolids collected from municipal wastewater treatment plants
Fish tissues from three species collected from the Great Lakes
The multi-laboratory validation study of Method 1628 met all of the goals that EPA set for this study.
The study generated initial precision and recovery data for aqueous, solid, and tissue matrices. Over 95%
of the spike recoveries for both the aqueous and tissue samples fell between 50 - 150%, while 87% of the
biosolids samples and 63% of the sediment samples had recoveries between 50 - 150%. The percentage
of false negatives for aqueous samples was less than 0.2%, for solids samples (sediments and biosolids) it
was less than 3%, and for tissue samples it was less than 0.1%. While a particularly difficult matrix may
cause an interference that invalidates the results for one congener, it is unlikely that the matrix will cause
the same type of interference for all the PCB congeners in the sample.
The performance of Method 1628 and the quality control requirements incorporated into it make it more
sensitive and accurate compared to the currently approved Method 608.3, as outlined in Section 11.
Method 1628 Multi-Laboratory Validation Study Report
m
April 2021
-------
Table of Contents
Disclaimer i
Acknowledgements ii
Executive Summary iii
1. Introduction 1
Background 1
EPA Workgroup 2
SOP Selection 2
Congener Prioritization 3
Quantification 3
Summary of the Results of the Single-laboratory Study 4
Multi-laboratory Study Goals and Design 5
2. Identification and Selection of Laboratories 6
3. Study Samples 8
Wastewater Matrices 8
Biosolids Matrices 9
Sediment Matrices 9
Fish Tissue Matrices 10
Reconnaissance Analyses 10
4. Approaches to Calibration and Quantification 12
Multi-point Initial Calibration 16
Response Ratios and Response Factors 18
Mass-to-charge Ratios Monitored for Each Analyte 19
Area Subtraction of Higher Homolog Interference 24
Calibration Linearity and Stability 24
5. Initial Precision and Recovery 27
Aqueous IPR/OPR Criteria 29
Solids IPR and OPR Results 33
Tissue IPR and OPR Results 37
6. Method Detection Limits 42
Aqueous Sample Pooled MDL Determinations 43
Soil/Sediment and Biosolids Sample MDL Determinations 49
Tissue Sample MDL Determinations 55
7. Unspiked Sample Analyses 61
Wastewater Samples 61
Sediment Samples 63
Biosolids Samples 67
Tissue Samples 70
8. Matrix Spike Analyses 75
Aqueous Sample MS/MSD Results 76
Sediment Sample MS/MSD Results 82
Biosolids Sample MS/MSD Results 86
Tissue Sample MS/MSD Results 89
Method 1628 Multi-Laboratory Validation Study Report iv April 2021
-------
9. Labeled Compound Results 92
Aqueous Sample Labeled Compound Results 92
Sediment Sample Labeled Compound Results 95
Biosolids Sample Labeled Compound Results 97
Tissue Sample Labeled Compound Results 99
10. Data Review and Validation 102
11. Conclusions 104
Study Goals 104
Method Performance 105
Summary 108
12. References 109
Appendices
Appendix A Labeled PCB Congeners to be used as Quantitation Standards
Appendix B Interim Quality Control Acceptance Criteria Arising from the Method Validation Study
Appendix C Study Plan for Multi-laboratory Validation of the EAD PCB Congener Method
List of Tables
Table 1. List of Participating Laboratories 6
Table 2. Bulk Wastewater Matrix Sources 8
Table 3. Water Quality Characteristics of the Study Samples (in mg/L) 9
Table 4. Reconnaissance Analyses Results for Study Samples 11
Table 5. Quantification Reference and Calibration Approach for the 65 Congeners with Direct
Calibration Data 12
Table 6. Congeners Not Calibrated Directly 14
Table 7. Composition of the Initial Calibration Standards 16
Table 8. Ions Monitored for each Congener with Direct Calibration Data and the Labeled
Compounds, in Retention Time Order 20
Table 9. Ions Monitored for the 144 Other Congeners, in Retention Time Order 21
Table 10. Ions Monitored for Correcting for Interferences from Higher Homologues 24
Table 11. Summary of Response Ratios, Response Factors, and Relative Standard Deviations for
32 Initial Calibrations 25
Table 12. Aqueous IPR and OPR Calculated QC Acceptance Criteria for Target Analytes 29
Table 13. IPR and OPR QC Acceptance Criteria for Labeled Compounds in Aqueous Matrices 31
Table 14. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance
Criteria for Aqueous Matrix IPR and OPR 32
Table 15. IPR and OPR QC Acceptance Criteria for Target Analytes in Solid Matrices 34
Table 16. IPR and OPR QC Acceptance Criteria for Labeled Congeners in Solid Matrices 35
Table 17. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance
Criteria for Solid Matrix IPR and OPR 36
Table 18. IPR and OPR QC Acceptance Criteria for Target Analytes in Tissue Matrices 37
Table 19. IPR and OPR QC Acceptance Criteria for Labeled Compounds in Tissue Matrices 39
Table 20. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance
Criteria for Tissue Matrix IPR and OPR 40
Table 21. Aqueous Sample Pooled MDL and ML Results (ng/L) 43
Table 22. Frequency of Aqueous MDLb values by Lab 47
Method 1628 Multi-Laboratory Validation Study Report
v
April 2021
-------
Table 23. Solid Sample Pooled MDL and ML Results (ng/g) 49
Table 24. Frequency of Solid MDLb values by Lab 53
Table 25. Tissue Sample Pooled MDL and ML Results (ng/g) 55
Table 26. Frequency of Tissue MDLb values by Lab 58
Table 27. Summary of Wastewater Sample Cleanups 61
Table 28. Unspiked Wastewater Sample Results in ng/L 61
Table 29. Summary of Sediment Sample Cleanups 63
Table 30. Unspiked Sediment Sample Results in ng/g (dw) 64
Table 31. Summary of Biosolids Sample Cleanups 67
Table 32. Unspiked Biosolids Sample Results in ng/g (dw) 67
Table 33. Unspiked Tissue Sample Results in ng/g 70
Table 34. Results for Congeners Detected by Four Labs for Tissue Sample #1 in ng/g 73
Table 35. Results for Congeners Detected by Five Labs for Tissue Sample #2 in ng/g 73
Table 36. Results for Congeners Detected by Five Labs for Tissue Sample #3 in ng/g 74
Table 37. Composition of Matrix Spiking Solutions 75
Table 38. Matrix Spike Recoveries for Wastewater Samples #3 to #6 (%) 77
Table 39. Matrix Spike Recoveries for Wastewater Samples #7 to #10 (%) 79
Table 40. Matrix Spike Recoveries for Sediment Samples 82
Table 41. Matrix Spike Recoveries for Biosolids Samples 86
Table 42. Matrix Spike Recoveries for Tissue Samples 89
Table 43. Observed Aqueous Labeled Compound Recoveries and Calculated Acceptance Criteria 92
Table 44. Observed Aqueous Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria 94
Table 45. Observed Sediment Labeled Compound Recoveries and Calculated Acceptance Criteria 95
Table 46. Observed Sediment Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria 96
Table 47. Observed Biosolids Labeled Compound Recoveries and Calculated Acceptance Criteria 97
Table 48. Observed Biosolids Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria 98
Table 49. Observed Tissue Labeled Compound Recoveries and Calculated Acceptance Criteria 99
Table 50. Observed Tissue Labeled Compound Recovery Failure Rates for Potential Acceptance Criteria
100
Table 51. Comparison of Method 608.3 and Method 1628 for Aqueous Samples 105
Table 52. Estimated Aroclor 1242 Concentrations Using 5 Most Prevalent Congeners in
Aqueous Matrices 106
List of Figures
Figure 1. PCB Structure and Congener Counts 1
Figure 2. Pooled Aqueous MDLS Values in Elution Order 48
Figure 3. Pooled Solid MDLS Values in Elution Order 54
Figure 4. Pooled Tissue MDLS Values in Elution Order 59
Figure 5. Mean Matrix Spike Wastewater Recoveries by Sample, in Elution Order
(without Sample #2) 81
Figure 6. Mean Matrix Spike Sediment Recoveries by Sample, in Elution Order 85
Figure 7. Mean Matrix Spike Biosolids Recoveries by Sample, in Elution Order 88
Figure 8. Mean Matrix Spike Tissue Recoveries by Sample, in Elution Order 91
Method 1628 Multi-Laboratory Validation Study Report vi April 2021
-------
1. Introduction
The goal of this project is to complete multi-laboratory validation of a polychlorinated biphenyl (PCB)
congener method that is a significant improvement to Method 608.3, the only approved PCB method for
wastewater. Method 608.3 only measures the seven common Aroclor mixtures, not the total amount of
PCB contamination, nor the specific PCB congeners known to be most toxic. PCB production has been
banned since 1977, therefore, most PCB contamination in the environment is more than 40 years old.
Environmental PCB contamination is usually made up of weathered PCBs that do not resemble the
original Aroclor mixtures. Furthermore, not all PCB contamination is from Aroclor mixtures. This leads
to false negatives and artificially low results. A new PCB congener method must meet the following
criteria:
1. The method identifies and quantifies PCB contamination using individual congeners, not an estimated
quantity based off patterns generated from Aroclor mixtures.
2. The method is more sensitive than the currently approved Method 608.3, but is not so sensitive that it
is adversely affected by typical laboratory background contamination.
3. The method can be implemented at a typical mid-sized full-service environmental laboratory.
This multi-laboratory validation study was a follow-on to the successful single-laboratory study
conducted in 2017 (CSRA 2018).
Background
PCB contamination and its properties are well documented, so only a brief summary is provided here. A
biphenyl is 12 carbon atoms, made up of two 6 carbon aromatic rings that are bonded together at one
location (Figure 1). Chlorinated biphenyls are a group of compounds made up of a biphenyl with one to
ten chlorine atoms attached to the 10 possible bond locations. There are 209 possible molecules,
depending on the number and the placement of the chlorine atoms. Each of these 209 possibilities is
called a congener. All of the congeners that contain the same number of chlorine atoms are called a
homolog (e.g., all of the congeners that contain 4 chlorine atoms are one homolog, this homolog is often
referred to as the tetrachlorobiphenyls or the tetrachloro homolog).
Figure 1. PCB Structure and Congener Counts
o 9 o' o'
\Jp i,/
para
\\ //
<7 v
rneta
ortho
# Chlorine Atoms -
Homolog
# of
Congeners
Congener
Numbers
1 - Monochloro
3
1 -3
2 - Dichloro
12
4-15
3 - Trichloro
24
16- 39
4 - Tetrachloro
42
40-81
5 - Pentachloro
46
82-127
6 - Hexachloro
42
128 -169
7 - Heptachloro
24
170 -193
8 - Octachloro
12
194-205
9 - Nonachloro
3
206 -208
10 - Decachloro
1
209
Most PCBs were produced in the U.S. by a batch process of exposing biphenyl to chlorine gas until the
desired percentage of chlorinated compounds were produced. The commercial mixtures were sold under
the name Aroclor. There were seven commonly produced Aroclors: 1016, 1221, 1232, 1242, 1248, 1254,
and 1260, and many other less common mixtures. The last two digits of the Aroclor mixture indicate the
percent chlorine by weight (except 1016, which is 41% chlorine by weight). The higher the number, the
Method 1628 Multi-Laboratory Validation Study Report
1
April 2021
-------
higher the chlorine content. Aroclors occurred as oily liquids or waxy solids, depending on the specific
Aroclor mixture. PCBs are known to occur from other sources aside from the Aroclor mixtures and are
sometimes produced as unintentional by-products of various combustion processes involving chlorine-
containing materials (Erickson, 1997).
Historically, PCBs were used for many industrial applications, most commonly in electrical capacitors
and transformers. Domestic manufacture of PCBs occurred from 1927 until 1977, when production was
banned. PCBs are very stable in the environment, nonflammable, insulating, and thermally stable. They
can be destroyed by incineration, but only at very high temperatures. PCBs are relatively non-volatile and
insoluble in water. Most PCBs found in water are believed to be attached to particulate matter. They
primarily exist in air as particulates, which can wash out during rain events and end up in surface waters.
PCBs bioaccumulate as they move up the food chain, but some congeners bioaccumulate more than
others.
EPA Workgroup
Prior to beginning the single-laboratory validation study, the Engineering and Analysis Division, which is
part of the Office of Science and Technology within the Office of Water (OW/OST/EAD), assembled an
EPA workgroup. The workgroup included representatives of all three divisions of OW/OST, the Office
of Research and Development (ORD), OW's Office of Wastewater Management, the Office of Land and
Emergency Management, EPA Regions 3, 4, 5, and 10, two subject matter experts from outside of EPA,
and EAD's support contractor staff, CSRA.
SOP Selection
Prior to the single-laboratory validation study, the workgroup met to review available laboratory standard
operating procedures (SOPs) for analysis of PCB congeners. The SOPs and published articles reviewed
can be summarized into three categories.
Dual-Column Gas Chromatography Electron Capture Detector (GC-ECD): This is the same analytical
technique used for Method 608.3. EPA and other organizations have developed methods and SOPs that
measure Aroclor mixtures and specific congeners using GC-ECD. The addition of the specific congeners
makes the technique better at detecting weathered PCB congeners, but the overall sensitivity is not
improved dramatically. Because ECD is not a selective detector and has the ability to detect many types
of organic compounds (halogens, organometallic compounds, nitriles, or nitro compounds), matrices that
are high in organic contamination can cause significant interference, and often require extensive sample
cleanup. Since many municipal wastewater treatment facilities treat high organic content wastewater with
chlorine, which can produce interfering background peaks when using an ECD, the workgroup did not
recommend these GC-ECD SOPs for further testing.
Gas Chromatography with Tandem Mass Spectrometers (GC-MS-MS): Some literature and instrument
SOPs exist for both GC-MS-MS and GC with triple quadrupole MS. Most of these are instrument SOPs
and no published SOP was found for environmental samples. The testing that has been done on PCBs is
with very small lists of congeners (22 or fewer of the 209). The European Union (EU) published
guidance on this type of testing, and some laboratories in Europe run methods with this technology, but it
is mainly used for environmental and feed screening. The EU guidance only mentions 22 congeners.
This technology looks promising and appears to have excellent sensitivity, but there are not any fully
developed lab SOPs currently available to EAD. The workgroup did not recommend this technology due
to a lack of any documented laboratory SOPs and because very few commercial environmental labs own
this instrumentation, making it difficult to implement on a national scale in the next few years.
Gas Chromatography with Mass Spectrometry and Selection Monitoring (<7 C-MS-SIM): Two
promising laboratory SOPs were reviewed that use GC-MS-SIM. Both addressed all 209 congeners, but
Method 1628 Multi-Laboratory Validation Study Report
2
April 2021
-------
some of the congeners co-eluted into one peak. One SOP resolved all 209 congeners into 189 peaks (22
co-eluting peaks), while the other resolved all 209 congeners in 160 peaks (49 co-eluting peaks). The
workgroup agreed to pursue the SOP with more co-elutions. The rationale was that this SOP had a run
time that was 20 minutes shorter (45 minutes instead of 65 minutes), which was better for
implementation, and had better quantitation and quality control. It used 12 Carbon-13 (13C) labeled
standards (one for each PCB homolog and 2 additional surrogates), while the other SOP only had 2
labeled standards and 4 internal standards (oddly, some were unlabeled target analytes - albeit congeners
that are almost never detected in the environment). The SOP selected by the workgroup has been run for
well over 10 years at the laboratory that submitted the SOP in a wide variety of matrices, and it requires
equipment that most environmental laboratories already own.
Congener Prioritization
Prior to the single-laboratory validation study, the workgroup met to discuss and prioritize which
congeners were a high priority for this method. There was general consensus that the more important
congeners to monitor have the following characteristics:
Most common in the environment
Congeners known to be most prevalent in human tissue
Congeners present in the largest quantities within the manufactured Aroclor mixtures
The 12 toxic congeners identified by the World Health Organization (WHO)
There is significant overlap between the first three of these four categories.
OW/OST/EAD assembled several databases of PCB congener data prior to the meeting. Databases were
selected that had PCB congener data from Method 1668, a highly sensitive high-resolution gas
chromatography mass spectrometry (HRGC/HRMS) method.
Wastewater data from the Delaware River Basin Commission (DRBC) (2005 - 2013)
EPA National Lake Fish Tissue Survey data (2000 - 2004)
EPA National Sewage Sludge Survey data (2001)
Upper Trenton Channel sediment data from the Great Lakes National Program Office
ORD provided a list of the most common congeners detected in tissues. The workgroup also consulted
the Aroclor formulation data compiled by Frame, G. M., Cochran, J. W., and Boewadt, S. S. in the
Journal of High Resolution Chromatography, Vol. 19, pp 657-668 (1996).
The selection process for how EPA chose which congeners to be used as isotope dilution standards and
internal standards is detailed in Appendix A, "Labeled PCB Congeners to be used as Quantitation
Standards." Through that selection process, EPA identified 48 congeners that are a high priority for the
new PCB method because of their prevalence in environmental samples, high concentrations in Aroclors,
and their toxicities (see the introduction to that report for further details). To facilitate proper
identification of the congeners in each homolog, EPA included the first and last eluting congeners in each
homolog in the target analytes list for the procedure. Using common GC columns and conditions, some
of those 48 targeted congeners coelute with 17 other congeners. Therefore, EPA included those 17 other
congeners in the calibration process of the draft procedure, and for the purposes of this report, EPA is
referring to these 65 congeners as the congeners with direct calibration data.
Quantification
In order to quantify those 65 congeners, EPA selected commercially available 13Ci2-labeled analogs of 29
of those congeners that are to be used as isotope dilution standards. Three additional 13Ci2-labeled
standards are used as "recovery standards" that are spiked after extraction and used to calculate the
Method 1628 Multi-Laboratory Validation Study Report
3
April 2021
-------
recoveries of the isotope dilution standards. The details of calibration and quantification approach are
described in Section 4 of this report.
Each laboratory will still need to run standards of all 209 congeners on occasion to establish retention
times and method detection limits, but not during routine analysis. This strikes a balance between
enabling the laboratory to detect and quantify all 209 congeners, without making the method too arduous.
If a laboratory or discharger wants to quantify one or more specific congeners more accurately, the
method allows the laboratory to expand the list of congeners used for calibration and/or the list of isotope
dilution standards. The method flexibility codified at 40 CFR Part 136.6 allows a laboratory to add
analytes and quantification standards to an approved wastewater method if they can demonstrate adequate
performance.
Summary of the Results of the Single-laboratory Study
The single-laboratory validation was performed by SGS AXYS Analytical, the developer of the original
laboratory SOP that was selected by the EPA workgroup and that study was deemed a success. As noted
in the report on that study (CSRA 2018), the single-laboratory validation study of the draft PCB congener
method met two of EPA's three criteria.
1. The method identifies and quantifies PCB contamination using individual congeners, not an
estimated quantity based offpatterns generated from Aroclor mixtures.
The study generated initial precision and recovery data for aqueous, solid, and tissue matrices. Of the
over two hundred matrix spike samples analyzed during the single-laboratory study:
Almost all of the sediment samples achieved recoveries between 60-115% (1,243 out 1,248
results, or 99.6%).
Almost all of the biosolid samples achieved recoveries between 75 -150% (1,334 out of 1,344
results, or 99.3%).
All of the fish tissue samples achieved recoveries between 88-115%.
A majority of the wastewater samples extracted by separatory funnel achieved recoveries between
60 - 130% (3588 out of 3648 results, or 98.4%)
Almost all of the wastewater samples extracted by disk-based SPE achieved recoveries between
75 - 130% (3448 out of 3456 results, or 99.8%).
A majority of the wastewater samples extracted by cartridge-based SPE achieved recoveries
between 60 - 130% (3,267 out of 3,360 results, or 97.2%).
The single-laboratory validation results demonstrate that this method can identify and quantify individual
PCB congeners.
2. The method is more sensitive than the currently approved Method 608.3.3, but is not so sensitive that
it is adversely affected by typical laboratory background contamination.
The only published sensitivity data for Method 608.3.3 are for Aroclor 1242, which has an MDL of 65
ng/L in aqueous samples. According to the 1996 Frame et al. data, the main constituents of Aroclor 1242
are PCB Congeners 8 (7.05%), 18 (8.53%), 28 (6.86%), 31 (7.34%), and 33 (5.01%). The highest
aqueous MDL calculated in the single-laboratory study for PCB-18, the largest component of Aroclor
1242, was 0.96 ng/L. Assuming that all of the PCB-18 came from unweathered Aroclor 1242, and the
other congeners were present at the proportions described by Frame et al., the PCB-18 MDL suggests that
Aroclor 1242 could be present at approximately 11 ng/L (0.96 ng/L/0.0853).
Method 1628 Multi-Laboratory Validation Study Report
4
April 2021
-------
The method detection limits generated during the single-laboratory study demonstrate that this method is
more sensitive than Method 608.3 and is not subject to any significant blank contamination.
Given that the single-laboratory study was conducted by the laboratory that developed the method, one of
the major goals of the multi-laboratory study was to evaluate EPA's third criterion for a new PCB
congener method.
3. The method can be implemented at a typical mid-sized full-service environmental laboratory.
The remainder of this report includes information that allowed EPA to address that goal. All of the
required instrumentation for this method is available in any typical full-service environmental laboratory.
Multi-laboratory Study Goals and Design
The goals of the multi-laboratory validation study were to:
Obtain data from matrices that are representative of the method's intended use
Obtain data from laboratories that are representative of those likely to use the approved method,
but that were not directly involved in its development
Obtain feedback from laboratory users on the specifics of the draft method (e.g., is it clear and
easy to understand, or are changes to the method text needed?)
Use study data to characterize performance of the method
Develop statistically derived QC acceptance criteria that will reflect method performance
capabilities in real-world situations
The design of the multi-laboratory study is described in a formal study plan that is included as Appendix
B to this report. The design is based on the specifications in EPA's Protocol for Review and Validation
of New Methods for Regulated Organic and Inorganic Analytes in Wastewater Under EPA 's Alternate
Test Procedure Program (USEPA 2018). Briefly, the design involved:
At least nine laboratories, with a goal of complete wastewater data from at least six laboratories
Nine wastewater samples from a variety of sources
Determination of retention times for all 209 PCB congeners and labeled analytes
Multi-point calibration of the target analytes
Initial demonstration of capability (IDC) by each laboratory
Determination of method detection limits (MDLs) by each laboratory
Analyses of matrix spike and matrix spike duplicate (MS/MSD) samples prepared from each of the
nine wastewater samples
In addition, the study involved similar analyses of three samples each of sediment, biosolids, and fish
tissue. Because these matrices are not subject to the same requirements for analyses by methods
approved at 40 CFR Part 136 as are wastewater samples, fewer samples and laboratories were required
for those portions of the study.
Method 1628 Multi-Laboratory Validation Study Report
5
April 2021
-------
2. Identification and Selection of Laboratories
Prior to completing the single-laboratory study, potential participants in the anticipated multi-laboratory
study were identified. By March 2018, through a combination of established relationships, review of an
EPA database of laboratory capabilities, internet searches, and telephone calls, over 100 potential
participants were identified, including commercial environmental laboratories, state laboratories, and
utility laboratories. EPA identified potential Regional and emergency response laboratories within the
Agency as well. Additional interest in the study was generated through poster and platform presentations
by EPA at national and regional meetings.
Between June and October 2018, EPA conducted teleconference calls and a webinar with the potential
participants to firm up the list of laboratories. Many of the potential participants decided that they would
not be interested in participating, either because of time and staff constraints, unavailability of suitable
instrumentation for the duration of the study, or for the utility laboratories in particular, an inability to
contract with EPA's support contractor, CSRA, for the study. Ultimately, EPA decided to target a
maximum of 20 participant laboratories, including those contracted and those who could volunteer.
CSRA developed a lengthy contractual statement of work (SOW) covering all aspects of the study, sent a
formal solicitation to 12 commercial laboratories, many with multiple locations, received bids from 9 of
those, and ultimately selected 8 commercial laboratories to receive purchase orders for participation.
EPA arranged for 4 volunteer participants and used the CSRA SOW as the basis for a memorandum of
understanding with each of the volunteer laboratories. The list of all 12 original participants is provided
in Table 1.
Table 1. List of Participating Laboratories
Alpha Analytical Inc.
Mansfield, MA
Southwest Research Institute
San Antonio, TX
Apex Laboratories
Tigard, OR
Week Laboratories
Industry, CA
Agriculture and Priority Pollutant Laboratories
Clovis, CA
Department of Toxic Substances Control
Pasadena, CA
Battelle Memorial Institute
Norwell, MA
US EPA/OSWER OEM/CBRN CMAT
Edison, NJ
Eurofins Lancaster Laboratories
Lancaster, PA
CSS
Castle Rock, CO
SGS North America Inc.
Wilmington, NC
US EPA Region 4
Athens, GA
The primary focus of the study was on the analyses of wastewater samples and all 12 laboratories agreed
to perform those analyses, by either separatory funnel extraction procedures, or solid-phase extraction
procedures. A few laboratories agreed to perform wastewater analyses using both extraction procedures.
The other three matrix types (sediment, biosolids, and fish tissue) were the secondary focus of the study.
Of the 12 laboratories:
8 agreed to analyze sediments by Soxhlet extraction, and 3 of those agreed to also analyze the
sediments by another extraction procedure,
6 agreed to analyze the biosolids by Soxhlet extraction, and 3 of those agreed to also analyze the
biosolids by another extraction procedure,
6 agreed to analyze the fish tissues by Soxhlet extraction, and 2 of those agreed to also analyze the
fish tissues by another extraction procedure,
Method 1628 Multi-Laboratory Validation Study Report
6
April 2021
-------
Immediately prior to the start of the study, EPA held kick-off calls with all of the participating
laboratories to discuss the specifics on the study as a group. In order to accommodate the schedules, two
kick-off calls were held in early February 2019. A summary of both calls, including answers to any
questions raised by the participants, was circulated to all of the laboratories after the second call.
EPA subsequently held biweekly conference calls from February 2019 through October 2019. Because
not all of the participants were able to attend every call, the discussion and any critical points were
summarized and circulated by email to all of the participants after each call.
Unfortunately, despite EPA's efforts, not all twelve laboratories completed the study, or completed all of
the analyses that they originally agreed to perform. In the end, only seven laboratories provided full data
sets for all aspects of the wastewater portion of the study. The other five laboratories provided some data
for the initial start-up phases of the study, but did not analyze any of the actual study samples. These five
laboratories all cited time and resource constraints as the reason for dropping out of the study. None of
the five laboratories dropped out of the study because they were unable to perform the analysis. Where
practical, the data from those five laboratories were considered for use in the study. However, having
data from seven laboratories met EPA's study design goals of acquiring data from at least six laboratories
for the wastewater matrices.
Other than the list of laboratories in Table 1 above, the remainder of this report does not associate specific
results with a named laboratory. Rather, each laboratory that completed any portion of the study was
randomly assigned an identifying number between 1 and 12.
Method 1628 Multi-Laboratory Validation Study Report
7
April 2021
-------
3. Study Samples
Wastewater Matrices
The wastewater samples used in the study were selected to meet the specifications in EPA's new method
protocol (USEPA 2018), namely, that at least one of the wastewater matrix types should have one of the
following characteristics, such that each criterion below is represented by at least one wastewater:
Total suspended solids (TSS) greater than 40 mg/L
Total dissolved solids (TDS) greater than 100 mg/L
Oil and grease greater than 20 mg/L
NaCl greater than 120 mg/L
CaCC>3 greater than 140 mg/L
EPA worked to obtain large volumes of actual wastewaters and sufficient masses of soils/sediments,
biosolids, and fish tissues. EPA contacted three major wastewater treatment operations: Massachusetts
Water Resources Authority (MWRA), Los Angeles Sanitation, and the Hampton Roads Sanitation
District (HRSD) to obtain bulk volumes of wastewater effluents and influents, as well as bulk samples of
aqueous matrices from indirect dischargers to those systems. EPA also worked with the National Council
for Air and Stream Improvement (NCASI) and the Delaware River Basin Commission (DRBC) to obtain
bulk wastewaters representing a pulp and paper discharge and an industrial discharge.
EPA planned to prepare enough samples for up to 20 laboratories to participate in the study, so
approximately 140 liters each of 10 aqueous matrices were collected by those organizations and shipped
to ERA, a commercial preparer of performance testing samples, that was contracted to homogenize and
aliquot the bulk samples into study-specific sizes and distribute them to each laboratory. During transit,
containers of one of the bulk samples developed leaks. As a result, that wastewater matrix did not have
enough volume to be used in the study. Fortunately, the other 9 bulk wastewaters were sufficient in
quantity and characteristics to meet EPA's study design specifications. Table 2 contains a list of the bulk
wastewater samples provided. (Because the numbering scheme for the wastewaters was assigned ahead
of time and the samples were collected and shipped over the course of several weeks, when Matrix 1 did
not provide sufficient volume because of leakage, the original numbering scheme was retained.)
Table 2. Bulk Wastewater Matrix Sources
Industry Type
Wastewater Matrix
Landfill leachate
2
Metal finisher
3
POTW Effluent
4
Hospital
5
POTW Influent
6
Bus washing station
7
Power Plant
8
Pulp and paper effluent
9
POTW Effluent
10
Once received at ERA, the bulk samples were homogenized and tested for the five water quality
characteristics listed above. Three replicate analyses of each wastewater were performed and the mean
result for each of the parameters was calculated. Each of those characteristic specifications were met by
at least one of the nine bulk samples, and EPA deemed the nine samples suitable for use in the study. A
summary of the characteristics is provided in Table 3. The shaded cells in Table 3 indicate that the
sample met the requirements for that parameter. All of the bulk samples met the requirements for TDS
and Conductivity. Four of the bulk samples met the requirements for TSS, and one each met the
requirements for oil and grease, and hardness, respectively.
Method 1628 Multi-Laboratory Validation Study Report
8
April 2021
-------
Table 3. Water Quality Characteristics of the Study Samples (in mg/L)
Wastewater
TSS
TDS
O&G
NaCI, as Conductivity
Hardness, as CaC03
2
168
4564.3
0.6667
7163
330
3
188
3681
1.07
4530
20.4
4
244
403
10.9
838
23.5
5
5.51
384
0.967
777
93.6
6
72
772
3.93
1708
67.5
7
29.0
509
23
688
23.2
8
8.97
143
0.333
256
31.8
9
37
1992
187
2815
205
10
9.69
893
0.0
1839
127
Following EPA's acceptance, the bulk samples were homogenized by placing the entire volume of each
matrix lot in a 200-L bulk tank with a mechanical stirrer and thoroughly mixed for 30 minutes. The
samples were aliquoted into 1-L glass containers and stored at ERA until they were shipped to the
laboratories. Each laboratory received four 1-L bottles of each of the nine wastewaters, providing enough
volume for analyses of the unspiked sample, an MS/MSD pair, and a spare bottle in case of issues in
transit or during preparation at the laboratory. Additional samples were shipped to the laboratories which
had agreed to extract the wastewater samples by both separatory funnel and solid-phase procedures. The
remaining bottles were stored at ERA in case they were needed at a later date.
Biosolids Matrices
Two of the municipal treatment facilities provided bulk quantities of the finished biosolids from their
treatment operations. Two biosolids were wet and one was in the form of dried pellets that are sold as a
soil amendment. The bulk samples were also sent to ERA, where they were homogenized by placing the
entire bulk volume of each matrix lot into a large Pyrex dish and stirred to mix. ERA determined the
percent solids using a portion of the three bulk samples and aliquoted the samples into 2-oz screw cap jars
with the required mass based on the determined dry weight (at least 5 grams dry weight). Biosolid
sample #2 required augmentation with Ottawa sand to meet the volume demand. This sample was re-
homogenized by mixing and % solids was re-performed prior to aliquoting. The jars were stored under
refrigeration at ERA until shipment to the laboratories.
The percent solids data for each bulk sample were provided to the participating laboratories to be used to
report dry-weight concentrations of the PCB congeners. That approach eliminated the need to ship
additional material to each laboratory just for the solids determination, and reduced the variability in the
study results that would have occurred by using the different solids contents determined in each
individual laboratory in the study.
Sediment Matrices
The sediment samples used for the study were prepared from excess archived material maintained by
CSRA. The samples were collected in 2011 as grab samples in the Raisin River in Monroe County,
Michigan, as part of an EPA Great Lakes National Program Office (GLNPO) remedial assessment effort
under the Great Lakes Legacy Act. GLNPO collected samples from a large area, sent some for analysis
immediately after collection and had CSRA store other samples in the event that additional analyses were
required.
CSRA had stored the archived samples in their original 1-gallon self-sealing plastic bags frozen at
approximately -20 °C since 2011. After GLNPO determined that all the archived samples could be sent
for disposal, CSRA retained a small number for possible use in studies such as this one. Aroclor 1242
was reported as present in many of the other samples in the original remedial assessment effort, and
therefore, CSRA proposed to use a number of the samples for this PCB method validation study. Using
Method 1628 Multi-Laboratory Validation Study Report
9
April 2021
-------
the results from the other samples collected near the archived samples, CSRA grouped the available
archived samples into those likely to contain low, medium, and high levels of PCBs.
As with the biosolids, the bulk sediment samples were sent to ERA, where they were processed and
stored as described above. In the case of the presumed low-level sediment sample (sediment #3), EPA
agreed to augment that sample with Ottawa sand to provide a sufficient number of 10-g (dry weight)
aliquots for 20 possible laboratories in the study. As with the biosolids samples, the percent solids data
for each sediment sample were provided to participating laboratories to be used to report dry-weight
concentrations of the PCB congeners.
Fish Tissue Matrices
The fish tissue samples used for the study were prepared from excess archived material maintained by
CSRA. The samples were field duplicate samples collected as part of the National Lake Fish Tissue
Study between 2000 and 2004, homogenized and stored, but never analyzed. CSRA had stored the
archived samples in their original 500-mL glass jars, frozen at approximately -20 °C. As with the
sediment samples, when EPA released these field duplicate samples for disposal, CSRA retained a
selection of the excess jars for other uses.
CSRA selected samples representing three common freshwater species: white sucker, largemouth bass,
and common carp, and used PCB congener results generated for the National Lake Fish Tissue Study
from the samples of these species collected from the same site, to group the samples together by
concentration level. Ultimately:
Four jars of homogenized fillet tissue from white sucker specimens that were collected from four sites
around the U.S. were composited to create a low-level PCB sample,
Four jars of homogenized fillet tissue from largemouth bass specimens that were collected from four
other sites around the U.S. were composited to create a medium-level PCB sample, and
Three jars of homogenized fillet tissue from common carp specimens that were collected from three
sites around the U.S. were composited to create a high-level PCB sample.
These three fish species represent both predator species (large-mouth bass) and bottom-dwelling species
(white sucker and common carp), as well as a range of lipid contents (~1 to 7%).
The jars of fish tissue were sent to ERA for compositing and further homogenization. The individual jars
of the matrix lot were combined into one large glass dish and stirred to mix. Each homogenized study
sample was divided into 10-g (wet weight) aliquots in 2-oz screw-top jars and stored frozen until shipped
to a participating laboratory.
Reconnaissance Analyses
As noted earlier, the study design called for analyses of MS/MSD samples. In order to provide
information to each participant lab about the concentrations at which to spike those MS/MSD pairs of
each sample, EPA sent single aliquots of each study sample to SGS AXYS Analytical, the developer of
the original laboratory SOP used as the basis for the draft method. The results from those analyses were
used by EPA and CSRA to develop spiking concentrations for the MS/MSD aliquots of each of the study
samples. Those results that passed the identification criteria in the draft method are summarized in Table
4 and they became the basis for study-specific instructions that were distributed to each laboratory before
sample analyses began. Additional peaks were present in many of those reconnaissance analyses that met
most, but not all, of the identification criteria. Those peaks are not counted in Table 4, but the results
were used to guide the spiking levels for the congeners involved (i.e., the instructions considered that
those congeners might be present and were candidates for the spiking instructions).
Method 1628 Multi-Laboratory Validation Study Report
10
April 2021
-------
The reconnaissance results of the sediment and tissue samples also confirmed CSRA's characterization of
the components of each of the composite samples to create study samples that contained low, medium,
and high concentration of PCBs. The reconnaissance results were not used to assess performance of the
participant laboratories in this study, nor were they used as "true values" in any way.
Additional aliquots of each of the study samples were sent to another commercial laboratory and analyzed
for Aroclors, using EPA Method 608.3, a GC/ECD method that is currently approved at 40 CFR Part
136.3 for NPDES compliance monitoring of PCBs as Aroclors. The purpose of those analyses was to be
able to contrast the results for samples in which Aroclors were not originally reported with the PCB
congener results from the draft procedure.
The laboratory contracted for the Aroclor analyses had method detection limits (MDLs) for Aroclor 1016
and Aroclor 1260 in aqueous matrices at 4.8 ng/L and 3.9 ng/L, and in solid matrices at 0.41 ng/g and
0.36 ng/g. Those aqueous MDLs are well below the published Method 608.3 Aroclor MDL of 65 ng/L.
Although Method 608.3 does not address solid samples directly, the laboratory's solid MDLs are well
below the value one would obtain by converting the method's aqueous MDL into an estimate of the
detection limit in a solid sample.
Table 4. Reconnaissance Analyses Results for Study Samples
Study Sample
# Peaks
Detected*
Sum of Detected Analyte Concentrations (ng/L or ng/g)
Gross
Characterization
as PCB congeners
as Aroclors
Wastewater 2
40
84.9
0
NA
Wastewater 3
0
0
0
NA
Wastewater 4
0
0
0
NA
Wastewater 5
0
0
0
NA
Wastewater 6
5
2.3
0
NA
Wastewater 7
0
0
0
NA
Wastewater 8
0
0
0
NA
Wastewater 9
0
0
0
NA
Wastewater 10
0
0
0
NA
Sediment 1
135
1208
210
Medium
Sediment 2
92
248
12
Low
Sediment 3
135
1454
460
High
Biosolids 1
86
99.2
25
NA
Biosolids 2
12
6.7
0
NA
Biosolids 3
119
246
110
NA
Tissue 1
46
2.7
0.9
Low
Tissue 2
76
15.4
2.5
Medium
Tissue 3
105
98.6
41
High
*Peaks in the congener analysis that met the identification criteria. Some peaks represent more than one congener.
NA = Not applicable. No information was available with which to characterize the wastewater or biosolids samples prior to the start of the
study.
Method 1628 Multi-Laboratory Validation Study Report
11
April 2021
-------
4. Approaches to Calibration and Quantification
The draft procedure calibrates and quantifies 65 target PCB congeners by one of three different
approaches:
True isotope dilution quantification (ID), whereby the response of the target congener is compared to
the response of its 13Ci2-labeled analog. 23 target congeners are quantified in this way.
Modified isotope dilution (mID), when one or more congeners in the same level of chlorination
(LOC) coelute with a congener that has a 13Ci2-labeled analog. 14 target congeners are quantified in
this way (6 with 13Ci2-labeled analogs and 8 that coelute with one of those 6).
Extracted internal standard quantification (EIS), whereby the response of the target congener (or one
or more congeners in the same level of chlorination that coelute) is compared to the response of the
13Ci2-labeled analog of another congener in the same level of chlorination (LOC) with which it
coelutes. 28 target congeners are quantified in this way.
Of these 65 congeners, 48 are those that EPA chose as high priorities for the procedure because of their
retention times (e.g., first and last eluting congeners in a LOC), prevalence in environmental samples,
high concentrations in Aroclors, and their toxicities (see the introduction to this report for further details).
The other 17 congeners coelute with one of the 48 high priority congeners.
The remaining 144 congeners are quantified indirectly using isotope dilution standards of similar
congeners with the same level of chlorination. The response factor is assumed to be the same as the
reference isotope dilution standard. This approach may produce less accurate results for these congeners
than using any of the three approaches described above, but calibrating all of the congeners would make
the level of effort more burdensome for the laboratories that are the intended users of this procedure.
These congeners were seen less often and/or at lower concentrations in the environmental databases
surveyed and the original Aroclor formulations. For the purposes of this report, EPA is referring to these
144 congeners as the congeners without direct calibration data.
Note: If any of those 144 congeners are a priority to a specific data user, the laboratory is welcome and
encouraged to calibrate additional congeners using any of the three approaches listed above. This
is allowed under the flexibility of the method.
During analysis of samples, the labeled compound is added to the sample before any other processing or
analysis steps and the final result for the target congener is corrected for any losses (or apparent gains) of
the labeled analogue during the entire analytical process. This recovery correction is inherent in the
calculations and is applied to the results for all of the congeners, regardless of the specific calibration and
quantification approach described above.
Table 5 provides the list of the 65 congeners with direct calibration data as well as the approach to
calibration and quantification used for each.
Table 5. Quantification Reference and Calibration Approach for the 65 Congeners with
Direct Calibration Data
Target Congener
LOC
Quantification Reference
Calibration Approach
Isotope Dilution and Modified Isotope Dilution Quantification
PCB-1
Mono
13Ci2-PCB-1
ID
PCB-3
13Ci2-PCB-3
ID
PCB-4+10
Di
13Ci2-PCB-4
mID
PCB-11
13Ci2-PCB-11
ID
PCB-15
13Ci2-PCB-15
ID
Method 1628 Multi-Laboratory Validation Study Report
12
April 2021
-------
Table 5. Quantification Reference and Calibration Approach for the 65 Congeners with
Direct Calibration Data
Target Congener
LOC
Quantification Reference
Calibration Approach
PCB-19
Tri
13Ci2-PCB-19
ID
PCB-28
13Ci2-PCB-28
ID
PCB-37
13Ci2-PCB-37
ID
PCB-52+73
Tetra
13Ci2-PCB-52
mID
PCB-54
13Ci2-PCB-54
ID
PCB-70
13Ci2-PCB-70
ID
PCB-77
13Ci2-PCB-77
ID
PCB-85+120
Penta
13Ci2-PCB-85
mID
PCB-89+90+101
13Ci2-PCB-101
mID
PCB-104
13Ci2-PCB-104
ID
PCB-106+118
13Ci2-PCB-118
mID
PCB-126
13Ci2-PCB-126
ID
PCB-138+163+164
Hexa
13Ci2-PCB-138
mID
PCB-153
13Ci2-PCB-153
ID
PCB-155
13Ci2-PCB-155
ID
PCB-169
13Ci2-PCB-169
ID
PCB-180
Hepta
13Ci2-PCB-180
ID
PCB-188
13Ci2-PCB-188
ID
PCB-189
13Ci2-PCB-189
ID
PCB-202
Octa
13Ci2-PCB-202
ID
PCB-205
13Ci2-PCB-205
ID
PCB-206
Nona
13Ci2-PCB-206
ID
PCB-208
13Ci2-PCB-208
ID
PCB-209
Deca
13Ci2-PCB-209
ID
Extracted Internal Standard Quantification
PCB-5+8
Di
13Ci2-PCB-11
EIS
PCB-18
Tri
13Ci2-PCB-28
EIS
PCB-31
13Ci2-PCB-28
EIS
PCB-41+64
Tetra
13Ci2-PCB-70
EIS
PCB-44
13Ci2-PCB-52
EIS
PCB-66+80
13Ci2-PCB-70
EIS
PCB-61+74
13Ci2-PCB-70
EIS
PCB-93+95
Penta
13Ci2-PCB-101
EIS
PCB-99
13Ci2-PCB-101
EIS
PCB-105+127
13Ci2-PCB-118
EIS
PCB-110
13Ci2-PCB-118
EIS
PCB-132+168
Hexa
13Ci2-PCB-153
EIS
PCB-147
13Ci2-PCB-153
EIS
PCB-139+149
13Ci2-PCB-153
EIS
PCB-156
13Ci2-PCB-153
EIS
PCB-166
13Ci2-PCB-153
EIS
PCB-177
Hepta
13Ci2-PCB-180
EIS
PCB-182+187
13Ci2-PCB-180
EIS
PCB-199
Octa
13Ci2-PCB-202
EIS
ID = Isotope dilution quantitation
mID = Modified isotope dilution quantitation
EIS = Extracted internal standard quantitation
Table 6 provides the lists of the congeners not calibrated directly and the quantification references that are
used to estimate their concentrations in samples.
Method 1628 Multi-Laboratory Validation Study Report
13
April 2021
-------
Table 6. Congeners Not Calibrated Directly
Congener
LOC
Quantification Reference
PCB-2
Mono
PCB-3
PCB-6
PCB-11
PCB-7+9
Di
PCB-11
PCB-12+13
PCB-11
PCB-14
PCB-11
PCB-16+32
PCB-18
PCB-17
PCB-18
PCB-20+21+33
PCB-28+31 (Average)1
PCB-22
PCB-28+31 (Average)1
PCB-23+34
PCB-28+31 (Average)1
PCB-24+27
PCB-28+31 (Average)1
PCB-25
Tri
PCB-28+31 (Average)1
PCB-26
PCB-28+31 (Average)1
PCB-29
PCB-28+31 (Average)1
PCB-30
PCB-28+31 (Average)1
PCB-35
PCB-28+31 (Average)1
PCB-36
PCB-28+31 (Average)1
PCB-38
PCB-28+31 (Average)1
PCB-39
PCB-28+31 (Average)1
PCB-40
PCB-44
PCB-42
PCB-44
PCB-43+49
PCB-44
PCB-45
PCB-44
PCB-46
PCB-44
PCB-47+48+75
PCB-52+73
PCB-50
PCB-52+73
PCB-51
PCB-52+73
PCB-53
PCB-52+73
PCB-55
PCB-70
PCB-56+60
PCB-70
PCB-57
PCB-70
PCB-58
Tetra
PCB-70
PCB-59
PCB-41+64
PCB-62
PCB-41+64
PCB-63
PCB-70
PCB-65
PCB-41+64
PCB-67
PCB-70
PCB-68
PCB-70
PCB-69
PCB-41+64
PCB-71
PCB-41+64
PCB-72
PCB-70
PCB-76
PCB-70
PCB-78
PCB-70
PCB-79
PCB-70
PCB-81
PCB-77
PCB-82
PCB-89+90+101
PCB-83+109
PCB-89+90+101
PCB-84
PCB-93+95
PCB-86+97
PCB-89+90+101
PCB-87+115+116
Penta
PCB-89+90+101
PCB-88+121
PCB-93+95
PCB-91
PCB-93+95
PCB-92
PCB-89+90+101
PCB-94
PCB-93+95
Method 1628 Multi-Laboratory Validation Study Report
14
April 2021
-------
Table 6. Congeners Not Calibrated Directly
Congener
LOC
Quantification Reference
PCB-96
PCB-104
PCB-98+102
PCB-93+95
PCB-100
PCB-93+95
PCB-103
PCB-93+95
PCB-107+108
PCB-106+118
PCB-111 + 117
PCB-110
PCB-112
Penta
PCB-110
PCB-113
PCB-110
PCB-114
PCB-106+118
PCB-119
PCB-110
PCB-122
PCB-106+118
riPCB-123
PCB-106+118
PCB-124
PCB-106+118
PCB-125
PCB-110
PCB-128
PCB-132+168
PCB-129
PCB-132+168
PCB-130
PCB-132+168
PCB-131 + 142
PCB-132+168
PCB-133
PCB-132+168
PCB-134
PCB-132+168
PCB-135+144
PCB-147
PCB-136
PCB-153
PCB-137
PCB-132+168
PCB-140
PCB-147
PCB-141
PCB-132+168
PCB-143
PCB-147
PCB-145
Hexa
PCB-153
PCB-146
PCB-153
PCB-148
PCB-147
PCB-150
PCB-153
PCB-151
PCB-147
PCB-152
PCB-153
PCB-154
PCB-147
PCB-157
PCB-156
PCB-158+160
PCB-166
PCB-159
PCB-156
PCB-161
PCB-166
PCB-162
PCB-156
PCB-165
PCB-153
PCB-167
PCB-156
PCB-170+190
PCB-180
PCB-171
PCB-181
PCB-172+192
PCB-180
PCB-173
PCB-181
PCB-174
PCB-181
PCB-175
PCB-180
PCB-176
Hepta
PCB-188
PCB-178
PCB-181
PCB-179
PCB-188
PCB-181
PCB-180
PCB-183
PCB-180
PCB-184
PCB-188
PCB-185
PCB-181
PCB-186
PCB-188
Method 1628 Multi-Laboratory Validation Study Report
15
April 2021
-------
Table 6. Congeners Not Calibrated Directly
Congener
LOC
Quantification Reference
PCB-191
Hepta
PCB-189
PCB-193
PCB-189
PCB-194
Octa
PCB-199
PCB-195
PCB-199
PCB-196+203
PCB-199
PCB-197
PCB-202
PCB-198
PCB-199
PCB-200
PCB-202
PCB-201
PCB-202
PCB-204
PCB-202
PCB-207
Nona
PCB-208
1 The quantification reference for these 12 congeners is the average of the response ratio for
PCB-28 and the response factor for PCB-31, which are calibrated as individual congeners
by isotope dilution and extracted internal standard, respectively. In contrast, other
congeners in this table use the single response factor of a coeluting pair of congeners that
are calibrated as that pair (e.g., PCB-52+73).
The 29 13Ci2-labeled analogs themselves tire present in each standard at a constant concentration that
reflects the concentration of the label that is added to each sample. All of the 13Ci2-labeled analogs that
are added to the samples before extraction are quantified by the traditional EPA non-extracted internal
standard (NIS) approach, whereby three other labeled compounds (13Ci2-PCB-8,13Ci2-PCB-79, and
13Ci2-PCB-162) are added to each sample extract shortly before GC/MS analysis and the responses of
those three compounds are used to quantify the other 13Ci2-labeled analogs. In some procedures, those
last three labeled compounds may be referred to as "recovery standards" because they are used to
determine the recovery of the other labeled compounds. In Table 7 below, they are referred to as the non-
extracted internal standards.
Multi-point Initial Calibration
The GC/MS instrument was calibrated using a series of six calibration standards designated as CS1 to
CS6. The concentrations of six calibration standards are shown in Table 7 below, along with the
approach used for calibration.
Table 7. Composition of the Initial Calibration Standards
Analyte
Calibration Standards (ng/mL)
Coeluting Congeners
Calibration
Approach
CS-1
CS-2
CS-3
CS-4
CS-5
CS-6
Target Congeners
PCB-1
10
20
40
160
400
2000
ID
PCB-3
10
20
40
160
400
2000
ID
PCB-4
10
20
40
160
400
2000
PCB-101
mID
PCB-8
10
20
40
160
400
2000
PCB-5 2
EIS
PCB-11
10
20
40
160
400
2000
ID
PCB-15
10
20
40
160
400
2000
ID
PCB-18
10
20
40
160
400
2000
EIS
PCB-19
10
20
40
160
400
2000
ID
PCB-28
10
20
40
160
400
2000
ID
PCB-31
10
20
40
160
400
2000
EIS
PCB-37
10
20
40
160
400
2000
ID
PCB-44
10
20
40
160
400
2000
EIS
PCB-52
10
20
40
160
400
2000
PCB-73 1
mID
PCB-54
10
20
40
160
400
2000
ID
PCB-64
10
20
40
160
400
2000
PCB-41 2
EIS
PCB-66
10
20
40
160
400
2000
PCB-80 2
EIS
PCB-70
10
20
40
160
400
2000
ID
Method 1628 Multi-Laboratory Validation Study Report
16
April 2021
-------
Table 7. Composition of the Initial Calibration Standards
Analyte
Calibration Standards (ng/mL)
Coeluting Congeners
Calibration
Approach
CS-1
CS-2
CS-3
CS-4
CS-5
CS-6
PCB-74
10
20
40
160
400
2000
PCB-61 2
EIS
PCB-77
10
20
40
160
400
2000
ID
PCB-85
10
20
40
160
400
2000
PCB-1201
mID
PCB-95
10
20
40
160
400
2000
PCB-93 2
EIS
PCB-99
10
20
40
160
400
2000
EIS
PCB-101
10
20
40
160
400
2000
PCB-89+0 1
mID
PCB-104
10
20
40
160
400
2000
ID
PCB-105
10
20
40
160
400
2000
PCB-127 2
EIS
PCB-110
10
20
40
160
400
2000
EIS
PCB-118
10
20
40
160
400
2000
PCB-106 1
mID
PCB-126
10
20
40
160
400
2000
ID
PCB-132
10
20
40
160
400
2000
PCB-168 2
EIS
PCB-138
10
20
40
160
400
2000
PCB-163+164 1
mID
PCB-147
10
20
40
160
400
2000
EIS
PCB-149
10
20
40
160
400
2000
PCB-139 2
EIS
PCB-153
10
20
40
160
400
2000
ID
PCB-155
10
20
40
160
400
2000
ID
PCB-156
10
20
40
160
400
2000
EIS
PCB-166
10
20
40
160
400
2000
EIS
PCB-169
10
20
40
160
400
2000
ID
PCB-177
10
20
40
160
400
2000
EIS
PCB-180
10
20
40
160
400
2000
ID
PCB-187
10
20
40
160
400
2000
PCB-182 2
EIS
PCB-188
10
20
40
160
400
2000
ID
PCB-189
10
20
40
160
400
2000
ID
PCB-199
10
20
40
160
400
2000
EIS
PCB-202
10
20
40
160
400
2000
ID
PCB-205
10
20
40
160
400
2000
ID
PCB-206
10
20
40
160
400
2000
ID
PCB-208
10
20
40
160
400
2000
ID
PCB-209
10
20
40
160
400
2000
ID
Labeled Congeners
13Ci2-PCB-1
400
400
400
400
400
400
NIS
13Ci2-PCB-3
400
400
400
400
400
400
NIS
13Ci2-PCB-4
400
400
400
400
400
400
NIS
13Ci2-PCB-11
400
400
400
400
400
400
NIS
13Ci2-PCB-15
400
400
400
400
400
400
NIS
13Ci2 PCB-19
400
400
400
400
400
400
NIS
13Ci2-PCB-28
400
400
400
400
400
400
NIS
13Ci2-PCB-37
400
400
400
400
400
400
NIS
13Ci2-PCB-52
400
400
400
400
400
400
NIS
13Ci2-PCB-54
400
400
400
400
400
400
NIS
13Ci2-PCB-70
400
400
400
400
400
400
NIS
13Ci2-PCB-77
400
400
400
400
400
400
NIS
13Ci2-PCB-85
400
400
400
400
400
400
NIS
13Ci2-PCB-101
400
400
400
400
400
400
NIS
13Ci2-PCB-104
400
400
400
400
400
400
NIS
13Ci2-PCB-118
400
400
400
400
400
400
NIS
13Ci2-PCB-126
400
400
400
400
400
400
NIS
13Ci2-PCB-138
400
400
400
400
400
400
NIS
13Ci2-PCB-153
400
400
400
400
400
400
NIS
13Ci2-PCB-155
400
400
400
400
400
400
NIS
Method 1628 Multi-Laboratory Validation Study Report
17
April 2021
-------
Table 7. Composition of the Initial Calibration Standards
Analyte
Calibration Standards (ng/mL)
Coeluting Congeners
Calibration
Approach
CS-1
CS-2
CS-3
CS-4
CS-5
CS-6
13Ci2-PCB-169
400
400
400
400
400
400
NIS
13Ci2-PCB-180
400
400
400
400
400
400
NIS
13Ci2-PCB-188
400
400
400
400
400
400
NIS
13Ci2-PCB-189
400
400
400
400
400
400
NIS
13Ci2-PCB-202
400
400
400
400
400
400
NIS
13Ci2-PCB-205
400
400
400
400
400
400
NIS
13Ci2-PCB-206
400
400
400
400
400
400
NIS
13Ci2-PCB-208
400
400
400
400
400
400
NIS
13Ci2-PCB-209
400
400
400
400
400
400
NIS
Non-extracted Internal Standards
13Ci2-PCB-8
400
400
400
400
400
400
NA
13Ci2-PCB-79
400
400
400
400
400
400
NA
13Ci2-PCB-162
400
400
400
400
400
400
NA
1 These coeluting congeners are not included in the calibration standard, but the responses in the samples for all of the congeners that elute
together at a given retention time are quantified by modified isotope dilution, based on the response ratio derived for the single congener in
the calibration standard and its corresponding labeled analogue.
2 These coeluting congeners are not included in the calibration standard, but the responses in the samples for all of the congeners that elute
together at a given retention time are quantified by extracted internal standard, based on the response factor derived for the single congener in
the calibration standard and the labeled analogue for another congener in the same level of chlorination.
ID = Isotope dilution quantitation
mID = Modified isotope dilution quantitation
EIS = Extracted internal standard quantitation
NIS = Non-extracted internal standard quantitation
NA = Not applicable - these congeners are not quantified
For the purpose of the multi-laboratory study, EPA procured full sets of all of the standards employed in
the method from commercial vendors. These standards were ordered as custom mixtures of the six
calibration standards (CS1 to CS6), as well as a spiking solution of the 65 native congeners of primary
interest, the non-extracted internal standard solution, and a series of commercially available mixtures of
all 209 native PCB congeners that was used to establish the retention time of each congener in each
laboratory.
By providing these standards to all of the laboratories, EPA reduced the variability in the study results
that would have resulted from having each laboratory prepare all of the standards from neat materials.
This approach also reduced the direct costs to each laboratory for their participation, allowing more
laboratories to participate. It also expanded the pool of potential participants because not all commercial
laboratories are willing or able to prepare standards from neat materials.
EPA anticipates that if this method comes into routine use, the vendor community will continue to
provide these method-specific standards as routine commercial products, as they do now for many other
EPA monitoring methods.
Response Ratios and Response Factors
The response ratio (RR) for each congener calibrated by isotope dilution is calculated according to the
equation below, separately for each of the calibration standards, using the areas of the ions with the mass-
to-charge ratios (m/z) shown in Table 8.
RR= AreanCl
AreaiCn
where:
Arean = The measured area of the primary m/z for the native (unlabeled) PCB
Areai = The measured area at the primary m/z for the labeled PCB
Method 1628 Multi-Laboratory Validation Study Report
18
April 2021
-------
Ci
c
= The concentration of the labeled compound in the calibration standard
= The concentration of the native compound in the calibration standard
This response ratio is used for the 23 congeners quantified by true isotope dilution and the 14 congeners
quantified by modified isotope dilution.
Similarly, the response factor (RF) for each unlabeled congener calibrated by extracted internal standard
is calculated according to the equation below.
Areas CejS
RF =
Areaeis Cs
where:
Areas = The measured area of the primary m/z for the target (unlabeled) PCB
Aries = The measured area at the primary m/z for the labeled PCB used as the extracted
internal standard
Cris = The concentration of the labeled compound used as the extracted internal standard in
the calibration standard
Cs = The concentration of the target compound in the calibration standard
This response factor is used for the 28 congeners quantified by extracted internal standard.
Similarly, a response factor is calculated for each labeled compound added before extraction using the
following equation:
Areat Cnis
RF =
Areanis Ct
where:
Areai = The measured area of the primary m/z for the labeled PCB standard added to the
sample before extraction
Arenas = The measured area at the primary m/z for the labeled PCB used as the non-extracted
internal standard
Cms = The concentration of the labeled compound used as the non-extracted internal standard
in the calibration standard
Ci = The concentration of the labeled PCB standard added to the sample before extraction
This response factor is used for the 29 labeled congeners quantified by non-extracted internal standard.
Mass-to-charge Ratios Monitored for Each Analyte
The equations above for the response ratio and the response factor are based on the area of the more
intense of the two characteristic ions produced by each congener under the mass spectrometer electron
impact (EI) operating conditions described in the draft method. For the purposes of this method, the
"primary ion" is the ion with the most intense response and the "confirmation ion" is the next most
intense response. For all of the analytes, the mass difference between the two m/zs is 2 Daltons and
represents the presence or absence of an atom of the less common isotope 37C1 in the m/z versus the more
common 35C1 isotope. For some congeners, the higher m/z is the primary ion, but in most cases, the lower
m/z provides the most intense response.
The draft method employs "single-ion quantitation," whereby the area response of the primary m/z for
each analyte is used to calculate a response ratio (RR) or response factor (RF) for each calibration
standard. The response of the confirmation m/z is not used to determine the RR or RF value, or to
quantify the analyte in a sample. However, the confirmation ion must be present and the ratio of the
abundance of the primary m/z to the confirmation m/z must meet an acceptance limit centered around the
theoretical abundance of all of the atoms making up the analyte in order to confirm the identification of
the analyte.
Method 1628 Multi-Laboratory Validation Study Report
19
April 2021
-------
Table 8 presents the ions monitored for the 65 congeners with direct calibration data and all of the labeled
compounds, while Table 9 presents similar data for the 144 other congeners that are analyzed by the draft
method using the indirect calibration approach described above.
Table 8. Ions Monitored for each Congener with Direct Calibration Data
and the Labeled Compounds, in Retention Time Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
13C12- PCB-8 (NIS)4
234
236
65.6
13Ci2- PCB-1
200
202
33.2
PCB-1
188
190
33.2
13Ci2- PCB-3
200
202
33.2
PCB-3
188
190
33.2
13Ci2- PCB-4
234
236
65.6
PCB-4+10
222
224
65.6
13Ci2-PCB-11
234
236
65.6
PCB-8+5
222
224
65.6
PCB-11
222
224
65.6
13Ci2-PCB-15
234
236
65.6
PCB-15
222
224
65.6
13Ci2-PCB-19
268
270
98.0
PCB-19
256
258
98.0
13Ci2-PCB-28
268
270
98.0
PCB-18
256
258
98.0
PCB-31
256
258
98.0
PCB-28
256
258
98.0
13Ci2-PCB-37
268
270
98.0
PCB-37
256
258
98.0
13Ci2-PCB-79 (NIS)4
304
302
76.7
13Ci2-PCB-54
304
302
76.7
PCB-54
292
290
76.7
13Ci2-PCB-52
304
302
76.7
PCB-18
292
290
76.7
PCB-52+73
292
290
76.7
PCB-44
292
290
76.7
13Ci2-PCB-70
304
302
76.7
PCB-41+64
292
290
76.7
PCB-74+61
292
290
76.7
PCB-70
292
290
76.7
PCB-66+80
292
290
76.7
13Ci2-PCB-77
304
302
76.7
PCB-77
292
290
76.7
13Ci2-PCB-162 (NIS)4
372
374
81.5
13Ci2-PCB-104
338
340
65.3
PCB-104
326
328
65.3
13Ci2-PCB-101
338
340
65.3
PCB-95+93
326
328
65.3
PCB-90+101+89
326
328
65.3
PCB-99
326
328
65.3
13Ci2-PCB-118
338
340
65.3
PCB-110
326
328
65.3
PCB-118+106
326
328
65.3
PCB-105+127
326
328
65.3
13Ci2-PCB-85
338
340
65.3
PCB-85+120
326
328
65.3
13Ci2-PCB-126
338
340
65.3
PCB-126
326
328
65.3
Method 1628 Multi-Laboratory Validation Study Report
20
April 2021
-------
Table 8. Ions Monitored for each Congener with Direct Calibration Data
and the Labeled Compounds, in Retention Time Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
13Ci2-PCB-155
372
374
81.5
PCB-155
360
362
81.5
13Ci2-PCB-153
372
374
81.5
PCB-147
360
362
81.5
PCB-149+139
360
362
81.5
PCB-153
360
362
81.5
PCB-132+168
360
362
81.5
PCB-166
360
362
81.5
PCB-156
360
362
81.5
13Ci2-PCB-138
372
374
81.5
PCB-138+163+164
360
362
81.5
13Ci2-PCB-169
372
374
81.5
PCB-169
360
362
81.5
13Ci2-PCB-188
406
408
97.7
PCB-188
394
396
97.7
13Ci2-PCB-180
406
408
97.7
PCB-187+182
394
396
97.7
PCB-177
394
396
97.7
PCB-180
394
396
97.7
13Ci2-PCB-189
406
408
97.7
PCB-189
394
396
97.7
13Ci2-PCB-202
442
440
87.8
PCB-202
430
428
87.8
PCB-199
430
428
87.8
13Ci2-PCB-205
442
440
87.8
PCB-205
430
428
87.8
13Ci2-PCB-208
476
474
76.9
PCB-208
464
462
76.9
13Ci2-PCB-206
476
474
76.9
PCB-206
464
462
76.9
13Ci2-PCB-209
510
512
86.7
PCB-209
498
500
86.7
1 The primary ion is the more intense ion of the two ions monitored for each analyte. Its area is used in
calculating the RR or RF values and for calculating the concentration of the analyte in samples.
2 The confirmation ion is the less intense ion of the two ions monitored for each analyte. Its area is
not used in the calculation of RR or RF values, or calculating the concentration of the analyte in
samples. However, it is used as part of the qualitative identification criteria for demonstrating that
the analyte is present.
3 The expected ratio is the area of the confirmation ion divided by the area of the primary ion. All
values are shown in percent and are less than 100%, indicating that the primary ion has the more
intense response.
4 Labeled congeners 8, 79, and 162 are added to the final extract immediately before injection and their
responses are used to quantify the other labeled compounds added to the sample prior to extraction.
These are termed "non-extracted internal standard" or "NIS"
Table 9. Ions Monitored for the 144 Other Congeners, in Retention Time
Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
PCB-2
188
190
33.2
PCB-7+9
222
224
65.6
PCB-6
222
224
65.6
PCB-14
222
224
65.6
PCB-12+13
222
224
65.6
PCB-30
256
258
98.0
PCB-17
256
258
98.0
PCB-24+27
256
258
98.0
Method 1628 Multi-Laboratory Validation Study Report
21
April 2021
-------
Table 9. Ions Monitored for the 144 Other Congeners, in Retention Time
Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
PCB-16+32
256
258
98.0
PCB-34+23
256
258
98.0
PCB-29
256
258
98.0
PCB-26
256
258
98.0
PCB-25
256
258
98.0
PCB-33+20+21
256
258
98.0
PCB-22
256
258
98.0
PCB-36
256
258
98.0
PCB-39
256
258
98.0
PCB-38
256
258
98.0
PCB-35
256
258
98.0
PCB-50
292
290
76.7
PCB-51
292
290
76.7
PCB-45
292
290
76.7
PCB-46
292
290
76.7
PCB-49+43
292
290
76.7
PCB-47+48+75
292
290
76.7
PCB-42
292
290
76.7
PCB-40
292
290
76.7
PCB-69
292
290
76.7
PCB-65
292
290
76.7
PCB-62
292
290
76.7
PCB-59
292
290
76.7
PCB-72
292
290
76.7
PCB-71
292
290
76.7
PCB-68
292
290
76.7
PCB-57
292
290
76.7
PCB-67
292
290
76.7
PCB-58
292
290
76.7
PCB-63
292
290
76.7
PCB-76
292
290
76.7
PCB-55
292
290
76.7
PCB-56+60
292
290
76.7
PCB-79
292
290
76.7
PCB-78
292
290
76.7
PCB-81
292
290
76.7
PCB-96
326
328
65.3
PCB-103
326
328
65.3
PCB-100
326
328
65.3
PCB-94
326
328
65.3
PCB-98+102
326
328
65.3
PCB-88+121
326
328
65.3
PCB-91
326
328
65.3
PCB-92
326
328
65.3
PCB-84
326
328
65.3
PCB-83+109
326
328
65.3
PCB-97+86
326
328
65.3
PCB-87+115+116
326
328
65.3
PCB-82
326
328
65.3
PCB-113
326
328
65.3
PCB-119
326
328
65.3
PCB-112
326
328
65.3
PCB-125
326
328
65.3
PCB-111 + 117
326
328
65.3
Method 1628 Multi-Laboratory Validation Study Report
22
April 2021
-------
Table 9. Ions Monitored for the 144 Other Congeners, in Retention Time
Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
PCB-124
326
328
65.3
PCB-107+108
326
328
65.3
PCB-123
326
328
65.3
PCB-114
326
328
65.3
PCB-122
326
328
65.3
PCB-150
360
362
81.5
PCB-152
360
362
81.5
PCB-145
360
362
81.5
PCB-148
360
362
81.5
PCB-136
360
362
81.5
PCB-154
360
362
81.5
PCB-151
360
362
81.5
PCB-144+135
360
362
81.5
PCB-140
360
362
81.5
PCB-143
360
362
81.5
PCB-134
360
362
81.5
PCB-133
360
362
81.5
PCB-131 + 142
360
362
81.5
PCB-165
360
362
81.5
PCB-146
360
362
81.5
PCB-161
360
362
81.5
PCB-141
360
362
81.5
PCB-137
360
362
81.5
PCB-130
360
362
81.5
PCB-158+160
360
362
81.5
PCB-129
360
362
81.5
PCB-166
360
362
81.5
PCB-159
360
362
81.5
PCB-162
360
362
81.5
PCB-128
360
362
81.5
PCB-167
360
362
81.5
PCB-156
360
362
81.5
PCB-157
360
362
81.5
PCB-184
394
396
97.7
PCB-179
394
396
97.7
PCB-176
394
396
97.7
PCB-186
394
396
97.7
PCB-178
394
396
97.7
PCB-175
394
396
97.7
PCB-183
394
396
97.7
PCB-185
394
396
97.7
PCB-174
394
396
97.7
PCB-181
394
396
97.7
PCB-171
394
396
97.7
PCB-173
394
396
97.7
PCB-172+192
394
396
97.7
PCB-170+190
394
396
97.7
PCB-193
394
396
97.7
PCB-191
394
396
97.7
PCB-201
430
428
87.8
PCB-204
430
428
87.8
PCB-197
430
428
87.8
PCB-200
430
428
87.8
PCB-198
430
428
87.8
Method 1628 Multi-Laboratory Validation Study Report
23
April 2021
-------
Table 9. Ions Monitored for the 144 Other Congeners, in Retention Time
Order
Congener
Primary1
Confirmation2
Expected Ratio (%)3
PCB-196+203
430
428
87.8
PCB-195
430
428
87.8
PCB-194
430
428
87.8
PCB-207
464
462
76.9
1 The primary ion is the more intense ion of the two ions monitored for each analyte. Its area is
used in the calculating the RR or RF values and for calculating the concentration of the analyte
in samples.
2 The confirmation ion is the less intense ion of the two ions monitored for each analyte. Its area
is not used in the calculation of RR or RF values, or calculating the concentration of the analyte
in samples. However, it is used as part of the qualitative identification criteria for demonstrating
that the analyte is present.
3 The expected ratio is the area of the confirmation ion divided by the area of the primary ion. All
values are shown in percent and are less than 100%, indicating that the primary ion has the more
intense response.
Area Subtraction of Higher Homolog Interference
Six congeners: PCB-35, PCB-77, PCB-81, PCB-123, PCB-126, and PCB-157, coelute with congeners
from higher homologues when using a DB-5 capillary column and the AXYS SOP GC instrument
parameters. Those higher homologue congeners can lose one or two chlorines during mass
fragmentation, producing the same ions that are used as the primary quantification ion for one of these
five congeners. This results in an artificial increase for the areas of the quantification ions of the six
congeners. The quantification ion areas for PCB-35, PCB-77, PCB-81, PCB-123, PCB-126, and PCB-
157 are corrected by multiplying the area of the quantification ion (Ql) of the higher (interfering)
homologue by an experimentally determined correction factor (see Table 10) and subtracting the product
from the area of Q1 of the co-eluting lower homologue congener. The recalculation is performed by the
quantification software. Therefore, the areas provided in the raw data (quantitation report) are post-
corrected areas. Of the six congeners exhibiting interferences, only PCB-77 and PCB-126 are congeners
with direct calibration data.
Table 10. Ions Monitored for Correcting for Interferences from Higher Homologues
Congener
Q1 of Congener
HH Interference
Q1 of HH interference
Correction Factor
PCB-35
256
PCB-104
326
0.4971
PCB-81
292
PCB-87+115+116
326
0.0141
PCB-77
292
PCB-110
326
0.0567
PCB-123
326
PCB149+139
360
0.0460
PCB-126
326
PCB-178
394
0.5122
PCB-157
360
PCB-201
430
0.451
Note: The ions listed as the Ql for the higher homologue interference are not necessarily those that
produce the interference with the five congeners, but they are the ions used for the correction
calculation. For example, PCB-110 fragments to produce an ion that is the same mass as the
primary quantitation ion for PCB-77. The difference between the quantitation ions for the two
congeners is 34 Daltons (326 minus 292). Given that chlorine has a mass of 35 Daltons, the
interfering peak probably did not originate from the quantification ion peak for PCB-110 at 326
Daltons, but from another mass fragment produced by PCB-110. The PCB-77 interference is
known to be caused by PCB-110, because the interference is seen when PCB-110 is spiked into a
reference solution.
Calibration Linearity and Stability
One of the tasks for each laboratory during the earliest portion of the study was to perform an initial
calibration of their instrument using the six calibration standards provided by EPA for the study. All 12
of the original laboratories submitted initial calibration data during that early phase. The 7 laboratories
Method 1628 Multi-Laboratory Validation Study Report
24
April 2021
-------
that completed the study performed another 20 initial calibrations during the course of their analyses of
the actual study samples. The results of those 32 calibrations of the target congeners (over 8,900
observations) are summarized in Table 11, in terms of the mean response ratio (RR), mean response
factor (RF), and the relative standard deviation (RSD) of the RR or RF values within each calibration.
For example, the low Mean RR of 0.799 for PCB-1 was the lowest mean RR from all 32 calibrations.
Table 11. Summary of Response Ratios, Response Factors, and Relative
Standard Deviations for 32 Initial Calibrations
Target Congener
LOC
Quantification
Reference
Mean RR or RF*
RSD (%)*
Low
High
Low
High
Isotope Dilution and Modified Isotope Dilution Quantification
PCB-1
Mono
13Ci2-PCB-1
0.799
1.302
0.5
7.2
PCB-3
13Ci2-PCB-3
0.781
1.300
0.6
5.5
PCB-4
Di
13Ci2-PCB-4
0.762
1.208
0.3
7.1
PCB-11
13Ci2-PCB-11
0.794
1.230
0.3
5.8
PCB-15
13Ci2-PCB-15
0.768
1.203
0.3
6.8
PCB-19
Tri
13Ci2-PCB-19
0.775
1.223
0.7
5.3
PCB-28
13Ci2-PCB-28
0.503
1.272
1.0
19.7
PCB-37
13Ci2-PCB-37
0.851
1.234
0.8
5.4
PCB-52
Tetra
13Ci2-PCB-52
0.797
1.178
0.5
5.0
PCB-54
13Ci2-PCB-54
0.823
1.250
0.3
5.7
PCB-70
13Ci2-PCB-70
0.850
1.190
0.5
26.0
PCB-77
13Ci2-PCB-77
0.893
1.215
0.3
5.7
PCB-85
Penta
13Ci2-PCB-85
0.772
1.092
0.4
6.7
PCB-101
13Ci2-PCB-101
0.852
1.135
0.6
7.4
PCB-104
13Ci2-PCB-104
0.805
1.139
0.3
4.1
PCB-118
13Ci2-PCB-118
0.860
1.624
0.6
6.4
PCB-126
13Ci2-PCB-126
0.869
1.133
0.5
5.8
PCB-138
Hexa
13Ci2-PCB-138
0.744
1.116
0.6
6.2
PCB-153
13Ci2-PCB-153
0.881
1.187
0.4
11.8
PCB-155
13Ci2-PCB-155
0.720
1.069
0.3
4.6
PCB-169
13Ci2-PCB-169
0.705
1.210
0.6
9.2
PCB-180
Hepta
13Ci2-PCB-180
0.698
1.149
0.4
8.1
PCB-188
13Ci2-PCB-188
0.715
1.139
0.3
11.1
PCB-189
13Ci2-PCB-189
0.696
1.227
0.5
9.3
PCB-202
Octa
13Ci2-PCB-202
0.636
1.173
0.3
9.6
PCB-205
13Ci2-PCB-205
0.831
1.337
0.8
27.0
PCB-206
Nona
13Ci2-PCB-206
0.496
1.208
0.5
28.4
PCB-208
13Ci2-PCB-208
0.512
1.297
0.5
19.8
PCB-209
Deca
13Ci2-PCB-209
0.207
1.156
0.6
18.5
Extracted Internal Standard Quantification
PCB-8
Di
13Ci2-PCB-11
0.722
1.218
2.3
10.7
PCB-18
Tri
13Ci2-PCB-28
0.437
0.736
0.7
7.6
PCB-31
13Ci2-PCB-28
0.716
1.399
0.8
16.1
PCB-41
Tetra
13Ci2-PCB-70
0.741
1.095
0.4
11.4
PCB-44
13Ci2-PCB-52
0.641
1.008
1.0
7.8
PCB-66
13Ci2-PCB-70
0.807
1.179
0.6
7.0
PCB-74
13Ci2-PCB-70
0.739
1.139
1.0
7.1
PCB-95
Penta
13Ci2-PCB-101
0.746
1.037
1.2
11.7
PCB-99
13Ci2-PCB-101
0.830
1.174
1.1
7.6
PCB-105
13Ci2-PCB-118
0.787
1.558
1.4
8.2
PCB-110
13Ci2-PCB-118
0.816
1.540
0.9
6.8
PCB-132
Hexa
13Ci2-PCB-153
0.424
0.912
1.1
21.6
PCB-147
13Ci2-PCB-153
0.636
0.876
1.4
10.9
PCB-149
13Ci2-PCB-153
0.672
0.938
1.3
12.1
PCB-156
Hexa
13Ci2-PCB-153
0.722
1.590
2.5
25.7
Method 1628 Multi-Laboratory Validation Study Report
25
April 2021
-------
Table 11. Summary of Response Ratios, Response Factors, and Relative
Standard Deviations for 32 Initial Calibrations
Target Congener
LOC
Quantification
Reference
Mean RR or RF*
RSD (%)*
Low
High
Low
High
PCB-166
13Ci2-PCB-153
0.705
1.223
1.9
15.0
PCB-177
Hepta
13Ci2-PCB-180
0.598
0.894
1.5
10.9
PCB-187
13Ci2-PCB-180
0.681
1.200
0.7
14.8
PCB-199
Octa
13Ci2-PCB-202
0.434
1.023
0.0
20.2
* The mean RR, mean RF, and RSD values are those calculated within a single initial calibration, and not across
calibrations, nor across laboratories.
The RR and RF values varied across the 12 laboratories, but within a given calibration at each laboratory,
the instrument responses were generally quite consistent. The mean RR values within each calibration
ranged from roughly 0.750 to 1.250 for most of the congeners quantified by isotope dilution. Four of
those congeners had mean RR values in a calibration that ranged much lower than 0.750 (PCBs 28, 206,
208, and 209), but those lower RR values tended to occur consistently in several calibrations from a given
laboratory for each congener, suggesting that those low values are not a pervasive concern and not issue
of a random variation in the response in a single standard among the six calibration points. This
observation is supported by the fact that the RSD values for those congeners in the laboratories with the
lower than expected response ratios are not noticeably different from the RSDs for other congeners in that
calibration, nor from the RSDs for those congeners in other laboratories. Whatever may be responsible
for the lower response ratios for those four congeners in certain calibrations, it is occurring consistently
across all six standards, such that the calibration still meets the linearity criterion in the draft procedure.
The ranges of mean RF values tend to have lower upper limits than the ranges of the mean RR values,
which is expected, because these congeners are not calibrated using isotope dilution, where the native
analyte and its label have identical structures and fragmentation patterns.
Table 11 also contains the range of RSD values for all 32 calibrations, which are a measure of the
variability in the actual RR or RF values for the analyte in each initial calibration. The RSD is used as a
metric of linearity and assumes that the calibration relationship can be represented by a straight line that
runs through the origin. EPA methods that employ the RSD as a linearity metric generally specify QC
limits on the order of 15% to 25%. The lower RSD values in Table 11 are all below 2%, with two
exceptions at 2.3% and 2.5%. The upper RSD values are below 20%, with six exceptions that range as
high as 28.4%. Four of those six exceptions are for congeners that are quantified by the extracted internal
standard approach, while the other two are quantified by isotope dilution. For two of the six exceptions,
the high RSD value was driven by the RR or RF value in either the lowest or highest of the six calibration
standards. In both of those cases, the laboratory could have dropped the offending standard from the
calibration, met a 20% criterion, and adjusted their calibration range accordingly.
Overall, the study data demonstrate that calibration standards specified in the draft procedure exhibit
excellent linearity for the target analytes and do not require the use of more involved calibration models
such as a linear regression that does not pass through the origin, or a quadratic equation. Moreover, the
commonly used linearity metric of RSD < 20% is appropriate for the target analytes in this procedure.
A similar examination of the calibration data was performed for the 29 labeled compounds. The mean RF
values ranged from 0.145 to 1.933 across all of the calibrations. Across all 32 calibration, the RSDs for
the labeled compounds in each laboratory were below 20%, with one exception at 34.8% for 13Ci2-PCB-
209 in the only calibration performed by one of the volunteer laboratories that did not complete the study.
The low RSD values are not unexpected, because the labeled compounds are present in the calibration
standards at a single concentration (400 ng/mL) across all six solutions (see Table 7).
Method 1628 Multi-Laboratory Validation Study Report
26
April 2021
-------
5. Initial Precision and Recovery
EPA required that each laboratory perform initial precision and recovery (IPR) studies in each of the
matrix types that they agreed to analyze: aqueous, solids (e.g., sediment and biosolids), and tissue. For
each IPR study, four aliquots of a clean reference matrix were spiked with the 65 PCB congeners of
primary interest for the method. The reference matrices and aliquot sizes were:
1 liter of reagent water for water matrices
10 grams of clean sand for solid matrices
10 grams of a 10:90 w/w mixture of canola oil and Ottawa sand
The mass of the 48 native PCB congeners added to the IPR study samples was 16 ng per sample and was
equivalent to the on-column concentration of the CS-4 standard in the 6-point initial calibration range.
The native compound spiking solution described in the method and provided to the laboratories by EPA
was used for spiking. For the aqueous IPR samples, the concentration of each congener was 16 ng/L. For
the solid and tissue IPR samples, the concentration of each congener was 1.6 ng/g. The labeled
compounds were spiked into the IPR aliquots separately, at the level used for all sample analyses, 40 ng
of each labeled compound in each sample.
For each set of four IPR aliquots, each laboratory calculated and reported the mean concentration and
mean recovery, the standard deviation of the recoveries of each target analyte, and the relative standard
deviations (RSDs) of the recoveries.
Each laboratory also prepared a single ongoing precision and recovery (OPR) sample with each batch of
study samples prepared and analyzed. The OPR aliquots were spiked at the same concentrations as the
IPR aliquots. Each laboratory calculated and reported the recovery of the spiked analytes in each OPR
aliquot.
The IPR and OPR results from the laboratories were used to calculate quality control (QC) acceptance
criteria for target as well as labeled compounds, using the statistical procedures described in the study
plan and EPA's new method protocol (USEPA 2018). Separate QC acceptance criteria were calculated
for the aqueous, solid, and tissue matrix types. Those criteria are presented in Tables 12 through 17
below.
One of the EPA Office of Water's objectives in conducting a multi-laboratory validation study is to
generate data from which the Office of Water can derive multi-laboratory QC acceptance criteria for the
various performance tests in the method, or to evaluate the ability of the method to meet commonly
applied acceptance criteria for some performance tests. In this study, the Office of Water calculated QC
acceptance criteria for initial precision and recovery (IPR) tests, ongoing precision and recovery (OPR)
tests, and labeled compound recoveries. The derivations of those limits were based on the processes and
equations in Appendix G of the Protocol for Review and Validation of New Methods for Regulated
Organic and Inorganic Analytes in Wastewater Under EPA's Alternate Test Procedure Program
(USEPA 2018), with modifications to account for the actual number of laboratories, samples and
replicates in the study. To yield a more complete dataset, IPR and OPR data were combined when
calculating the criteria, using different formulas to generate IPR- and OPR-specific criteria that account
for how they would be evaluated in practice (i.e., mean and RSD of four IPR replicates, and OPRs
evaluated on an individual basis). Labeled congener recovery in samples was calculated by combining
recoveries for the unspiked and spiked samples for the given matrix, to ensure that within-laboratory
variability could be distinguished from between-sample variability for each laboratory. Briefly:
The QC acceptance criteria for recovery in the IPR test is calculated by constructing a prediction
interval around the mean percent recovery, using a Student's t value, with the degrees of freedom
Method 1628 Multi-Laboratory Validation Study Report
27
April 2021
-------
determined using the Satterthwaite estimation procedure (Satterthwaite, 1946), using the between-
and within-laboratory variance components calculated for that congener, weighted based on future
IPR usage (assuming means of four replicates per laboratory).
The maximum acceptable RSD for the four IPR aliquots is calculated by an upper confidence limit
around the observed RSD of the results from all of the laboratories. The RSDipr (computed as s
divided by X) is multiplied by the square root of a 95th percentile F value with 3 degrees of freedom
in the numerator and n j-m degrees of freedom in the denominator, where m = the number of
laboratories, and nT is the number of data points across all laboratories for the given congener.
The QC acceptance criteria for recovery in the OPR test is calculated by constructing a prediction
interval around the mean percent recovery, using a Student's t value, with the degrees of freedom
determined using the Satterthwaite estimation procedure, using the between- and within-laboratory
variance components calculated for that congener, weighted based on future OPR usage (assuming a
single replicate per laboratory).
The QC acceptance criteria for labeled sample recovery is calculated by constructing a prediction
interval around the mean percent recovery, using a Student's t value with the degrees of freedom
determined using the Satterthwaite estimation procedure, using the between-laboratory, between-
sample, and within-laboratory variance components calculated for that congener, weighted based on
future criterion usage (assuming a single replicate per laboratory).
Generally, these criteria would be calculated at the 95% confidence level (a 95th percentile /'-statistic for
the one-sided RSD upper bound, and a 97.5th percentile ^-statistic for the two-sided IPR and OPR
recovery bounds). This means that a laboratory performing the method properly would be assumed to
have a 5% probability of failing that criterion merely due to chance. However, the probability would be
5% for each of the 48 individual target and 29 labeled PCB congeners being evaluated, and as a result, the
probability of at least one of those congeners failing just by chance would be much higher. To ensure an
overall 5% probability of any of the congeners failing the criteria, each congener criterion was calculated
using t- and /'-statistics with much more stringent confidence levels (using 99.9th percentile /'-statistics
and 99.95th percentile ^-statistics).
Given the multi-analyte correction and because these criteria are designed to assess performance in
multiple laboratories, the calculations often result in acceptance limits that are fairly wide. Historically,
the Office of Water has been willing to accept the fact that limits derived from multi-laboratories studies
are wider than those that would be derived in a single-laboratory study, or from long-term data within any
given laboratory. However, in this study, many of the calculated lower limits for the IPR, OPR, and
labeled compound recoveries were negative numbers. As noted repeatedly in the body of this report,
negative recovery values have no physical meaning. (Even in the unlikely scenario that something in the
samples was actively destroying the analytes or that they were irreversibly removed from the sample
extracts, the calculated recoveries would bottom out at zero percent.)
The Office of Water's challenge in such situations is to balance its desire for practical acceptance criteria
that can be applied across all laboratories against the time and expense that would be required to collect
much more data from laboratories that have had significant time to practice the method before the study
begins. The solution that the Office of Water has successfully utilized in the past is a hybrid approach
that employs statistically calculated limits where such limits appear reasonable to most analysts, and rely
on simpler "consensus-style" round number limits in place of calculated limits that include negative lower
limits and/or exceptionally high upper limits.
The remainder of this section includes the results of such a hybrid approach. The Office of Water
anticipates including the limits in these tables into the draft method as an interim starting point. If
practical, the Office of Water may solicit additional performance data at a later date and revise these
limits accordingly.
Method 1628 Multi-Laboratory Validation Study Report
28
April 2021
-------
Aqueous IPR/OPR Criteria
The aqueous IPR and OPR data from eight of the nine laboratories that completed that phase of the study
were used to determine QC acceptance criteria for IPRs and OPRs. The data from the ninth laboratory
(Lab 5) were excluded from the statistical analyses because they prepared their IPR samples using a
volume of only 750 mL, instead of 1 L, but spiked the same mass of the PCB congeners, resulting in a
higher concentration in the final samples. A total of 45 sets of IPR and OPR data were provided by those
eight laboratories, and were used to calculate the acceptance criteria shown in Table 12, rounded to no
decimal places. The criteria are listed for each of the 48 spiked congeners, in congener order number.
Some of these congeners coelute with other congeners (as illustrated in Table 5), but because only these
48 congeners are spiked in the IPR and OPR aliquots, only the spiked congeners are listed.
Table 12. Aqueous IPR and OPR Calculated QC Acceptance Criteria for Target Analytes
Congener
# Labs
# Results
Calculated Acceptance Criteria
IPR Range (%)
Max RSD (%)
OPR Range (%)
PCB-1
8
45
78-130
18
71 -136
PCB-3
8
45
74-117
14
71 -120
PCB-4
8
45
77-112
14
72-117
PCB-8
8
45
42-120
18
43-119
PCB-11
8
45
62-125
9
63 -124
PCB-15
8
45
70-111
10
69-111
PCB-18
8
45
60-107
17
57-111
PCB-19
8
45
77-107
12
73-111
PCB-28
8
45
18-184
17
21 -180
PCB-31
8
45
46-129
19
46 -129
PCB-37
8
45
67-112
9
68-111
PCB-44
8
45
44-131
13
46 -130
PCB-52
8
45
61 -128
8
62 -127
PCB-54
8
45
67-112
8
68-111
PCB-64
8
45
74-108
10
73-110
PCB-66
8
45
64-118
8
65-117
PCB-70
8
45
55-127
8
57 -126
PCB-74
8
45
74-102
8
73 -103
PCB-77
8
45
58-118
9
59-116
PCB-85
8
45
68-106
7
69 -105
PCB-95
8
45
63-117
12
63-117
PCB-99
8
45
66-107
10
66 -107
PCB-101
8
45
64-118
9
65-117
PCB-104
8
45
64-117
8
65-116
PCB-105
8
45
64-120
10
65-119
PCB-118
8
45
61-119
10
62-118
PCB-110
8
45
63-106
12
62 -107
PCB-126
8
45
63-113
7
64-112
PCB-132
8
45
51 -133
11
53 -131
PCB-138
8
45
61-117
11
61-116
PCB-147
8
45
61-117
12
62-117
PCB-149
8
45
57-120
11
58-119
PCB-153
8
45
46-134
16
48 -132
PCB-155
8
45
64-116
10
65-115
PCB-156
8
45
46-149
23
45 -150
PCB-166
8
45
34-157
9
36 -156
PCB-169
8
45
50-122
10
52 -121
PCB-177
8
45
47-130
10
49 -128
PCB-180
8
45
52-124
11
53 -123
Method 1628 Multi-Laboratory Validation Study Report
29
April 2021
-------
Table 12. Aqueous IPR and OPR Calculated QC Acceptance Criteria for Target Analytes
Congener
# Labs
# Results
Calculated Acceptance Criteria
IPR Range (%)
Max RSD (%)
OPR Range (%)
PCB-187
8
45
36-138
17
38 -136
PCB-188
8
45
57-122
11
58 -121
PCB-189
8
45
56-119
11
58-118
PCB-199
8
45
-100 -281
59
-93 - 273
PCB-199 (w/o Lab 8)
7
39
42-164
57
14 -193
PCB-202
8
45
55-121
12
56 -120
PCB-205
8
45
52-118
18
51-119
PCB-206
8
45
35-135
17
37 -133
PCB-208
8
45
44-125
15
45 -124
PCB-209
8
45
31 -130
27
30 -131
The statistically determined recovery ranges for the IPRs and OPRs reflect the variability within each
laboratory, as well as the variability across all eight laboratories that completed that portion of the study
using 1-L samples. Generally speaking, the congeners that are quantified by isotope dilution have
narrower ranges than the congeners that are quantified by extracted internal standard. For example, the
observed IPR recoveries for PCB-4, quantified by isotope dilution, ranged from 86 to 103%, while, the
recoveries for PCB-8, quantified by extracted internal standard, ranged from 68 to 92%. The observed
OPR results were also affected by fact that some laboratories submitted more OPR data than others (e.g.,
there were 13 sets of OPR results for aqueous samples, from 8 labs) because some laboratories analyzed
the wastewater samples in more batches than other laboratories, and each sample preparation batch
contained an OPR aliquot.
The initial statistical calculations for PCB-199 resulted in exceptionally wide ranges that included a
negative lower value, as shown in Table 12. Those results were driven by the wildly variable results for
this congener in only one laboratory (Lab 8). CSRA staff examined the results for those four IPR aliquots
and two OPR aliquots in detail, as did the laboratory staff. When no obvious errors were identified
through either review, the statistical calculations were rerun without the results from Lab 8 for that one
congener, yielding the much more reasonable ranges for PCB-199 shown in red in Table 12 for the IPR,
although the OPR calculations still result in a fairly high upper limit of 193%.
The calculated maximum RSD values for target analytes in Table 12 are below 20% for 45 of the 48
target analytes. The exceptions are PCB-156 at 23%, PB-209 at 27%, and PCB-199 at 57% even after
removal of the results from Laboratory 8 for that congener. Except for the calculated maximum RSD for
PCB-199, all of the values in Table 12 are well within reasonable expectations.
The labeled compound data from the same eight laboratories were used to calculate the IPR and OPR
acceptance criteria. The range of the observed mean recoveries of the labeled compounds in the IPRs was
from 35 to 81%. The observed mean RSD was less than 20% for 26 of the 29 labeled congeners. The
mean RSDs for the two monochlorinated labeled congeners and the first dichlorinated labeled congener
(labeled PCBs 1, 3, and 4) were 29%, 28%, and 27%, respectively.
The calculated ranges for IPR/OPR labeled compound recoveries and RSD are presented in Table 13,
rounded to no decimal places. All of the calculated ranges for the IPR and OPR are much wider than the
observed values from the eight laboratories in the study. This is a function of the statistical calculations,
which incorporate not only the variability within a given laboratory, but the variability across all of the
laboratories, and an allowance for testing multiple analytes at the same time.
Method 1628 Multi-Laboratory Validation Study Report
30
April 2021
-------
Table 13. IPR and OPR QC Acceptance Criteria for Labeled Compounds in Aqueous Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-1
-39 -110
91
-40 -111
15 -130
40
15-130
13Ci2-PCB-3
-27 -111
75
-29 -113
15 -130
40
15-130
13Ci2-PCB-4
-18 -102
72
-22 -105
15 -130
40
15-130
13Ci2-PCB-11
-24 -136
47
-22 -133
15 -130
40
15-130
13Ci2-PCB-15
-29 -144
44
-25 -140
15 -130
40
15-130
13Ci2-PCB-19
-16 -116
53
-16 -115
15 -130
40
15-130
13Ci2-PCB-28
-21 -143
39
-18 -139
16 -130
40
15-130
13Ci2-PCB-37
-38 -179
36
-33 -174
15 -130
40
15-130
13Ci2-PCB-52
-3-117
34
-1-115
15 -130
40
15-130
13Ci2-PCB-54
-4-103
43
-4-102
15 -130
40
15-130
13Ci2-PCB-70
-9-138
30
-6-135
15 -130
40
15-130
13Ci2-PCB-77
LO
CM
29
-9-148
15 -130
40
15-130
13Ci2-PCB-85
5 -129
31
7 -127
15 -130
40
15-130
13Ci2-PCB-101
11-97
40
8-99
15 -130
40
15-130
13Ci2-PCB-104
8-118
34
8-118
15 -130
40
15-130
13Ci2-PCB-118
-1 -141
30
1 -138
15 -130
40
15-130
13Ci2-PCB-126
-13 -159
32
-9-155
15 -130
40
15-130
13Ci2-PCB-138
17-98
39
13 -102
15 -130
40
15-130
13Ci2-PCB-153
14-122
30
15 -122
15 -130
40
15-130
13Ci2-PCB-155
3 -142
28
6 -139
15 -130
40
15-130
13Ci2-PCB-169
-92 - 253
39
-85 - 246
15 -130
40
15-130
13Ci2-PCB-180
17-110
36
15-113
15 -130
40
15-130
13Ci2-PCB-188
7 -139
27
9 -136
15 -130
40
15-130
13Ci2-PCB-189
-5-157
30
-2-153
15 -130
40
15-130
13Ci2-PCB-202
10-120
28
11-118
15 -130
40
15-130
13Ci2-PCB-205
-9-153
37
LO
CO
15 -130
40
15-130
13Ci2-PCB-206
-12 -147
35
-9-144
15 -130
40
15-130
13Ci2-PCB-208
-18 -156
41
-15 -154
15 -130
40
15-130
13Ci2-PCB-209
-30 -165
39
CO
CO
C\l
15 -130
40
15-130
As can be seen in Table 13, the lower limits of the IPR acceptance criteria are less than zero for 20 of the
29 labeled analytes, and for the OPR criteria, 19 of the 29 labeled analytes have lower limits less than
zero. Negative recovery values have no physical basis. Rather, as noted here, those calculated limits are a
function of the variability of the data from the study. Had more laboratories chosen to complete the
study, or had all of the laboratories had more time to practice the method, one would expect that the
observed recoveries and their precision would have improved, yielding narrower ranges for the IPR and
OPR acceptance criteria.
In contrast, to the calculated lower limits, the calculated upper limits for the IPR ranged from 97% to
253%, with 21 of 29 labeled analytes having calculated upper limits less than or equal to 150%.
Likewise, the calculated upper limits for the OPR ranged from 99% to 246%, with 22 of 29 labeled
analytes having calculated upper limits less than or equal to 150%. Recoveries well over 100% are a
function of the uncertainty in the quantitation of the labeled compounds added prior to sample extraction
using the internal standards injected into the final extract immediately prior to the instrumental analysis.
Other isotope dilution methods from EPA have used 150% as an upper limit for labeled compound
recovery, and many non-isotope dilution methods allow the areas of the injected internal standards used in
those procedures to range up to 200% of their corresponding areas in the most recent calibration
verification standard. Therefore, the upper limits of IPR and OPR ranges in Table 13 for many of the
labeled compounds are not unprecedented by any means.
Method 1628 Multi-Laboratory Validation Study Report
31
April 2021
-------
The criterion for the maximum RSD for the labeled compounds in the IPR analyses ranged from 27% to
91%. The calculated maximum RSD values for 21 of the 29 labeled compounds are at or below 40%.
The highest calculated RSD values are for the mono- and dichlorinated labeled compounds. Those values
may reflect the known concerns with potential loss of these lightest of the labels during extract
concentration.
Given the nature of isotope dilution quantitation, and the fact that the labeled compounds are not
regulated parameters under the Clean Water Act, there is much merit to using simpler consensus-style
acceptance limits for the recoveries of the labeled compounds in the IPR and OPR aliquots. After EPA
reviewed the results of the validation study and the calculated QC acceptance criteria, the decision was
made to instead compile the interim acceptance criteria for the draft method shown (in green) in Table 13
above. The draft method used in the study employed limits of 15 - 130% for the recovery of the labeled
analogs of PCB-1, PCB-3, PCB-4, PCB-11, PCB-15, and PCB-19, and limits of 40 - 130% for all of the
other 23 labeled analogs (e.g., labeled PCB-28 to labeled PCB-209). However, those limits were based
on data from the single laboratory that developed the original procedure, and as such, are not expected to
be representative of multi-laboratory performance. Therefore, EPA examined the failure rates for the
labeled compound recoveries in the IPR and OPR analyses from the study using two sets of consensus-
style limits: 15 - 130% and 25 - 150%, as shown Table 14. These failures represent instances where the
mean IPR result from all four IPR aliquots or the single OPR aliquot fell outside of the stated limits.
Table 14. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance Criteria for
Aqueous Matrix IPR and OPR
Congener
Aqueous IPR (mean of 4 aliquots)
Aqueous OPR (1 aliquot)
Total #
Mean
Results
Observed Failure Rate (%)
Total #
Results
Observed Failure Rate (%)
LL
15%
UL
130%
LL
25%
UL
150%
LL
15%
UL
130%
LL
25%
UL
150%
13C12 PCB-1
8
25.0
0.0
50.0
0.0
13
7.7
0.0
15.4
0.0
13Ci2 PCB-3
8
0.0
0.0
25.0
0.0
13
0.0
0.0
15.4
0.0
13Ci2 PCB-4
8
0.0
0.0
25.0
0.0
13
0.0
0.0
7.7
0.0
13Ci2 PCB-11
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-15
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-19
8
0.0
0.0
12.5
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-28
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-37
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-52
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-54
8
0.0
0.0
3.1
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-70
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-77
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-85
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-101
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-104
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-118
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-126
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-138
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-153
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-155
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-169
8
0.0
12.5
0.0
0.0
13
7.7
0.0
7.7
0.0
13Ci2 PCB-180
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-188
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-189
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-202
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
Method 1628 Multi-Laboratory Validation Study Report
32
April 2021
-------
Table 14. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance Criteria for
Aqueous Matrix IPR and OPR
Congener
Aqueous IPR (mean of 4 aliquots)
Aqueous OPR (1 aliquot)
Total #
Mean
Results
Observed Failure Rate (%)
Total #
Results
Observed Failure Rate (%)
LL
15%
UL
130%
LL
25%
UL
150%
LL
15%
UL
130%
LL
25%
UL
150%
13C12 PCB-205
8
0.0
0.0
0.0
0.0
13
0.0
7.7
0.0
0.0
13Ci2 PCB-206
8
0.0
0.0
0.0
0.0
13
0.0
7.7
0.0
0.0
13Ci2 PCB-208
8
0.0
0.0
0.0
0.0
13
0.0
0.0
0.0
0.0
13Ci2 PCB-209
8
0.0
0.0
0.0
0.0
13
0.0
7.7
0.0
0.0
No labeled congeners failed the 150% upper limit for either the IPR or OPR analyses. Only one labeled
congener failed the 130% upper limit for the mean of the IPR and only three labeled congeners failed the
130% upper limit in the OPR. The labeled analogs of PCB-205, PCB-206, and PCB-209 each had one
failure out of 13 sets of OPR results (e.g., 7.7% of 13 = 1 failure). All three of those labeled analogs
failed in the same OPR from one laboratory (a second aqueous OPR run by that laboratory had all three
congeners pass the 130% upper limit). The mean recovery of the labeled analog of PCB-169 exceeded
the 130% upper limits for the IPR results once in the 8 sets of IPR results (e.g., 12.5% of 8 = 1 failure).
There were more failures of the lower limits for the IPR and OPR data. The labeled analogs of PCB-1,
PCB-3, PCB-4, and PCB-19 failed the 25% lower limit for the mean 12.5% to 50% of the time for the
sets of IPR data. The labeled analogs of PCB-1, PCB-3, and PCB-4 failed the 25% lower limit from 7.7%
to 15.4% for OPR results (e.g., 1 and 2 failures respectively). The labeled analog of PCB-169 also failed
in the OPR one time at the 25% lower limit.
Using a 15% lower limit reduced or eliminated all but one of the failures observed at the 25% IPR limit,
for the labeled analog of PCB-1, which still failed 25% of the time. This labeled congener and its native
counterpart have the lowest molecular weights of all of the congeners and are known to be susceptible to
evaporative losses during extract concentration. Such losses can be overcome by employing appropriate
care during extract concentration.
Therefore, based on the results in Table 14, EPA recommends the use of a single set of limits, namely
15 - 130%, for all of the labeled compound recoveries in the IPR and OPR aliquots.
The maximum RSD final criteria are based on the mean of the calculated RSDs from all 29 labeled
compounds, which was 40% RSD.
Solids IPR and OPR Results
The solids IPR and OPR data from all six laboratories that completed the sediment portion of the study
were used to determine QC acceptance criteria for IPRs and OPRs. The solids IPR samples were
prepared in Ottawa sand. (While tissues are a "solid," as opposed to a "liquid," in the method and this
report, the term "solid" refers to soils, sediments, or biosolids.) A total of 31 sets of IPR and OPR data
were provided by those six laboratories and were used to calculate the acceptance criteria shown in Table
15, rounded to no decimal places. The criteria are listed for each of the 48 spiked congeners, in congener
order number. Some of these congeners coelute with other congeners (as illustrated in Table 5), but
because only these 48 congeners are spiked in the IPR and OPR aliquots, only the spiked congeners are
listed.
Method 1628 Multi-Laboratory Validation Study Report
33
April 2021
-------
Table 15. IPR and OPR QC Acceptance Criteria for Target Analytes in Solid Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
IPR Mean
Max RSD
OPR
PCB-1
61 -154
59
19 -196
61 -154
59
19-196
PCB-3
40-156
42
29 -167
40-156
42
29-167
PCB-4
48-144
61
13 -179
48-144
61
13-179
PCB-8
-21 -196
42
-11 -187
35-150
40
25-180
PCB-11
-35 - 255
35
-20 - 239
35-150
40
25-160
PCB-15
36-150
44
25 -162
36-150
44
25-162
PCB-18
20-148
40
18 -149
20-148
40
18-149
PCB-19
26-157
32
28 -156
26-157
32
28-156
PCB-28
-5 - 202
41
2 -195
25-150
35
30-150
PCB-31
38-147
37
32 -153
38-147
37
32-153
PCB-37
38-147
38
31 -155
38-147
38
31 -155
PCB-44
23-153
34
24 -151
23-153
34
24-151
PCB-52
57-138
42
36 -159
57-138
42
36-159
PCB-54
56-132
56
21 -167
56-132
56
21 -167
PCB-64
29-153
29
31 -150
29-153
29
31 -150
PCB-66
50-138
25
49 -140
50-138
25
49-140
PCB-70
43-144
27
42 -144
43-144
27
42-144
PCB-74
41 -135
30
38 -138
41 -135
30
38-138
PCB-77
42-134
40
30 -145
42-134
40
30-145
PCB-85
57-121
27
50 -128
57-121
27
50-128
PCB-95
55-125
29
47 -133
55-125
29
47-133
PCB-99
33-140
34
30 -143
33-140
34
30-143
PCB-101
57-125
26
51 -132
57-125
26
51 -132
PCB-104
52-128
44
32 -148
52-128
44
32-148
PCB-105
65-122
17
63 -124
65-122
17
63-124
PCB-118
48-133
19
50 -131
48-133
19
50-131
PCB-110
31 -142
20
36 -137
31 -142
20
36-137
PCB-126
48-129
14
52 -124
48-129
14
52-124
PCB-132
42-146
18
47 -141
42-146
18
47-141
PCB-138
60-123
19
58 -125
60-123
19
58-125
PCB-147
58-126
25
53 -132
58-126
25
53-132
PCB-149
51 -129
28
46 -134
51 -129
28
46-134
PCB-153
76-109
25
61 -124
76-109
25
61 -124
PCB-155
60-122
37
41 -140
60-122
37
41 -140
PCB-156
76-119
25
62 -133
76-119
25
62-133
PCB-166
71 -122
21
64 -128
71 -122
21
64-128
PCB-169
56-130
55
23 -164
56-130
55
23-164
PCB-177
71-114
29
55 -130
71-114
29
55-130
PCB-180
72-112
25
58 -125
72-112
25
58-125
PCB-187
64-114
23
56 -122
64-114
23
56-122
PCB-188
61-118
27
52 -128
61-118
27
52-128
PCB-189
67-117
24
58 -126
67-117
24
58-126
PCB-199
62-126
22
58 -130
62-126
22
58-130
PCB-202
51 -127
24
49 -129
51 -127
24
49-129
PCB-205
54-116
31
44 -126
54-116
31
44-126
PCB-206
52-129
49
27 -154
52-129
49
27-154
PCB-208
45-131
21
47 -129
45-131
21
47-129
PCB-209
67-111
19
62-117
67-111
19
62-117
As with the aqueous sample portion of the study, the statistically determined recovery ranges for the
solids IPRs and OPRs reflect the variability within each laboratory, as well as the variability across all six
laboratories that completed that portion of the study. The observed mean IPR recoveries for the 48
Method 1628 Multi-Laboratory Validation Study Report
34
April 2021
-------
congeners ranged from about 86 to 114%. The calculated IPR ranges, while wider than the ranges for the
sample congeners in the aqueous IPR samples, are generally reasonable for solid samples.
As shown in Table 15, the IPR ranges for 45 of the 48 target analytes have calculated lower limits well
above zero. The three target analytes with calculated lower limits that are negative values are PCB-8,
PCB-11, and PCB-28 (in red). The calculated upper limits of those three congeners are also well above
150% (in red). The OPR ranges for those three congeners (in red) are also notably wider than for the
other 45 target analytes. Except for those three congeners, although generally wider than the calculated
limits in the aqueous matrix, the study data for the solid IPR and OPR analyses demonstrate reasonable
reproducibility.
EPA examined the failure rates for PCB-8, PCB-11, and PCB-28 for various consensus-style acceptance
lower and upper limits, including 15%, 25%, 35%, 130%, 150%, and 160%. None of the mean IPR
results from the study failed at any of those potential lower limits. All three of those congeners had
failures of the upper limit at 130%, with failure rates of 4% to 20% for the IPR results. Relative to the
150% upper limit, only PCB-11 had any failures, with 4% of the study results above that limit (e.g.,
1 mean IPR). Based on those failure rates, EPA replaced the calculated criteria for PCB-8 and PCB-11
with interim acceptance criteria (in green) based on the calculated results for PCB-15, a closely eluting
dichlorobiphenyl, but rounded to multiples of 5. PCB-28 is a trichlorobiphenyl, and the final criteria (in
green) are based on the results for PCB-18 and PCB-19, similarly rounded. The final method will
encourage each laboratory to employ control charts and to develop in-house statistical quality control
limits, as long as those limits are no wider than the limits in the published method.
The labeled compound data from the six laboratories that completed the solid sample portion of the study
were used to calculate the IPR and OPR acceptance criteria presented in Table 16. The derivation of the
interim acceptance criteria is described after the table.
Table 16. IPR and OPR QC Acceptance Criteria for Labeled Congeners in Solid Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-1
-105 -186
89
-90-171
15-130
60
15-130
13Ci2-PCB-3
-81 -170
83
-68-158
15-130
60
15-130
13Ci2-PCB-4
-71 -163
81
-60-153
15-130
60
15-130
13Ci2-PCB-11
-34-151
70
-31 -148
15-130
60
15-130
13Ci2-PCB-15
-29-145
73
-29-145
15-130
60
15-130
13Ci2-PCB-19
-46-152
71
-39-145
15-130
60
15-130
13Ci2-PCB-28
-18-146
72
-23-151
15-130
60
15-130
13Ci2-PCB-37
0-142
71
-14-156
15-130
60
15-130
13Ci2-PCB-52
-41 -140
58
-32-132
15-130
60
15-130
13Ci2-PCB-54
-27-142
53
-21 -136
15-130
60
15-130
13Ci2-PCB-70
-16-144
50
-11 -140
15-130
60
15-130
13Ci2-PCB-77
-7 -148
52
-6 -147
15-130
60
15-130
13Ci2-PCB-85
co
51
-23-133
15-130
60
15-130
13Ci2-PCB-101
-14-139
49
-9 -135
15-130
60
15-130
13Ci2-PCB-104
-12-143
47
-8 -138
15-130
60
15-130
13Ci2-PCB-118
-9 -145
48
-5 -142
15-130
60
15-130
13Ci2-PCB-126
-10-155
51
-8 -153
15-130
60
15-130
13Ci2-PCB-138
-18-139
51
-13-134
15-130
60
15-130
13Ci2-PCB-153
-2 -139
49
-1 -139
15-130
60
15-130
13Ci2-PCB-155
3-141
50
1 -142
15-130
60
15-130
13Ci2-PCB-169
-16-158
61
-17-159
15-130
60
15-130
13Ci2-PCB-180
-7 -142
50
-5 -141
15-130
60
15-130
13Ci2-PCB-188
-2 -151
50
-2 -151
15-130
60
15-130
13Ci2-PCB-189
-22-168
44
-14-160
15-130
60
15-130
13Ci2-PCB-202
-1 -143
52
-2 -145
15-130
60
15-130
Method 1628 Multi-Laboratory Validation Study Report
35
April 2021
-------
Table 16. IPR and OPR QC Acceptance Criteria for Labeled Congeners in Solid Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-205
2-149
49
1 -150
15-130
60
15-130
13Ci2-PCB-206
5-136
53
1 -141
15-130
60
15-130
13Ci2-PCB-208
-11 -155
51
-8 -153
15-130
60
15-130
13Ci2-PCB-209
4-138
52
1 -141
15-130
60
15-130
In marked contrast to the calculated ranges for the target analytes, the vast majority of the calculated IPR
and OPR ranges for the labeled compounds extend below zero. Only five of the labeled compounds have
non-negative calculated lower IPR limits, and those five range from 0 to 5%. Somewhat surprisingly, the
upper IPR limits for the labeled compound are not as extreme. The upper limits for 19 of the labeled
compounds are at or below 150%, and six of the ten labeled compounds are between 151% and 160%.
The calculated OPR ranges for the labeled compounds are similarly affected, with only 4 labeled
compounds with calculated lower limits above zero (all four at 1%), and the upper OPR limits are
generally below 150%.
EPA used a similar approach in evaluating potential labeled compound acceptance criteria as was used for
the aqueous IPR and OPR results. Table 17 presents the IPR and OPR failure rates relative to the same
two consensus-style acceptance criteria of 15% to 130% and 25% to 150%.
Table 17. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance Criteria for
Solid Matrix IPR and OPR
Congener
Solid IPR (mean of 4 aliquots)
Solid OPR (1 aliquot)
Total #
Mean
Results
Observed Failure Rate (%)
Total #
Results
Observed Failure Rate (%)
LL
15%
UL
130%
LL
25%
UL
150%
LL
15%
UL
130%
LL
25%
UL
150%
13C12 PCB-1
6
0.0
0.0
33.3
0.0
7
28.6
0.0
28.6
0.0
13Ci2 PCB-3
6
0.0
0.0
0.0
0.0
7
28.6
0.0
28.6
0.0
13Ci2 PCB-4
6
0.0
0.0
0.0
0.0
7
28.6
0.0
28.6
0.0
13Ci2 PCB-11
6
0.0
0.0
0.0
0.0
7
14.3
0.0
28.6
0.0
13Ci2 PCB-15
6
0.0
0.0
0.0
0.0
7
14.3
0.0
28.6
0.0
13Ci2 PCB-19
6
0.0
0.0
0.0
0.0
7
28.6
0.0
28.6
0.0
13Ci2 PCB-28
6
0.0
0.0
0.0
0.0
7
14.3
0.0
28.6
0.0
13Ci2 PCB-37
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-52
6
0.0
0.0
0.0
0.0
7
0.0
0.0
14.3
0.0
13Ci2 PCB-54
6
0.0
0.0
0.0
0.0
7
14.3
0.0
28.6
0.0
13Ci2 PCB-70
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-77
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-85
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-101
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-104
6
0.0
0.0
0.0
0.0
7
0.0
0.0
14.3
0.0
13Ci2 PCB-118
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-126
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-138
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-153
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-155
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-169
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-180
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-188
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
Method 1628 Multi-Laboratory Validation Study Report
36
April 2021
-------
Table 17. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance Criteria for
Solid Matrix IPR and OPR
Congener
Solid IPR (mean of 4 aliquots)
Solid OPR (1 aliquot)
Total #
Mean
Results
Observed Failure Rate (%)
Total #
Results
Observed Failure Rate (%)
LL
15%
UL
130%
LL
25%
UL
150%
LL
15%
UL
130%
LL
25%
UL
150%
13C12 PCB-189
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-202
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-205
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-206
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-208
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
13Ci2 PCB-209
6
0.0
0.0
0.0
0.0
7
0.0
0.0
0.0
0.0
As can be seen in Table 17, neither upper acceptance limit was exceeded by the study IPR or OPR results.
The smaller number of IPR and OPR data sets for the solid matrices, compared to the aqueous matrices,
means that a 14.3% failure rate for the OPR represents 1 of the 7 OPR sets and 28.6% represents 2 of the
7 OPR sets.
Only the IPR mean for one labeled congener, PCB-1, failed the 25% lower recovery limit and using a
lower limit of 15% recovery for the labeled compounds in the IPR analyses eliminates that failure. The
use of the 15% lower limit for the OPR data is not as effective, but it still reduces the overall number of
failures across all of the labeled congeners from 10 labeled analogs to 8, 4 of which only had one failure
each at 15%.
Based on these failure rates, EPA compiled the interim acceptance criteria shown in Table 16 (in green)
for the draft method. Because the same issues of negative lower limits and very high upper limits arose
with the calculated limits for the labeled compounds in the solid sample IPR and OPR aliquots as
occurred for labeled compounds in the aqueous matrix IPR and OPR aliquots, the solution was to apply
the same 15 - 130% limits for each labeled compound. In the case of the maximum RSD, the value was
based on rounding the mean of the calculated RSDs from all 29 labeled compounds of 57% RSD to 60%.
Tissue IPR and OPR Results
The tissue IPR and OPR data from all four laboratories that completed the tissue portion of the study were
used to determine the QC acceptance criteria for IPRs and OPRs. The tissue IPR samples were prepared
in a 10:90 mixture of canola oil and sand, which EPA used as a simulant for fish tissues containing 10%
lipids. A total of 20 sets of IPR and OPR data were provided by those four laboratories and were used to
establish the acceptance criteria shown in Table 18, rounded to no decimal places. The criteria are listed
for each of the 48 spiked congeners, in congener order number. Some of these congeners coelute with
other congeners (as illustrated in Table 5), but because only these 48 congeners are spiked in the IPR and
OPR aliquots, only the spiked congeners are listed. The derivation of the interim acceptance criteria is
described after the table.
Table 18. IPR and OPR QC Acceptance Criteria for Target Analytes in Tissue Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Range
Max RSD
OPR Range
IPR Mean
Max RSD
OPR Range
PCB-1
-28 - 227
22
6 -194
25-150
25
25-150
PCB-3
-62 - 253
24
-23-214
25-150
25
25-150
PCB-4
-602 - 876
110
-405 - 679
25-150
25
25-150
PCB-8
-34 - 220
29
0 -186
25-150
25
25-150
PCB-11
-18-212
21
12 -182
25-150
25
25-150
PCB-15
-40 - 225
21
-7-193
25-150
25
25-150
PCB-18
-38-215
26
00
25-150
25
25-150
Method 1628 Multi-Laboratory Validation Study Report
37
April 2021
-------
Table 18. IPR and OPR QC Acceptance Criteria for Target Analytes in Tissue Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Range
Max RSD
OPR Range
IPR Mean
Max RSD
OPR Range
PCB-19
-56 - 240
22
-20 - 204
25-150
25
25-150
PCB-28
-82 - 292
19
-45 - 254
25-150
25
25-150
PCB-31
-62 - 239
36
-22 -199
25-150
25
25-150
PCB-37
12-179
23
33 -157
25-150
25
25-150
PCB-44
-60 - 234
25
-23 -197
25-150
25
25-150
PCB-52
-11 -199
20
17 -171
25-150
25
25-150
PCB-54
11-177
20
33 -155
25-150
25
25-150
PCB-64
-49 - 232
22
-14 -198
25-150
25
25-150
PCB-66
-30-213
18
0 -184
25-150
25
25-150
PCB-70
-27 - 207
21
3 -177
25-150
25
25-150
PCB-74
-39-215
14
-16 -192
25-150
25
25-150
PCB-77
-4-182
13
17 -161
25-150
25
25-150
PCB-85
6 -171
12
25 -152
25-150
25
25-150
PCB-95
5 -175
13
25 -155
25-150
25
25-150
PCB-99
15-160
16
15 -160
25-150
25
25-150
PCB-101
-1 -185
15
23 -162
25-150
25
25-150
PCB-104
-33-214
17
-5-185
25-150
25
25-150
PCB-105
11-172
17
33 -151
25-150
25
25-150
PCB-118
-6-181
13
15 -161
25-150
25
25-150
PCB-110
20-156
13
38 -138
25-150
25
38-138
PCB-126
20-157
15
38 -139
25-150
25
38-139
PCB-132
32-156
15
48 -140
32-156
25
48-140
PCB-138
27-150
13
43 -134
27-150
25
43-134
PCB-147
-2-182
14
20 -160
25-150
25
25-150
PCB-149
-23 - 204
17
4 -177
25-150
25
25-150
PCB-153
33-142
14
47 -127
33-142
25
47-127
PCB-155
-15 -194
13
8 -171
25-150
25
25-150
PCB-156
-13-206
22
17 -177
25-150
25
25-150
PCB-166
-3-191
18
23 -165
25-150
25
25-150
PCB-169
14-167
14
34 -147
25-150
25
25-150
PCB-177
16-164
14
36 -144
25-150
25
25-150
PCB-180
42-137
18
52 -127
42-137
25
52-127
PCB-187
39-137
19
49 -126
39-137
25
49-126
PCB-188
7 -170
13
27 -150
25-150
25
25-150
PCB-189
o
CM
O
CO
17
-3-175
25-150
25
25-150
PCB-199
34-153
24
45 -142
34-153
25
45-142
PCB-202
33-145
16
47 -131
33-145
25
47-131
PCB-205
22-158
25
37 -144
25-150
25
37-144
PCB-206
4 -166
20
26 -144
25-150
25
25-150
PCB-208
-23 -191
16
1 -167
25-150
25
25-150
PCB-209
-198-396
25
-150-348
25-150
25
25-150
The calculated limits for the target analytes in the tissue matrix IPR and OPR aliquots exhibited
significantly more issues with negative lower limits and very high upper limits than either the aqueous
matrix or the solid matrix. This may be due, in part to the smaller number of laboratories that completed
that portion of the study, but also to the challenges and potential interferences in the tissue matrix. The
calculated tissue limits also included many positive lower limits that were below 25%.
EPA used a similar approach in evaluating potential target analyte acceptance criteria as was used for the
aqueous IPR and OPR results, evaluating the IPR and OPR failures rates in tissues relative to the same
two consensus-style acceptance criteria of 15% to 130% and 25% to 150%. The failures for the IPR and
Method 1628 Multi-Laboratory Validation Study Report
38
April 2021
-------
OPR in tissues were quite limited for the target analytes. No target analytes had IPR or OPR failures at
15% and 25%. One laboratory exceeded both the 130% and 150% upper limits for PCB-4+10 in all of
their IPR aliquots, equating to a 25% failure rate for the four laboratories in the tissue phase of the study.
That same laboratory has three IPR aliquot failures for PCB-209 at the 130% upper limit and one failure
at the 150% upper limit, which equates to 18.75% and 6.25% rates respectively. However, the mean
recovery of PCB-209 in the four IPR aliquots was 135%, which would pass relative to an upper limit of
150%. The only OPR failure was in that same laboratory, for PCB-209, coincidentally with a recovery of
135%. Given that those failures were limited to one laboratory, the EPA recommendation for the IPR
limits for the target analytes in tissues is a range of 25 - 150%. Therefore, any of the calculated limits that
began below 25% were replaced with 25 -150% (in green above), and any calculated limits that began
above 25% were retained and shown in black above.
In contrast to the overly wide recovery limits for many of the target analytes, the calculated maximum
RSD values were at or below 25% for 45 of the 48 target analytes, and the mean of all of the calculated
maximum RSD values was 21%. That suggests that while the recoveries across all of the labs varied
greatly, within each laboratory, the recoveries were more consistent. However, for simplicity, a
maximum RSD limit of 25% was applied to all of the target analytes.
As with the other matrices, the labeled compound data from the four laboratories that completed the tissue
portion of the study were used to calculate the IPR and OPR acceptance criteria presented in Table 19.
The derivation of the interim acceptance criteria is described after the table.
Table 19. IPR and OPR QC Acceptance Criteria for Labeled Compounds in Tissue Matrices
Congener
Calculated Acceptance Criteria (%)
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-1
-198-291
76
-138 -251
15 -130
60
15 -130
13Ci2-PCB-3
-218-313
73
-157 -208
15 -130
60
15 -130
13Ci2-PCB-4
-160-261
63
-107 -204
15 -130
60
15 -130
13Ci2-PCB-11
-145-257
71
-91 -182
15 -130
60
15 -130
13Ci2-PCB-15
-115-226
78
-71 -207
15 -130
60
15 -130
13Ci2-PCB-19
-152-261
65
-97-190
15 -130
60
15 -130
13Ci2-PCB-28
-113-234
74
-69-170
15 -130
60
15 -130
13Ci2-PCB-37
-59 -183
90
-46 - 240
15 -130
60
15 -130
13Ci2-PCB-52
-186-289
51
-137-229
15 -130
60
15 -130
13Ci2-PCB-54
-170-283
53
-117-218
15 -130
60
15 -130
13Ci2-PCB-70
-154-274
60
-98 - 222
15 -130
60
15 -130
13Ci2-PCB-77
-158-279
65
-100-226
15 -130
60
15 -130
13Ci2-PCB-85
-166-284
62
-107-216
15 -130
60
15 -130
13Ci2-PCB-101
-157-272
62
-101 -215
15 -130
60
15 -130
13Ci2-PCB-104
-162-267
57
-110-211
15 -130
60
15 -130
13Ci2-PCB-118
-145-265
64
-91 - 220
15 -130
60
15 -130
13Ci2-PCB-126
-159-278
68
-101 -219
15 -130
60
15 -130
13Ci2-PCB-138
-150-275
63
-93-214
15 -130
60
15 -130
13Ci2-PCB-153
-157-269
62
-102-214
15 -130
60
15 -130
13Ci2-PCB-155
-148-269
61
-93 - 204
15 -130
60
15 -130
13Ci2-PCB-169
-146-258
72
-92 - 242
15 -130
60
15 -130
13Ci2-PCB-180
-171 -300
54
-112-215
15 -130
60
15 -130
13Ci2-PCB-188
-150-270
61
-95 - 263
15 -130
60
15 -130
13Ci2-PCB-189
-191 -318
46
-135 -208
15 -130
60
15 -130
13Ci2-PCB-202
-139-261
61
-86 - 385
15 -130
60
15 -130
13Ci2-PCB-205
-312-446
47
-252 - 293
15 -130
60
15 -130
13Ci2-PCB-206
-222 - 346
45
-168 -385
15 -130
60
15 -130
13Ci2-PCB-208
-315-459
54
-241 -630
15 -130
60
15 -130
13Ci2-PCB-209
-600 - 758
85
-472 - 630
15 -130
60
15 -130
Method 1628 Multi-Laboratory Validation Study Report
39
April 2021
-------
The same issues of negative lower limits and very high upper limits arose with the calculated limits for
the labeled compounds in the tissue sample IPR and OPR aliquots as occurred for labeled compounds in
the other matrices, only with negative lower IPR limits for all of the labeled compounds, and negative
lower OPR limits for all but three of the labeled compounds.
As with the IPR/OPR data for the other matrices, EPA examined the labeled compound failures rates
relative to the same two potential sets of acceptance criteria. Table 20 presents the IPR and OPR failures
rates relative to the same two consensus-style acceptance criteria of 15% to 130% and 25% to 150%.
Table 20. Observed Labeled Compound Recovery Failure Rates for Two Potential Acceptance Criteria for
Tissue Matrix IPR and OPR
Congener
Tissue IPR (mean of 4 aliquots)
Tissue OPR (1 aliquot)
Total #
Mean
Results
Observed Failure Rate (%)
Total #
Results
Observed Failure Rate (%)
LL
15%
UL
130%
LL
25%
UL
150%
LL
15%
UL
130%
LL
25%
UL
150%
13C12 PCB-1
4
0.0
0.0
25.0
0.0
4
0.0
0.0
25.0
0.0
13Ci2 PCB-3
4
0.0
0.0
25.0
0.0
4
0.0
0.0
25.0
0.0
13Ci2 PCB-4
4
0.0
0.0
0.0
0.0
4
0.0
0.0
25.0
0.0
13Ci2 PCB-11
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-15
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-19
4
0.0
0.0
0.0
0.0
4
0.0
0.0
25.0
0.0
13Ci2 PCB-28
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-37
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-52
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-54
4
0.0
0.0
0.0
0.0
4
0.0
0.0
25.0
0.0
13Ci2 PCB-70
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-77
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-85
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-101
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-104
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-118
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-126
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-138
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-153
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-155
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-169
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-180
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-188
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-189
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-202
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-205
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-206
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-208
4
0.0
0.0
0.0
0.0
4
0.0
0.0
0.0
0.0
13Ci2 PCB-209
4
0.0
25.0
0.0
25.0
4
0.0
0.0
25.0
0.0
None of the means for the labeled compounds in the IPR aliquots failed a lower limit of 15%. Only two
labeled congener means failed at 25%, the labeled analogs for PCB-1 and PCB-3 each failed once. On
the upper end of the IPR recovery ranges, the label for PCB-209 failed the 130% upper limit once (e.g.,
25% of the 4 IPR sets) and it still failed using a 150% upper limit. Both of those failures of labeled PCB-
209 were in a single laboratory.
Method 1628 Multi-Laboratory Validation Study Report
40
April 2021
-------
For the four sets of OPR data, six labeled congeners failed once each at the 25% lower recovery limit.
Labeled congener PCB-209 failed in one laboratory and the other five labeled compounds all failed in a
second laboratory. All six of those failures were eliminated using the 15% lower recovery limit.
Based on these failure rates, EPA compiled the interim acceptance criteria for the draft method shown (in
green) in Table 19. The solution was to apply the same 15 - 130% limits for each label. In the case of the
maximum RSD, the average RSD (which was 60%) was used for all 29 labeled compounds.
Method 1628 Multi-Laboratory Validation Study Report
41
April 2021
-------
6. Method Detection Limits
EPA required that each laboratory determine the method detection limits (MDLs) for all 209 PCB
congeners in each of the matrix types that they agreed to analyze: aqueous, solids (e.g., sediment and
biosolids), and tissue. MDLs were determined using the newly revised MDL procedure promulgated by
EPA in 2017.
The revised procedure defines the MDL as:
"... the minimum measured concentration of a substance that can be reported with 99%
confidence that the measured concentration is distinguishable from method blank results. "
The procedure consists of two parts: determination of the MDL based on method blanks (called MDLb),
and determination of the MDL based on spiked samples (called MDLS). Both MDLb and MDLS are
determined in a reference matrix, using at least seven replicates prepared and analyzed on three non-
consecutive days.
The MDLb is calculated as:
MDLb=X+t(n.li i-oc=0.99)Sb
where:
X = mean of the method blank results (use zero in place of the mean if the mean is negative)
Vi i-a = o 99) = t'lc Student's /-value appropriate for the single-tailed 99th percentile t statistic and a
standard deviation estimate with n-1 degrees of freedom
Sb = sample standard deviation of the replicate method blank sample analyses
Note: The equation above is used when all of the method blanks for an individual analyte give
numerical results. If some (but not all) of the method blank results give numerical results, then
the MDLb is set to be equal to the highest method blank result.
The MDLS is calculated as:
MDLs=t(n.li i-oc=0.99)Ss
where:
t(n j l a = 0 99)= the Student's lvalue appropriate for a single-tailed 99th percentile t statistic and a standard
deviation estimate with n-1 degrees of freedom.
Ss = sample standard deviation of the replicate spiked sample analyses.
For aqueous sample MDL determinations, the matrix was reagent water. For solid sample MDL
determinations, the matrix was Ottawa sand. For tissue sample MDL determinations, the matrix was the
same 10:90 mixture of canola oil and sand used for the IPR samples. After both an MDLb and MDLS
have been determined, the laboratory sets their initial MDL as the greater of the MDLb and MDLS values.
The laboratories provided all of the results to CSRA and CSRA independently performed the calculations
of MDLb and MDLS after subjecting all of the results to a formal data review process. The results from all
of the laboratories were then pooled by matrix type and used to calculate pooled MDLs and pooled
Minimum Levels (MLs), based on the statistical procedures outlined in EPA's new method protocol
(USEPA 2018). The purpose of those pooled MDL and ML values is to provide data users with a
conservative overall assessment of the sensitivity of the analytical method. In practice, each laboratory
utilizing the method will determine its own MDL values for each matrix of interest and utilize those
laboratory-specific MDLs in assessing method blanks and determining when an analyte is detected.
Method 1628 Multi-Laboratory Validation Study Report 42 April 2021
-------
Aqueous Sample Pooled MDL Determinations
CSRA provided all of the laboratories with instructions for preparing their aqueous MDL samples based
on the MDL results from the single-laboratory validation study. Using the sets of nine retention time
standards provided by EPA, the laboratories prepared a spiking solution that contained 100 ng/mL of each
of the 209 PCB congeners in a water-miscible solvent. Adding 20 |_iL of that solution to each 1-L reagent
water aliquot yielded a mass of 2 ng of each congener in the 1-L sample. Given the known coelutions of
some congeners, this means that some ""analytes" were spiked at 4 ng, and some at 6 ng, but 130 of the
167 analytes were spiked a 2 ng. Laboratories were permitted to increase that concentration if their initial
attempts to determine the MDL did not yield detectable results for each of the congeners.
CSRA determined the pooled MDLs and pooled MLs for aqueous samples using the data from 7
laboratories. The results are summarized in Table 21, below. The analytes listed in the shaded rows in
Table 21 are the congeners and coeluting congeners that represent the analytes that have direct calibration
data for this method.
Table 21. Aqueous Sample Pooled MDL and ML Results (ng/L)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-1
7
1.75
5.97
1
0.53
5
20
PCB-2
7
0.71
2.33
0.00
2
5
PCB-3
7
0.69
2.23
1.14
2
5
PCB-4+10
7
1.90
6.10
0.00
5
20
PCB-8+5
7
1.00
2.73
1
0.12
2
10
PCB-6
7
0.57
1.55
1.05
2
5
PCB-7+9
7
1.17
2.87
0.07
5
10
PCB-11
7
0.72
2.34
1
0.14
2
5
PCB-12+13
7
1.11
3.56
0.08
5
10
PCB-14
7
0.64
1.99
0.06
2
5
PCB-15
7
0.44
1.34
1
0.02
1
5
PCB-16+32
7
0.80
1.58
1
0.02
2
5
PCB-17
7
0.49
1.01
1
0.18
2
2
PCB-18
7
0.46
0.80
1
0.09
1
2
PCB-19
7
0.63
1.90
1
0.01
2
5
PCB-33+20+21
7
1.11
3.15
1
0.52
5
10
PCB-22
7
0.39
0.96
0.00
1
2
PCB-34+23
7
1.00
2.88
0.00
2
10
PCB-24+27
7
0.64
1.24
0.08
2
5
PCB-25
7
0.46
1.37
0.23
1
5
PCB-26
7
0.43
1.30
1
0.02
1
5
PCB-28
7
0.69
1.71
1
0.08
2
5
PCB-29
7
0.49
1.40
1
0.02
2
5
PCB-30
7
0.61
1.38
0.00
2
5
PCB-31
7
0.50
1.36
1
0.30
2
5
PCB-35
7
0.89
2.54
1
0.69
2
10
PCB-36
7
0.54
1.57
0.15
2
5
PCB-37
7
0.44
1.27
0.20
1
5
PCB-38
7
1.66
3.64
0.00
5
10
PCB-39
7
0.53
1.47
0.29
2
5
PCB-40
7
1.12
3.84
0.43
5
10
PCB-41+64
7
0.97
1.97
0.00
2
5
PCB-42
7
0.73
1.38
0.14
2
5
PCB-49+43
7
1.06
3.11
1
0.01
2
10
PCB-44
7
0.40
0.75
0.30
1
2
PCB-45
7
0.31
0.66
0.00
1
2
Method 1628 Multi-Laboratory Validation Study Report
43
April 2021
-------
Table 21. Aqueous Sample Pooled MDL and ML Results (ng/L)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-46
7
0.36
0.79
0.21
1
2
PCB-47+48+75
7
1.71
5.39
0.00
5
20
PCB-50
7
0.58
1.42
0.15
2
5
PCB-51
7
0.48
1.26
0.08
2
5
PCB-52+73
7
0.97
2.60
1
0.08
2
10
PCB-53
7
0.33
0.72
0.11
1
2
PCB-54
7
0.58
1.80
1
0.01
2
5
PCB-55
7
0.39
1.02
0.53
1
2
PCB-56+60
7
0.74
1.62
0.17
2
5
PCB-57
7
0.47
1.30
0.00
1
5
PCB-58
7
0.46
1.20
0.00
1
5
PCB-59
7
0.60
1.19
0.22
2
5
PCB-74+61
7
0.96
2.68
0.00
2
10
PCB-62
7
0.49
0.99
0.09
2
2
PCB-63
7
0.38
1.11
0.13
1
5
PCB-65
7
0.57
0.89
0.16
2
2
PCB-66+80
7
0.91
2.33
1
0.02
2
5
PCB-67
7
0.45
1.34
0.17
1
5
PCB-68
7
0.66
1.88
0.00
2
5
PCB-69
7
0.53
1.18
1
0.08
2
5
PCB-70
7
1.32
4.20
2
0.10
5
10
PCB-71
7
1.09
3.65
0.00
2
10
PCB-72
7
0.50
1.37
0.17
2
5
PCB-76
7
0.53
1.31
0.14
2
5
PCB-77
7
0.50
1.40
0.16
2
5
PCB-78
7
0.51
0.93
0.00
2
2
PCB-79
7
0.48
1.28
0.00
2
5
PCB-81
7
0.49
1.43
0.10
2
5
PCB-82
7
0.61
1.31
0.12
2
5
PCB-83+109
7
0.76
1.47
0.07
2
5
PCB-84
7
2.53
8.66
0.16
10
20
PCB-85+120
7
1.19
3.08
0.00
5
10
PCB-97+86
7
1.67
4.07
1
0.16
5
10
PCB-87+115+116
7
2.23
5.08
1
0.15
5
20
PCB-88+121
7
0.93
2.23
0.20
2
5
PCB-90+101+89
7
3.36
11.03
1
0.23
10
50
PCB-91
7
0.39
0.92
0.00
1
2
PCB-92
7
0.53
1.63
0.00
2
5
PCB-95+93
7
2.01
5.20
1
0.10
5
20
PCB-94
7
0.32
0.56
0.00
1
2
PCB-96
7
0.37
0.78
0.00
1
2
PCB-98+102
7
0.77
1.38
0.00
2
5
PCB-99
7
1.30
4.39
1
0.09
5
10
PCB-100
7
0.50
1.09
0.00
2
2
PCB-103
7
0.48
0.85
0.00
2
2
PCB-104
7
0.51
1.31
0.06
2
5
PCB-105+127
7
1.23
3.90
1
0.11
5
10
PCB-118+106
7
3.21
11.04
1
0.27
10
50
PCB-107+108
7
0.86
2.56
1
0.03
2
10
PCB-110
7
3.94
13.18
1
0.49
10
50
PCB-111 + 117
7
1.33
2.54
0.15
5
10
PCB-112
7
0.34
0.74
1
0.30
1
2
PCB-113
7
0.34
0.74
0.10
1
2
Method 1628 Multi-Laboratory Validation Study Report
44
April 2021
-------
Table 21. Aqueous Sample Pooled MDL and ML Results (ng/L)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-114
7
0.28
0.68
0.00
1
2
PCB-119
7
0.42
0.87
1
0.16
1
2
PCB-122
7
0.19
0.35
0.14
0.5
1
PCB-123
7
0.31
0.84
0.19
1
2
PCB-124
7
0.35
0.76
0.11
1
2
PCB-125
7
0.81
2.17
0.08
2
5
PCB-126
7
0.42
1.07
0.00
1
2
PCB-128
7
1.27
4.19
0.00
5
10
PCB-129
7
0.33
0.68
0.19
1
2
PCB-130
7
0.35
0.67
0.00
1
2
PCB-131 + 142
7
1.46
4.73
0.00
5
20
PCB-132+168
7
1.91
5.83
2
0.63
5
20
PCB-133
7
0.39
0.77
0.00
1
2
PCB-134
7
0.75
2.33
0.12
2
5
PCB-144+135
7
1.26
3.75
1
0.09
5
10
PCB-136
7
1.39
4.62
0.12
5
10
PCB-137
7
0.38
0.68
0.19
1
2
PCB-138+163+164
7
3.95
11.56
2
0.26
10
50
PCB-149+139
7
4.98
16.27
2
0.77
20
50
PCB-140
7
4.00
14.06
0.21
10
50
PCB-141
7
1.35
4.42
1
0.23
5
10
PCB-143
7
0.40
0.82
0.00
1
2
PCB-145
7
0.43
0.82
0.00
1
2
PCB-146
7
0.57
1.28
0.00
2
5
PCB-147
7
0.30
0.61
0.00
1
2
PCB-148
7
0.44
0.81
0.00
1
2
PCB-150
7
0.46
0.86
0.00
1
2
PCB-151
7
1.97
6.70
0.00
5
20
PCB-152
7
0.50
1.05
0.00
2
2
PCB-153
7
3.90
12.59
6
1.51
10
50
PCB-154
7
0.42
0.90
0.00
1
2
PCB-155
7
0.43
1.00
0.08
1
2
PCB-156
7
0.37
0.61
0.24
1
2
PCB-157
7
0.60
1.67
0.46
2
5
PCB-158+160
7
0.73
1.55
0.00
2
5
PCB-159
7
0.51
1.38
0.12
2
5
PCB-161
7
0.43
1.20
0.00
1
5
PCB-162
7
0.60
1.60
2
0.17
2
5
PCB-165
7
1.51
5.20
0.06
5
20
PCB-166
7
0.37
0.74
0.16
1
2
PCB-167
7
0.94
3.11
0.00
2
10
PCB-169
7
0.34
0.93
0.00
1
2
PCB-170+190
7
1.95
6.68
0.00
5
20
PCB-171
7
0.60
1.78
1
0.10
2
5
PCB-172+192
7
0.59
1.09
0.00
2
2
PCB-173
7
0.33
0.56
0.21
1
2
PCB-174
7
3.12
10.92
0.00
10
20
PCB-175
7
0.33
0.56
0.00
1
2
PCB-176
7
0.56
1.58
0.09
2
5
PCB-177
7
1.57
5.44
0.18
5
20
PCB-178
7
0.50
1.03
0.14
2
2
PCB-179
7
1.54
5.27
0.10
5
20
PCB-180
7
0.37
0.92
1
0.07
1
2
Method 1628 Multi-Laboratory Validation Study Report
45
April 2021
-------
Table 21. Aqueous Sample Pooled MDL and ML Results (ng/L)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-181
7
3.10
10.86
0.09
10
20
PCB-187+182
7
2.28
7.75
1
0.25
5
20
PCB-183
7
0.92
3.03
2
0.25
2
10
PCB-184
7
0.49
1.16
0.00
2
5
PCB-185
7
0.36
0.78
0.00
1
2
PCB-186
7
0.35
0.74
0.00
1
2
PCB-188
7
0.39
0.90
0.00
1
2
PCB-189
7
0.26
0.54
0.00
1
2
PCB-191
7
0.22
0.48
0.00
0.5
2
PCB-193
7
0.39
1.11
1
0.13
1
5
PCB-194
7
3.16
10.31
0.00
10
20
PCB-195
7
0.43
0.81
0.36
1
2
PCB-196+203
7
1.13
3.51
0.00
5
10
PCB-197
7
0.43
0.90
0.06
1
2
PCB-198
7
0.80
2.27
1
0.34
2
5
PCB-199
7
0.88
2.84
0.20
2
10
PCB-200
7
0.44
1.30
0.00
1
5
PCB-201
7
0.59
1.81
0.23
2
5
PCB-202
7
0.26
0.46
0.00
1
1
PCB-204
7
0.65
1.81
0.08
2
5
PCB-205
7
0.75
2.49
0.00
2
10
PCB-206
7
0.64
1.92
1
0.09
2
5
PCB-207
7
0.62
1.97
0.00
2
5
PCB-208
7
0.90
3.05
0.00
2
10
PCB-209
7
0.50
1.34
2
3.12
2
5
The analytes listed in the shaded rows are the congeners and coeluting congeners that represent the analytes that have direct
calibration data for this method.
The results from Lab 5 were not used in the pooled MDL calculations shown in Table 21 for two reasons.
First, the laboratory only prepared 750-mL samples for their MDL studies, instead of a full 1 L. Second,
their MDL results differed widely from those of the other seven laboratories and did not appear to be
representative of the community as a whole.
Lab 8 was the only laboratory that completed the study that had agreed to analyze the aqueous samples
using both separatory funnel extraction and solid-phase extraction (SPE). Their separatory funnel MDL
results were included in the pooled MDLs shown in Table 21. However, their SPE MDL results were not
included because they would be the only SPE results, which would not be representative of the MDLs
that other laboratories might achieve.
For the 167 analytes, the separatory funnel results from Lab 8 accounted for 125 of the maximum MDLS
values in Table 21. The maximum MDLS values for the remaining 42 analytes were contributed by five
of the other six laboratories, with one laboratory never having the highest MDLS value for any analyte.
Overall, the pooled MDLS values ranged from 0.19 ng/L to 4.98 ng/L, while the maximum MDLS values
ranged from 0.35 ng/L to 16.27 ng/L. In total, 23 of the maximum MDLS values in Table 20 are greater
than 5 ng/L, and 10 of those are greater than 10 ng/L. However, as noted in Section 1, the published
MDL value for Aroclor 1242 in EPA Method 608.3 is 65 ng/L. Thus, the highest MDLS value in this
study, 16.27 ng/L, is still 3.995 times lower than the MDL in Method 608.3. The majority of the pooled
MDL values are far lower, with the highest pooled MDL of 4.98 for the coeluting pair of PCB-149+139
being 13 times lower. The pooled MDLS values for all 167 analytes are presented graphically in Figure 2.
The x-axis is in the same order as the analytes are listed in Table 21.
Method 1628 Multi-Laboratory Validation Study Report
46
April 2021
-------
The pooled ML for each of the 167 analytes was calculated as a multiplier of 3.18 times the pooled MDL
value, and then rounded to the nearest multiple of 1, 2, or 5, in order to facilitate future preparation of
calibration standards at the ML. The 167 pooled ML values ranged from 0.5 ng/L to 20 ng/L. In total,
127 pooled ML values are less than or equal to 2 ng/L, and 156 pooled ML values are less than or equal
to 5 ng/L. Another 10 pooled ML values are 10 ng/L, and only one analyte had a pooled ML value of 20
ng/L. The pooled MDLS values for aqueous samples are presented in Figure 2.
Although the SPE MDL results from Lab 8 were not used in the pooled MDL calculations, CSRA did
perform a gross comparison of those SPE MDLs to the pooled separatory funnel MDLs. In general, the
SPE MDLs from Lab 8 were significantly higher than the pooled separatory funnel MDLs. Of the 167
analytes, 123 analytes had SPE MDLs that were more than 1.5 times the pooled MDLs, with 29 SPE
MDLs lower than the pooled MDLs. Those 123 SPE MDLs ranged up to 160 times the pooled MDLs,
with 8 SPE MDLs at least 25 times higher than the pooled MDLs. Lab 8 noted significant blank levels
for several congeners that drove their use of MDLb over MDLS for 6 of the 167 analytes. Lab 8 also
performed the SPE manually, rather than with an automated system, which appears to have introduced
much larger variability into their MDLS results and would explain their much higher MDL values,
especially compared to the automated SPE MDLs from the single-laboratory study of the draft method.
Because of the nature of the MDLb value, those were not used in the calculation of the pooled MDLS
values shown in Table 21 and Figure 2. However, Table 21 includes a column that provides the number
of times that any one of the seven laboratories in the aqueous portion of the study reported a non-zero
MDLb value. EPA evaluated those MDLb values and the frequencies at which they were reported. In
total, aqueous MDLb values were reported by one or more laboratories for 46 of the 167 analytes.
As anyone familiar with PCB analyses can attest, "blanks happen." Therefore, the fact that a given
laboratory reported non-zero MDLb values is not surprising. Perhaps equally important is that reporting
an MDLb value itself is not evidence of problems in an individual laboratory. In fact, the reason that EPA
promulgated the revised MDL procedure that includes the MDLb concept was to acknowledge the role of
method blanks in assessing the sensitivity (detection limit) of any method.
What is evident in the study data is that the necessity of reporting MDLb values is decidedly a lab-specific
issue. Table 22 includes a summary of the frequencies at which the seven laboratories reported aqueous
MDLb values.
Table 22. Frequency of Aqueous MDLb values by Lab
Lab#
1
3
4
6
7
8
9
# MDLb values
26
0
2
2
24
4
1
More specifically, of the 26 MDLb values reported by Laboratory 1, 21 of them were exclusive to that
laboratory (in other words, no other laboratory reported an MDLb value for those 21 congeners). Of the
24 values reported by Laboratory 7, 16 of those were exclusive to that laboratory. In contrast, while
Laboratory 4 reported MDLb values for four analytes, at least one other laboratory also reported an MDLb
value for those four analytes.
The only congener where the MDLb was prevalent was for PCB-153, where six of the seven laboratories
reported an MDLb. Even for this congener, the actual MDLb values in those six laboratories ranged from
0.27 ng/L to 1.21 ng/L. The MDLS value reported for PCB-153 by the seventh laboratory was 0.38 ng/L.
The pooled MDLS value in Table 20 and calculated from the results from all seven laboratories was 3.90
ng/L, reflecting the variability in the MDLS values across all of the laboratories. None of those MDL
values (whether an MDLb or an MDLS) was an impediment to analyzing the real-world samples in the
study.
Method 1628 Multi-Laboratory Validation Study Report
47
April 2021
-------
Pooled MDLs (Aqueous)
VIII
IX X
CD
Coeluting group
Single Congener
in
CO
CN
o
Figure 2.
Pooled Aqueous MDLS Values in Elution Order
Dotted lines and Roman numerals delineate the levels of chlorination. The red triangle symbols denote a group of two or more coeluting
congeners, while the blue diamond symbols denote single congeners.
Method 1628 Multi-Laboratory Validation Study Report
48
April 2021
-------
Through these MDL data and the routine method blank results generated during the course of the
validation study, EPA has demonstrated that background levels in typical laboratories are not a limiting
factor in the application of this method, but that some laboratories do have much better control of
background levels than others.
Soil/Sediment and Biosolids Sample MDL Determinations
The laboratories also determined the MDLs for the 209 PCB congeners in solid matrices, using Ottawa
sand as the reference matrix. Separate MDLs were not determined for the soil/sediment and biosolids
matrices in this study. Rather, the solid sample MDLs were applied to the biosolids samples for the
purposes of the study. However, in practice, each laboratory performing biosolids analyses using this
procedure would determine its own MDLs for the biosolids matrix.
CSRA provided all of the laboratories with instructions for preparing their solid MDL samples based on
the MDL results from the single-laboratory validation study. The same 100 ng/mL spiking solution used
for the aqueous MDL was used for the solid MDL determination. Laboratories were permitted to increase
that concentration if their initial attempts to determine the MDL did not yield detectable results for each
of the congeners.
CSRA determined the pooled MDLs for solid samples using the data from six laboratories. The results
are summarized in Table 23, below. The analytes listed in the shaded rows in Table 23 are the congeners
and coeluting congeners that represent the analytes that have direct calibration data for this method.
Table 23. Solid Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-1
6
0.63
1.72
0.00
2
5
PCB-2
6
0.06
0.13
0.00
0.2
0.5
PCB-3
6
0.10
0.22
0.00
0.2
0.5
PCB-4+10
6
0.15
0.34
0.00
0.5
1
PCB-8+5
6
0.22
0.62
0.00
0.5
2
PCB-6
6
0.09
0.23
0.00
0.2
0.5
PCB-7+9
6
0.24
0.65
0.00
1
2
PCB-11
6
0.42
1.27
1
1.46
1
5
PCB-12+13
6
0.21
0.59
0.00
0.5
2
PCB-14
6
0.11
0.30
0.00
0.2
1
PCB-15
6
0.09
0.23
0.00
0.2
0.5
PCB-16+32
6
0.14
0.32
0.00
0.5
1
PCB-17
6
0.07
0.18
0.00
0.2
0.5
PCB-18
6
0.07
0.13
1
0.05
0.2
0.5
PCB-19
6
0.08
0.21
0.00
0.2
0.5
PCB-33+20+21
6
0.30
0.75
0.00
1
2
PCB-22
6
0.08
0.19
0.00
0.2
0.5
PCB-34+23
6
0.11
0.27
0.00
0.2
1
PCB-24+27
6
0.09
0.21
0.00
0.2
0.5
PCB-25
6
0.08
0.19
0.00
0.2
0.5
PCB-26
6
0.09
0.24
0.00
0.2
1
PCB-28
6
0.15
0.41
0.00
0.5
1
PCB-29
6
0.06
0.16
0.00
0.2
0.5
PCB-30
6
0.08
0.21
0.00
0.2
0.5
PCB-31
6
0.07
0.16
0.00
0.2
0.5
PCB-35
6
0.21
0.41
0.00
0.5
1
PCB-36
6
0.17
0.46
0.00
0.5
1
PCB-37
6
0.18
0.50
0.00
0.5
2
PCB-38
6
0.14
0.33
0.00
0.5
1
Method 1628 Multi-Laboratory Validation Study Report
49
April 2021
-------
Table 23. Solid Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-39
6
0.10
0.23
0.00
0.2
0.5
PCB-40
6
0.16
0.38
0.00
0.5
1
PCB-41+64
6
0.17
0.48
0.00
0.5
2
PCB-42
6
0.10
0.22
0.00
0.2
0.5
PCB-49+43
6
0.24
0.65
0.00
1
2
PCB-44
6
0.11
0.29
0.00
0.5
1
PCB-45
6
0.09
0.24
0.00
0.2
1
PCB-46
6
0.06
0.16
0.00
0.2
0.5
PCB-47+48+75
6
0.24
0.60
1
0.02
1
2
PCB-50
6
0.07
0.20
0.00
0.2
0.5
PCB-51
6
0.06
0.17
0.00
0.2
0.5
PCB-52+73
6
0.17
0.43
0.00
0.5
1
PCB-53
6
0.05
0.13
0.00
0.2
0.5
PCB-54
6
0.06
0.16
0.00
0.2
0.5
PCB-55
6
0.08
0.21
0.00
0.2
0.5
PCB-56+60
6
0.13
0.26
0.00
0.5
1
PCB-57
6
0.10
0.29
0.00
0.2
1
PCB-58
6
0.11
0.33
0.00
0.5
1
PCB-59
6
0.07
0.18
0.00
0.2
0.5
PCB-74+61
6
0.14
0.36
0.00
0.5
1
PCB-62
6
0.11
0.32
0.00
0.5
1
PCB-63
6
0.08
0.21
0.00
0.2
0.5
PCB-65
6
0.10
0.29
0.00
0.2
1
PCB-66+80
6
0.19
0.50
0.00
0.5
2
PCB-67
6
0.11
0.31
0.00
0.2
1
PCB-68
6
0.16
0.45
0.00
0.5
1
PCB-69
6
0.12
0.32
0.00
0.5
1
PCB-70
6
0.08
0.14
1
0.06
0.2
0.5
PCB-71
6
0.14
0.39
0.00
0.5
1
PCB-72
6
0.10
0.24
0.00
0.2
1
PCB-76
6
0.11
0.30
0.00
0.2
1
PCB-77
6
0.07
0.16
0.00
0.2
0.5
PCB-78
6
0.10
0.23
0.00
0.2
0.5
PCB-79
6
0.08
0.17
0.00
0.2
0.5
PCB-81
6
0.09
0.22
0.00
0.2
0.5
PCB-82
6
0.06
0.13
0.00
0.2
0.5
PCB-83+109
6
0.14
0.28
0.00
0.5
1
PCB-84
6
0.07
0.16
0.00
0.2
0.5
PCB-85+120
6
0.15
0.31
0.00
0.5
1
PCB-97+86
6
0.11
0.18
0.00
0.2
0.5
PCB-87+115+116
6
0.37
0.82
1
0.09
1
2
PCB-88+121
6
0.12
0.29
0.00
0.5
1
PCB-90+101+89
6
0.24
0.49
2
0.15
1
2
PCB-91
6
0.05
0.13
0.00
0.2
0.5
PCB-92
6
0.06
0.10
0.00
0.2
0.2
PCB-95+93
6
0.12
0.29
1
0.06
0.5
1
PCB-94
6
0.06
0.18
0.00
0.2
0.5
PCB-96
6
0.06
0.14
0.00
0.2
0.5
PCB-98+102
6
0.12
0.29
0.00
0.5
1
PCB-99
6
0.10
0.20
0.00
0.2
0.5
PCB-100
6
0.17
0.51
0.00
0.5
2
PCB-103
6
0.15
0.44
0.00
0.5
1
PCB-104
6
0.05
0.12
0.00
0.2
0.5
Method 1628 Multi-Laboratory Validation Study Report 50
April 2021
-------
Table 23. Solid Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-105+127
6
0.19
0.38
1
0.06
0.5
1
PCB-118+106
6
0.39
0.91
1
0.15
1
2
PCB-107+108
6
0.16
0.39
0.00
0.5
1
PCB-110
6
0.31
0.72
1
0.18
1
2
PCB-111 + 117
6
0.21
0.51
0.00
0.5
2
PCB-112
6
0.09
0.25
0.00
0.2
1
PCB-113
6
0.08
0.22
0.00
0.2
0.5
PCB-114
6
0.06
0.14
0.00
0.2
0.5
PCB-119
6
0.08
0.22
0.00
0.2
0.5
PCB-122
6
0.07
0.17
0.00
0.2
0.5
PCB-123
6
0.09
0.24
0.00
0.2
1
PCB-124
6
0.08
0.19
0.00
0.2
0.5
PCB-125
6
0.07
0.15
0.00
0.2
0.5
PCB-126
6
0.07
0.17
0.00
0.2
0.5
PCB-128
6
0.08
0.14
0.00
0.2
0.5
PCB-129
6
0.07
0.13
0.00
0.2
0.5
PCB-130
6
0.07
0.12
0.00
0.2
0.5
PCB-131 + 142
6
0.10
0.26
0.00
0.2
1
PCB-132+168
6
0.18
0.35
1
0.05
0.5
1
PCB-133
6
0.07
0.18
0.00
0.2
0.5
PCB-134
6
0.08
0.17
0.00
0.2
0.5
PCB-144+135
6
0.19
0.52
0.00
0.5
2
PCB-136
6
0.06
0.16
0.00
0.2
0.5
PCB-137
6
0.08
0.17
0.00
0.2
0.5
PCB-138+163+164
6
0.34
0.72
2
0.12
1
2
PCB-149+139
6
0.20
0.42
2
0.10
0.5
1
PCB-140
6
0.06
0.13
0.00
0.2
0.5
PCB-141
6
0.09
0.16
0.00
0.2
0.5
PCB-143
6
0.07
0.15
0.00
0.2
0.5
PCB-145
6
0.08
0.22
0.00
0.2
0.5
PCB-146
6
0.07
0.15
0.00
0.2
0.5
PCB-147
6
0.08
0.20
0.00
0.2
0.5
PCB-148
6
0.07
0.19
0.00
0.2
0.5
PCB-150
6
0.07
0.19
0.00
0.2
0.5
PCB-151
6
0.08
0.19
0.00
0.2
0.5
PCB-152
6
0.07
0.17
0.00
0.2
0.5
PCB-153
6
0.20
0.48
5
0.15
0.5
2
PCB-154
6
0.08
0.19
0.00
0.2
0.5
PCB-155
6
0.05
0.10
0.00
0.1
0.2
PCB-156
6
0.06
0.10
0.00
0.2
0.2
PCB-157
6
0.07
0.12
0.00
0.2
0.5
PCB-158+160
6
0.12
0.20
0.00
0.5
0.5
PCB-159
6
0.06
0.13
0.00
0.2
0.5
PCB-161
6
0.07
0.14
0.00
0.2
0.5
PCB-162
6
0.06
0.12
0.00
0.2
0.5
PCB-165
6
0.07
0.15
0.00
0.2
0.5
PCB-166
6
0.08
0.15
0.00
0.2
0.5
PCB-167
6
0.06
0.10
0.00
0.2
0.2
PCB-169
6
0.10
0.26
0.00
0.2
1
PCB-170+190
6
0.14
0.24
0.00
0.5
1
PCB-171
6
0.07
0.13
0.00
0.2
0.5
PCB-172+192
6
0.13
0.26
0.00
0.5
1
PCB-173
6
0.07
0.17
0.00
0.2
0.5
Method 1628 Multi-Laboratory Validation Study Report 51
April 2021
-------
Table 23. Solid Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDLs
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-174
6
0.09
0.22
0.00
0.2
0.5
PCB-175
6
0.07
0.16
0.00
0.2
0.5
PCB-176
6
0.06
0.12
0.00
0.2
0.5
PCB-177
6
0.07
0.16
0.00
0.2
0.5
PCB-178
6
0.09
0.22
0.00
0.2
0.5
PCB-179
6
0.05
0.10
0.00
0.2
0.2
PCB-180
6
0.07
0.14
2
0.05
0.2
0.5
PCB-181
6
0.07
0.16
0.00
0.2
0.5
PCB-187+182
6
0.15
0.35
0.00
0.5
1
PCB-183
6
0.08
0.17
0.00
0.2
0.5
PCB-184
6
0.05
0.11
0.00
0.2
0.2
PCB-185
6
0.07
0.15
0.00
0.2
0.5
PCB-186
6
0.07
0.16
0.00
0.2
0.5
PCB-188
6
0.06
0.12
0.00
0.2
0.5
PCB-189
6
0.06
0.11
0.00
0.2
0.2
PCB-191
6
0.07
0.12
0.00
0.2
0.5
PCB-193
6
0.07
0.12
0.00
0.2
0.5
PCB-194
6
0.18
0.32
0.00
0.5
1
PCB-195
6
0.93
3.16
0.00
2
10
PCB-195 (w/o Lab 8)
5
0.07
0.12
0.00
0.2
0.5
PCB-196+203
6
0.15
0.35
1
0.15
0.5
1
PCB-197
6
0.06
0.10
0.00
0.2
0.2
PCB-198
6
0.10
0.19
0.00
0.2
0.5
PCB-199
6
0.08
0.19
1
0.17
0.2
0.5
PCB-200
6
0.06
0.13
0.00
0.2
0.5
PCB-201
6
0.06
0.11
0.00
0.2
0.5
PCB-202
6
0.05
0.08
0.00
0.2
0.2
PCB-204
6
0.07
0.12
0.00
0.2
0.5
PCB-205
6
0.06
0.12
0.00
0.2
0.5
PCB-206
6
0.06
0.10
1
0.28
0.2
0.2
PCB-207
6
0.06
0.10
0.00
0.2
0.2
PCB-208
6
0.05
0.09
0.00
0.2
0.2
PCB-209
6
0.26
0.87
3
0.41
1
2
The analytes listed in the shaded rows are the congeners and coeluting congeners that represent the analytes that have direct
calibration data for this method.
Overall, the pooled MDLS values ranged from 0.05 ng/g to 0.93 ng/g. However, the pooled MDLS of 0.93
ng/g for PCB-195 was dramatically influenced by the results from Lab 8, at 3.16 ng/g. Review of that
MDLS value at CSRA and the laboratory did not uncover any calculation errors, but their MDLS was over
25 times the next highest MDLS for that congener in any of the other laboratories (0.12 ng/g). Therefore,
EPA decided to report the MDL values in Table 23 both with and without the PCB-195 result from Lab 8.
Without the MDLS from Lab 8, the pooled MDLS for PCB-195 is only 0.07 ng/g.
In total, 109 of the pooled MDLS values in Table 23 are less than or equal to 0.1 ng/g, 42 more are
between 0.1 and 0.2 ng/g, and 16 pooled MDLS values are greater than 0.2 ng/g. The pooled MDLS values
for all 167 analytes are presented graphically in Figure 3. The x-axis is in the same order as the analytes
are listed in Table 23.
The pooled ML for each of the 167 analytes was calculated as a multiplier of 3.18 times the pooled MDL
value, and then rounded to the nearest multiple of 1, 2, or 5, in order to facilitate future preparation of
calibration standards at the ML.
Method 1628 Multi-Laboratory Validation Study Report
52
April 2021
-------
The 167 pooled ML values ranged from 0.5 ng/g to 2.0 ng/g. In total, 1 pooled ML value is at 0.1 ng/g,
113 pooled ML values are at 0.2 ng/g, 41 pooled ML values are at 0.5 ng/g, 11 pooled ML values are at
1.0 ng/g, and 1 value is at 2.0 (for PCB-1).
Table 23 includes a column that provides the number of times that any one of the six laboratories in the
solids portion of the study reported a non-zero MDLb value. EPA evaluated those MDLb values and the
frequencies at which they were reported in a similar manner as for the aqueous MDL values. In total,
solid MDLb values were reported by one or more laboratories for 19 of the 167 analytes. Table 24
includes a summary of the frequencies at which the six laboratories reported solid MDLb values.
Table 24. Frequency of Solid MDLb values by Lab
Lab#
3
4
6
7
8
9
# MDLb values
0
1
7
11
9
1
As with the aqueous MDLs, the need to employ an MDLb was lab-specific. However, the patterns of
occurrences within each laboratory often differed between the aqueous and solid phases of the study. For
example, of the 9 solid MDLb values reported by Laboratory 8, only three of them overlapped with the
four MDLb values that they reported in the aqueous phase (PCB-149+139, PCB-153, and PCB-209).
Conversely, in Laboratory 7, 10 of the 11 solid MDLb values were for congeners that also had aqueous
MDLb values in that laboratory. This may be an indication that there is a common source in Laboratory 7
for those 10 congeners that affected both the aqueous and solid portion of the study, while a separate
source may be responsible for the other 14 congeners with MDLb values in the aqueous portion of the
study.
PCB-153 was the one congener for which five of the six laboratories in the solid phase of the study
reported an MDLb value. Those five MDLb values ranged from about 0.047 to 0.153 ng/g. The pooled
MDLS calculated from the study results for PCB-153 was 0.200 ng/g.
Overall, these MDL data demonstrate that background levels in typical laboratories are not a limiting
factor in the application of this method, but that some laboratories do have much better control of
background levels than others.
Method 1628 Multi-Laboratory Validation Study Report
53
April 2021
-------
Pooled MDLs (Solids)
CD
o
VIII
Coeluting group
Single Congener
CO
O
O
CN
O
O
O
Congener
Figure 3. Pooled Solid MDLS Values in Elution Order
Dotted lines and Roman numerals delineate the levels of chlorination. The red triangle symbols denote a group of two or more coeluting
congeners, while the blue diamond symbols denote single congeners.
Multi-Laboratory Validation Study Report
54
April 2021
-------
Tissue Sample MDL Determinations
The laboratories also determined the MDLs for the 209 PCB congeners in tissue matrices, using the 90:10
mixture of Ottawa sand and canola oil as the reference matrix.
CSRA provided all of the laboratories with instructions for preparing their tissue MDL samples based on
the MDL results from the single-laboratory validation study. The same 100 ng/mL spiking solution used
for the aqueous MDL was used for the aqueous and solid MDL determinations. Laboratories were
permitted to increase that concentration if their initial attempts to determine the MDL did not yield
detectable results for each of the congeners.
CSRA determined the pooled MDLs for tissue samples using the data from four laboratories. The results
are summarized in Table 25, below. The analytes listed in the shaded rows are the congeners and
coeluting congeners that represent the analytes that have direct calibration data for this method.
Table 25. Tissue Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDL
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-1
4
0.11
0.18
0.00
0.2
0.5
PCB-2
4
0.13
0.29
0.00
0.5
1.0
PCB-3
4
0.11
0.23
0.00
0.2
0.5
PCB-4+10
3
0.23
0.43
0.00
0.5
1.0
PCB-8+5
4
0.18
0.28
0.00
0.5
1.0
PCB-6
4
0.10
0.18
0.00
0.2
0.5
PCB-7+9
3
0.22
0.38
0.00
0.5
1.0
PCB-11
4
0.06
0.09
0.00
0.2
0.2
PCB-12+13
4
0.13
0.24
0.00
0.5
1.0
PCB-14
4
0.07
0.14
0.00
0.2
0.5
PCB-15
4
0.06
0.11
0.00
0.2
0.2
PCB-16+32
4
0.18
0.32
0.00
0.5
1.0
PCB-17
4
0.08
0.13
0.00
0.2
0.5
PCB-18
4
0.09
0.15
0.00
0.2
0.5
PCB-19
4
0.07
0.10
0.00
0.2
0.2
PCB-33+20+21
4
0.20
0.34
0.00
0.5
1.0
PCB-22
4
0.10
0.22
0.00
0.2
0.5
PCB-34+23
4
0.13
0.25
0.00
0.5
1.0
PCB-24+27
4
0.11
0.18
0.00
0.5
0.5
PCB-25
4
0.08
0.16
0.00
0.2
0.5
PCB-26
4
0.07
0.13
0.00
0.2
0.5
PCB-28
4
0.14
0.33
0.00
0.5
1.0
PCB-29
4
0.08
0.14
0.00
0.2
0.5
PCB-30
4
0.08
0.12
0.00
0.2
0.5
PCB-31
4
0.09
0.18
0.00
0.2
0.5
PCB-35
4
0.14
0.32
0.00
0.5
1.0
PCB-36
4
0.10
0.22
0.00
0.2
0.5
PCB-37
4
0.12
0.24
0.00
0.5
1.0
PCB-38
4
0.13
0.31
0.00
0.5
1.0
PCB-39
4
0.06
0.10
0.00
0.2
0.2
PCB-40
4
0.13
0.28
0.00
0.5
1.0
PCB-41+64
4
0.15
0.25
0.00
0.5
1.0
PCB-42
4
0.09
0.18
0.00
0.2
0.5
PCB-49+43
4
0.22
0.53
0.00
0.5
2.0
PCB-44
4
0.09
0.18
0.00
0.2
0.5
PCB-45
4
0.07
0.13
0.00
0.2
0.5
PCB-46
4
0.07
0.14
0.00
0.2
0.5
PCB-47+48+75
4
0.23
0.51
0.00
0.5
2.0
PCB-50
4
0.07
0.13
0.00
0.2
0.5
PCB-51
4
0.07
0.13
0.00
0.2
0.5
Multi-Laboratory Validation Study Report
55
April 2021
-------
Table 25. Tissue Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDL
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-52+73
4
0.24
0.48
0.00
1.0
2.0
PCB-53
4
0.05
0.08
0.00
0.2
0.2
PCB-54
4
0.06
0.09
0.00
0.2
0.2
PCB-55
4
0.10
0.24
0.00
0.2
1.0
PCB-56+60
4
0.09
0.16
1
0.03
0.2
0.5
PCB-57
4
0.07
0.15
0.00
0.2
0.5
PCB-58
4
0.09
0.22
0.00
0.2
0.5
PCB-59
4
0.08
0.13
0.00
0.2
0.5
PCB-74+61
4
0.12
0.25
0.00
0.5
1.0
PCB-62
4
0.06
0.09
0.00
0.2
0.2
PCB-63
4
0.08
0.19
0.00
0.2
0.5
PCB-65
4
0.07
0.12
0.00
0.2
0.5
PCB-66+80
4
0.16
0.34
0.00
0.5
1.0
PCB-67
4
0.07
0.14
0.00
0.2
0.5
PCB-68
4
0.10
0.22
0.00
0.2
0.5
PCB-69
4
0.06
0.12
0.00
0.2
0.5
PCB-70
4
0.09
0.20
0.00
0.2
0.5
PCB-71
4
0.07
0.11
0.00
0.2
0.5
PCB-72
4
0.11
0.26
0.00
0.2
1.0
PCB-76
4
0.08
0.16
0.00
0.2
0.5
PCB-77
4
0.09
0.22
0.00
0.2
0.5
PCB-78
4
0.11
0.26
0.00
0.5
1.0
PCB-79
4
0.06
0.10
1
0.02
0.2
0.2
PCB-81
4
0.07
0.14
0.00
0.2
0.5
PCB-82
4
0.08
0.19
0.00
0.2
0.5
PCB-83+109
4
0.10
0.18
0.00
0.2
0.5
PCB-84
4
0.06
0.14
0.00
0.2
0.5
PCB-85+120
4
0.17
0.37
0.00
0.5
1.0
PCB-97+86
4
0.07
0.13
0.00
0.2
0.5
PCB-87+115+116
4
0.22
0.39
0.00
0.5
1.0
PCB-88+121
4
0.13
0.23
0.00
0.5
0.5
PCB-90+101+89
4
0.10
0.20
0.00
0.2
0.5
PCB-91
4
0.05
0.08
0.00
0.2
0.2
PCB-92
4
0.05
0.11
0.00
0.2
0.2
PCB-95+93
4
0.10
0.18
1
0.02
0.2
0.5
PCB-94
4
0.03
0.06
0.00
0.1
0.2
PCB-96
4
0.05
0.08
0.00
0.1
0.2
PCB-98+102
4
0.12
0.20
0.00
0.5
0.5
PCB-99
4
0.06
0.11
0.00
0.2
0.2
PCB-100
4
0.06
0.10
0.00
0.2
0.2
PCB-103
4
0.06
0.10
0.00
0.2
0.2
PCB-104
4
0.05
0.08
0.00
0.2
0.2
PCB-105+127
4
0.14
0.32
0.00
0.5
1.0
PCB-118+106
4
0.12
0.24
0.00
0.5
1.0
PCB-107+108
4
0.13
0.28
0.00
0.5
1.0
PCB-110
4
0.06
0.11
0.00
0.2
0.5
PCB-111 + 117
4
0.16
0.32
0.00
0.5
1.0
PCB-112
4
0.06
0.11
0.00
0.2
0.5
PCB-113
4
0.04
0.08
0.00
0.1
0.2
PCB-114
4
0.07
0.16
0.00
0.2
0.5
PCB-119
4
0.09
0.18
0.00
0.2
0.5
PCB-122
4
0.05
0.11
0.00
0.2
0.2
PCB-123
4
0.06
0.12
0.00
0.2
0.5
PCB-124
4
0.06
0.15
0.00
0.2
0.5
PCB-125
4
0.05
0.11
0.00
0.2
0.5
PCB-126
4
0.10
0.19
0.00
0.2
0.5
Multi-Laboratory Validation Study Report
56
April 2021
-------
Table 25. Tissue Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDL
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-128
4
0.08
0.16
1
0.02
0.2
0.5
PCB-129
4
0.08
0.18
0.00
0.2
0.5
PCB-130
4
0.06
0.11
0.00
0.2
0.5
PCB-131 + 142
4
0.19
0.44
0.00
0.5
1.0
PCB-132+168
4
0.14
0.26
0.00
0.5
1.0
PCB-133
4
0.07
0.15
0.00
0.2
0.5
PCB-134
4
0.06
0.11
0.00
0.2
0.2
PCB-144+135
4
0.11
0.21
0.00
0.5
0.5
PCB-136
4
0.05
0.08
0.00
0.1
0.2
PCB-137
4
0.06
0.12
0.00
0.2
0.5
PCB-138+163+164
4
0.17
0.32
1
0.02
0.5
1.0
PCB-149+139
4
0.12
0.25
0.00
0.5
1.0
PCB-140
4
0.06
0.10
0.00
0.2
0.2
PCB-141
4
0.07
0.14
0.00
0.2
0.5
PCB-143
4
0.07
0.15
0.00
0.2
0.5
PCB-145
4
0.06
0.10
1
0.04
0.2
0.2
PCB-146
4
0.04
0.07
0.00
0.1
0.2
PCB-147
4
0.07
0.13
0.00
0.2
0.5
PCB-148
4
0.06
0.10
0.00
0.2
0.2
PCB-150
4
0.11
0.23
0.00
0.2
0.5
PCB-151
4
0.05
0.08
0.00
0.2
0.2
PCB-152
4
0.06
0.09
0.00
0.2
0.2
PCB-153
4
0.09
0.15
3
0.07
0.2
0.5
PCB-154
4
0.06
0.09
0.00
0.2
0.2
PCB-155
4
0.05
0.11
0.00
0.2
0.5
PCB-156
4
0.07
0.13
1
0.04
0.2
0.5
PCB-157
4
0.08
0.14
1
0.01
0.2
0.5
PCB-158+160
4
0.13
0.23
0.00
0.5
0.5
PCB-159
4
0.06
0.10
0.00
0.2
0.2
PCB-161
4
0.07
0.13
0.00
0.2
0.5
PCB-162
4
0.05
0.09
1
0.01
0.2
0.2
PCB-165
4
0.04
0.08
1
0.01
0.1
0.2
PCB-166
4
0.09
0.21
0.00
0.2
0.5
PCB-167
4
0.06
0.10
0.01
0.2
0.2
PCB-169
4
0.06
0.13
0.00
0.2
0.5
PCB-170+190
3
0.15
0.28
0.00
0.5
1.0
PCB-171
4
0.13
0.32
0.00
0.5
1.0
PCB-172+192
4
0.12
0.20
0.00
0.5
0.5
PCB-173
4
0.21
0.52
0.00
0.5
2.0
PCB-174
4
0.08
0.13
0.00
0.2
0.5
PCB-175
4
0.08
0.17
0.00
0.2
0.5
PCB-176
4
0.07
0.14
0.00
0.2
0.5
PCB-177
4
0.06
0.12
0.00
0.2
0.5
PCB-178
4
0.09
0.18
0.00
0.2
0.5
PCB-179
4
0.05
0.09
0.00
0.2
0.2
PCB-180
4
0.19
0.45
1
0.03
0.5
1.0
PCB-181
4
0.08
0.17
0.00
0.2
0.5
PCB-187+182
4
0.13
0.24
0.00
0.5
1.0
PCB-183
4
0.07
0.15
0.00
0.2
0.5
PCB-184
4
0.06
0.10
0.00
0.2
0.2
PCB-185
4
0.06
0.10
0.00
0.2
0.2
PCB-186
4
0.05
0.08
0.00
0.2
0.2
PCB-188
4
0.05
0.09
0.00
0.2
0.2
PCB-189
4
0.07
0.12
0.00
0.2
0.5
PCB-191
4
0.05
0.10
0.00
0.2
0.2
PCB-193
4
0.06
0.11
0.00
0.2
0.2
Multi-Laboratory Validation Study Report
57
April 2021
-------
Table 25. Tissue Sample Pooled MDL and ML Results (ng/g)
Analyte
# Labs
Pooled MDL
Max. MDLs
# MDLb
Max. MDLb
Pooled ML
Max. ML
PCB-194
3
0.10
0.17
0.00
0.2
0.5
PCB-195
4
0.08
0.15
0.00
0.2
0.5
PCB-196+203
4
0.17
0.32
0.00
0.5
1.0
PCB-197
4
0.04
0.08
0.00
0.1
0.2
PCB-198
3
0.06
0.09
0.00
0.2
0.2
PCB-199
4
0.10
0.20
0.00
0.2
0.5
PCB-200
4
0.05
0.10
0.00
0.2
0.2
PCB-201
4
0.14
0.33
0.00
0.5
1.0
PCB-202
4
0.05
0.09
0.00
0.2
0.2
PCB-204
4
0.09
0.20
0.00
0.2
0.5
PCB-205
4
0.11
0.27
0.00
0.5
1.0
PCB-206
4
0.06
0.10
0.00
0.2
0.2
PCB-207
4
0.06
0.10
0.00
0.2
0.2
PCB-208
4
0.05
0.10
0.00
0.2
0.2
PCB-209
4
0.09
0.20
0.00
0.2
0.5
The analytes listed in the shaded rows are the congeners and coeluting congeners that represent the analytes that have direct
calibration data for this method.
Initial examination of the MDLS values from all four laboratories showed good consistency for the
majority of the congeners, with pooled MDLS values ranging from about 0.035 to 0.60 for 162 of the 167
analytes. The initial pooled MDLS values for PCB-4+10, PCB-7+9, PCB-170+190, PCB-194 and PCB-
198 were significantly higher than those for all the other congeners. Closer examination revealed that
those higher values were driven by the MDLS values from Lab 6, which ranged from 1.55 to 10.8 ng/g for
those five analytes. All of those results were reviewed in detail, and while no obvious issues were
identified, the project team opted to recalculate the pooled MDLS values for those five analytes without
the results from Lab 6. Those recalculated values are the ones presented in Table 25. The pooled MDLs
values ranged from about 0.035 to 0.23 ng/g for all 167 analytes. The pooled MDLS values for all 167
analytes are presented graphically in Figure 4. The x-axis is in the same order as the analytes are listed in
Table 25.
The pooled ML for each of the 167 analytes was calculated as a multiplier of 3.18 times the pooled MDL
value, and then rounded to the nearest multiple of 1, 2, or 5, in order to facilitate future preparation of
calibration standards at the ML.
The 167 pooled ML values ranged from 0.1 ng/g to 1.0 ng/g. In total, 7 pooled ML values are at 0.1 ng/g,
116 pooled ML values are at 0.2 ng/g, 43 pooled ML values are at 0.5 ng/g, and 1 pooled ML value is at
1.0 ng/g.
Table 25 includes a column includes that provides the number of times that any one of the four
laboratories in the tissue portion of the study reported a non-zero MDLb value. EPA evaluated those
MDLb values and the frequencies at which they were reported in a similar manner as for the other two
matrices. In total, tissue MDLb values were reported by one or more laboratories for 13 of the 167
analytes. Table 26 includes a summary of the frequencies at which the four laboratories reported tissue
MDLb values.
Table 26. Frequency of Tissue MDLb values by Lab
Lab#
2
4
6
9
# MDLb values
0
1
6
8
Multi-Laboratory Validation Study Report
58
April 2021
-------
Pooled MDLs (Tissue)
IX X
VIII
Coeluting group
Single Congener
_i
O
2
Congener
Figure 4. Pooled Tissue MDLS Values in Elution Order
Dotted lines and Roman numerals delineate the levels of chlorination. The red triangle symbols denote a group of two or more coeluting
congeners, while the blue diamond symbols denote single congeners.
Multi-Laboratory Validation Study Report
59
April 2021
-------
The patterns of tissue MDLb values differed across the four laboratories, as was found for the other two
matrices. The patterns also differed between those matrices within some of the four laboratories in the
tissue portion of the study. For example, Laboratory 9 only reported one MDLb in the aqueous and solid
portions of the study, for PCB-153 in both cases. They also reported a tissue MDLb value for PCB-153,
but they reported seven more MDLb values in tissue, all of them occurring only at Laboratory 9.
In contrast, Laboratory 6 reported seven MDLb values in the solid portion of the study and six in the
tissue portion. Four of those six MDLb values in tissue were unique to Laboratory 6, but other than
PCB-153, none of them overlapped with the solid MDLb values from that laboratory.
Once again, PCB-153 was the one congener in common across the majority of the laboratories in the
study, with three of the four laboratories reporting an MDLb value for this congener. Those three MDLb
values ranged from about 0.04 to 0.07 ng/g, and the pooled MDLS values for PCB-153 in Table 25 is
0.09 ng/g.
As with the other two matrices, these MDL data demonstrate that background levels in typical
laboratories are not a limiting factor in the application of this method, but that some laboratories have
better control of background levels than others.
Multi-Laboratory Validation Study Report
60
April 2021
-------
7. Unspiked Sample Analyses
The results for all of the unspiked sample analyses are summarized in the section below.
Wastewater Samples
Each of the seven laboratories that completed the wastewater portion of the study analyzed all nine of the
wastewater samples as received, unspiked. As described in the draft procedure, all of the wastewater
samples received the copper and Florisil cleanups. Wastewater #2 presented significant analytical
challenges to all of the laboratories. Four of the laboratories applied additional cleanups, and two other
laboratories diluted an aliquot of the final extract of wastewater #2 before analysis. The various cleanups
applied by each laboratory as presented in Table 27.
Table 27. Summary of Wastewater Sample Cleanups
Sample Clean-up
Lab 1
Lab 3
Lab 42
Lab 6
Lab 7
Lab 8
Lab 9
Recon
Acid - Sulfuric
X
X
Alumina
X
X
Copper
X
X
X
X
X
X
X
X
FlorisilŽ
X
X
X
X
X
X
X
X
GPC
X
X
X
Silica
X
Dilution1
X
X
1 Dilution was only applied to Sample #2
2 Full list of cleanups only applied to Sample #2
The results of the wastewater sample analyses are summarized in Table 28. Given the total number of
analytes in the draft method and the fact that the purpose of the analyses of the unspiked samples was to
allow the laboratories to calculate their matrix spike recoveries and not to characterize the unspiked
samples otherwise, we have limited the results in the table to those congeners that were spiked into the
matrix spike samples later (see Section 8 of this report). We also only included those detected congeners
that met the identification criteria in the draft procedure and where the results were at least 5 times the
results for the associated method blank. The results from the reconnaissance analyses are also included in
Table 28.
Table 28. Unspiked Wastewater Sample Results in ng/L
Analyte
Lab 1
Lab 3
Lab 4
Lab 6
Lab 7
Lab 8
Lab 9
Recon
Wastewater #2
PCB-4+10
2.80
2.43
PCB-8+5
3.64
PCB-18
17.95
10.96
10.8
10.5
PCB-19
0.675
2.13
2.21
2.07
PCB-28
33.2
11.64
7.35
5.68
PCB-31
35.9
7.54
4.97
PCB-36
163
PCB-41+64
23.59
3.34
2.45
PCB-44
5.53
7.14
5.12
PCB-52+73
12.81
9.31
9.99
7.73
PCB-70
2.10
PCB-90+101+89
2.09
2.81
2.93
1.91
PCB-95+93
3.49
2.81
3.96
2.37
PCB-99
1.36
1.27
0.827
PCB-105+127
6.62
PCB-110
2.23
2.38
1.69
PCB-132+168
1.07
0.419
Multi-Laboratory Validation Study Report
61
April 2021
-------
Table 28. Unspiked Wastewater Sample Results in ng/L
Analyte
Lab 1
Lab 3
Lab 4
Lab 6
Lab 7
Lab 8
Lab 9
Recon
PCB-138+163+164
1.77
1.08
PCB-149+139
1.74 B
1.55
2.00
0.875
PCB-153
3.38 B*
2.04 B*
12.28
1.78 B*
0.909
PCB-177
0.319
PCB-180
0.776
0.825
0.415
PCB-182+187
0.28
PCB-199
0.582
0.954
PCB-206
0.388
0.495
PCB-209
0.235
Wastewater #3 - no congeners found
Wastewater #4
PCB-28
1.24
PCB-180
3.88
Wastewater #5
PCB-209
0.097
Wastewater #6
PCB-1
1.98 B*
PCB-11
0.822 B
5.08
PCB-70
1.26 B
PCB-90+101+89
0.9
0.998
PCB-99
0.358
PCB-110
0.756
0.9
0.788
0.777
0.623
PCB-118+106
0.690
0.622
PCB-138+163+164
0.954
1.18
0.959
PCB-149+139
0.653 B
0.72
0.464
PCB-153
2.12 B*
1.18
1.02
0.535
PCB-180
0.464
0.51
0.459
0.413
PCB-182+187
0.215
Wastewater #7
PCB-3
13.37
PCB-31
3.32
PCB-66+80
11.67 B
PCB-110
0.412
1.58
PCB-153
2.51 B*
2.67
0.863
PCB-177
0.693
PCB-180
1.29
0.719
PCB-189
0.869
PCB-199
0.908
PCB-205
0.913
PCB-206
0.931
PCB-208
0.543
PCB-209
0.820
Wastewater #8 - no congeners found
Wastewater #9
PCB-153
0.882
PCB-180
2.91
Wastewater #10
PCB-153
0.438
B* Congener found in the associated method blank. Sample result is between 5x and lOx the blank value.
B Congener found in the associated method blank. Sample result is >10x the blank value.
The results for the unspiked samples varied across both the samples and the laboratories. Two of the
samples (#3 and #8) had no congeners detected by any laboratory, including the reconnaissance lab. Four
Multi-Laboratory Validation Study Report
62
April 2021
-------
of the samples had only one or two congeners reported by any laboratory. Those results are not
surprising, since PCBs are only slightly soluble in water, and are more likely to be associated with the
suspended particulates in the wastewater (IARC, 2016). The aqueous samples distributed for this study
were thoroughly mixed, but it is impossible to perfectly distribute the amount of particulates present in
hundreds of 1-L samples. Samples #6 and #7 had 12 and 13 congeners detected by any of the
laboratories. Sample #2 presented the greatest analytical challenge, but had the largest number congeners
detected, 39 in all, including the coeluting congeners.
Labs 3, 7, and 8 found the smallest numbers of congeners across all nine samples. The performance of
Labs 7 and 8 was most affected by their dilutions of Wastewater #2, which decreased their overall
sensitivity to the point that many of the congeners present at low levels could not be detected. Lab 3 only
detected three congeners in Wastewater #2, and none in any of the other eight samples. Lab 1 had more
congeners with potential method blank issues than the other laboratories, however, even then, many of the
concentration that they reported were similar to the results from the other labs that reported those same
congeners. Labs 1, 6, and 9 reported results for many of these congeners that were quite similar to one
another and similar to the results from the reconnaissance analyses by the laboratory that developed the
draft procedure. In many of the cases where a minority of laboratories detected a congener, the congener
was present at a concentration near the lowest detection limits of any of the laboratories and therefore
laboratories with higher detection limits could not detect the congener. That said, the purpose of the
study was not to assess the performance of the individual laboratories.
Sediment Samples
Each of the six laboratories that completed the sediment portion of the study analyzed the three sediment
samples as received, unspiked. As described in the draft procedure, all of the sediment samples received
the copper and Florisil cleanups. All of the laboratories applied additional cleanups, and four of the
laboratories diluted an aliquot of the final extract of sediment #1 and #3 before analysis. The various
cleanups applied by each laboratory are presented in Table 29.
Table 29. Summary of Sediment Sample Cleanups
Sample Clean-up
Lab 31
Lab 42
Lab 6
Lab 7
Lab 84
Lab 9
Recon3
Acid - Sulfuric
X
X
Alumina
X
X
X
X
X
X
X
Copper
X
X
X
X
X
X
X
FlorisilŽ
X
X
X
X
X
X
X
GPC
X
X
X
Silica
X
X
X
X
X
X
Dilution
X
X
X
X
1 Sample #1 only had alumina, silica and acid cleanup performed in addition to copper and Florisil
2 Sample #1 had all procedures except acid cleanup
3 Sample #1 had GPC and sample #3 had dilution in addition to the other cleanup procedures
4 Dilution on samples #1 and #3
The results of the sediment sample analyses as summarized in Table 30. As with the wastewater sample
results, we have limited the results in the table to those congeners that were spiked into the matrix spike
samples later. We also only included those detected congeners that met the identification criteria in the
draft procedure and where the results were at least 5 times the results for the associated method blank.
The results from the reconnaissance analyses are also included in Table 30.
Note: For ease of comparison, as well as preparation of the table, all of the sediment sample results are
reported to three decimal places. However, the results have at most three significant figures, so
trailing zeroes and the figures past one decimal place in results larger than 10 are not significant.
Multi-Laboratory Validation Study Report
63
April 2021
-------
Table 30. Unspiked Sediment Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Lab 8
Lab 9
Recon
Sediment #1
PCB-1
4.216
5.978
8.030
29.300
PCB-3
3.018
2.430
5.920
PCB-4+10
55.164
24.752
16.085
46.241
32.900
82.200
PCB-8+5
12.096
8.367
4.357
11.994
0.292
11.100
61.600
PCB-11
7.148
1.139
8.630
PCB-15
7.214
4.320
17.615
0.544
12.300
25.600
PCB-18
16.350
17.023
4.688
20.374
0.813
12.400
21.900
PCB-19
7.497
2.904
8.857
6.780
14.300
PCB-28
30.527
14.550
6.526
4.221
14.600
13.700
PCB-31
23.602
26.389
6.507
30.437
5.789
20.000
32.300
PCB-37
3.371
3.820
PCB-41+64
24.067
2.665
1.280
5.360
6.840
PCB-44
32.261
17.811
6.857
20.459
1.295
15.800
19.200
PCB-52+73
61.312
15.129
44.642
8.984
35.700
45.100
PCB-54
0.055
0.152
0.212
PCB-66+80
14.545
6.507
19.108
3.445
12.300
14.400
PCB-70
28.439
16.489
8.290
23.275
4.023
18.300
18.500
PCB-72
3.357
2.992
0.911
0.936
PCB-74+61
13.308
2.537
1.884
5.450
5.970
PCB-77
1.359
PCB-85+120
6.470
2.794
6.533
1.464
6.080
6.410
PCB-90+101+89
62.736
40.945
17.665
47.169
12.772
38.300
40.300
PCB-95+93
41.459
31.133
12.390
34.713
6.975
28.200
31.700
PCB-96
3.923
0.129
0.281
0.342
PCB-98+102
1.623
0.460
1.544
1.350
1.500
PCB-99
32.689
21.236
8.934
24.619
5.981
18.800
20.300
PCB-104
5.320
PCB-105+127
9.991
4.651
11.417
2.211
9.650
10.400
PCB-110
63.369
42.933
18.971
52.013
8.024
42.200
44.300
PCB-118+106
35.405
15.938
44.139
10.387
34.600
37.200
PCB-132+168
12.357
12.356
7.555
3.084
10.500
12.800
PCB-138+163+164
48.072
41.722
15.956
42.536
11.589
35.400
38.700
PCB-147
6.616
1.270
0.331
1.098
0.785
0.845
PCB-149+139
32.112
25.609
10.110
28.504
10.199
23.000
23.500
PCB-153
35.076
29.746
12.224
35.575
27.700
28.900
PCB-155
4.653
PCB-156
4.194
1.544
4.714
1.582
3.640
4.290
PCB-169
11.519
PCB-177
8.534
4.094
1.324
4.018
3.120
3.260
PCB-180
24.060
18.796
6.728
20.305
5.997
15.900
16.700
PCB-184
3.736
PCB-187+182
23.063
14.771
5.221
16.835
5.449
12.000
12.200
PCB-188
5.296
PCB-189
13.448
0.211
PCB-199
16.843
13.649
6.654
22.104
11.800
10.800
PCB-202
3.336
1.305
5.778
1.491
2.660
2.740
PCB-205
0.110
0.357
5.000
PCB-206
22.865
13.933
5.074
28.552
5.976
11.700
11.700
PCB-208
4.282
1.857
8.653
3.400
3.600
PCB-209
19.345
11.983
4.210
16.062
6.331
10.700
10.500
Multi-Laboratory Validation Study Report
64
April 2021
-------
Table 30. Unspiked Sediment Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Lab 8
Lab 9
Recon
Sediment #2
PCB-1
12.218
2.379
1.790
8.440
PCB-3
7.279
2.477
1.052
1.652
1.277
1.020
3.310
PCB-4+10
25.143
7.643
7.391
13.956
8.529
5.960
25.600
PCB-8+5
7.174
3.388
4.253
1.810
1.600
6.550
PCB-11
0.109
PCB-15
3.124
3.001
2.533
9.260
PCB-18
4.394
3.530
2.538
3.553
1.910
1.460
7.840
PCB-19
3.725
2.256
1.416
3.099
1.503
1.140
5.720
PCB-28
2.644
2.258
1.527
1.540
4.970
PCB-31
6.041
5.291
4.306
8.195
3.591
2.350
11.900
PCB-37
0.192
0.691
PCB-41+64
3.800
0.410
1.930
PCB-44
2.365
2.738
1.725
3.068
1.528
1.140
4.780
PCB-52+73
6.984
4.934
3.717
6.286
3.841
2.620
11.300
PCB-54
0.046
0.131
PCB-66+80
0.898
1.316
0.593
2.300
PCB-70
1.197
0.729
1.188
0.777
0.716
2.050
PCB-74+61
0.309
0.334
0.336
1.270
PCB-90+101+89
1.696
0.827
0.901
1.130
2.410
PCB-95+93
2.266
1.928
1.052
2.328
1.110
1.200
3.640
PCB-96
0.068
0.117
PCB-98+102
0.194
0.354
PCB-99
0.951
0.393
0.413
0.514
1.170
PCB-105+127
0.309
0.357
0.726
PCB-107+108
1.556
PCB-110
2.166
1.248
3.102
1.360
3.980
PCB-118+106
1.727
2.067
PCB-126
17.378
PCB-132+168
1.369
0.287
0.401
0.643
PCB-138+163+164
1.908
2.201
1.178
2.586
1.022
1.390
PCB-147
1.888
PCB-149+139
2.228
1.856
0.996
2.299
1.270
1.950
PCB-153
1.721
1.252
1.206
3.156
1.042
1.630
2.220
PCB-169
4.488
PCB-177
0.546
1.290
0.210
0.448
0.575
PCB-180
2.795
3.104
1.851
3.448
2.850
2.820
PCB-187+182
2.645
2.568
2.937
1.270
2.430
2.440
PCB-199
1.892
2.859
1.487
3.615
1.652
3.000
2.110
PCB-202
0.622
0.365
0.372
0.718
PCB-204
0.266
0.556
PCB-206
2.993
1.529
3.536
3.190
2.590
PCB-208
0.870
0.561
0.897
0.808
PCB-209
1.796
2.479
0.603
3.060
2.040
2.150
Sediment #3
PCB-1
16.195
4.130
0.534
10.400
15.500
PCB-3
18.567
15.168
3.732
14.309
1.368
9.990
14.200
PCB-4+10
77.696
72.225
17.161
70.259
5.882
52.600
72.900
PCB-8+5
51.276
45.722
8.106
41.243
6.909
33.000
40.200
PCB-11
1.775
1.744
1.170
1.830
PCB-15
72.781
63.165
15.341
68.986
9.816
46.300
67.400
PCB-18
38.448
38.251
5.950
33.639
8.194
24.700
35.500
PCB-19
23.740
21.044
4.237
20.558
2.502
14.600
20.400
Multi-Laboratory Validation Study Report
65
April 2021
-------
Table 30. Unspiked Sediment Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Lab 8
Lab 9
Recon
PCB-28
90.003
70.738
13.398
67.861
10.654
47.700
64.700
PCB-31
61.234
66.985
12.465
63.961
10.212
43.000
62.100
PCB-37
10.918
2.065
10.314
1.675
6.790
9.940
PCB-41+64
30.154
30.629
29.137
6.131
19.400
25.700
PCB-44
54.141
54.151
8.810
51.757
9.454
35.100
50.600
PCB-52+73
78.735
79.641
12.695
75.627
11.961
52.800
77.400
PCB-54
0.990
0.964
0.184
0.901
0.151
0.665
0.909
PCB-66+80
53.547
55.274
8.932
53.301
7.753
36.700
55.100
PCB-70
48.891
50.594
8.014
49.165
8.262
34.600
46.000
PCB-72
1.452
0.184
0.752
1.080
PCB-74+61
31.221
31.620
5.216
30.469
4.087
20.900
28.400
PCB-77
6.592
6.505
0.713
4.030
0.987
3.840
5.450
PCB-79
0.332
PCB-85+120
8.019
7.212
1.101
6.901
1.161
4.890
6.700
PCB-90+101+89
24.290
22.425
3.258
21.691
3.848
15.300
20.300
PCB-95+93
20.054
19.563
2.432
18.310
5.178
13.400
18.300
PCB-96
0.558
0.643
0.092
0.601
0.197
0.437
0.578
PCB-98+102
2.971
2.635
0.352
2.368
0.561
15.300
2.500
PCB-99
17.003
15.582
2.340
14.895
2.287
10.500
13.900
PCB-104
0.118
0.085
PCB-105+127
12.408
11.445
1.728
11.293
2.414
7.790
9.840
PCB-107+108
2.023
0.260
1.785
1.730
PCB-110
33.250
30.734
3.824
28.527
20.400
27.300
PCB-118+106
24.848
22.864
3.288
22.351
3.682
15.400
20.500
PCB-132+168
4.971
3.257
0.382
3.067
1.256
2.220
3.070
PCB-138+163+164
4.042
10.409
1.514
10.159
1.821
7.030
9.780
PCB-147
0.505
0.410
0.061
0.296
0.365
PCB-149+139
8.447
7.635
0.979
7.341
2.003
5.200
6.830
PCB-152
0.039
PCB-153
7.376
7.866
1.269
8.167
1.528
5.630
7.150
PCB-156
1.091
1.133
0.153
1.065
0.361
0.727
0.952
PCB-169
0.639
PCB-177
1.908
1.803
0.229
1.612
1.250
1.660
PCB-180
6.794
6.599
0.994
6.118
1.198
4.980
6.230
PCB-187+182
4.143
3.744
0.489
3.323
0.802
2.750
3.510
PCB-189
0.193
0.031
0.115
0.142
PCB-199
1.774
2.026
0.306
1.894
1.500
1.710
PCB-202
0.274
0.331
0.031
0.348
0.230
0.273
PCB-205
0.163
0.112
0.129
PCB-206
0.890
0.866
0.671
0.750
PCB-208
0.179
0.131
0.179
PCB-209
0.234
0.203
0.236
In general, all of the laboratories who completed the sediment portion of the study reported more
congeners in these three sediment samples than they did in the wastewater samples, which would be
expected since the sediments came from contaminated sites. For example, Laboratory 3, that only
reported 3 congeners across all 8 wastewater samples, reported 90 results for the three sediment samples.
The other five laboratories reported between 84 and 119 congeners across all three samples.
As with the unspiked wastewater sample results, there is some variability across the laboratories. Some
of that variability may be due to the challenge of homogenizing a bulk sediment sample. In many of the
cases where a minority of laboratories detected a congener, the congener was present at a concentration
Multi-Laboratory Validation Study Report
66
April 2021
-------
near the lowest detection limits of any of the laboratories and therefore laboratories with higher detection
limits could not detect the congener. However, Laboratory 6 and Laboratory 8 often reported lower
values, and sometimes much lower values for some congeners than any other laboratory, suggesting that
they have a low bias. For example, compare their results in Table 29 for PCB-18, PCB-31, and PCB-44
in Sediment # 1. Similar comparisons for other congeners indicate that the differences between the results
from Laboratory 6 and Laboratory 8 and the other four laboratories are not related to concentration (for
example, see their results in Sediment #3 for PCB-4+10 and PCB-54). Despite a review of their results,
CSRA did not identify any obvious or likely causes.
Biosolids Samples
Each of the four laboratories that completed the biosolids portion of the study analyzed the three biosolids
samples as received, unspiked. As described in the draft procedure, all of the biosolids samples received
the copper and Florisil cleanups. All of the laboratories applied additional cleanups, and two of the
laboratories diluted an aliquot of the final extract before analysis. The various cleanups applied by each
laboratory are presented in Table 31.
Table 31. Summary of Biosolids Sample Cleanups
Sample Clean-up
Lab 3
Lab 41
Lab 6
Lab 72
Recon3
Alumina
X
X
X
X
Copper
X
X
X
X
X
FlorisilŽ
X
X
X
X
X
GPC
X
X
X
Silica
X
X
X
X
Dilution
X
X
1 Dilution and GPC was applied to Samples #1 and #3
2 Dilution for all 3 biosolid samples
3 GPC was performed on Samples #1 and #3
As with the wastewater sample results, we have limited the results in Table 32 to those congeners that
were spiked into the matrix spike samples later. We also only included those detected congeners that met
the identification criteria in the draft procedure and where the results were at least 5 times the results for
the associated method blank. The results from the reconnaissance analyses are also included in the table.
Table 32. Unspiked Biosolids Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Recon
Biosolids Sample #1
PCB-1
2.690
1.300# B
PCB-3
2.682#
2.610
PCB-11
3.465
2.180#
PCB-15
0.383#
PCB-18
1.475
0.848
PCB-28
3.023
1.270
PCB-31
2.138
1.270
PCB-41+64
1.040#
PCB-44
4.350
2.110#
PCB-52+73
3.347
6.857
9.490
3.540#
PCB-54
0.074
PCB-66+80
3.244
1.950#
PCB-70
5.013
3.040#
PCB-74+61
2.654
1.280#
PCB-77
0.227
PCB-79
14.331
PCB-85+120
1.180
0.958#
PCB-90+101+89
6.293#
7.602
9.216
13.840
6.140#
Multi-Laboratory Validation Study Report
67
April 2021
-------
Table 32. Unspiked Biosolids Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Recon
PCB-95+93
3.801
5.588
6.488
13.260
4.790#
PCB-96
0.147
PCB-98+102
0.295
0.149#
PCB-99
2.300#
3.187
3.686
6.550
2.380#
PCB-104
0.147
PCB-105+127
2.507
1.940#
PCB-107+108
0.369#
0.302#
PCB-110
9.510#
7.807
7.004
15.050
6.360#
PCB-118+106
6.142#
6.755
5.530
12.920
5.000#
PCB-126
0.118#
PCB-132+168
2.520
2.359
5.260
1.910#
PCB-138+163+164
7.343
7.434
6.857
16.320
6.490#
PCB-147
0.147
0.127#
PCB-149+139
3.658
5.233
4.719
11.630
3.810#
PCB-153
5.193#
6.997
5.308 B
15.14
4.570#
PCB-155
0.295
0.163#
PCB-156
0.811
0.687#
PCB-169
8.739#
PCB-177
0.516
2.450
0.464
PCB-180
1.926
2.375
1.991
1.970
PCB-184
0.442
0.330
PCB-187+182
0.699#
1.555
1.327
5.410
1.190
PCB-188
0.074
PCB-189
0.074
PCB-199
0.361 #
0.590
2.820
0.626#
PCB-202
0.147
0.150#
PCB-206
0.295
0.321 #
PCB-208
0.147
PCB-209
0.221
Biosolids Sample #2
PCB-1
1.790# B
PCB-3
15.700#
PCB-11
2.077#
PCB-18
0.651
PCB-28
0.855
PCB-31
0.651
PCB-44
1.344
PCB-52+73
2.239
PCB-54
0.081
PCB-66+80
0.774
PCB-70
1.140
PCB-72
0.244#
PCB-74+61
0.651
PCB-85+120
0.244
PCB-90+101+89
0.719#
1.995
0.971 #
PCB-95+93
1.425
0.733#
PCB-98+102
0.0814#
PCB-99
0.774
0.383#
PCB-104
0.081
PCB-105+127
0.448
PCB-107+108
0.081
PCB-110
0.538
1.221
0.984#
PCB-118+106
0.643#
0.896
0.814#
Multi-Laboratory Validation Study Report
68
April 2021
-------
Table 32. Unspiked Biosolids Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Recon
PCB-132+168
0.407
PCB-138+163+164
1.546#
1.099
1.180#
PCB-149+139
0.354
0.692
0.643#
PCB-152
0.041
PCB-153
0.258#
0.977 B
0.698#
PCB-155
0.122
PCB-156
0.122
PCB-180
0.285
0.301
PCB-184
PCB-187+182
0.204
0.180#
PCB-209
0.244
Biosolids Sample #3
PCB-1
2.34 #B
PCB-3
5.367
PCB-4+10
1.282#
0.998#
PCB-11
3.609#
2.890#
PCB-15
0.759#
PCB-18
2.172
2.280
PCB-19
0.677
0.594
PCB-28
3.811
3.110
PCB-31
2.885
2.990
PCB-36
0.670 #B*
PCB-41+64
2.244
1.770#
PCB-44
4.843
4.250#
PCB-52+73
11.133
9.473
8.760#
PCB-54
0.783
0.710#
PCB-66+80
16.010#
4.879
3.830#
PCB-70
8.728#
6.588
5.810#
PCB-72
2.665#
0.080#
PCB-74+61
2.350
2.100#
PCB-77
0.319#
PCB-85+120
2.101
2.040#
PCB-90+101+89
16.097
14.623
15.135
6.750
14.000#
PCB-95+93
10.841
9.490
10.292
7.840
9.900#
PCB-96
0.099#
PCB-98+102
10.760
0.677
0.570#
PCB-99
6.584
6.268
5.600#
PCB-104
0.214
0.205#
PCB-105+127
7.555#
4.701
4.410#
PCB-107+108
0.812#
PCB-110
21.134#
13.481
13.925
5.680
13.200#
PCB-118+106
12.278#
11.998
11.503
4.210
11.100#
PCB-132+168
6.054
4.760#
PCB-138+163+164
18.776
17.450
7.300
16.70#
PCB-147
0.783
0.648#
PCB-149+139
12.522
11.422
11.396
5.13
9.830#
PCB-152
0.071
0.047#
PCB-153
17.201
15.971
13.141 B
8.100
12.300#
PCB-155
0.178
0.175#
PCB-156
1.709
1.730#
PCB-166
1.608#
0.058#
PCB-169
9.737#
PCB-177
1.924
2.208
1.800
Multi-Laboratory Validation Study Report
69
April 2021
-------
Table 32. Unspiked Biosolids Sample Results in ng/g (dw)
Analyte
Lab 3
Lab 4
Lab 6
Lab 7
Recon
PCB-180
7.839
8.583
2.290
7.470
PCB-184
0.321
0.269
PCB-187+182
4.927
5.306
1.570
4.330
PCB-188
0.036
PCB-189
0.117
PCB-199
2.384
2.525
1.709
2.280#
PCB-202
0.499
0.458#
PCB-204
0.147
0.285
PCB-205
0.071
0.697#
PCB-206
1.229
1.318
1.200#
PCB-208
0.427#
0.391 #
PCB-209
0.344
0.997
0.520#
# = Analyte did not meet the ion abundance ratio criterion, but met all of the other identification criteria
B* = Analyte detected in the sample at a concentration between 5 and 10 times that in associated method blank
B = Analyte detected in the sample at a concentration above 10 that in associated method blank
The results for the unspiked biosolids samples are influenced by several factors. As part of the study
design, EPA instructed the laboratories to apply their solid sample MDLs to the biosolids analyses, rather
than developing separate MDLs for biosolids samples. Because the draft method calls for using a 5-g
sample size for biosolids, as opposed to a 10-g sample for sediments. Therefore, adjusting the MDLs for
the biosolids samples by a factor of 2 higher meant that the sensitivity would be less for biosolids.
In the case of Laboratory 3, their MDLs for the solid samples often were much higher than those for the
other laboratories, and this means that their biosolids MDLs were also much higher than the adjusted
MDLs from the other three laboratories. That issue was exacerbated by the fact that Laboratory 3 also
deviated from the draft method and only extracted 1 to 2 g of biosolids (dry weight).
In contrast, Laboratory 6 had many of the lowest MDL values for solid samples, so they reported many
more congeners in the unspiked samples than the other three laboratories. The results from Laboratories
4 and 7 do not follow an obvious pattern. Neither laboratory detected any congeners for Biosolids sample
#2, despite having lower MDLs than Laboratory 3. It may be that their applications of the cleanup
procedures were not as effective at reducing interferences in this sample as Laboratories 3 or 6.
Tissue Samples
Each of the four laboratories that completed the tissue portion of the study analyzed the three tissue
samples as received, unspiked. All four laboratories performed the FlorisilŽ and GPC cleanup described
in the draft method; however, two laboratories performed additional cleanup of the sample extracts,
including the use of copper, alumina, silica, and an acid wash. As with the wastewater sample results, we
have limited the results in Table 33 to those congeners that were spiked into the matrix spike samples
later. We also only included those detected congeners that met the identification criteria in the draft
procedure and where the results were at least 5 times the results for the associated method blank. The
results from the reconnaissance analyses are also included in Table 33.
Table 33. Unspiked Tissue Sample Results in ng/g
Analyte
Lab 3
Lab 4
Lab 6
Lab 9
Recon
Tissue sample #1
PCB-28
0.030
PCB-31
0.024
PCB-41+64
0.032
PCB-44
0.034
PCB-52+73
0.040
Multi-Laboratory Validation Study Report
70
April 2021
-------
Table 33. Unspiked Tissue Sample Results in ng/g
Analyte
Lab 3
Lab 4
Lab 6
Lab 9
Recon
PCB-70
0.088
PCB-85+120
0.050
0.037
PCB-90+101+89
0.200
0.163
0.153
PCB-95+93
0.072
0.061
PCB-99
0.128
0.113
0.126
0.105
PCB-107+108
0.019
PCB-105+127
0.074
0.048
PCB-110
0.164
0.138
0.172
0.137
PCB-118+106
0.214
0.163
0.201
0.174
PCB-132+168
0.038
PCB-138+163+164
0.344
0.300
0.336
0.296
PCB-149+139
0.144
0.125
0.107
PCB-153
0.393
0.350
0.312
PCB-156
0.021
PCB-177
0.039
0.028
PCB-180
0.171
0.166
0.135
PCB-187+182
0.120
0.138
0.094
PCB-199
0.066
PCB-206
0.058
0.054
PCB-208
0.017
PCB-209
0.030
0.032
Tissue Sample #2
PCB-28
0.046
PCB-31
0.027
PCB-41+64
0.044
PCB-44
0.086
0.068
PCB-52+73
0.119
0.227
0.179
PCB-66+80
0.122
0.170
0.234
0.185
PCB-70
0.132
0.150
0.208
0.153
PCB-72
0.128
PCB-74+61
0.070
0.101
PCB-85+120
0.097
0.160
0.217
0.162
PCB-90+101+89
2.080
0.421
0.630
0.986
0.749
PCB-95+93
0.162
0.260
0.333
0.254
PCB-99
1.890
0.403
0.660
0.764
0.582
PCB-105+127
0.170
0.230
0.328
0.239
PCB-110
0.348
0.560
0.713
0.538
PCB-118+106
2.090
0.477
0.670
1.070
0.795
PCB-132+168
0.170
0.182
0.145
PCB-138+163+164
4.360
1.110
1.700
2.240
1.800
PCB-147
0.031
PCB-149+139
1.890
0.463
0.740
0.832
0.623
PCB-153
5.720
1.352
2.060
2.450
PCB-156
0.086
0.130
0.151
0.117
PCB-177
0.358
0.078
0.138
0.099
PCB-180
2.270
0.545
0.830
0.973
0.762
PCB-187+182
1.340
0.451
0.720
0.748
0.585
PCB-189
0.012
PCB-199
0.158
0.292
0.230
PCB-202
0.056
0.100
0.090
0.078
PCB-206
0.132
0.180
0.250
0.225
PCB-208
0.136
0.068
0.090
0.145
0.130
PCB-209
0.955
0.165
0.290
0.417
0.350
Multi-Laboratory Validation Study Report
71
April 2021
-------
Table 33. Unspiked Tissue Sample Results in ng/g
Analyte
Lab 3
Lab 4
Lab 6
Lab 9
Recon
Tissue Sample #3
PCB-3
0.100
PCB-4+10
0.072
PCB-11
0.267
0.283
PCB-18
0.227
0.089
0.113
0.075
PCB-19
0.087
PCB-28
0.450
0.389
0.493
0.404
PCB-31
0.306
0.200
0.261
0.230
PCB-37
0.090
0.057
PCB-41+64
0.611
0.371
0.500
0.518
0.456
PCB-44
0.684
0.611
0.712
0.636
PCB-52+73
2.760
1.333
1.178
1.490
1.440
PCB-54
0.057
PCB-66+80
2.310
1.200
1.480
1.530
PCB-70
1.200
1.570
1.370
PCB-72
0.029
PCB-74+61
1.390
0.747
0.744
0.868
0.790
PCB-78
0.011
PCB-85+120
1.480
1.025
0.911
1.170
1.010
PCB-90+101+89
9.110
5.399
4.900
6.280
5.570
PCB-95+93
2.700
1.838
1.600
2.090
PCB-98+102
0.051
PCB-99
4.720
2.925
2.578
3.370
2.930
PCB-104
0.049
PCB-105+127
2.600
1.584
1.444
1.760
1.550
PCB-107+108
0.795
0.548
0.444
4.580
0.497
PCB-110
6.800
3.847
3.567
6.210
3.940
PCB-118+106
8.800
5.142
4.556
1.380
5.310
PCB-132+168
2.430
1.241
1.111
13.100
1.340
PCB-138+163+164
16.000
10.800
9.622
11.300
PCB-147
0.242
0.156
0.162
PCB-149+139
6.590
4.514
3.944
4.830
4.000
PCB-152
0.047
PCB-153
16.300
11.296
9.889
13.600
11.000
PCB-155
0.053
PCB-156
0.953
0.917
0.767
0.904
0.836
PCB-166
0.209
0.042
PCB-177
0.987
1.022
1.100
0.961
PCB-180
6.870
4.611
4.756
5.880
4.960
PCB-184
0.047
PCB-187+182
4.060
3.041
3.278
3.700
3.040
PCB-188
0.061
PCB-189
0.126
0.083
0.063
PCB-199
1.470
1.390
0.900
1.560
1.150
PCB-202
0.196
0.275
0.211
0.262
0.208
PCB-204
0.046
PCB-205
0.106
0.641
PCB-206
0.543
0.664
0.444
0.760
0.549
PCB-208
0.274
0.167
0.276
0.231
PCB-209
0.089
0.455
0.422
0.453
0.382
The unspiked tissue sample results varied across the three samples and the four laboratories. Laboratory
3 did not detect any congeners in Tissue Sample # 1 at all, whereas the other three laboratories reported
Multi-Laboratory Validation Study Report
72
April 2021
-------
between 6 and 16 congeners in that sample, and the reconnaissance laboratory reported 45 congeners.
Those differences in detection were largely driven by the MDLs in each of those laboratories. For
example, Laboratory 3 had the highest tissue MDLs for 127 of the 167 analytes, and the results reported
by the other laboratories were often at concentrations below the detection limits in Laboratory 3.
Laboratory 6 had the highest MDLs for 31 of the 167 analytes, which may explain their lower numbers of
detects in Tissue Sample #1 as well. In contrast, the reconnaissance laboratory had many of the lowest
MDLs, as a result of their experience with the method, and they detected far more congeners in Tissue
Sample #1, often at very low levels that were below the MDLs of the other laboratories. In many of the
cases where not all of the laboratories detected a congener, the congener was present at a concentration
near the lowest detection limits of any of the laboratories and therefore laboratories with higher detection
limits could not detect the congener.
As noted in Table 4, Tissue Sample # 1 was characterized as the low concentration sample in this study
and as is shown in Table 33, most of the concentrations reported by the laboratories were fairly low.
Setting aside the results from Laboratory 3, there were four analytes detected in Tissue Sample #1 by all
three of the other laboratories and the reconnaissance laboratory. The results for those four congeners are
remarkably similar across those four laboratories, as shown in Table 34.
Table 34. Results for Congeners Detected by Four Labs for
Tissue Sample #1 in ng/g
Analyte
Lab 4
Lab 6
Lab 9
Recon
PCB-99
0.128
0.113
0.126
0.105
PCB-110
0.164
0.138
0.172
0.137
PCB-118+106
0.214
0.163
0.201
0.174
PCB-138+163+164
0.344
0.300
0.336
0.296
Tissue Sample #2 was designed to be a medium level sample. As a result of the higher concentrations in
this sample compared to Tissue Sample #1, Laboratory 3 detected 11 analytes, even with their higher
detection limits, and the other three laboratories in the study detected 20 to 25 analytes each. Of the 11
analytes detected by Laboratory 3,10 also were detected in the other three laboratories in the study. For 9
of those 10 those analytes, Laboratory 3 reported markedly higher results than the other three laboratories,
or the reconnaissance laboratory, and the results from the other three laboratories were generally quite
similar, as shown in Table 35.
Table 35. Results for Congeners Detected by Five Labs for Tissue Sample #2
in ng/g
Analyte
Lab 3
Lab 4
Lab 6
Lab 9
Recon
PCB-90+101+89
2.080
0.421
0.630
0.986
0.749
PCB-99
1.890
0.403
0.660
0.764
0.582
PCB-118+106
2.090
0.477
0.670
1.070
0.795
PCB-138+163+164
4.360
1.110
1.700
2.240
1.800
PCB-149+139
1.890
0.463
0.740
0.832
0.623
PCB-153
5.720
1.352
2.060
2.450
PCB-177
0.358
0.078
0.138
0.099
PCB-180
2.270
0.545
0.830
0.973
0.762
PCB-187+182
1.340
0.451
0.720
0.748
0.585
PCB-208
0.136
0.068
0.090
0.145
0.130
PCB-209
0.955
0.165
0.290
0.417
0.350
Tissue Sample #3 was designed to be a higher concentration sample and that characterization held true.
Laboratory #3 detected 23 analytes, and the other three laboratories in the study detected 30 to 43 analytes
each. As in Tissue Sample #2, Laboratory 3 reported higher concentrations than the other laboratories or
the reconnaissance laboratory for many of the analytes, and the results from the other three laboratories
were generally quite similar. Although CSRA examined the results in detail, as well as the associated
Multi-Laboratory Validation Study Report
73
April 2021
-------
calibration data, no obvious reason for the higher results were immediately apparent. However, the
potential bias does appear to be consistent for Laboratory 3. A comparison of these results is shown in
Table 36.
Table 36. Results for Congeners Detected by Five Labs for Tissue Sample
#3 in ng/g
Analyte
Lab 3
Lab 4
Lab 6
Lab 9
Recon
PCB-41+64
0.611
0.371
0.500
0.518
0.456
PCB-52+73
2.760
1.333
1.178
1.490
1.440
PCB-74+61
1.390
0.747
0.744
0.868
0.790
PCB-85+120
1.480
1.025
0.911
1.170
1.010
PCB-90+101+89
9.110
5.399
4.900
6.280
5.570
PCB-99
4.720
2.925
2.578
3.370
2.930
PCB-105+127
2.600
1.584
1.444
1.760
1.550
PCB-107+108
0.795
0.548
0.444
4.580
0.497
PCB-110
6.800
3.847
3.567
6.210
3.940
PCB-118+106
8.800
5.142
4.556
1.380
5.310
PCB-132+168
2.430
1.241
1.111
1.380
1.340
PCB-149+139
6.590
4.514
3.944
4.830
4.000
PCB-153
16.300
11.296
9.889
13.600
11.000
PCB-156
0.953
0.917
0.767
0.904
0.836
PCB-180
6.870
4.611
4.756
5.880
4.960
PCB-187+182
4.060
3.041
3.278
3.700
3.040
PCB-199
1.470
1.390
0.900
1.560
1.150
PCB-202
0.196
0.275
0.211
0.262
0.208
PCB-206
0.543
0.664
0.444
0.760
0.549
PCB-209
0.089
0.455
0.422
0.453
0.382
Multi-Laboratory Validation Study Report
74
April 2021
-------
8. Matrix Spike Analyses
Isotope dilution methods generate recovery data for all of the labeled compounds spiked into every
sample, so EPA has not included the use of matrix spike (MS) and matrix spike duplicate (MSD) samples
as part of the routine per-sample-batch quality control operations in those methods. However, in order to
demonstrate the performance of the draft procedure in real-world samples that contain the native analytes,
EPA required each laboratory in the study to prepare an MS/MSD pair for each of the study samples that
they agreed to analyze (e.g., all nine wastewaters, three sediments, three biosolids, and three fish tissues).
Generation of the most useful MS/MSD data requires knowledge of the background levels of the analytes
in the unspiked samples so that appropriate spiking levels can be chosen (not so high as to be unrealistic
in the context of actual sample concentrations, and not so low that the spiked amount is difficult to
discern given the background concentration in the original sample). Rather than waiting for each
participant laboratory to analyze all of their unspiked samples and then go back and develop a customized
spiking scheme based on those results, EPA and CSRA used the reconnaissance analysis results described
in Section 3 to develop sample-specific spiking instructions for all of the samples in the study.
The basic approach was to spike all 48 high-priority congeners, using the native compound spiking
solution described in the draft procedure and provided to each laboratory in the study by EPA. The same
spiking level of 16 ng per sample used for the IPR samples was used for the MS/MSD samples where
practical.
However, in order to gather data on additional congeners, EPA and CSRA instructed the laboratory to
prepare an additional spiking solution by diluting one of the retention time standards by a factor of 200,
using acetonitrile to make the solution water miscible. Retention time Mix #7 (from the set of 9 standards
provided by EPA) contains a total of 14 congeners. Of those, 13 are not among the 48 high-priority
congeners in the native congener spiking solution. The one congener in common to both solutions is
PCB-166. Both solutions contain some congeners that coelute with other congeners in the analysis.
EPA and CSRA instructed the laboratories to spike two aliquots of each study sample with 200 |_iL of the
native compound spiking solution, and 200 |_iL of the diluted retention time Mix #7. The list of congeners
spiked into the samples is shown in Table 37. Based on the volumes of the spiking solutions, each
MS/MSD aliquot received 16 ng of each of the congeners, except PCB-166, for which the total mass
spiked was 32 ng. The same spiking scheme was used for all four matrix types.
Table 37. Composition of Matrix Spiking Solutions
Analyte
Source Solution
Analyte
Source Solution
PCB-1
Native spiking solution
PCB 105
Native spiking solution
PCB-3
Native spiking solution
PCB-106
RT solution #7
PCB-4
Native spiking solution
PCB-108
RT solution #7
PCB-8
Native spiking solution
PCB 110
Native spiking solution
PCB-11
Native spiking solution
PCB 118
Native spiking solution
PCB-15
Native spiking solution
PCB 126
Native spiking solution
PCB-18
Native spiking solution
PCB 132
Native spiking solution
PCB-19
Native spiking solution
PCB 138
Native spiking solution
PCB-28
Native spiking solution
PCB 147
Native spiking solution
PCB-31
Native spiking solution
PCB 149
Native spiking solution
PCB-36
RT solution #7
PCB-152
RT solution #7
PCB-37
Native spiking solution
PCB 153
Native spiking solution
PCB-44
Native spiking solution
PCB 155
Native spiking solution
PCB-52
Native spiking solution
PCB 156
Native spiking solution
PCB-54
Native spiking solution
PCB 166
Both
PCB-64
Native spiking solution
PCB 169
Native spiking solution
Multi-Laboratory Validation Study Report
75
April 2021
-------
Table 37. Composition of Matrix Spiking Solutions
Analyte
Source Solution
Analyte
Source Solution
PCB-66
Native spiking solution
PCB 177
Native spiking solution
PCB-70
Native spiking solution
PCB 180
Native spiking solution
PCB-72
RT solution #7
PCB-182
RT solution #7
PCB-74
Native spiking solution
PCB-184
RT solution #7
PCB-77
Native spiking solution
PCB 187
Native spiking solution
PCB-78
RT solution #7
PCB 188
Native spiking solution
PCB-79
RT solution #7
PCB 189
Native spiking solution
PCB-85
Native spiking solution
PCB 199
Native spiking solution
PCB-89
RT solution #7
PCB 202
Native spiking solution
PCB-95
Native spiking solution
PCB-204
RT solution #7
PCB-96
RT solution #7
PCB 205
Native spiking solution
PCB-98
RT solution #7
PCB 206
Native spiking solution
PCB-99
Native spiking solution
PCB 208
Native spiking solution
PCB-101
Native spiking solution
PCB 209
Native spiking solution
PCB-104
Native spiking solution
Each laboratory reported their unspiked sample results, as well as the MS and MSD concentrations,
recoveries, and relative percent differences (RPDs) for each matrix type that they analyzed. CSRA used
those data to evaluate method performance, as described in the subsections that follow.
Aqueous Sample MS/MSD Results
Each of the seven laboratories that completed the wastewater portion of the study analyzed MS/MSD
aliquots of all nine of the wastewater samples. Unfortunately, Lab 1 failed to follow the study
instructions and did not spike the congeners in RT Mix #7 into the MS/MSD aliquots. However, they did
spike the other analytes in the native spiking solution.
In addition, the analytical challenges evident in the analyses of the unspiked aliquots of Wastewater #2
were also present in the MS/MSD analyses. Although a few of the laboratories were able to generate
useful MS/MSD results for that sample, other laboratories provided results with calculated recoveries that
were negative numbers or well over 150%, and both those extremes are evidence of problems with the
choice of spiking levels or with interferences. As a result, EPA decided not to evaluate the MS/MSD
results for Wastewater #2 any further. That decision was based not only on the results from the study, but
also after considering that the sample represented a landfill leachate, which is not a matrix type that is
discharged to surface water without some level of treatment. Rather, the leachate in question is collected
at the landfill and sent to a POTW for treatment, and the effluent from the POTW is what is subject to a
discharge permit and NPDES monitoring requirements.
CSRA calculated the mean recovery of each spiked analyte by sample across all seven laboratories, along
with the minimum and maximum observed values. Given the number of samples and analytes, those
results are presented in Tables 38 and 39, for the MS/MSD analyses of wastewater samples #3 to #6 and
samples #7 to #10, respectively. The mean MS/MSD recoveries are presented graphically in Figure 5, for
all eight of the wastewater samples. The highlighted rows show the congeners that were quantified
indirectly, using labeled standards of similar congeners in the same level of chlorination. Values in
parentheses, next to zero values, represent the number of results which were reported as non-detects by
the laboratories. Of the 8 wastewater samples summarized in Tables 38 and 39, four samples
(Wastewater #3, #6, #7 and #10) had false negatives reported by the laboratories, with Wastewater #7
having the most reported false negatives by Labs 1, 6, 7 and 9. Although six wastewater samples had
false negatives reported, the percentage was less than 0.2% of the 7128 total data points.
Multi-Laboratory Validation Study Report
76
April 2021
-------
Table 38. Matrix Spike Recoveries for Wastewater Samples #3 to #6 (%)
Analyte
#
Results
Wastewater #3
Wastewater #4
Wastewater #5
Wastewater #6
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
PCB-1
14
99.9
90
123
97.1
86
111
101.1
88
123
101.6
74
126
PCB-3
14
94.6
87
103
93.1
84
109
97.6
86
129
95.4
73
113
PCB-4+10
14
120.0
86
250
96.1
86
111
97.1
88
115
84.0
10
110
PCB-8+5
14
83.8
76
94
84.9
74
115
85.4
76
97
87.4
68
106
PCB-11
14
94.4
86
115
102.4
84
145
103.1
87
147
101.9
62
186
PCB-15
14
86.3
56
103
85.4
79
100
89.3
84
101
85.1
62
95
PCB-18
14
86.1
76
96
91.1
75
149
86.9
78
103
81.9
62
92
PCB-19
14
89.6
79
102
89.0
79
119
92.9
84
109
93.8
64
133
PCB-28
14
99.4
79
123
102.4
80
158
97.9
82
118
94.8
67
131
PCB-31
14
83.3
64
97
82.4
63
139
86.9
78
94
77.4
57
90
PCB-36
12
102.9
0(3)
126
108.0
86
140
114.7
97
131
111.0
69
128
PCB-37
14
88.3
73
102
87.3
73
123
90.8
83
106
107.1
0(2)
298
PCB-41+64
14
90.2
80
117
102.4
64
324
93.4
81
113
90.3
54
147
PCB-44
14
99.9
86
176
89.3
75
116
95.2
87
110
91.9
59
115
PCB-52+73
14
90.1
76
120
86.5
73
104
90.9
83
106
89.4
58
108
PCB-54
14
90.0
83
107
88.1
76
107
92.1
84
101
88.9
58
103
PCB-74+61
14
91.4
82
104
83.4
11
98
93.8
84
108
84.6
15
123
PCB-66+80
14
91.2
76
118
124.3
73
585
90.2
75
110
109.9
52
307
PCB-70
14
86.3
76
96
87.6
73
131
88.2
82
96
91.1
54
155
PCB-72
12
99.7
85
110
103.8
86
114
104.3
98
113
103.7
62
120
PCB-77
14
87.1
78
114
88.6
74
149
87.1
78
108
88.7
50
135
PCB-78
12
101.9
90
114
101.6
75
119
107.2
89
117
109.5
57
129
PCB-79
12
104.6
80
125
107.0
68
124
114.5
88
155
117.4
65
164
PCB-85+120
14
85.4
70
98
84.1
68
96
87.9
79
100
89.0
46
125
PCB-90+101+89
14
84.8
60
99
85.5
67
100
90.2
82
101
88.3
47
106
PCB-95+93
14
87.2
69
100
82.6
45
113
86.9
63
101
88.1
53
107
PCB-96
12
97.2
87
107
93.6
76
103
102.4
88
114
94.0
60
105
PCB-98+102
12
100.7
80
111
98.6
73
112
102.3
78
115
86.9
16
115
PCB-99
14
87.7
72
99
88.2
69
103
91.3
79
107
92.7
49
109
PCB-104
14
87.3
74
102
87.4
69
108
91.6
81
105
86.8
50
104
PCB-105+127
14
86.6
72
102
85.6
67
103
89.8
76
110
92.7
47
115
PCB-118+106
14
87.1
71
106
86.8
65
98
91.0
82
103
88.6
45
103
PCB-107+108
12
97.3
81
108
94.0
57
108
102.3
92
113
102.0
55
114
PCB-110
14
87.4
62
109
84.8
65
116
88.1
76
103
90.2
44
138
Multi-Laboratory Validation Study Report
77
April 2021
-------
Table 38. Matrix Spike Recoveries for Wastewater Samples #3 to #6 (%)
Analyte
#
Results
Wastewater #3
Wastewater #4
Wastewater #5
Wastewater #6
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
PCB-128
14
85.4
71
106
88.5
67
107
87.3
77
100
87.9
41
117
PCB-132+168
14
95.6
76
214
91.6
68
137
90.6
78
106
88.4
46
109
PCB-138+163+164
14
81.6
60
102
86.6
62
120
86.5
73
105
88.0
39
116
PCB-149+139
14
85.4
58
102
89.7
64
117
88.0
72
108
89.1
48
121
PCB-147
14
83.5
61
95
88.8
63
102
88.3
75
103
87.4
44
106
PCB-152
12
101.7
54
125
105.0
81
122
105.3
72
126
100.8
63
124
PCB-153
14
85.3
48
142
88.1
61
117
86.9
65
110
85.8
39
126
PCB-155
14
80.2
58
92
84.6
61
100
88.4
71
107
84.7
39
105
PCB-156
14
79.4
47
96
88.6
47
111
87.1
64
117
91.4
44
114
PCB-166
14
90.6
68
104
97.3
68
113
97.5
81
113
99.7
48
114
PCB-169
14
77.9
48
101
83.3
64
103
85.6
69
103
88.2
42
115
PCB-177
14
77.7
50
96
89.1
60
116
86.1
68
107
89.3
40
120
PCB-180
14
71.7
44
86
83.7
59
124
81.8
63
112
84.5
39
130
PCB-187+182
14
84.9
56
102
94.9
65
113
93.6
77
108
96.3
43
124
PCB-184
12
83.6
56
98
88.3
63
107
93.0
80
103
88.8
43
100
PCB-188
14
73.7
50
88
84.4
58
107
84.2
64
108
83.4
36
109
PCB-189
14
70.5
48
85
83.5
56
102
81.5
60
111
84.2
36
114
PCB-199
14
59.7
19
93
74.2
14
97
73.2
2
96
123.6
40
419
PCB-202
14
65.9
40
84
81.1
55
104
78.6
61
99
80.4
36
104
PCB-204
12
80.6
47
98
89.3
63
102
92.8
78
104
90.8
40
105
PCB-205
14
61
40
82
79.9
54
99
77.0
56
104
81.1
48
103
PCB-206
14
58.4
30
84
76.5
54
99
81.0
58
123
84.5
37
114
PCB-208
14
59.6
31
101
77.7
53
98
78.0
56
98
79.8
33
102
PCB-209
14
54.6
25
95
81.1
52
122
78.4
56
103
82.8
36
107
Multi-Laboratory Validation Study Report
78
April 2021
-------
Table 39. Matrix Spike Recoveries for Wastewater Samples #7 to #10 (%)
Wastewater #7
Wastewater #8
Wastewater #9
Wastewater #10
#
Mean
Min.
Max.
Mean
Min.
Max.
Mean
Min.
Max.
Mean
Min.
Max.
Analyte
Results
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
Rec.
PCB-1
14
187.2
-33
453
105.2
93
143
106.5
96
163
94.5
0(1)
114
PCB-3
14
98.7
58
158
95.4
88
107
97.2
88
117
96.9
89
113
PCB-4+10
14
96.6
38
164
106.4
85
266
85.4
14
139
95.4
13
147
PCB-8+5
14
93.1
59
165
83.0
70
99
81.6
56
101
81.6
31
114
PCB-11
14
97.7
33
157
94.1
88
121
93.2
86
104
114.4
85
263
PCB-15
14
90.4
60
116
87.1
74
93
85.7
68
92
86.7
82
92
PCB-18
14
90.4
41
145
76.9
55
85
78.7
53
86
79.7
34
101
PCB-19
14
93.6
47
166
89.6
83
98
88.8
79
98
90.5
84
103
PCB-28
14
85.4
43
122
97.9
76
122
96.9
76
127
98.5
77
123
PCB-31
14
76.6
35
118
82.2
71
96
84.4
66
104
83.6
74
94
PCB-36
12
97.4
0(3)
174
114.5
74
138
109.8
72
138
116.8
98
144
PCB-37
14
74.5
0(2)
236
89.0
82
97
90.4
85
100
87.9
82
92
PCB-41+64
14
118.2
26
255
79.2
65
94
80.9
63
96
82.3
64
89
PCB-44
14
82.6
31
109
91.6
71
110
89.0
66
100
94.2
88
102
PCB-52+73
14
87.0
32
171
87.1
66
104
84.9
64
92
89.1
84
105
PCB-54
14
86.7
34
108
89.1
74
100
86.1
65
97
91.5
85
104
PCB-74+61
14
90.4
29
114
87.8
62
110
93.1
61
143
88.9
67
97
PCB-66+80
14
116.6
29
293
89.0
63
120
84.6
60
92
86.9
68
91
PCB-70
14
80.7
29
114
88.6
62
156
81.2
60
90
83.9
68
90
PCB-72
12
116.7
30
193
100.4
69
119
97.2
67
115
104.5
75
123
PCB-77
14
73.2
24
92
84.1
62
96
81.4
59
101
80.3
44
97
PCB-78
12
95.5
19
151
105.6
67
128
100.8
59
122
103.5
77
115
PCB-79
12
111.5
12
220
111.3
73
141
106.0
67
139
108.7
82
128
PCB-85+120
14
76.0
26
95
86.0
57
122
81.1
56
93
86.1
52
112
PCB-90+101+89
14
75.7
22
94
93.2
57
203
83.1
54
95
86.1
58
95
PCB-95+93
14
77.9
26
101
84.9
60
153
81.5
59
97
81.6
58
92
PCB-96
12
85.6
22
106
99.3
69
121
92.8
64
115
100.6
93
112
PCB-98+102
12
91.7
22
120
97.2
67
112
96.6
64
116
99.4
64
110
PCB-99
14
77.2
26
98
91.5
57
163
84.0
56
98
86.3
57
93
PCB-104
14
76.3
26
95
86.6
60
104
81.9
59
91
87.2
76
94
PCB-105+127
14
74.2
23
129
90.6
55
161
83.4
53
97
82.7
45
92
PCB-118+106
14
76.4
20
96
94.8
56
209
83.6
55
99
85.9
48
95
PCB-107+108
12
91.9
20
113
100.3
59
132
94.8
60
114
97.6
55
109
PCB-110
14
81.7
25
133
93.9
56
280
81.7
57
104
81.5
47
110
Multi-Laboratory Validation Study Report
79
April 2021
-------
Table 39. Matrix Spike Recoveries for Wastewater Samples #7 to #10 (%)
Analyte
#
Results
Wastewater #7
Wastewater #8
Wastewater #9
Wastewater #10
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
PCB-126
14
82.4
22
228
83.1
54
92
82.6
54
99
84.1
41
103
PCB-132+168
14
58.1
0(2)
103
88.9
53
135
83.6
54
101
84.6
42
97
PCB-138+163+164
14
84.2
24
141
91.7
50
197
82.9
49
97
82.6
38
96
PCB-149+139
14
82.2
23
112
89.8
50
179
84.0
51
100
83.5
39
97
PCB-147
14
77.8
21
103
84.1
49
95
82.3
49
102
83.7
37
99
PCB-152
12
98.6
18
130
99.5
67
121
99.9
68
130
100.3
51
124
PCB-153
14
93.1
23
159
91.2
53
180
82.6
49
109
83.3
38
99
PCB-155
14
67.9
22
91
85.6
50
99
78.6
49
89
83.6
46
91
PCB-156
14
73.1
23
139
88.1
53
113
82.4
50
100
84.4
41
100
PCB-166
14
86.7
20
142
97.4
54
112
91.9
53
108
94.5
44
109
PCB-169
14
75.5
0(1)
138
84.6
48
97
81.4
47
101
80.8
32
97
PCB-177
14
68.8
25
101
85.5
46
102
82.4
46
121
81.1
31
102
PCB-180
14
60.4
15
96
84.4
44
97
73.7
44
95
78.1
31
92
PCB-187+182
14
73.1
19
132
92.6
45
105
88.4
45
130
88.1
33
105
PCB-184
12
66.4
11
89
91.8
45
106
81.9
46
104
89.0
40
103
PCB-188
14
61.6
24
87
83.9
44
94
77.3
45
90
80.9
35
93
PCB-189
14
61.4
20
103
82.2
44
96
75.6
42
98
77.6
29
92
PCB-199
14
85.6
14
188
84.4
50
106
80.7
47
151
79.0
38
98
PCB-202
14
57.8
24
107
79.6
43
93
73.0
41
91
75.7
29
87
PCB-204
12
73.0
12
150
92.4
41
111
85.6
42
116
92.5
38
122
PCB-205
14
60.8
0(1)
135
74.3
40
92
74.0
42
96
75.5
29
96
PCB-206
14
57.5
16
100
80.1
38
118
73.7
37
94
76.7
29
111
PCB-208
14
47.8
15
76
76.9
38
91
71.2
37
93
77.9
27
118
PCB-209
14
79.9
13
356
77.9
36
92
77.1
36
122
79.1
29
99
Multi-Laboratory Validation Study Report
80
April 2021
-------
Aqueous Mean MS Recovery
o
o
CM
VIII
o WW7
o WW8
o
CD
o WW10
o
CD
O
o
CM
O
O
o
CD
O
CD
o
^r
o
CM
Congener
Figure 5. Mean Matrix Spike Wastewater Recoveries by Sample, in Elution Order (without Sample #2)
Dotted lines and Roman numerals delineate the levels of chlorination. The colored symbols and lines denote each of the eight
wastewater samples that were spiked.
Multi-Laboratory Validation Study Report
81
April 2021
-------
The majority of the mean matrix spike recoveries fall within the range of 60 to 120%. Of 464 mean
recoveries, only 8 were below 60% and only 3 were over 120%. The mean recovery of 187% for PCB-1
in wastewater #7 was driven by the results from Lab 1, Lab 6, and Lab 8, with reported recoveries
between approximately 200 and 500%. All of those results for PCB-1 also exhibited issues with the ion
abundance ratio for this congener, indicating an interference was present. The mean recovery for
PCB-199 in wastewater #6 was only slightly higher than 120% and was driven by the results from Lab 8,
which were reported between approximately 300 and 420%. The mean recovery for PCB-66+80 in
wastewater #4 was only slightly higher than 120% and was driven by the results from Lab 1, with a
recovery in the MSD aliquot of 585%. That result was likely driven by the very low recovery of the
labeled compound used to quantify this analyte in the MSD aliquot. In contrast, the MS recovery from
Lab 1 was 94%.
Sediment Sample MS/MSD Results
Each of the six laboratories that completed the sediment portion of the study analyzed MS/MSD aliquots
of the three sediment samples. CSRA calculated the mean recovery of each spiked analyte by sample
across all six laboratories, along with the minimum and maximum observed values and presented those
results in Table 40. Because the non-wastewater matrices in this study were a lower priority for EPA, and
because more of the laboratories experienced issues with qualitative identification of the target analytes in
these other matrices, if the results from a given laboratory did not meet the identification criteria for the
analyte, then the recoveries from that laboratory were not included in the mean, and the number of results
(#) for that sample in Table 40 will be less than 12. The highlighted rows show the congeners that were
quantified indirectly, using labeled standards of similar congeners in the same level of chlorination.
Values in parentheses, next to the zero values, represent the number of results which were reported as
non-detects by the laboratories. Although there were some false negative results reported for Sediment
#1, the percentage was less than 2.0% of the 2088 total data points. There were no false negatives for
Sediments #2 and #3.
The mean MS/MSD recoveries are presented graphically in Figure 6, for all three of the sediment
samples.
Table 40. Matrix Spike Recoveries for Sediment Samples
Analyte
#
Sediment #1
Sediment #2
Sediment #3
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
PCB-1
12
147
-63
380
125
36
260
176
-106
420
PCB-3
12
102
4
163
71
33
150
71
-86
170
PCB-4+10
12
107
-213
805
137
-67
971
121
-613
529
PCB-8+5
12
55
-29
145
89
18
368
24
-498
306
PCB-11
12
79
11
165
89
56
114
87
5
142
PCB-15
12
114
-158
334
157
-27
621
97
-456
542
PCB-18
12
142
-27
366
81
-35
389
34
-418
293
PCB-19
12
92
-5
239
89
16
322
74
-197
219
PCB-28
12
227
-112
1092
111
-2
279
87
-860
633
PCB-31
12
113
-154
392
88
-51
557
97
-744
573
PCB-36
12
59
0(4)
148
108
71
167
117
14
156
PCB-37
12
56
0(2)
117
88
0(2)
296
110
-61
294
PCB-41+64
12
98
-14
284
86
53
153
83
-346
285
PCB-44
12
51
-256
220
75
-2
290
71
-675
442
PCB-52+73
12
461
-222
2761
102
-27
565
90
-1005
591
PCB-54
12
72
4
125
80
46
97
77
-5
102
PCB-66+80
12
155
-73
525
115
66
182
103
-642
504
PCB-70
12
50
-188
291
85
31
182
92
-572
451
PCB-72
12
73
0(2)
121
109
83
143
122
20
258
Method 1628 Multi-Laboratory Validation Study Report
82
April 2021
-------
Table 40. Matrix Spike Recoveries for Sediment Samples
Analyte
#
Sediment #1
Sediment #2
Sediment #3
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
PCB-74+61
12
117
-141
325
99
44
146
87
-335
319
PCB-77
12
70
0(2)
134
92
75
110
78
-35
146
PCB-78
12
55
0(4)
133
109
51
172
109
15
151
PCB-79
12
262
0(2)
1129
118
62
166
104
15
160
PCB-85+120
12
69
-172
228
93
51
122
80
-14
139
PCB-90+101+89
12
57
-157
305
93
44
135
68
-103
168
PCB-95+93
12
100
-171
354
83
36
227
7
-500
174
PCB-96
12
62
-88
148
92
55
108
97
-8
217
PCB-98+102
12
146
-14
815
103
63
142
44
-783
448
PCB-99
12
50
-162
271
94
43
135
78
-126
190
PCB-104
12
45
-88
93
79
44
92
79
13
101
PCB-105+127
12
119
-44
283
96
61
122
87
-68
186
PCB-107+108
12
220
43
485
93
21
130
121
32
207
PCB-110
12
181
-183
472
99
41
231
107
-338
262
PCB-118+106
12
204
-118
638
101
63
174
83
-105
196
PCB-126
12
44
0(4)
99
75
-26
107
81
10
104
PCB-132+168
12
55
-77
212
90
46
130
81
-11
119
PCB-138+163+164
12
109
-230
424
96
57
210
91
-74
189
PCB-147
12
59
-12
100
73
-20
101
82
19
106
PCB-149+139
12
99
-213
334
98
60
191
76
-7
127
PCB-152
12
102
15
139
103
60
131
98
29
123
PCB-153
12
197
-107
615
95
46
236
79
-15
136
PCB-155
12
50
-72
92
81
45
93
80
15
105
PCB-156
12
94
4
162
90
58
114
80
10
122
PCB-166
12
92
0(1)
160
104
65
146
95
21
132
PCB-169
12
73
0(2)
117
115
70
278
74
12
99
PCB-177
12
73
8
163
90
49
127
80
4
113
PCB-180
12
75
-85
316
112
43
275
79
-9
125
PCB-184
12
61
-65
110
93
54
110
90
20
121
PCB-187+182
12
55
-45
157
107
53
149
90
13
125
PCB-188
12
59
-6
93
83
46
101
81
19
111
PCB-189
12
54
-82
130
93
48
115
78
19
101
PCB-199
12
91
-30
232
94
-12
232
85
10
117
PCB-202
12
77
-16
159
91
43
120
83
22
123
PCB-204
12
77
0(2)
141
95
47
128
93
6
130
PCB-205
12
74
0(2)
152
85
38
106
78
7
105
PCB-206
12
34
-117
224
130
-3
411
82
25
108
PCB-208
12
92
18
173
101
64
123
86
30
120
PCB-209
12
114
32
244
95
40
140
88
34
105
As can be seen in Table 40, there are many congeners with minimum recoveries that are negative
numbers. Although all of the laboratories reported some negative recoveries, four of the laboratories
reported 26 to 63 negative recoveries across all the 348 results for the three sediment samples, while the
other two laboratories reported only 3 and 4 negative recoveries. Although negative recoveries have no
""physical" meaning, in that the analytes are not in fact removed from the sample by the analytical
procedures, they can be indicative of issues with interferences of either the spiked sample analysis, or the
unspiked sample analysis that bias the results of one or both of those analyses. CSRA examined the
patterns of negative recoveries across the laboratories. Although Laboratory 8 had 63 negative
recoveries, most of those occurred in Sediment #1 and Sediment #3, and usually only in one of the two
spiked aliquots (i.e., the MS or the MSD, but not both). That pattern suggests an interference with one,
but not both of the spiked samples for the affected congeners. In contrast, Laboratory 3 had all of its
Method 1628 Multi-Laboratory Validation Study Report
83
April 2021
-------
negative recoveries in Sediments #1 and #2, and almost always for the same congeners in both aliquots.
That pattern suggests that the negative values may be due to issues with the results for the unspiked
aliquots, where a positive bias in the unspiked sample results leads to negative recoveries. The results for
the unspiked sample analyses in Table 30 indicate that Laboratory 3 often reported higher results for the
affected congeners than the other five laboratories. The patterns of negative recoveries from Laboratory 6
are similar, with all 26 of the negative recoveries occurring in Sediments #1 and #2, and almost always
for the same congeners in both aliquots. In contrast, Laboratory 4 reported 47 negative recoveries, 35
from Sediment #3, but with 28 of those in the MSD aliquot, and only 7 in the MS aliquot. That pattern
suggests a mix of both issue with the unspiked sample results and interferences in the MSD aliquot.
In addition to the negative recoveries, there were a number of very high positive recoveries reported, a
few over 2,000%. As with the negative recoveries, the high recoveries were distributed across all six
laboratories, but unevenly among the laboratories, as well as among the three sediment samples. In some
laboratories, the higher than expected recoveries were reported in both the MS and MSD aliquots for the
same congeners, suggesting that the issue may have been due to a low bias in the unspiked sample
analyses. In other cases, the high recoveries were only in one of the spiked aliquots, indicative of aliquot-
specific interferences.
Another potential source of the unusual recoveries is sample homogeneity, or lack thereof. Although the
vendor took extensive measures to thoroughly blend and homogenize the bulk sediments used to create
these study samples and divide them into aliquots for each laboratory, there still may be some
inhomogeneity from aliquot to aliquot. This is because PCBs are sorbed onto the sediment particles, and
it may not be possible to evenly distribute any more highly contaminated particles across the entire bulk
sample, despite careful preparation. Therefore, one of the aliquots of a given sample distributed to each
laboratory may have had more PCBs than the other two aliquots used for the unspiked and spiked sample
analyses. Such differences will affect the assumptions in the analyte recovery calculations.
As can be seen in Figure 6, the highest mean recoveries were reported from Sediment #1, and represent
PCB-28, PCB-52+73, PCB-66+80, PCB-79, PCB-107+108, PCB-110, PCB-118+106, and PCB-153.
Method 1628 Multi-Laboratory Validation Study Report
84
April 2021
-------
Sediment Mean MS Recovery
o
o
LO
VIII
o
UD
Sed1
Sed2
Sed3
O
O
^r
o
o
co
o
in
CM
o
o
CM
o
to
o
o
o
in
o
Congener
Figure 6. Mean Matrix Spike Sediment Reeoveries by Sample, in Elution Order
Dotted lines and Roman numerals delineate the levels of chiori nation. The colored symbols and lines denote each of the three sediment
samples that were spiked.
Method 1628 Multi-Laboratory Validation Study Report
85
April 2021
-------
Biosolids Sample MS/MSD Results
Each of the four laboratories that completed the biosolids portion of the study analyzed MS/MSD aliquots
of the three biosolids samples. CSRA calculated the mean recovery of each spiked analyte by sample
across all four laboratories, along with the minimum and maximum observed values and presented those
results in Table 41. As noted in the discussion of the sediment sample results, because the non-
wastewater matrices in this study were a lower priority for EPA, and because more of the laboratories
experienced issues with qualitative identification of the target analytes in these other matrices, if the
results from a given laboratory did not meet the identification criteria for the analyte, the recoveries from
that laboratory were not included in the mean, and the number of results (#) for that sample in Table 41
will be less than 8. The highlighted rows show the congeners that were quantified indirectly, using
labeled standards of similar congeners in the same level of chlorination.
Samples #1 and #3 had many false negatives with a percentage of 4.8% of the 928 combined data points
for those two samples for all four labs; however, all the false negative results are from Laboratory 4. The
laboratory performed GPC clean-up, plus an extra dilution, which may have caused loss of some target
analytes. The congeners affected were from the monochlorinated up to the heptachlorinated congeners.
The mean MS/MSD recoveries are presented graphically in Figure 7, for all three of the biosolids
samples.
Table 41. Matrix Spike Recoveries for Biosolids Samples
Analyte
#
Results1
Biosolid #1
Biosolid #2
Biosolid #3
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
PCB-1
8
79
-86
151
113
88
146
135
73
257
PCB-3
8
106
56
194
177
41
478
224
83
543
PCB-4+10
8
94
24
130
186
33
354
83
63
104
PCB-8+5
8
78
49
108
71
28
117
101
64
144
PCB-11
8
100
18
130
91
33
169
178
71
427
PCB-15
8
71
32
94
72
26
121
84
74
96
PCB-18
8
95
14
165
74
13
131
82
54
110
PCB-19
8
78
21
128
70
25
104
89
44
130
PCB-28
8
103
28
149
90
14
133
107
73
147
PCB-31
8
95
29
164
79
18
105
104
44
144
PCB-36
8(6)
86
0(2)
154
101
44
141
84
0(2)
126
PCB-37
8(6)
59
0(2)
109
89
39
128
92
52
124
PCB-41+64
8(6)
72
0(2)
108
78
37
120
68
0(2)
105
PCB-44
8(6)
70
0(2)
116
80
21
100
80
0(2)
129
PCB-52+73
8
93
25
129
71
4
100
94
54
152
PCB-54
8
83
24
118
72
20
91
85
54
110
PCB-66+80
8(6)
86
0(2)
160
99
46
154
94
0(2)
194
PCB-70
8(7)
111
0(1)
242
83
36
120
90
0(2)
149
PCB-72
8(6)
80
0(2)
132
104
42
144
94
0(2)
149
PCB-74+61
8
185
62
463
88
39
116
84
0(2)
154
PCB-77
8(6)
52
0(2)
86
80
61
97
101
0(2)
281
PCB-78
8(6)
84
0(2)
145
101
84
119
112
0(2)
233
PCB-79
8(6)
90
0(2)
157
120
79
189
101
0(2)
150
PCB-85+120
8
100
80
138
82
61
90
99
76
122
PCB-90+101+89
8
77
60
91
85
42
104
75
40
150
PCB-95+93
8
93
50
137
86
26
126
97
55
159
PCB-96
8
85
47
95
81
35
98
91
76
100
PCB-98+102
8(6)
74
0(2)
115
95
49
117
40
-52
129
PCB-99
8
93
37
122
85
51
98
109
72
133
Method 1628 Multi-Laboratory Validation Study Report
86
April 2021
-------
Table 41. Matrix Spike Recoveries for Biosolids Samples
Analyte
#
Results1
Biosolid #1
Biosolid #2
Biosolid #3
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
Mean
Rec.
Min
Rec.
Max
Rec.
PCB-104
8
73
36
85
70
23
86
76
59
91
PCB-105+127
8
121
97
189
90
81
100
81
0(2)
132
PCB-107+108
8(6)
89
0(2)
147
105
82
129
89
0(2)
130
PCB-110
8
101
74
150
92
58
113
121
67
228
PCB-118+106
8
90
63
117
96
72
107
108
68
155
PCB-126
8(6)
67
0(2)
97
89
80
98
78
0(2)
136
PCB-132+168
8
87
66
115
92
73
108
116
59
181
PCB-138+163+164
8
86
68
100
102
82
123
105
59
166
PCB-147
8
95
81
120
85
68
92
94
88
104
PCB-149+139
8
92
78
123
93
69
109
91
61
151
PCB-152
8
103
69
133
95
50
116
108
98
118
PCB-153
8
100
73
142
99
77
116
93
59
137
PCB-155
8
79
57
92
76
39
87
80
70
89
PCB-156
8(6)
70
0(2)
102
90
68
107
74
0(2)
122
PCB-166
8
105
71
144
103
98
108
98
73
114
PCB-169
8(6)
74
0(2)
136
94
74
132
58
0(2)
108
PCB-177
8
93
87
99
89
81
96
98
90
109
PCB-180
8
151
78
361
86
83
91
118
70
194
PCB-184
8
106
87
128
95
83
101
99
83
119
PCB-187+182
8
95
80
109
96
87
103
93
71
117
PCB-188
8
88
79
102
83
67
92
85
77
92
PCB-189
8
89
60
108
91
86
94
96
83
119
PCB-199
8
89
61
112
90
75
100
83
48
111
PCB-202
8
93
86
110
87
82
90
91
86
98
PCB-204
8
102
95
108
101
92
108
101
88
117
PCB-205
8
84
63
105
90
77
110
80
58
106
PCB-206
8
115
84
199
88
79
105
96
73
110
PCB-208
8
92
79
120
94
82
109
95
82
121
PCB-209
8
98
82
112
91
73
117
99
88
119
1 The value in parentheses is the number of results used to determine the mean recovery in biosolid sample #1
Overall, across all four laboratories and all three biosolids samples, the majority of the mean recoveries
ranged from 70% to 185%, with six mean recoveries below 70%, four mean recoveries between 150%
and 185%, and two mean recoveries above 185%.
Biosolids are well known as a challenging matrix for any analysis. In the case of PCB analyses, there are
many potential organic components present that can affect the sample extraction processes that may
require additional cleanup steps to remove from the extracts, which present instrumental interferences that
affect the chromatographic separation, or that affect the identification of a peak as a target analyte.
Despite such challenges, when compared to the sediment results, the biosolids results from this study
demonstrate better recoveries across all four of the laboratories and all of the congeners. For example,
there was only two instances of a congener with a minimum recovery less than 0%, and 22 congeners
with maximum recoveries greater than 150%, out of 435 results in Table 41. In comparison, there were
81 instances of negative recoveries and 97 instances of recoveries greater than 150% out of the 2054
sediment results in Table 40. It is unclear if the apparent differences between the results for the two
matrices are a function of issues in either of the two laboratories that performed sediment analyses but did
not perform biosolids analyses, or if other factors may be involved. As can be seen in Figure 7, mean
recoveries greater than 150% occurred in all three of the biosolids samples, but for different congeners in
different samples.
Method 1628 Multi-Laboratory Validation Study Report
87
April 2021
-------
Biosolids Mean MS Recovery
o
^r
CM
VIII
o Bsldl
o Bsld2
o Bsld3
o
o
OJ
o
-------
Tissue Sample MS/MSD Results
Each of the four laboratories that completed the tissue portion of the study analyzed MS/MSD aliquots of
the three tissue samples. CSRA calculated the mean recovery of each spiked analyte by sample across all
six laboratories, along with the minimum and maximum observed values and presented those results in
Table 42. The highlighted rows show the congeners that were quantified indirectly, using labeled
standards of similar congeners in the same level of chlorination. Values in parentheses next to the zero
values, represent the number of results which were reported at the same value of the unspiked samples by
the laboratories. The percentage of false negatives for tissue samples was less than 0.1% of the 1392 total
data points.
The mean MS/MSD recoveries are presented graphically in Figure 8, for all three of the tissue samples.
Table 42. Matrix Spike Recoveries for Tissue Samples
Analyte
#
Results
Tissue #1
Tissue #2
Tissue #3
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
PCB-1
8
101.8
94
111
95.0
87
104
95.5
92
101
PCB-3
8
95.8
89
105
91.9
86
98
89.8
83
97
PCB-4+10
8
95.3
87
106
87.8
80
95
89.4
84
96
PCB-8+5
8
112.0
75
194
84.0
73
99
67.9
48
97
PCB-11
8
97.9
89
118
95.5
85
113
106.0
86
118
PCB-15
8
84.9
77
93
90.9
85
110
88.5
82
92
PCB-18
8
85.8
73
94
83.0
71
102
87.8
69
106
PCB-19
8
89.6
86
97
88.6
82
101
85.6
76
93
PCB-28
8
100.1
82
130
109.4
86
153
102.4
79
142
PCB-31
8
87.6
76
101
87.4
78
109
83.4
71
91
PCB-36
8
115.6
101
130
114.6
101
146
112.9
99
123
PCB-37
8
90.4
81
112
91.5
83
123
90.6
74
108
PCB-41+64
8
86.6
73
92
90.3
79
129
80.5
70
91
PCB-44
8
94.9
92
102
99.9
81
155
93.8
83
106
PCB-52+73
8
90.1
75
106
95.1
75
152
80.1
28
114
PCB-54
8
89.3
87
91
89.6
82
116
84.5
67
93
PCB-74+61
8
95.3
92
101
96.9
81
149
82.3
71
102
PCB-66+80
8
95.1
84
102
93.9
77
153
175.9
65
557
PCB-70
8
194.8
84
554
89.8
47
149
224.6
84
699
PCB-72
8
112.0
94
136
114.5
84
143
111.4
89
142
PCB-77
8
88.4
80
103
85.4
80
91
87.9
75
102
PCB-78
8
110.3
103
120
111.3
94
175
228.5
109
669
PCB-79
8
105.1
94
119
117.3
89
196
106.3
88
134
PCB-85+120
8
89.1
84
99
85.8
75
104
84.8
77
97
PCB-90+101+89
8
85.3
57
95
84.1
56
152
77.4
45
110
PCB-95+93
8
93.0
80
153
87.8
69
137
79.5
65
99
PCB-96
8
104.8
93
150
99.1
81
147
96.0
91
105
PCB-98+102
8
102.1
93
113
103.4
90
143
80.0
0(2)
114
PCB-99
8
86.3
55
99
83.8
61
166
81.9
51
118
PCB-104
8
89.3
86
99
89.9
79
131
86.1
79
94
PCB-105+127
8
92.5
88
97
87.6
74
102
85.0
68
101
PCB-118+106
8
91.3
88
95
81.4
62
104
79.3
39
116
PCB-107+108
8
104.5
98
113
105.9
91
115
114.1
91
158
PCB-110
8
86.0
82
97
82.1
59
101
68.1
20
104
PCB-126
8
87.3
83
91
85.3
73
90
90.1
85
99
PCB-132+168
8
94.1
89
107
91.1
75
107
80.1
67
91
PCB-138+163+164
8
88.4
85
96
50.8
7
101
65.1
-12
143
Method 1628 Multi-Laboratory Validation Study Report
89
April 2021
-------
Table 42. Matrix Spike Recoveries for Tissue Samples
Analyte
#
Results
Tissue #1
Tissue #2
Tissue #3
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
Mean
Rec.
Min.
Rec.
Max.
Rec.
PCB-149+139
8
92.0
86
97
71.6
54
94
77.1
38
108
PCB-147
8
89.5
82
95
87.5
82
92
87.5
84
94
PCB-152
8
108.0
94
121
102.4
88
120
105.9
95
117
PCB-153
8
95.8
84
112
43.3
6
96
65.0
-11
150
PCB-155
8
87.6
85
90
90.8
82
129
85.8
79
94
PCB-156
8
97.3
91
105
92.1
80
99
89.1
62
108
PCB-166
8
103.5
97
112
99.5
90
107
103
99
108
PCB-169
8
92.9
86
106
82.3
71
88
90.3
82
104
PCB-177
8
92.8
90
100
81.1
75
92
90.1
81
99
PCB-180
8
91.8
85
101
62.9
49
87
79.4
50
118
PCB-187+182
8
102.9
99
113
81.5
67
93
93.0
86
103
PCB-184
8
95.8
87
106
92.6
83
101
95.4
89
101
PCB-188
8
87.8
83
91
86.3
81
92
86.3
81
90
PCB-189
8
82.5
61
94
82.4
61
92
79.4
61
98
PCB-199
8
90.5
71
98
88.5
76
101
81.1
55
97
PCB-202
8
88.5
85
92
80.8
67
87
86.3
81
92
PCB-204
8
99.6
90
104
94.3
78
104
104.8
97
112
PCB-205
8
80.3
61
91
78.6
61
86
78.9
63
91
PCB-206
8
85.4
69
96
80.4
68
88
80.3
65
89
PCB-208
8
83.9
68
93
78.9
60
87
81.6
70
91
PCB-209
8
91.5
87
101
75.9
65
94
88.4
79
94
The matrix spike recoveries in the tissue samples were generally much more consistent than for the
sediment samples. Across all four laboratories and all three tissue samples, the mean recoveries range
from about 43% to 229%, with only 4 congeners with mean recoveries over 120% (PCB-66+80, PCB-70,
and PCB-78) in Tissue Samples #1 and #3. Those four recoveries over 120% were driven by the very
high reported recoveries from Laboratory 6 in those two samples. In contrast, Laboratory 6 reported
recoveries of those three congeners in Tissue Sample #2 that ranged from about 84% to 104%, which
suggests that whatever the cause of the very high recoveries in the other samples, it may have been
sample-specific, and not an issue of laboratory bias or other error.
Laboratory 6 reported no recoveries for PCB-98+102 in Tissue Sample #3 in either the MS or MSD
aliquots. Laboratory 3 also reported negative recoveries in Tissue Sample #3 for PCB-138+163+164 and
PCB-153 in both the MS and MSD aliquots. Beyond those analytes, no other negative recoveries were
reported.
As can be seen in Figure 8, the highest mean recoveries were reported from Tissue #1 and Tissue #3, and
represent PCB-66+80, PCB-70, and PCB-78.
Method 1628 Multi-Laboratory Validation Study Report
90
April 2021
-------
Tissue Mean MS Recovery
o
\r
Csl
VIII
Tissl
Tiss2
Tiss3
O
CN
CM
O
O
CM
O
CO
o
CD
O
-sT
O
CM
O
O
o
00
o
CD
O
Congener
Figure 8. Mean Matrix Spike Tissue Recoveries by Sample, in Elution Order
Dotted lines and Roman numerals delineate the levels of chlorination. The colored symbols and lines denote each of the three tissue
samples that were spiked.
Method 1628 Multi-Laboratory Validation Study Report
91
April 2021
-------
9. Labeled Compound Results
One of the most important aspects of the draft PCB method study is its use of isotope dilution quantitation
to determine the concentrations of the targeted analytes. As described in Section 4 of this report, each
sample to be analyzed is spiked with a suite of 29 13Ci2-labeled analogs of the target PCBs that are used as
quantitation reference standards for both true isotope dilution quantitation and a modified form of isotope
dilution for other target congeners in the same level of chlorination as the labeled compound.
This process results in an inherent correction of the target analyte concentration for the loss (or apparent
gain) of the labeled compound throughout the entire analytical process, including the extraction steps as
well as the many extract cleanup steps that are often necessary. Relative to the more commonly
employed internal standards that are injected into the final sample extract shortly before the instrumental
analysis, isotope dilution quantitation yields data that are both more accurate (less bias) and more precise.
Methods that rely on the analysis of MS/MSD samples to estimate accuracy and precision as a QC
measure typically limit those MS/MSD analyses to a small subset of all the samples prepared together,
with the typical frequency of 5%, or 1 in every 20 field samples. Whatever accuracy and precision
information is generated is often assumed to apply to the entire sample batch, even when samples from
different sources or locations are prepared and analyzed together.
In contrast, the labeled isotope dilution standards are added to every sample in the batch, so the analysis
generates sample-specific accuracy data, in the form of the measured recovery of each of the labeled
compounds in each sample.
CSRA and EPA compiled the labeled compound recovery data from all of analyses of spiked and
unspiked study samples of wastewaters, sediments, biosolids, and fish tissues. Those results are
discussed in the sections below, by matrix type.
Aqueous Sample Labeled Compound Results
The labeled compound recoveries from the seven laboratories that completed the aqueous sample portion
of the study are summarized in Table 43 below. These data represent the analyses of the unspiked, MS,
and MSD aliquots of all eight wastewater samples, for a total of 168 observations for each labeled
compound (4,872 observations in all). The table contains the observed mean, minimum, and maximum
recoveries from those 168 observations for each labeled compound, across all of the seven laboratories.
Values in parentheses, next to zero values, represent the number of results which were reported as non-
detects by the laboratories. The table also contains the QC acceptance criteria that CSRA calculated from
those results.
Table 43. Observed Aqueous Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-1
7
168
41.8
0.0 (1)
142.6
-31
115
13Ci2-PCB-3
7
168
46.0
0.0 (1)
110.1
-21
114
13Ci2-PCB-4
7
168
46.5
1.4
102.4
-20
113
13Ci2-PCB-11
7
168
52.0
2.2
88.8
-11
115
13Ci2-PCB-15
7
168
56.6
3.1
104.1
-10
123
13Ci2-PCB-19
7
168
48.7
2.1
82.4
-13
111
13Ci2-PCB-28
7
168
57.0
9.0
92.5
-8
122
13Ci2-PCB-37
7
168
61.6
8.9
104.4
-11
134
13Ci2-PCB-52
7
168
40.9
1.5
78.8
-18
99
13Ci2-PCB-54
7
168
46.0
6.9
85.5
-16
108
13Ci2-PCB-70
7
168
50.2
3.4
91.0
-20
120
Method 1628 Multi-Laboratory Validation Study Report
92
April 2021
-------
Table 43. Observed Aqueous Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-77
7
168
56.6
6.0
117.7
-15
128
13Ci2-PCB-85
7
168
53.5
5.5
95.0
-12
119
13Ci2-PCB-101
7
168
52.7
5.6
95.0
-13
118
13Ci2-PCB-104
7
168
45.8
5.6
86.5
-12
104
13Ci2-PCB-118
7
168
55.3
5.5
95.5
-14
125
13Ci2-PCB-126
7
168
56.2
5.3
94.5
-17
130
13Ci2-PCB-138
7
168
54.5
4.9
96.5
-21
130
13Ci2-PCB-153
7
168
48.7
4.7
93.8
-14
111
13Ci2-PCB-155
7
168
53.5
5.0
97.3
-20
127
13Ci2-PCB-169
7
168
52.3
4.5
106.5
-31
135
13Ci2-PCB-180
7
168
53.5
4.2
97.8
-31
138
13Ci2-PCB-188
7
168
50.4
4.3
95.3
-24
125
13Ci2-PCB-189
7
168
52.0
3.3
123.5
-46
150
13Ci2-PCB-202
7
168
50.4
3.6
93.0
-34
135
13Ci2-PCB-205
7
168
51.5
1.1
143.5
-64
167
13Ci2-PCB-206
7
168
50.3
1.1
140.3
-63
164
13Ci2-PCB-208
7
168
51.0
0.6
150.3
-69
171
13Ci2-PCB-209
7
168
49.2
0.3
166.8
-69
168
Overall, the observed labeled compound recoveries are typical of what one would expect from a method
with multiple cleanup procedures. The minimum recoveries ranged from 0% to 9% across the 29 labeled
compounds, while the maximum recoveries ranged from about 79% to 167%. The mean recoveries
across all seven laboratories, and all eight wastewater matrix types, both unspiked and spiked, ranged
from about 41% to 62%.
Since the development of EPA's first isotope dilution methods for wastewater and other matrices in the
late 1970s, the Office of Water has used two approaches to establishing QC acceptance limits for labeled
compound recovery. Early efforts derived limits from the results of an interlaboratory method validation
study, using the same statistical procedures used to derive the acceptance limits for the target analytes in
IPR and OPR analyses. The second approach has been to set the acceptance limits as simpler whole
number ranges and then evaluate those limits using the results of an interlaboratory method validation
study.
The calculated acceptance criteria in Table 43 represent the first of those two approaches. As can be seen
in Table 43, the calculated lower recovery limits for all 29 labeled congeners are negative numbers. As
discussed previously in this report, negative recovery values have no physical meaning for either labeled
compounds, or for spiked target analytes. Rather, the calculated acceptance criteria are a function of the
variability in the observed recovery data across all of the laboratories and samples in the study.
The calculated upper recovery limits for all 29 labeled congeners ranged from 115% to 171%, with the
upper limits for 25 of the 29 labeled compounds between 115% and 150%. The four labeled compounds
with the highest upper limits are the four labeled compounds with the highest levels of chlorination, e.g.,
the labeled analogs of PCBs 205 to 209.
Both the lower and upper limits are driven by the variability within each laboratory and across all
laboratories and all samples. Thus, if one or more of the laboratories have highly variable recoveries, the
width of the acceptance limits for the labeled compounds can be very wide, even if no actual recoveries
approach those limits.
Method 1628 Multi-Laboratory Validation Study Report
93
April 2021
-------
Clearly, the calculated negative recovery values in Table 43 are not useful as lower limits. Therefore,
CSRA examined the observed recovery data and compared them to two potential consensus-based
acceptance limits to determine the frequencies at which the results from the study would fail to meet those
potential acceptance limits. The draft method used acceptance limits of 15 - 130% for the labeled
compounds, so those limits were used as a potential set of limits for the final method. However, EPA
also evaluated the study results using two additional lower limits of 10% and 25%, and one additional
upper limit of 150%. Table 44 contains the rates at which the results from the study failed to meet those
potential lower and upper acceptance limits.
Table 44. Observed Aqueous Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-1
168
6.5
10.7
19.0
1.2
0.0
13Ci2-PCB-3
168
5.4
8.3
14.9
0.0
0.0
13Ci2-PCB-4
168
3.0
7.1
14.3
0.0
0.0
13Ci2-PCB-11
168
0.6
3.0
8.3
0.0
0.0
13Ci2-PCB-15
168
0.6
3.0
6.5
0.0
0.0
13Ci2-PCB-19
168
1.8
6.0
9.5
0.0
0.0
13Ci2-PCB-28
168
1.2
3.0
6.5
0.0
0.0
13Ci2-PCB-37
168
0.6
2.4
5.4
0.0
0.0
13Ci2-PCB-52
168
4.8
6.5
13.7
0.0
0.0
13Ci2-PCB-54
168
1.8
3.6
8.9
0.0
0.0
13Ci2-PCB-70
168
4.2
6.0
8.9
0.0
0.0
13Ci2-PCB-77
168
1.8
3.6
7.7
0.0
0.0
13Ci2-PCB-85
168
1.8
3.6
7.7
0.0
0.0
13Ci2-PCB-101
168
1.8
3.6
7.7
0.0
0.0
13Ci2-PCB-104
168
2.4
3.6
9.5
0.0
0.0
13Ci2-PCB-118
168
1.8
3.6
7.7
0.0
0.0
13Ci2-PCB-126
168
1.8
3.6
7.1
0.0
0.0
13Ci2-PCB-138
168
2.4
3.6
8.3
0.0
0.0
13Ci2-PCB-153
168
3.0
3.6
8.9
0.0
0.0
13Ci2-PCB-155
168
2.4
3.6
8.9
0.0
0.0
13Ci2-PCB-169
168
1.8
6.0
12.5
0.0
0.0
13Ci2-PCB-180
168
2.4
6.0
9.5
0.0
0.0
13Ci2-PCB-188
168
3.6
4.2
11.3
0.0
0.0
13Ci2-PCB-189
168
4.2
8.9
15.5
0.0
0.0
13Ci2-PCB-202
168
3.6
7.1
13.1
0.0
0.0
13Ci2-PCB-205
168
8.3
11.9
16.1
3.0
0.0
13Ci2-PCB-206
168
8.3
11.9
17.9
3.0
0.0
13Ci2-PCB-208
168
11.9
13.7
16.7
1.2
0.0
13Ci2-PCB-209
168
11.9
13.1
18.5
1.8
0.6
The rate at which the study's aqueous results failed a 25% lower limit ranged from 5.4% to 19%, with the
higher rates occurring for the lightest and heaviest of the labeled congeners. Using the 15% lower
recovery limit, the failures rates ranged from 2.4% to 13.7%. The majority of those failures were
concentrated in two of the laboratories (Labs 1 and 7 accounted for 232 of 293 failures at 15%, about 80%
of the failures came from 2 of the 7 laboratories). Using a 10% lower recovery limit further reduced the
failure rates to 0.6% to 11.9%, which were still concentrated in Labs 1 and 7.
The rate at which the study results failed a 130% upper recovery limit was trivial in comparison. Only 5
of the 29 labeled compounds exhibited any failures of the upper limit, and those failure rates ranged from
Method 1628 Multi-Laboratory Validation Study Report
94
April 2021
-------
1.2% to 3.0%. The evaluation of the study result against the 150% upper recovery limit led to 0%
failures.
EPA has previously shown that isotope dilution quantitation functions well even when the observed
recovery of the labeled compound drops as low as 5%. Therefore, based on the observed recoveries in
this study, EPA recommends retaining the 15 - 130% labeled compound recovery limits for aqueous
samples in the final method. Given that there are 29 labeled compounds that are being tested
simultaneously, a laboratory will be allowed to have up to three labeled compounds in a sample that do
not meet the acceptance criterion, provided that those compounds have at least 5% recovery. EPA will
also add language in the final version of the method to advise laboratories how to develop and utilize in-
house limits for the recoveries of labeled compounds in all matrices (see Appendix B of this report).
Sediment Sample Labeled Compound Results
The labeled compound recoveries from the six laboratories that completed the sediment sample portion of
the study are summarized in Table 45 below. These data represent the analyses of the unspiked, MS, and
MSD aliquots of all three sediment samples, for a total of 54 observations for each labeled compound.
Table 45. Observed Sediment Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-1
6
54
48.4
1.7
122.4
-99
196
13Ci2-PCB-3
6
54
55.0
2.8
128.7
-112
222
13Ci2-PCB-4
6
54
50.6
2.5
103.4
-84
186
13Ci2-PCB-11
6
54
69.3
5.3
190.2
-121
260
13Ci2-PCB-15
6
54
65.1
5.9
144.3
-94
224
13Ci2-PCB-19
6
54
56.0
3.9
118.2
-101
213
13Ci2-PCB-28
6
54
65.0
6.8
149.9
-97
227
13Ci2-PCB-37
6
54
74.7
10.8
185.9
-97
247
13Ci2-PCB-52
6
54
56.6
4.4
128.8
-156
270
13Ci2-PCB-54
6
54
58.4
6.5
140.2
-92
208
13Ci2-PCB-70
6
54
59.7
8.2
146.4
-82
201
13Ci2-PCB-77
6
54
63.8
10.3
142.2
-75
202
13Ci2-PCB-85
6
54
66.9
00
CO
161.0
-114
248
13Ci2-PCB-101
6
54
63.8
8.2
156.0
-105
233
13Ci2-PCB-104
6
54
60.7
5.6
146.3
-127
248
13Ci2-PCB-118
6
54
65.9
9.5
160.9
-102
234
13Ci2-PCB-126
6
54
66.6
10.6
144.6
-108
241
13Ci2-PCB-138
6
54
66.2
9.7
157.3
-101
233
13Ci2-PCB-153
6
54
59.4
7.2
154.6
-96
215
13Ci2-PCB-155
6
54
64.1
9.4
159.7
-96
224
13Ci2-PCB-169
6
54
61.8
0.0
140.0
-92
216
13Ci2-PCB-180
6
54
62.0
10.4
158.8
-89
213
13Ci2-PCB-188
6
54
61.1
8.2
154.6
-95
217
13Ci2-PCB-189
6
54
55.8
9.1
131.6
-104
215
13Ci2-PCB-202
6
54
60.1
8.7
152.2
-92
213
13Ci2-PCB-205
6
54
53.4
9.5
125.5
-92
199
13Ci2-PCB-206
6
54
55.4
9.1
172.3
-97
208
13Ci2-PCB-208
6
54
50.7
10.3
118.1
-97
199
13Ci2-PCB-209
6
54
55.0
10.2
178.0
-120
230
Overall, the observed labeled compound recoveries are typical of what one would expect from a method
with multiple cleanup procedures. The minimum recoveries ranged from 1.7% to 10.8% across the 29
Method 1628 Multi-Laboratory Validation Study Report
95
April 2021
-------
labeled compounds, while the maximum recoveries ranged from about 103% to 190%. The mean
recoveries across all 6 laboratories, 3 samples, both unspiked and spiked, ranged from about 48% to 75%.
The calculated acceptance criteria in Table 45 represent the same approach described for the aqueous
sample results. As can be seen in Table 45, the calculated lower recovery limits for all 29 labeled
congeners are negative numbers, and are more negative than for the calculated aqueous limits.
The calculated upper recovery limits for all 29 labeled congeners ranged from 186% to 270%, much
greater than the upper limits for the aqueous samples.
However, EPA also evaluated the study results using two additional lower limits of 10% and 25%, and
one additional upper limit of 150%. Table 46 contains the rates at which the results from the study failed
to meet those potential lower and upper acceptance limits.
Table 46. Observed Sediment Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-1
54
16.7
22.2
29.6
0.0
0.0
13Ci2-PCB-3
54
11.1
14.8
27.8
0.0
0.0
13Ci2-PCB-4
54
9.3
14.8
25.9
0.0
0.0
13Ci2-PCB-11
54
5.6
13.0
22.2
7.4
7.4
13Ci2-PCB-15
54
5.6
13.0
22.2
5.6
0.0
13Ci2-PCB-19
54
7.4
16.7
25.9
0.0
0.0
13Ci2-PCB-28
54
5.6
13.0
22.2
5.6
0.0
13Ci2-PCB-37
54
0.0
11.1
18.5
5.6
5.6
13Ci2-PCB-52
54
5.6
16.7
25.9
0.0
0.0
13Ci2-PCB-54
54
5.6
13.0
22.2
1.9
0.0
13Ci2-PCB-70
54
1.9
9.3
22.2
5.6
0.0
13Ci2-PCB-77
54
0.0
7.4
16.7
3.7
0.0
13Ci2-PCB-85
54
1.9
11.1
24.1
5.6
3.7
13Ci2-PCB-101
54
1.9
13.0
24.1
5.6
1.9
13Ci2-PCB-104
54
5.6
16.7
25.9
5.6
0.0
13Ci2-PCB-118
54
1.9
9.3
22.2
5.6
3.7
13Ci2-PCB-126
54
0.0
5.6
18.5
3.7
0.0
13Ci2-PCB-138
54
0.0
7.4
20.4
5.6
1.9
13Ci2-PCB-153
54
3.7
14.8
24.1
5.6
1.9
13Ci2-PCB-155
54
1.9
9.3
24.1
5.6
3.7
13Ci2-PCB-169
54
7.4
9.3
22.2
1.9
0.0
13Ci2-PCB-180
54
0.0
7.4
20.4
5.6
3.7
13Ci2-PCB-188
54
1.9
11.1
24.1
5.6
1.9
13Ci2-PCB-189
54
1.9
11.1
24.1
1.9
0.0
13Ci2-PCB-202
54
1.9
11.1
24.1
5.6
1.9
13Ci2-PCB-205
54
0.0
11.1
22.2
0.0
0.0
13Ci2-PCB-206
54
1.9
13.0
29.6
3.7
3.7
13Ci2-PCB-208
54
0.0
7.4
25.9
0.0
0.0
13Ci2-PCB-209
54
0.0
14.8
31.5
1.9
1.9
The rate at which the study's sediment results failed a 25% lower recovery limit ranged from 16.7% to
31.5% with rates above 20% spread across most of the labeled congeners. The results failed a 15% lower
recovery limit ranged from 5.6% to 22.2%, with the higher rates occurring for the lightest and heaviest of
the labeled congeners. The failure rates for the sediments are almost double that for the aqueous samples.
However, the vast majority of those failures were concentrated in two of the laboratories (Labs 4 and 6
Method 1628 Multi-Laboratory Validation Study Report
96
April 2021
-------
accounted for 196 of 198 failures, and both those laboratories had many fewer failures for the aqueous
samples in the study). The rate at which the study results failed a 10% lower recovery limit ranged from
0.0% to 16.7%, with 27 of the 29 labeled congeners having failure rates below 10% at that 10% recovery
limit.
The rate at which the study results failed a 130% upper recovery limit ranged from 0% to 7.4%. Of the
29 labeled compounds, 15 compounds exhibited more than a 5% failure rate at the upper recovery limit,
although most of those 15 congeners had a rate of 5.6%.
Given the very high numbers of failures observed for Labs 4 and 6 but moderately better performance of
the other four laboratories that completed the sediment portion of the study, EPA recommends retaining
the 15 - 130% labeled compound recovery limits for sediment samples in the final method. Given that
there are 29 labeled compounds that are being tested simultaneously, a laboratory will be allowed to have
up to three labeled compounds in a sample that do not meet the acceptance criterion, provided that those
compounds have at least 5% recovery. As noted for the aqueous sample labeled compound recoveries,
EPA will also add language in the final version of the method to advise laboratories how to develop and
utilize in-house limits for the recoveries of labeled compounds in all matrices (see Appendix B of this
report).
Biosolids Sample Labeled Compound Results
The labeled compound recoveries from the four laboratories that completed the biosolids sample portion
of the study are summarized in Table 47 below. These data represent the analyses of the unspiked, MS,
and MSD aliquots of all three biosolids samples, for a total of 36 observations for each labeled
compound.
Table 47. Observed Biosolids Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-1
4
36
39.4
10.0
67.0
-17
96
13Ci2-PCB-3
4
36
46.8
11.0
75.0
-21
114
13Ci2-PCB-4
4
36
58.9
14.0
117.4
-35
153
13Ci2-PCB-11
4
36
70.3
21.5
104.5
-6
146
13Ci2-PCB-15
4
36
76.4
22.5
126.5
-27
180
13Ci2-PCB-19
4
36
57.2
18.0
90.0
-15
129
13Ci2-PCB-28
4
36
69.9
17.9
113.5
-53
193
13Ci2-PCB-37
4
36
88.8
22.4
201.5
-141
318
13Ci2-PCB-52
4
36
52.2
17.0
85.0
-21
125
13Ci2-PCB-54
4
36
61.2
22.7
99.5
-20
142
13Ci2-PCB-70
4
36
69.6
25.8
113.6
-10
149
13Ci2-PCB-77
4
36
69.7
0.0
117.9
-100
239
13Ci2-PCB-85
4
36
73.6
21.3
118.0
-22
170
13Ci2-PCB-101
4
36
67.3
18.8
112.0
-24
159
13Ci2-PCB-104
4
36
63.4
22.5
108.0
-22
148
13Ci2-PCB-118
4
36
72.7
18.3
119.0
-51
197
13Ci2-PCB-126
4
36
73.4
16.8
122.5
-126
272
13Ci2-PCB-138
4
36
75.8
22.5
119.0
-62
214
13Ci2-PCB-153
4
36
67.2
22.2
113.5
-19
153
13Ci2-PCB-155
4
36
71.9
19.4
118.5
-57
201
13Ci2-PCB-169
4
36
62.7
0.0
122.5
-129
254
13Ci2-PCB-180
4
36
75.2
19.9
117.0
-56
207
13Ci2-PCB-188
4
36
68.7
17.9
113.0
-42
180
13Ci2-PCB-189
4
36
71.6
16.2
113.7
-59
202
13Ci2-PCB-202
4
36
70.0
17.1
105.5
-53
193
Method 1628 Multi-Laboratory Validation Study Report
97
April 2021
-------
Table 47. Observed Biosolids Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-205
4
36
66.6
18.2
105.5
-46
180
13Ci2-PCB-206
4
36
63.2
17.4
101.6
-30
156
13Ci2-PCB-208
4
36
65.3
14.3
103.5
-40
171
13Ci2-PCB-209
4
36
62.6
14.9
104.9
-43
168
Overall, the observed labeled compound recoveries are typical of what one would expect from a method
with multiple cleanup procedures. The minimum recoveries ranged from 0% to 25.8% across the 29
labeled compounds, while the maximum recoveries ranged from about 67% to 202%. The mean
recoveries across all 4 laboratories, 3 samples, both unspiked and spiked, ranged from about 39% to 89%.
The calculated acceptance criteria in Table 47 represent the same approach described for the aqueous
sample results. As can be seen in Table 47, the calculated lower recovery limits for all 29 labeled
congeners are negative numbers. Somewhat surprisingly, given the known analytical challenges of
biosolids, the calculated lower recovery limits are not as extreme as those for the sediment samples
(e.g., in Table 45).
The calculated upper recovery limits for all 29 labeled congeners ranged from 96% to 318%, much
greater than the upper limits for the aqueous samples, but often lower than the corresponding upper limits
for sediments.
However, EPA also evaluated the study results using two additional lower limits of 10% and 25%, and
one additional upper limit of 150%. Table 48 contains the rates at which the results from the study failed
to meet those potential lower and upper acceptance limits.
Table 48. Observed Biosolids Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-1
36
0.0
2.8
16.7
0.0
0.0
13Ci2-PCB-3
36
0.0
2.8
13.9
0.0
0.0
13Ci2-PCB-4
36
0.0
2.8
2.8
0.0
0.0
13Ci2-PCB-11
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-15
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-19
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-28
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-37
36
0.0
0.0
2.8
11.1
8.3
13Ci2-PCB-52
36
0.0
0.0
8.3
0.0
0.0
13Ci2-PCB-54
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-70
36
0.0
0.0
0.0
0.0
0.0
13Ci2-PCB-77
36
16.7
16.7
16.7
0.0
0.0
13Ci2-PCB-85
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-101
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-104
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-118
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-126
36
0.0
0.0
8.3
0.0
0.0
13Ci2-PCB-138
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-153
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-155
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-169
36
16.7
16.7
25.0
0.0
0.0
13Ci2-PCB-180
36
0.0
0.0
2.8
0.0
0.0
Method 1628 Multi-Laboratory Validation Study Report
98
April 2021
-------
Table 48. Observed Biosolids Labeled Compound Recovery Failure Rates for Potential Acceptance
Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-188
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-189
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-202
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-205
36
0.0
0.0
8.3
0.0
0.0
13Ci2-PCB-206
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-208
36
0.0
2.8
11.1
0.0
0.0
13Ci2-PCB-209
36
0.0
0.0
8.3
0.0
0.0
The rate at which the study's biosolids results failed a 25% lower recovery limit ranged from 0.0% to
25.0% with 17 of the 29 labeled compounds exhibiting failures rates of less than 5%. For a 15% lower
recovery limit, the failure rates ranged from 0% to 16.7% (for labeled congeners 77 and 169). At the 10%
lower recovery limit, only two labeled congeners failed at all, 77and 169, both at 16.7%. The failure rates
for the biosolids are surprisingly low and the majority of those failures were concentrated in one of the
laboratories (Lab 4 accounted for 13 of 16 failures).
Only the labeled analog of PCB-37 failed the 130% upper recovery limit, at a rate of 11.1% (4 of 36
samples) and no labeled congeners failed the 150% upper recovery limit.
Given the data from this portion of the study, EPA recommends retaining the 15 - 130% labeled
compound recovery limits for biosolids in the final method. Given that there are 29 labeled compounds
that are being tested simultaneously, a laboratory will be allowed to have up to three labeled compounds
in a sample that do not meet the acceptance criterion, provided that those compounds have at least 5%
recovery. As noted for the aqueous and sediment sample labeled compound recoveries, EPA will also
add language in the final version of the method to advise laboratories how to develop and utilize in-house
limits for the recoveries of labeled compounds in all matrices (see Appendix B of this report).
Tissue Sample Labeled Compound Results
The labeled compound recoveries from the four laboratories that completed the tissue sample portion of
the study are summarized in Table 49 below. These data represent the analyses of the unspiked, MS, and
MSD aliquots of all three fish tissue samples, for a total of 36 observations for each labeled compound.
Table 49. Observed Tissue Labeled Compound Recoveries and Calculated Acceptance Criteria
Congener
# Labs
# Results
Observed Recoveries
Calculated Acceptance Criteria
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-1
4
36
50.5
13.5
103.7
-93
194
13Ci2-PCB-3
4
36
54.7
17.5
110.0
-89
198
13Ci2-PCB-4
4
36
54.9
17.8
113.0
-96
206
13Ci2-PCB-11
4
36
63.2
23.3
118.8
-89
215
13Ci2-PCB-15
4
36
63.5
20.6
120.6
-97
224
13Ci2-PCB-19
4
36
58.3
21.0
113.7
-86
202
13Ci2-PCB-28
4
36
66.8
24.0
116.8
-77
211
13Ci2-PCB-37
4
36
71.7
23.3
127.2
-55
199
13Ci2-PCB-52
4
36
49.6
19.0
81.0
-69
169
13Ci2-PCB-54
4
36
53.7
20.8
73.7
-48
155
13Ci2-PCB-70
4
36
59.8
24.1
103.3
-26
145
13Ci2-PCB-77
4
36
59.8
21.7
79.4
1
118
13Ci2-PCB-85
4
36
59.3
20.7
79.1
-16
134
13Ci2-PCB-101
4
36
57.6
20.3
77.4
-27
142
Method 1628 Multi-Laboratory Validation Study Report
99
April 2021
-------
Table 49. Observed Tissue Labeled Compound Recoveries and Calculated Acceptance Criteria
Observed Recoveries
Calculated Acceptance Criteria
Congener
# Labs
# Results
Mean
Min.
Max.
Lower Limit
Upper Limit
13Ci2-PCB-104
4
36
52.6
18.7
74.7
-62
168
13Ci2-PCB-118
4
36
61.2
21.2
81.3
-6
128
13Ci2-PCB-126
4
36
59.7
15.9
84.1
-18
138
13Ci2-PCB-138
4
36
63.3
22.5
83.4
3
124
13Ci2-PCB-153
4
36
56.2
19.8
75.8
-40
152
13Ci2-PCB-155
4
36
61.7
22.4
79.5
-6
129
13Ci2-PCB-169
4
36
56.7
10.3
95.7
-113
226
13Ci2-PCB-180
4
36
65.1
24.4
92.6
-8
138
13Ci2-PCB-188
4
36
60.2
21.6
79.2
-15
136
13Ci2-PCB-189
4
36
66.7
23.5
92.0
-7
140
13Ci2-PCB-202
4
36
61.8
23.1
81.0
-13
137
13Ci2-PCB-205
4
36
65.6
24.2
88.3
-27
158
13Ci2-PCB-206
4
36
63.1
23.9
89.5
-68
195
13Ci2-PCB-208
4
36
64.4
21.6
89.3
-31
160
13Ci2-PCB-209
4
36
60.9
20.5
86.2
-93
215
Overall, the observed labeled compound recoveries in fish tissue were quite good, especially for a method
with multiple cleanup procedures. The minimum recoveries ranged from 10.3% to 24.4% across the 29
labeled compounds, while the maximum recoveries ranged from about 73% to 127%. The mean
recoveries across all 4 laboratories, 3 samples, both unspiked and spiked, ranged from about 50% to 72%.
The calculated acceptance criteria in Table 49 represent the same approach described for the aqueous
sample results. As can be seen in Table 49, all but one of the calculated lower recovery limits are
negative numbers, and the one positive value is at 1%. The calculated upper recovery limits for all 29
labeled congeners ranged from 118% to 226%.
However, EPA also evaluated the study results using two additional lower limits of 15% and 25%, and
one additional upper limit of 150%. Table 50 contains the rates at which the results from the study failed
to meet those potential lower and upper acceptance limits.
Table 50. Observed Tissue Labeled Compound Recovery Failure Rates for Potential Acceptance Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-1
36
0.0
2.8
30.6
0.0
0.0
13Ci2-PCB-3
36
0.0
0.0
16.7
0.0
0.0
13Ci2-PCB-4
36
0.0
0.0
16.7
0.0
0.0
13Ci2-PCB-11
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-15
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-19
36
0.0
0.0
11.1
0.0
0.0
13Ci2-PCB-28
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-37
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-52
36
0.0
0.0
11.1
0.0
0.0
13Ci2-PCB-54
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-70
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-77
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-85
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-101
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-104
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-118
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-126
36
0.0
0.0
2.8
0.0
0.0
Method 1628 Multi-Laboratory Validation Study Report
100
April 2021
-------
Table 50. Observed Tissue Labeled Compound Recovery Failure Rates for Potential Acceptance Criteria
Congener
Total #
Results
Observed Failure Rate (%)
If Lower
Limit = 10%
If Lower Limit
= 15%
If Lower
Limit = 25%
If Upper Limit
= 130%
If Upper Limit
= 150%
13Ci2-PCB-138
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-153
36
0.0
0.0
5.6
0.0
0.0
13Ci2-PCB-155
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-169
36
0.0
2.8
5.6
0.0
0.0
13Ci2-PCB-180
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-188
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-189
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-202
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-205
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-206
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-208
36
0.0
0.0
2.8
0.0
0.0
13Ci2-PCB-209
36
0.0
0.0
5.6
0.0
0.0
The rate at which the study's tissue results failed a 25% lower recovery limit ranged from 2.8% to 30.6%,
with 24 of the 29 labeled congeners having failure rates below 10%. The 5 labeled compounds that failed
the 25% lower recovery limit were concentrated in two of the laboratories (Labs 3 and 6 accounted for 62
of the 65 failures).
For the 15% lower recovery limit, there were only two congeners that failed at all, at 2.8% for the labeled
analogs of PCB-1 and PCB-169 (i.e., 1 out of 36 samples). Those two failures occurred in two different
laboratories. No labeled congeners failed the 10% lower recovery limit. No labeled congeners failed the
130% upper recovery limit or the 150% upper recovery limit.
EPA recommends retaining the 15 - 130% labeled compound recovery limits for tissue in the final
method. Given that there are 29 labeled compounds that are being tested simultaneously, a laboratory
will be allowed to have up to three labeled compounds in a sample that do not meet the acceptance
criterion, provided that those compounds have at least 5% recovery. As noted for the aqueous, sediment,
and biosolid sample labeled compound recoveries, EPA will also add language in the final version of the
method to advise laboratories how to develop and utilize in-house limits for the recoveries of label
compounds in all matrices (see Appendix B of this report).
Method 1628 Multi-Laboratory Validation Study Report
101
April 2021
-------
10. Data Review and Validation
The results for all of the analyses in this study were submitted as electronic data deliverables (EDDs) in
Excel format, and supported by raw data and reporting forms provided in PDF format equivalent to a
hardcopy data package. Both the electronic data deliverables and the supporting raw data were reviewed
for completeness and for data quality. Data were evaluated based on the preliminary method performance
criteria described in the draft procedure, but because one purpose of the study is to develop formal
acceptance criteria, all of the results were retained for further consideration. The formal method
performance criteria will be established as a result of this multi-laboratory validation study. The data
review process was patterned after that used for PCB results from various other Office of Water studies.
Completeness check - The supporting data provided in the "hardcopy package" were compared to the
results in the EDD. The data report narratives in the hardcopy package were reviewed and any quality
control or performance related issues were noted. The data was verified to be consistent with the
narrative and appropriate validation qualifiers were applied. Electronic data deliverable (EDD) elements
and results were checked for completeness and consistency with the hardcopy data. Elements checked
included sample number and laboratory sample number identifiers, analysis date and time, Chemical
Abstract Service Registry Number (CASRN), congener number identifier, coeluting congeners,
laboratory qualifiers, found concentrations, detection limits, ion abundance ratios, relative retention times,
sample sizes (volume or weight), dilution factors, spiked amounts, percent recoveries, extraction date,
sample receipt date, and concentration units.
Hardcopy and EDD data were checked to ensure that no data were missing or inconsistent for all samples
and blanks. Hardcopy data were checked to ensure that all chromatograms and quantitation reports were
available for all analyses and that all samples were reported on the sample pretreatment, sample
preparation, extraction worksheet records.
Instrument Sensitivity - Results between the MDL and the laboratory's quantitation or nominal
reporting limit were checked to confirm they were flagged by the laboratory. Any flagged results were
checked to confirm the concentration warranted the qualifier (e.g., was between the MDL and the
quantitation limit). All reported non-detects were checked to confirm no signal was present, or that they
were detects below the MDL value.
Sample Dilutions - Results were checked to ensure they were within the range of the calibration curve.
If results were above the curve, then a diluted reanalysis should have been performed and results were to
be reported from the dilution.
Ion abundance ratios - Ion abundance ratios (IARs) were checked for all calibration standards, samples
and blanks to determine if they were within the preliminary method control limits. Deviations were
flagged, but the data retained.
Blank Contamination - Found concentrations in the method blank were compared to the associated
samples. If a congener concentration in a sample was greater than 10 times the same congener
concentration in the method blank, then the sample result was considered to be unaffected and the
following validation flag was applied, "B, RNAF" to indicate that the blank contamination was at a low
enough level to not significantly affect sample results.
IPR and OPR Recovery - The percent recoveries for native congeners in the OPR were checked to
determine if they were within preliminary control limits (i.e., 60 - 130% for mono- through tri-
chlorinated congeners in tissues and all other matrices, respectively; 70 - 130% for tetra- through deca-
chlorinated congeners in all matrices). Results outside of those preliminary criteria were noted, but
retained because one purpose of the study is to develop IPR and OPR acceptance criteria that are based on
Method 1628 Multi-Laboratory Validation Study Report
102
April 2021
-------
actual study data. Percent recovery calculations for a few congeners were independently performed to
ensure they were within 1% of the reported values.
Labeled Compound Recoveries - Labeled compound recoveries were checked to determine if they were
within the preliminary control limits in all samples and blanks (i.e., 15 - 130% for 13C-PCB-1 through
13C-PCB-19 and 40 - 130% for the remaining labeled congeners in all matrices). Results outside of those
preliminary criteria were noted, but retained because one purpose of the study is to develop labeled
compound acceptance criteria that are based on actual study data.
Quantification Check - Concentration result calculations for several native and labeled congeners in
each sample were independently performed by CSRA using equations provided in the draft method to
ensure they were within 1% percent of the reported values.
Resolution of questions and issues - Any issues of missing data, calculation errors, or questions about
specific results were communicated to the laboratory by email or telephone. The final resolution of the
issues may have involved resubmission of the EDD or the raw data, or missing portions thereof.
Correction of simpler issues, such as transposition errors in the EDD could be corrected by CSRA after
receiving concurrence from the laboratory by email, in which case a copy of the email was retained in the
project files.
Method 1628 Multi-Laboratory Validation Study Report
103
April 2021
-------
11. Conclusions
The Office of Water's Engineering and Analysis Division completed a multi-laboratory validation study
of a method for PCB congeners in wastewater, biosolids, sediments, and fish tissue. The multi-laboratory
validation study achieved all its intended goals, as outlined below.
Study Goals
1. Obtain data from matrices that are representative of the method's intended use
As described in Section 3, the wastewater matrices were a diverse selection of wastewaters from multiple
parts of the country with different physical parameters, as demonstrated in Table 3. The matrices chosen
are typical of what might be analyzed by a laboratory performing NPDES compliance monitoring and
included some pretreatment samples that would be more challenging than a typical NPDES compliance
final effluent. The three biosolid, sediment, and fish tissue samples were all reasonably typical samples
that might be analyzed for data gathering and monitoring in support of the Clean Water Act.
2. Obtain data from laboratories that are representative of those likely to use the approved
method, but that were not directly involved in its development
The laboratories that participated in this method validation study were mostly commercial laboratories
that routinely perform NPDES compliance monitoring analyses. One state and several EPA laboratories
also participated. Commercial laboratories are those most likely to use this method for NPDES
compliance monitoring, so the laboratories in this study are representative of the laboratories most likely
to use the method.
3. Obtain feedback from laboratory users on the specifics of the draft method (e.g., is it clear and
easy to understand, or are changes to the method text needed?)
Participant laboratories were all encouraged to provide feedback on the method, and most did. EPA has
revised the method in response to such feedback.
4. Use study data to characterize performance of the method
All of the data collected during this study were reviewed and evaluated to characterize the performance of
this method, as summarized in detail in Sections 4-9. This includes data on calibration, initial precision
and recovery, method detection limits, performance in real-world matrices, and labeled compound
recoveries.
5. Develop statistically derived QC acceptance criteria that will reflect method performance
capabilities in real-world situations
Sections 4, 5, 6, and 9 contain statistically derived QC limits that were calculated from the data collected
during this study. The laboratories that participated are representative of the real-world laboratories that
would potentially run this method, and the matrices are typical of matrices that a laboratory using this
method would analyze.
Method 1628 Multi-Laboratory Validation Study Report
104
April 2021
-------
Method Performance
Method performance is summarized below for each matrix type tested.
Wastewater
Since Method 608.3 is the most commonly used EPA method approved at 40 CFR Part 136 for PCBs in
wastewater, it was used as the basis of comparison for PCB congener Method 1628. The criteria used for
developing Method 1628, which could replace Method 608.3, included: ability to identify and quantify
individual PCB congeners instead of Aroclor mixtures, a higher sensitivity without being adversely
affected by typical laboratory background contamination, and implementation at a typical mid-sized full-
service environmental laboratory.
Before comparing Method 608.3 and Method 1628, it is important to state that Method 608 (subsequently
revised as Method 608.3) was the best available technology in the late 1970s when it was validated. The
dedicated laboratories, analysts, EPA employees, and contractors that developed Method 608 were using
the best tools available to them at the time. Their efforts pioneered one of the first validated EPA
analytical methods that is the foundation upon which Method 1628 and every other EPA method is built.
Analytical technology, laboratory information management systems, and EPA method quality control
monitoring have all improved significantly over the last 40 years, and these improvements make Method
1628 far superior to Method 608. Table 51 provides a side-by-side comparison of the two methods,
which illuminates this point.
Table 51. Comparison of Method 608.3 and Method 1628 for Aqueous Samples
Method 608.3
Method 1628
Calibration
3 to 5 calibration points for 2 Aroclors, one
calibration point for the other 5
6-point calibration with 48 congeners that are representative
of the 209 congeners.
Quantitation: Surrogates vs. Internal Standards
One surrogate is required, no specific surrogates
are specified, no criteria attached to the
performance of the surrogate(s), nor has any
testing been performed by EPA to validate any
surrogates.
29 13C-labeled isotope dilution standards, representing each
homolog and including the labeled analogs of the most
commonly detected congeners in the environment.
3 13C-labeled non-extracted internal standards are used to
calculate the recovery of the 29 isotope dilution standards.
The performance of these standards was tested in a variety
of wastewaters at seven laboratories to generate statistically
derived performance criteria.
Initial Precision and Recovery
Between 28-197% among the 7 Aroclor mixtures
Mostly between 50-130% recovery for the 48 calibrated
congeners, with some outliers
Method Detection Limits
Aroclor 1242 - 65 ng/L, no other Aroclor mixtures
were tested.
MDLs values ranged from 0.19 ng/L to 4.98 ng/L among the
209 congeners. None of the MDLb values were higher than
the pooled MDLs values, so blank contamination was not a
significant issue.
Ongoing Precision and Recovery
Aroclor-specific recovery criteria vary from as
narrow as 50-114%, to as wide as 10-215%
Congener-specific criteria for the 48 calibrated congeners
vary from as narrow as about 70-120%, to as wide as 14-
193%.
Wastewater Matrix Performance
Only MS/MSD reproducibility is stated in the
method. No recovery data are presented.
Mostly between 60-120% recovery for the 60 spiked
congeners, with a few outliers. False negatives in less than
0.2% of the 7,128 total data points.
Method 1628 Multi-Laboratory Validation Study Report
105
April 2021
-------
Calibration and Quantitation
Method 1628 is superior to Method 608.3 in its calibration and quantitation approach. Method 608.3 uses
a multi-point calibration for two of the seven Aroclors, while the other Aroclor mixtures are quantified
using a 1-point calibration. Method 608.3 also mentions that at least one surrogate should be used to
represent all the pesticides and Aroclors in the method, yet it does not provide any criteria for this
surrogate. On the other hand, Method 1628 has a 6-point calibration containing 48 congeners,
representing every homolog of the 209 congeners, plus 29 isotope dilution standards and 3 non-extracted
internal standards to track the measurement quality in every sample.
Method Detection Limits
When it comes to comparing Aroclor MDLs to individual congener MDLs, the process is not
straightforward. Based on the data from Frame el al. (1996), the main constituents of Aroclor 1242 are
PCB-8 (7.05%), PCB-18 (8.53%), PCB-28 (6.86%), PCB-31 (7.34%), and PCB-33 (5.01%). The pooled
aqueous MDL calculated in this study for PCB-18, the largest component of Aroclor 1242, was 0.46
ng/L. Assuming that all of the PCB-18 came from unweathered Aroclor 1242, detecting 0.46 ng/L in an
aqueous sample would equate to a concentration of 5.39 ng/L of Aroclor 1242. Similar estimates derived
for the other major components of Aroclor 1242 are shown in Table 52.
Table 52. Estimated Aroclor 1242 Concentrations Using 5 Most Prevalent Congeners
in Aqueous Matrices
Congener
% Contribution
Pooled MDL in
Estimated Concentration of Aroclor
to Aroclor-1242
this study (ng/L)
1242 at the Pooled MDL (ng/L)
8
7.05%
1.00
14.18
18
8.53%
0.46
5.39
28
6.86%
0.69
10.06
31
7.34%
0.50
6.81
33
5.01%
1.11
22.16
The published sensitivity data for Method 608.3 is only for Aroclor 1242, with an MDL of 65 ng/L in
aqueous samples. The highest estimate of 22.16 ng/L in Table 52, derived from the pooled MDL for
PCB-33, is roughly three times below the published MDL for Aroclor 1242. In fact, even the highest
reported MDLS value for these five congeners from any of the laboratories in the study, 3.15 ng/L for the
coeluting congeners PCB-33+20+21 (see Table 21), would yield an estimated Aroclor 1242 concentration
of 62.87 ng/L, which is still below the published MDL in Method 608.3.
Admittedly, the published MDL data for Aroclor 1242 date to the original version of Method 608.
However, the analyses for Aroclors performed as part of this study reported no Aroclors in the original
nine wastewater matrices. The MDLS values used by that laboratory ranged from 2.8 to 9.5 ng/L for the
seven common Aroclor mixtures, and were similar to MDL values provided by other laboratories
solicited for the effort. The MDLS for Aroclor 1242 was 9.5 ng/L, which is comparable to the Aroclor
1242 concentrations that were estimated from the congener results shown in Table 52. As shown in Table
4, while two of the wastewater matrices contained measurable concentrations of 5 and 40 of the
congeners in this method, no Aroclors were reported, even with more recent GC/ECD instrumentation.
This indicates that method sensitivity did not limit the ability of the laboratory to determine Aroclor 1242
using Method 608.3, but rather, it was the result of the lack of a recognizable Aroclor pattern, and
supports the position that analysis of congeners is superior to analysis of Aroclors because it provides a
direct measurement of the PCB contamination.
Performance in Wastewater
For wastewater analyses, Method 1628 is a more advanced method than the currently approved Method
608.3 by virtually any manner of comparison. In the original interlaboratory validation report for Method
Method 1628 Multi-Laboratory Validation Study Report
106
April 2021
-------
608 (published in June 1984), recoveries of <20% for Aroclors were very common among the six test
matrices, causing a significant quantity of data for the Aroclors to be rejected (e.g., over 15% of the data
for Aroclor 1254). Also, the matrices used in that validation study were significantly less challenging
(reagent water, a drinking water, a surface water, and three final effluents).
The validation study for Method 1628 used several pretreatment matrices that have not undergone any
treatment (refer to Table 2). Pretreatment matrices are generally more challenging than treated final
effluents. Ninety-eight percent of the mean recoveries for the matrix spike samples fell between 60% and
120%, with a false negative rate below 0.2%, demonstrating the ruggedness of the method across a range
of wastewaters.
Method 1628 uses a mass spectrometer, which is less prone to interferences than the electron capture
detector used in Method 608.3. The use of isotope dilution standards in Method 1628 corrects the target
analyte concentration in every sample for the recovery of the labeled standards in that sample, thus
accounting for matrix effects, and improving the accuracy and precision of the results, which is especially
important in challenging matrices.
Most importantly, Method 608.3 does not actually measure PCBs, but instead it measures seven Aroclor
mixture patterns. This is an indirect measurement that is prone to false negatives and low bias.
Manufacturing of PCBs has been banned in the U.S. for over 40 years. While Aroclor contamination is
an important legacy source of PCB contamination in the environment, much of the PCB contamination in
the environment is now so weathered that it no longer matches the original Aroclor mixture when
analyzed. Method 1628 addresses this issue by directly measuring the 209 PCB congeners. Measurement
of individual congeners has an added advantage because PCB contamination rarely involves just one
congener. A particularly difficult matrix may cause an interference that invalidates a low-level detect for
one congener, but it is unlikely that the sample will cause the same of interference for all of the PCB
congeners in the sample.
Sediments andBiosolids
The reference matrix used for solid samples was Ottawa sand. Since the reference matrix represented
solid samples as a category, the same reference matrix performance (IPRs, OPRs, and MDLs) was applied
for sediments and biosolids. The observed mean IPR recoveries for the 48 spiked congeners ranged from
about 86 to 114%. The calculated IPR ranges, while wider than the ranges for the same congeners in the
aqueous IPR samples, are generally reasonable for solid samples. Almost all the calculated OPR criteria
were between 25-160% recovery. The pooled MDLS values ranged from about 0.05 to 0.93 ng/g. None
of the MDLb values were higher than the pooled MDLS values; therefore, blank contamination was not a
significant issue.
For real-world samples, the sediments and biosolids analyses differed in the weight of sample used. For
sediments, the laboratories used 10 g, while 5 g was used for biosolids. Solids are well known for being
an overall challenging matrix, with many potential interfering organic components, as well as being a
difficult matrix to homogenize. Since PCBs are sorbed onto particles, it may not be possible to evenly
distribute more highly contaminated particles across the entire bulk sample, despite careful preparation.
Therefore, some aliquots of a given sample may have had more PCBs than other aliquots for the unspiked
and spiked sample analyses. Such differences affected the assumptions in the analyte recovery
calculations. Considering all the difficulties presented by the matrix, the method performed well overall
for the sediment and biosolid matrices that were analyzed.
Mean recoveries among the sediment matrices mostly fell between 30% and 200%, with one laboratory
reporting up to 79% of the congeners for two samples with negative values for either the matrix spike
(MS) or the matrix spike duplicate (MSD), suggesting interference or homogeneity issues with one but
Method 1628 Multi-Laboratory Validation Study Report
107
April 2021
-------
not both of the spiked samples. Less than 2% false detects were observed, with most of them coming
from two of the six laboratories.
Mean recoveries ranged between 70% to 185% for 54 of the 60 spiked congeners in real-world biosolid
samples. The other six congeners ranged between 40% and 230%, which may have been caused by
interferences that were not completely removed with the sample clean-ups utilized, or by a lack of sample
homogeneity. The false negative percentage was 4.8% and was only observed in two biosolids samples.
All the observed false negatives among the 928 results were from one laboratory that used GPC cleanup,
which was an optional cleanup, and an additional sample dilution for those two samples. While GPC is a
very robust clean-up procedure, it is recommended that the laboratories become familiar with the
procedure before implementing it for use in biosolids. Overall, laboratory performance was typical for
organic analytes in a challenging matrix like biosolid.
The Clean Water Act does not approve methods for either sediment or biosolids; therefore, EPA used
fewer laboratories and matrices than for wastewater. The performance criteria that will be listed in the
method will be noted as advisory.
Tissue
The reference matrix used for tissue was a 90:10 mixture of Ottawa sand and canola oil to mimic a lipid
level of 10% in tissue samples. The IPR and OPR calculated limits for most of the congeners were
unrealistically wide and did not resemble the range seen among the laboratories. A recovery limit of 25 -
150% for IPR and OPR was adopted for most of the congeners, giving a low failure rate among the data
collected. The pooled MDLS values ranged from about 0.035 to 0.23 ng/g. None of the MDLb values
were higher than the pooled MDLS values, so blank contamination was not a significant issue.
Among the four laboratories that analyzed three real-world fish tissue sample types, mean matrix spike
recoveries ranged from about 43% to 229%, with only 4 congeners with mean recoveries over 120%. The
high recoveries were observed only for two samples in one laboratory, suggesting a sample-specific issue
most likely due to lack of homogenization. The percentage of false negatives for tissue samples was less
than 0.1% of the 1392 total data points.
The Clean Water Act does not approve methods for fish tissue; therefore, EPA used fewer laboratories
and matrices than for wastewater. The performance criteria that will be listed in the method will be noted
as advisory.
Summary
EPA has demonstrated that Method 1628 is effective in all the matrices tested, and is far superior to the
currently approved EPA method for PCBs in wastewater, Method 608.3. This multi-laboratory validation
study also demonstrated that this method can be implemented at a typical full-service environmental
laboratory. Currently, EPA's only other PCB congener method uses high-resolution mass spectrometry
instrumentation, which many full-service laboratories do not own. This method provides access to PCB
congener analysis to any laboratory using a typical gas chromatograph/mass spectrometer (GC/MS)
instrument that is used for many other EPA methods.
Method 1628 Multi-Laboratory Validation Study Report
108
April 2021
-------
12. References
CSRA, 2018, PCB Congener Clean Water Act Method Single-Laboratory Validation Study Report,
January 22, 2018.
Erickson, Mitchell D. 1997. Analytical Chemistry ofPCBs, Second Edition, CRC Press LLC, Florida.
Frame, G. M., Cochran, J. W., and Boewadt, S. S. in the Journal of High Resolution Chromatography,
Vol. 19, pp 657-668 (1996).
IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, No. 107, Poly chlorinated
Biphenyls and Polybrominated Biphenyls, Lyon (FR): International Agency for Research on Cancer;
2016. https://www.ncbi.nlm.nih.gov/books/NBK361696
Satterthwaite, F. E. (1946), An Approximate Distribution of Estimates of Variance Components,
Biometrics Bulletin, 2: 110-114.
USEPA, 2018, Protocol for Review and Validation of New Methods for Regulated Organic and Inorganic
Analytes in Wastewater Under EPA's Alternate Test Procedure Program, February 2018.
Method 1628 Multi-Laboratory Validation Study Report
109
April 2021
-------
Appendix A
Labeled PCB Congeners to be used as Quantitation Standards
Method 1628 Multi-Laboratory Validation Study Report
A-l
April 2021
-------
Appendix A
Labeled PCB Congeners to be used as Quantitation Standards
March 2016 (Revised April 2021)
This document provides information about PCB congeners in order to help EPA select appropriate
13C-labeled PCB standards to be used for quantitation purposes in the low-resolution GC/MS PCB
method. These recommendations are based on information from a number of sources. Compiled below is
a table that lists all 209 congeners with information on:
Congener number
Level of chlorination (LOC)
Retention times of the DB-5 GC column specified in the AXYS SOP
40 congeners that were most prevalent in database queries
Risk Priority based on abundance in fish and human tissues and the availability of toxicological data
Labels used by AXYS at present
Labels available as individual standards from commercial vendors
Comments on appropriate choices
The retention time data are from the AXYS SOP, which uses a DB-5 column (primarily made of diphenyl
dimethyl polysiloxane), congeners are sorted by level of chlorination (LOC) and then retention time.
Co-eluting congeners are listed individually, but are listed with the same retention time in that column.
Please note that the first and last eluters for each LOC are also the first and last eluters for a DB-1 and
SPB-octyl column (the two other most common columns used for PCBs). The 13C-labeled PCB standards
have very similar retention times to their parent congeners, usually within a second or less. The first and
last eluting congeners were selected as isotope dilution standards (e.g., used for quantitation). If
13C-labeled PCB standards are present for both the first and last eluting congener of each LOC, the
analyst will know that the selected ion monitoring descriptors of the mass spectrometer are set
appropriately.
The "Top 40" congener list is a summary of the congeners that are most present in the environment, from
querying the following databases of PCB congener data 1) Wastewater data from the Delaware River
Basin Commission (DRBC) (2005 - 2013), 2) EPA National Lake Fish Tissue Survey data (2000 - 2004),
3) EPA National Sewage Sludge Survey (NSSS) data (2001), 4) Upper Trenton Chanel sediment data
from the Great Lakes National Program Office.
Priority risk congeners are from Geniece Lehmann (ORD), based upon congeners that are known to be
present in human blood and fish tissue, and whether risk data are available for these congeners; lower
numbers indicate higher priorities.
The challenge with any choices for labels is assigning each label to a native congener for quantitation
purposes. In traditional EPA full-scan GC/MS methods, the assignments of internal standards and target
analytes simply is based on retention times, with the internal standard always associated with target
analytes that elute at the same retention time or later, not with analytes that elute significantly before the
internal standard.
For these labels, associations are made by LOC, as well as retention time. Thus, using the label for the
last eluting congener means "reaching back" chromatographically to make some of those associations.
While that can be done, it adds some level of arbitrariness to the process. Once the selected standards are
run, EPA will need to determine whether native congeners are assigned to individual isotope dilution
standards with a relative response factor, or if the responses from multiple isotopes should be used and
averaged for the congeners in a given LOC.
Method 1628 Multi-Laboratory Validation Study Report
A-2
April 2021
-------
The comments section contains some information gained from researching supplier catalogs for the
labeled congeners and some recommendations reflect the apparent availability of individual labels versus
existing mixtures, and other relevant details.
The goal is to select the labeled analogs of the first and last eluter of each LOC, then select additional
labeled standards that are commercially available, spread out over the retention times for each LOC, and
represent congeners that are believed to be the most abundant in the environment. Too many labeled
standards will be overly burdensome for the laboratory community, so the number of labeled standards
will be limited to about 30. The labeled congeners believed to be most appropriate to act as isotope
dilution standards are in bold italics text in the table below.
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 1
1
13.63
The most abundant monochloro congener in Aroclor
mixtures (-0.5% of 1016 and 1242). Usually has poor
retention during analysis. Available as an individual C13
standard
PCB 2
1
14.91
PCB 3
1
15.04
Last eluter. Available as C13 standard
PCB 4
2
15.86
Y
5
It is the first eluter, and was the most abundant
congener (by mean % contribution) detected in WW
discharged to the Delaware River. One concern may
be that labeled PCB 4 will elute very close to both its
native and PCB 10. Available as C13 standard.
PCB 10
2
15.86
PCB 7
2
16.84
PCB 9
2
16.84
PCB 6
2
17.28
PCB 5
2
17.52
PCB 8
2
17.52
A good choice, since it's the most abundant dichloro
congener in Aroclor mixtures, but PCB 11 is more
abundant in the environmental databases reviewed, so
it was chosen instead.
PCB 14
2
18.12
PCB 11
2
18.98
Y
The most abundant congener in the NSSS, and 6th
most common congener detected in the Delaware River
WW data. Not a component of Aroclor mixtures, found
in some inks. Available as C13 standard.
PCB 12
2
19.25
PCB 13
2
19.25
PCB 15
2
19.50
Last eluter, also makes up about 2% of 1016 and 1242.
Available as C13 standard
PCB 19
3
18.40
Y
First eluting tri-CB. Available as C13 standard.
PCB 30
3
18.77
PCB 18
3
19.40
PCB 17
3
19.50
PCB 24
3
19.88
PCB 27
3
19.88
PCB 16
3
20.27
PCB 32
3
20.27
PCB 23
3
20.73
PCB 34
3
20.73
PCB 29
3
20.90
PCB 26
3
21.13
PCB 25
3
21.51
Method 1628 Multi-Laboratory Validation Study Report
April 2021
-------
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 31
3
21.58
Y
PCB28
3
22.03
Y
5
Available as an individual C13 standard.
PCB 20
3
22.40
Y
PCB 21
3
22.40
PCB 33
3
22.40
PCB 22
3
22.69
PCB 36
3
23.12
PCB 39
3
23.58
PCB 38
3
24.08
PCB 35
3
24.45
PCB 37
3
24.50
Last eluter, and available as a standard. Present at
-1% in Aroclor 1016, 1242, and 1248. Available as an
individual C13 standard
PCB 54
4
20.91
5
First eluter, toxicity data is available, and it is available
as C13 standard.
PCB 50
4
21.53
PCB 53
4
22.09
PCB 51
4
22.32
5
PCB 45
4
22.66
PCB 46
4
23.05
PCB 69
4
23.18
4
PCB 52
4
23.29
Y
3
One of the most commonly occurring congeners in the
environment. Present at ~4-7% in every Aroclor
mixture but 1260. Available as C13 standard.
PCB 73
4
23.29
PCB 43
4
23.49
PCB 49
4
23.49
4
PCB 47
4
23.65
Y
3
A good backup choice since it breaks up the tetras by
RT. Commonly occurring congener, but not as common
as PCB 52. Available as a C13 standard.
PCB 48
4
23.65
PCB 75
4
23.65
PCB 62
4
23.82
PCB 65
4
23.82
Y
4
PCB 44
4
24.30
Y
4
PCB 42
4
24.40
PCB 59
4
24.40
PCB 72
4
24.73
PCB 41
4
24.92
PCB 64
4
24.92
Y
PCB 68
4
24.92
PCB 71
4
24.92
PCB 40
4
25.31
PCB 57
4
25.39
PCB 67
4
25.62
PCB 58
4
25.79
PCB 63
4
25.91
PCB 61
4
26.11
Y
4
PCB 74
4
26.11
Y
4
Method 1628 Multi-Laboratory Validation Study Report
A-4
April 2021
-------
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 70
4
26.28
Y
4
Commonly detected in the environment, and a major
component (4-7%) of 1242, 1248, and 1254. Available
as a C13 standard
PCB 76
4
26.28
Y
4
A good choice, but cannot find a commercially available
C13 standard.
PCB 66
4
26.45
Y
4
A good choice, but cannot find a commercially available
C13 standard.
PCB 80
4
26.45
5
PCB 55
4
26.90
PCB 56
4
27.32
PCB 60
4
27.32
PCB 79
4
28.03
PCB 78
4
28.55
PCB 81
4
29.06
0.000
3
PCB 77
4
29.56
5
0.000
1
Last eluter and a WHO toxic. Available as a C13
standard.
PCB 104
5
24.08
5
The first eluter for the pentachloro congeners, and
available as a C13 standard.
PCB 96
5
25.14
PCB 103
5
25.34
PCB 100
5
25.62
4
Not present in the original Aroclor mixtures.
PCB 94
5
26.03
PCB 102
5
26.39
4
Barely present in the original Aroclor mixtures (<0.2%).
PCB 98
5
26.39
4
Not present in the original Aroclor mixtures.
PCB 93
5
26.51
4
Barely present in the original Aroclor mixtures (<0.1 %).
PCB 95
5
26.51
3
Present at a few percent in some of the Aroclor
mixtures, but very close in retention time to congener
104.
PCB 121
5
26.66
PCB 88
5
26.66
PCB 91
5
26.83
PCB 92
5
27.41
PCB 84
5
27.58
PCB 101
5
27.70
Y
3
Good choice for mid-RT penta-CBs. Detected regularly
in the environment, and one of the main components of
Aroclor 1248, 1254, and 1260). Available as a C13
standard.
PCB 89
5
27.70
PCB 90
5
27.70
Y
4
PCB 113
5
27.88
Y
4
PCB 99
5
27.97
Y
2
A good choice, but elutes close to 101, and is not
usually detected in as high quantities as 101. Also, it's
not available as C13 standard.
PCB 119
5
28.30
4
PCB 112
5
28.41
PCB 83
5
28.52
Y
PCB 108
5
28.54
4
PCB 86
5
28.80
4
PCB 97
5
28.80
4
PCB 125
5
28.92
4
PCB 111
5
29.00
Method 1628 Multi-Laboratory Validation Study Report
April 2021
-------
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 117
5
29.00
2
PCB 115
5
29.08
Y
PCB 116
5
29.08
2
PCB 87
5
29.08
4
PCB 120
5
29.28
PCB 85
5
29.28
2
Present at 1-2% in Aroclor 1248 and 1254, and known
to be present in fish and human tissue. Available as a
C13 standard.
PCB 110
5
29.58
Y
3
110 is an ideal candidate for an isotopically labeled
standard. It is a major component of Aroclor 1248,
1254, and 1260, and is present regularly in the
environment. Unfortunately, a commercial C13
standard is not currently available.
PCB 82
5
30.14
PCB 124
5
30.49
PCB 107
5
30.67
PCB 109
5
30.67
4
PCB 123
5
30.80
0.000
03
PCB 106
5
30.92
PCB 118
5
30.92
Y
1
0.000
03
It's a WHO toxic, and a major component of Aroclor
1248 and 1254. It is also one of the most common
congeners in the environment, fish tissue, and human
tissue. Available as a C13 standard.
PCB 114
5
31.50
5
0.000
03
PCB 122
5
31.63
PCB 105
5
32.30
Y
3
0.000
03
A good choice, but elutes close to 118, and is not
usually detected in as high quantities as 118.
PCB 127
5
32.30
PCB 126
5
33.99
5
0.1
The last eluter, and the most toxic congener according
to WHO. It is rarely detected in the environment.
Available as a C13 standard.
PCB 155
6
27.23
The first eluter for the hexa congeners. Available as a
C13 standard.
PCB 150
6
28.33
PCB 152
6
28.71
PCB 145
6
29.04
PCB 148
6
29.26
PCB 136
6
29.41
PCB 154
6
29.64
4
PCB 151
6
30.22
4
PCB 135
6
30.46
4
PCB 144
6
30.46
PCB 147
6
30.64
Y
PCB 139
6
30.84
PCB 149
6
30.84
Y
3
A good candidate, commonly detected in the
environment, and is at a good retention time to
distribute the hexa standards. Currently not available
as a C13 standard.
PCB 140
6
30.99
PCB 134
6
31.31
PCB 143
6
31.31
Method 1628 Multi-Laboratory Validation Study Report
A-6
April 2021
-------
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 133
6
31.50
PCB 131
6
31.63
PCB 142
6
31.63
PCB 165
6
31.72
PCB 146
6
31.82
Y
2
A good choice, but too close to the retention time of
153, which is usually detected in environmental
standards at higher concentrations. Currently not
available as a C13 standard.
PCB 161
6
31.91
PCB 153
6
32.12
Y
1
One of the most abundant congeners in the
environment and in the Aroclor mixtures. Available as a
C13 standard.
PCB 132
6
32.25
Y
3
All good choices, but too close to the retention time of
153, which is usually detected in environmental
standards at higher concentrations.
PCB 168
6
32.25
Y
2
PCB 141
6
32.77
Y
4
PCB 137
6
33.12
PCB 130
6
33.26
PCB 138
6
33.52
Y
1
One of the more abundant congeners in the
environment and Aroclor 1254 and 1260. Present in
human and fish tissue and toxicity data is available.
Available as a C13 standard.
PCB 163
6
33.52
Y
2
PCB 164
6
33.52
4
PCB 158
6
33.66
PCB 160
6
33.66
Y
2
PCB 129
6
33.95
Y
2
PCB 166
6
34.27
4
PCB 159
6
34.44
PCB 162
6
34.71
PCB 128
6
34.93
3
PCB 167
6
35.01
0.000
03
PCB 156
6
36.17
6
0.000
03
It's a WHO toxic congener, and available as a C13
standard. It was not selected since the RT is so close
to 169.
PCB 157
6
36.46
0.000
03
PCB 169
6
37.91
5
0.03
The last eluter, and the second most toxic congener
according to the WHO. It is rarely detected in the
environment. Available as a C13 standard.
PCB 188
7
31.77
First eluter and available as an individual standard.
PCB 184
7
32.12
PCB 179
7
32.85
PCB 176
7
33.22
PCB 186
7
33.67
PCB 178
7
34.00
Available as a C13 standard.
PCB 175
7
34.29
PCB 182
7
34.42
PCB 187
7
34.42
Y
2
A good choice, but not available as a C13 standard.
PCB 183
7
34.68
4
Not available as a C13 standard.
PCB 185
7
35.23
PCB 174
7
35.68
Y
4
Not available as a C13 standard.
Method 1628 Multi-Laboratory Validation Study Report
A-7
April 2021
-------
Congener
LOC
RT
(DB-5)
Top
40?
Risk
Priority
WHO
TEF
Comments
PCB 181
7
35.68
PCB 177
7
35.89
Y
2
Not available as a C13 standard.
PCB 171
7
36.12
PCB 173
7
36.42
PCB 172
7
36.75
PCB 192
7
36.75
PCB 180
7
37.06
Y
1
The most abundant congener in the Aroclor 1260
mixture, and regularly detected in the environment.
Available as a C13 standard.
PCB 193
7
37.19
Y
PCB 191
7
37.39
PCB 170
7
38.22
Y
1
A good choice, and available as a standard, but too
close to the retention time of 180, which is usually
detected in environmental standards at higher
concentrations.
PCB 190
7
38.22
PCB 189
7
38.98
5
0.000
03
Last eluter, a WHO toxic congener, and available as a
C13 standard.
PCB 202
8
36.08
First eluter, and available as a C13 standard.
PCB 201
8
36.48
PCB 204
8
36.58
PCB 197
8
36.86
PCB 200
8
37.61
PCB 198
8
38.35
4
Not available as a C13 standard.
PCB 199
8
38.45
4
Not available as a C13 standard.
PCB 196
8
38.60
Not available as a C13 standard.
PCB 203
8
38.60
2
Not available as a C13 standard.
PCB 195
8
39.37
PCB 194
8
39.90
2
Good choice and available as a C13 standard, but too
close in retention time to 205.
PCB 205
8
40.03
Last eluter, and available as an individual C13
standard.
PCB 208
9
39.35
First eluter, and available as a C13 standard.
PCB 207
9
39.56
Least abundant nona-congener in Aroclor mixtures.
PCB 206
9
40.82
2
Last eluter, and available as a C13 standard. Most
abundant nona-congener in Aroclor mixtures.
PCB 209
10
41.51
Only choice. Not abundant in Aroclor mixtures, but has
been detected in the environment. May be present
from non-Aroclor sources. Available as a C13
standard.
Non-extracted Internal Standards (NIS)
PCB 79
4
28.03
Available as a C13 standard.
PCB 162
6
34.71
Available as an individual C13 standard
Non-extracted internal standards are not being used for quantification of the target analytes, but to determine the
recovery of the other labeled standards added to each sample prior to extraction. As such, they are sometimes
called "recovery standards" in other methods. The C13 congeners that were selected are not commonly present in
the environment, nor are they abundant in Aroclor mixtures.
Method 1628 Multi-Laboratory Validation Study Report
A-8
April 2021
-------
Summary of Isotope Dilution Standards Selected by EPA
Level of Chlorination
Labeled Congeners
1
1, 3
2
4, 11, 15
3
19, 28, 37
4
54, 52, 70, 77
5
104, 101, 85, 118, 126
6
155, 153, 138, 169
7
188, 180, 189
8
202, 205
9
208, 206
10
209
Priority Congeners by Risk (1=highest)
Priority
Native Congeners
Rationale
1
118, 138, 153, 170, 180
Abundant in fish and human tissues,
and tox data available
2
85, 99, 116, 117, 129, 146, 160, 163, 168, 177, 187, 194,
203, 206
Abundant in fish and human tissues
3
47, 52, 95, 101, 105, 110, 128, 132, 149
Abundant in fish, and tox data available
4
44, 49, 61, 65, 66, 69, 70, 74, 76, 86, 87, 90, 93, 97, 98,
100, 102, 108, 109, 113, 119, 125, 135, 141, 151, 154, 166,
174, 183, 198, 199
Abundant in fish
5
4, 28, 51, 54, 77, 80, 104, 114, 126, 169, 189
Toxicity data available
6
156
Abundant in human tissues and tox
data available
Method 1628 Multi-Laboratory Validation Study Report
A-9
April 2021
-------
Appendix B
Interim Quality Control Acceptance Criteria
Arising from the Method Validation Study
Method 1628 Multi-Laboratory Validation Study Report
B-l
April 2021
-------
Interim Quality Control Acceptance Criteria Arising from the Method Validation Study
The tables below present the interim QC acceptance criteria that EPA anticipates including in the draft
method. The derivations of these criteria are described in the body of this report.
Aqueous Matrix IPR and OPR QC Acceptance Criteria for Target Analytes
Congener
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
PCB-1
78-130
18
71 -136
PCB-3
74-117
14
71 -120
PCB-4
77-112
14
72-117
PCB-8
42-120
18
43-119
PCB-11
62-125
9
63-124
PCB-15
70-111
10
69-111
PCB-18
60-107
17
57-111
PCB-19
77-107
12
73-111
PCB-28
18-184
17
21 -180
PCB-31
46-129
19
46-129
PCB-37
67-112
9
68-111
PCB-44
44-131
13
46-130
PCB-52
61-128
8
62-127
PCB-54
67-112
8
68-111
PCB-64
74-108
10
73-110
PCB-66
64-118
8
65-117
PCB-70
55-127
8
57-126
PCB-74
74-102
8
73-103
PCB-77
58-118
9
59-116
PCB-85
68-106
7
69-105
PCB-95
63-117
12
63-117
PCB-99
66-107
10
66-107
PCB-101
64-118
9
65-117
PCB-104
64-117
8
65-116
PCB-105
64-120
10
65-119
PCB-118
61-119
10
62-118
PCB-110
63-106
12
62-107
PCB-126
63-113
7
64-112
PCB-132
51-133
11
53-131
PCB-138
61-117
11
61-116
PCB-147
61-117
12
62-117
PCB-149
57-120
11
58-119
PCB-153
46-134
16
48-132
PCB-155
64-116
10
65-115
PCB-156
46-149
23
45-150
PCB-166
34-157
9
36-156
PCB-169
50-122
10
52-121
PCB-177
47-130
10
49-128
PCB-180
52-124
11
53-123
PCB-187
36-138
17
38-136
PCB-188
57-122
11
58-121
PCB-189
56-119
11
58-118
PCB-199
42-164
57
14-193
PCB-202
55-121
12
56-120
PCB-205
52-118
18
51-119
PCB-206
35-135
17
37-133
PCB-208
44-125
15
45-124
PCB-209
31-130
27
30-131
Method 1628 Multi-Laboratory Validation Study Report
B-2
April 2021
-------
Aqueous Matrix IPR and OPR QC Acceptance Criteria for Labeled Compounds
Congener
Interim Acceptance Criteria (%)
IPR (each
aliquot)
Max RSD
OPR
13Ci2-PCB-1
15 -130
40
15-130
13Ci2-PCB-3
15 -130
40
15-130
13Ci2-PCB-4
15 -130
40
15-130
13Ci2-PCB-11
15 -130
40
15-130
13Ci2-PCB-15
15 -130
40
15-130
13Ci2-PCB-19
15 -130
40
15-130
13Ci2-PCB-28
15 -130
40
15-130
13Ci2-PCB-37
15 -130
40
15-130
13Ci2-PCB-52
15 -130
40
15-130
13Ci2-PCB-54
15 -130
40
15-130
13Ci2-PCB-70
15 -130
40
15-130
13Ci2-PCB-77
15 -130
40
15-130
13Ci2-PCB-85
15 -130
40
15-130
13Ci2-PCB-101
15 -130
40
15-130
13Ci2-PCB-104
15 -130
40
15-130
13Ci2-PCB-118
15 -130
40
15-130
13Ci2-PCB-126
15 -130
40
15-130
13Ci2-PCB-138
15 -130
40
15-130
13Ci2-PCB-153
15 -130
40
15-130
13Ci2-PCB-155
15 -130
40
15-130
13Ci2-PCB-169
15 -130
40
15-130
13Ci2-PCB-180
15 -130
40
15-130
13Ci2-PCB-188
15 -130
40
15-130
13Ci2-PCB-189
15 -130
40
15-130
13Ci2-PCB-202
15 -130
40
15-130
13Ci2-PCB-205
15 -130
40
15-130
13Ci2-PCB-206
15 -130
40
15-130
13Ci2-PCB-208
15 -130
40
15-130
13Ci2-PCB-209
15 -130
40
15-130
Solid Matrix IPR and OPR QC Acceptance Criteria for Target Analytes
Congener
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
PCB-1
61 -154
59
19-196
PCB-3
40-156
42
29-167
PCB-4
48-144
61
13-179
PCB-8
35-150
40
25-160
PCB-11
35-150
40
25-160
PCB-15
36-150
44
25-162
PCB-18
20-148
40
18-149
PCB-19
26-157
32
28-156
PCB-28
25-150
35
30-150
PCB-31
38-147
37
32-153
PCB-37
38-147
38
31 -155
PCB-44
23-153
34
24-151
PCB-52
57-138
42
36-159
PCB-54
56-132
56
21 -167
PCB-64
29-153
29
31 -150
PCB-66
50-138
25
49-140
PCB-70
43-144
27
42-144
Method 1628 Multi-Laboratory Validation Study Report
B-3
April 2021
-------
Solid Matrix IPR and OPR QC Acceptance Criteria for Target Analytes
Congener
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR
PCB-74
41 -135
30
38-138
PCB-77
42-134
40
30-145
PCB-85
57-121
27
50-128
PCB-95
55-125
29
47-133
PCB-99
33-140
34
30-143
PCB-101
57-125
26
51 -132
PCB-104
52-128
44
32-148
PCB-105
65-122
17
63-124
PCB-118
48-133
19
50-131
PCB-110
31 -142
20
36-137
PCB-126
48-129
14
52-124
PCB-132
42-146
18
47-141
PCB-138
60-123
19
58-125
PCB-147
58-126
25
53-132
PCB-149
51 -129
28
46-134
PCB-153
76-109
25
61 -124
PCB-155
60-122
37
41 -140
PCB-156
76-119
25
62-133
PCB-166
71 -122
21
64-128
PCB-169
56-130
55
23-164
PCB-177
71-114
29
55-130
PCB-180
72-112
25
58-125
PCB-187
64-114
23
56-122
PCB-188
61-118
27
52-128
PCB-189
67-117
24
58-126
PCB-199
62-126
22
58-130
PCB-202
51 -127
24
49-129
PCB-205
54-116
31
44-126
PCB-206
52-129
49
27-154
PCB-208
45-131
21
47-129
PCB-209
67-111
19
62-117
Solid Matrix IPR and OPR QC Acceptance Criteria for Labeled Congeners
Congener
Interim Acceptance Criteria (%)
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-1
15 -130
60
15 -130
13Ci2-PCB-3
15 -130
60
15 -130
13Ci2-PCB-4
15 -130
60
15 -130
13Ci2-PCB-11
15 -130
60
15 -130
13Ci2-PCB-15
15 -130
60
15 -130
13Ci2-PCB-19
15 -130
60
15 -130
13Ci2-PCB-28
15 -130
60
15 -130
13Ci2-PCB-37
15 -130
60
15 -130
13Ci2-PCB-52
15 -130
60
15 -130
13Ci2-PCB-54
15 -130
60
15 -130
13Ci2-PCB-70
15 -130
60
15 -130
13Ci2-PCB-77
15 -130
60
15 -130
13Ci2-PCB-85
15 -130
60
15 -130
13Ci2-PCB-101
15 -130
60
15 -130
13Ci2-PCB-104
15 -130
60
15 -130
13Ci2-PCB-118
15 -130
60
15 -130
Method 1628 Multi-Laboratory Validation Study Report
B-4
April 2021
-------
Solid Matrix IPR and OPR QC Acceptance Criteria for Labeled Congeners
Congener
Interim Acceptance Criteria (%)
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-126
15 -130
60
15 -130
13Ci2-PCB-138
15 -130
60
15 -130
13Ci2-PCB-153
15 -130
60
15 -130
13Ci2-PCB-155
15 -130
60
15 -130
13Ci2-PCB-169
15 -130
60
15 -130
13Ci2-PCB-180
15 -130
60
15 -130
13Ci2-PCB-188
15 -130
60
15 -130
13Ci2-PCB-189
15 -130
60
15 -130
13Ci2-PCB-202
15 -130
60
15 -130
13Ci2-PCB-205
15 -130
60
15 -130
13Ci2-PCB-206
15 -130
60
15 -130
13Ci2-PCB-208
15 -130
60
15 -130
13Ci2-PCB-209
15 -130
60
15 -130
Tissue Matrix IPR and OPR QC Acceptance Criteria for Target Analytes
Congener
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR Range
PCB-1
25-150
25
25-150
PCB-3
25-150
25
25-150
PCB-4
25-150
25
25-150
PCB-8
25-150
25
25-150
PCB-11
25-150
25
25-150
PCB-15
25-150
25
25-150
PCB-18
25-150
25
25-150
PCB-19
25-150
25
25-150
PCB-28
25-150
25
25-150
PCB-31
25-150
25
25-150
PCB-37
25-150
25
25-150
PCB-44
25-150
25
25-150
PCB-52
25-150
25
25-150
PCB-54
25-150
25
25-150
PCB-64
25-150
25
25-150
PCB-66
25-150
25
25-150
PCB-70
25-150
25
25-150
PCB-74
25-150
25
25-150
PCB-77
25-150
25
25-150
PCB-85
25-150
25
25-150
PCB-95
25-150
25
25-150
PCB-99
25-150
25
25-150
PCB-101
25-150
25
25-150
PCB-104
25-150
25
25-150
PCB-105
25-150
25
25-150
PCB-118
25-150
25
25-150
PCB-110
25-150
25
38-138
PCB-126
25-150
25
38-139
PCB-132
32-156
25
48-140
PCB-138
27-150
25
43-134
PCB-147
25-150
25
25-150
PCB-149
25-150
25
25-150
PCB-153
33-142
25
47-127
PCB-155
25-150
25
25-150
Method 1628 Multi-Laboratory Validation Study Report
B-5
April 2021
-------
Tissue Matrix IPR and OPR QC Acceptance Criteria for Target Analytes
Congener
Interim Acceptance Criteria (%)
IPR Mean
Max RSD
OPR Range
PCB-156
25-150
25
25-150
PCB-166
25-150
25
25-150
PCB-169
25-150
25
25-150
PCB-177
25-150
25
25-150
PCB-180
42-137
25
52-127
PCB-187
39-137
25
49-126
PCB-188
25-150
25
25-150
PCB-189
25-150
25
25-150
PCB-199
34-153
25
45-142
PCB-202
33-145
25
47-131
PCB-205
25-150
25
37-144
PCB-206
25-150
25
25-150
PCB-208
25-150
25
25-150
PCB-209
25-150
25
25-150
Tissue Matrix IPR and OPR QC Acceptance Criteria for Labeled Compounds
Congener
Interim Acceptance Criteria (%)
IPR (each aliquot)
Max RSD
OPR
13Ci2-PCB-1
15 -130
60
15 -130
13Ci2-PCB-3
15 -130
60
15 -130
13Ci2-PCB-4
15 -130
60
15 -130
13Ci2-PCB-11
15 -130
60
15 -130
13Ci2-PCB-15
15 -130
60
15 -130
13Ci2-PCB-19
15 -130
60
15 -130
13Ci2-PCB-28
15 -130
60
15 -130
13Ci2-PCB-37
15 -130
60
15 -130
13Ci2-PCB-52
15 -130
60
15 -130
13Ci2-PCB-54
15 -130
60
15 -130
13Ci2-PCB-70
15 -130
60
15 -130
13Ci2-PCB-77
15 -130
60
15 -130
13Ci2-PCB-85
15 -130
60
15 -130
13Ci2-PCB-101
15 -130
60
15 -130
13Ci2-PCB-104
15 -130
60
15 -130
13Ci2-PCB-118
15 -130
60
15 -130
13Ci2-PCB-126
15 -130
60
15 -130
13Ci2-PCB-138
15 -130
60
15 -130
13Ci2-PCB-153
15 -130
60
15 -130
13Ci2-PCB-155
15 -130
60
15 -130
13Ci2-PCB-169
15 -130
60
15 -130
13Ci2-PCB-180
15 -130
60
15 -130
13Ci2-PCB-188
15 -130
60
15 -130
13Ci2-PCB-189
15 -130
60
15 -130
13Ci2-PCB-202
15 -130
60
15 -130
13Ci2-PCB-205
15 -130
60
15 -130
13Ci2-PCB-206
15 -130
60
15 -130
13Ci2-PCB-208
15 -130
60
15 -130
13Ci2-PCB-209
15 -130
60
15 -130
Method 1628 Multi-Laboratory Validation Study Report
B-6
April 2021
-------
QC Acceptance Criteria for Recovery of Labeled
Compounds in Samples
Congener
Interim QC Acceptance Criteria (%)
13Ci2-PCB-1
15-130
13Ci2-PCB-3
15-130
13Ci2-PCB-4
15-130
13Ci2-PCB-11
15-130
13Ci2-PCB-15
15-130
13Ci2-PCB-19
15-130
13Ci2-PCB-28
15-130
13Ci2-PCB-37
15-130
13Ci2-PCB-52
15-130
13Ci2-PCB-54
15-130
13Ci2-PCB-70
15-130
13Ci2-PCB-77
15-130
13Ci2-PCB-85
15-130
13Ci2-PCB-101
15-130
13Ci2-PCB-104
15-130
13Ci2-PCB-118
15-130
13Ci2-PCB-126
15-130
13Ci2-PCB-138
15-130
13Ci2-PCB-153
15-130
13Ci2-PCB-155
15-130
13Ci2-PCB-169
15-130
13Ci2-PCB-180
15-130
13Ci2-PCB-188
15-130
13Ci2-PCB-189
15-130
13Ci2-PCB-202
15-130
13Ci2-PCB-205
15-130
13Ci2-PCB-206
15-130
13Ci2-PCB-208
15-130
13Ci2-PCB-209
15-130
Practical Application of These QC Acceptance Criteria
One of the challenges in developing statistically based QC acceptance criteria is that the limits are
sometimes seen by users as overly wide, and thus not providing some preconceived level of "control"
over laboratory performance. However, those concerns should be weighed against the risk and cost of
rejecting results from samples and QC operations such as the IPR and OPR based on random variability.
Every laboratory performing analyses in support of Clean Water Act compliance monitoring must have an
effective quality management system in place. Such systems must include assessment of all results
against the various QC acceptance limits in a given analytical method, but also should include procedures
for longer-term internal evaluations of laboratory performance. In fact, most EPA methods include
discussions of the use of control charts and the development of in-house performance criteria. EPA
expects that responsible laboratories will perform such evaluations and develop and apply in-house
criteria, which by virtue of being from a single laboratory, will be narrower than the acceptance criteria
listed in this appendix and incorporated in the final PCB method.
Method 1628 Multi-Laboratory Validation Study Report
B-7
April 2021
-------
Appendix C
Study Plan for Multi-laboratory Validation of the
EAD PCB Congener Method
This appendix contains the study plan developed by EPA for the multi-laboratory method validation
study. The pagination, footers, and formatting remain the same as in the original study plan.
-------
Study Plan for Multi-laboratory Validation of the EAD PCB Congener Method
Prepared for
Adrian Hanley
U.S. Environmental Protection Agency
Office of Water
Office of Science and Technology
Engineering and Analysis Division (4303T)
1200 Pennsylvania Avenue, NW
Washington DC 20460
Prepared by
CSRA LLC
6361 Walker Lane
Suite 300
Alexandria, Virginia 22310
Prepared under
EPA Scientific and Technical Support Contract EP-C-17-024
January 17,2018
Approvals
Brian D" \rnico, EPA Branch Chief
Date
Adriiin Hanley. FPA Piqjcct Manager
Date
2/15/18
MiSon Kelly. EPA Q.\ Coordinator
Date
J-L
I \nť WjIicix ( SRA Program
Date
Dak-
l'18'ZOli
-------
AC KN OW LEDGMEN TS
This study plan was prepared under the direction of Adrian Hanley of the Engineering and Analysis
Division (EAD) within the U.S. Environmental Protection Agency (EPA) Office of Water. It was
prepared by CSRA LLC under EPA Contract No. EP-C-17-024.
Disclaimer
This document has been reviewed and approved by the Engineering and Analytical Support Branch of
EAD. Mention of company names, trade names, or commercial products does not constitute endorsement
or recommendation for use.
PCB Congener Multi-lab Validation Study Plan
1
January 2018
-------
Table of Contents
Page
Acknowledgments i
Disclaimer i
Distribution List iii
Section 1. Background 1
Section 2. Study Objectives 2
Section 3. Study Management 3
Section 4. Technical Approach 4
4.1 Phase 1 - Soliciting Laboratories 4
4.2 Phase 2 - Procuring Standards and Study Samples 5
4.3 Phase 3 - Calibration and Demonstration of Capability 6
4.4 Phase 4 - Analyses of Study Samples 8
4.5 Phase 5 - Data Validation 9
4.6 Phase 6 - Development of QC Acceptance Criteria 10
Section 5. Quality Control Procedures 10
Section 6. Data Reporting 10
Section 7. Evaluation of Method Performance 12
Section 8. References 12
Appendix A Standards Required for the Validation Study
Appendix B Sample Types for the Validation Study
Appendix C Procedures for Derivation of QC Acceptance Criteria from the Validation Study Results
Appendix D Draft GC-LRMS-SIM Method
PCB Congener Multi-lab Validation Study Plan
li
January 2018
-------
Distribution List
Brian D'Amico, EPA EAD
Adrian Hanley, EPA EAD
Marion Kelly, EPA EAD
Lynn Walters, CSRA
Harry McCarty, CSRA
Marguerite Jones, CSRA
Eric Boring, CSRA
Participant laboratories (TBD)
PCB Congener Multi-lab Validation Study Plan
m
January 2018
-------
Study Plan for Multi-laboratory Validation of the EAD PCB Congener Method
Section 1. Background
From the 1930s into the early 1980s, PCBs were manufactured under several trade names, most
predominantly "Aroclor" in the U.S. The Aroclor name was accompanied by a four-digit number
indicating the degree of chlorination of the commercial mixture (e.g., Aroclor 1016, Aroclor 1260, etc.).
In general, the higher the number, the higher the degree of chlorination. From the late 1950s through the
1970s, PCBs were determined as Aroclors by low resolution (packed column) gas chromatography (GC)
with an electron capture detector (ECD). During this time period, EPA developed Method 608 for
Aroclor determination in wastewater (Reference 8.1). The method detects seven of the most common
Aroclor mixtures by comparing a peak pattern created by the congeners that make up each mixture.
Method 608 does not target every known Aroclor mixture.
Aroclors are not the only source of PCB contamination, and the composition of Aroclors and other PCB
sources within the environment changes over time due to variations in the stability, solubility, volatility,
and other properties of the individual congeners. This weathering of PCB contamination often causes
false nondetects and poor accuracy when measured as an Aroclor mixture. In the late 1970s and early
1980s, heightened interest in PCBs and ambiguities in PCB identification led several researchers to
separate and identify all 209 PCB congeners using high resolution (open tubular capillary) GC columns
coupled with low resolution mass spectrometry (LRMS). Detecting the individual congeners instead of
the peak pattern of Aroclor mixtures provides a more accurate result of how much PCB contamination is
in the environmental sample. By the 1990's PCB analysis research became focused on the development
of high resolution GC (HRGC) high resolution MS (HRMS) methods.
In 1995, EPA developed Method 1668, which uses HRGC combined with HRMS for determination of 13
dioxin-like PCBs that the World Health Organization (WHO) designated as "toxic" in 1994. Method
1668 was based on data from studies conducted at Pacific Analytical, Inc., Carlsbad, California). In 1997,
interest in additional congeners led EPA to investigate determination of as many congeners as possible in
a single HRGC/HRMS run. This led to draft Revision A of EPA Method 1668. EPA subsequently drafted
Revisions B and C of Method 1668, covering all 209 PCB congeners (Reference 8.2).
While Method 1668 is considered to be a highly sensitive and accurate method for the determination of
individual PCB congeners, it is not without its limitations. HRGC/HRMS instrumentation is expensive
and few environmental laboratories possess the capabilities to perform these analyses. In addition,
sample analyses by Method 1668 are rather costly to perform and require highly technical personnel to
interpret the data. The fact that Method 1668 is a highly sensitive method can also be problematic
because PCBs are routinely detected below ambient background levels. For all these reasons, an EPA
GC-LRMS method that quantifies individual congeners and is more sensitive than the GC-ECD
procedure in Method 608, but more accessible to the laboratory community than Method 1668, would be
highly useful for the determination of PCB congeners in environmental samples.
In 2016 and early 2017, the Engineering and Analysis Division (EAD) within the Office of Science and
Technology, Office of Water conducted a single-laboratory study of a proprietary method from AXYS
Analytical Services, Ltd (Reference 8.3). That single-laboratory study was successful in demonstrating
the applicability of the GC-LRMS SIM method to monitoring PCB congeners in wastewater,
soil/sediment, biosolids, and fish tissue samples relative to the needs of the Clean Water Act.
The next step is to conduct a multi-laboratory validation study of the draft EPA method that resulted from
the single-laboratory study. This study plan presents EAD's approach to that study.
PCB Congener Multi-lab Validation Study Plan
1
January 2018
-------
Section 2. Study Objectives
The goals of the multi-laboratory validation study are to:
Obtain data from matrices that are representative of the method's intended use
Obtain data from laboratories that are representative of those likely to use the approved method,
but that were not directly involved in its development
Obtain feedback from laboratory users on the specifics of the draft method (e.g., is it clear and
easy to understand, or are changes to the method text needed?)
Use study data to characterize performance of the method
Develop statistically derived QC acceptance criteria that will reflect method performance
capabilities in real-world situations
In addition to the overall objective described above, EAD has two general quality objectives for this
study:
1. Except where otherwise directed, all samples and data must be generated according to the
analytical and quality assurance/quality control (QA/QC) procedures specified in this study plan
and the GC-LRMS-SIM procedure. Alternatively, the data must be the result of pre-approved
and documented changes to these procedures. This will allow EAD to collect data that accurately
reflects the performance capabilities of the methodology and to use study results to identify the
need for any further revisions to the procedure.
2. All data produced must be capable of being verified by an independent person reviewing the
analytical data package.
To meet these quality objectives, EPA and CSRA will employ the following QA/QC strategies:
All CSRA activities will be performed in accordance with this study plan, which also serves as a
Quality Assurance Project Plan (QAPP) for the study.
The vendor responsible for preparing the study samples must have demonstrated experience in
performing work of a similar nature and must have a comprehensive QA program in place and
operating throughout their study operations.
The vendor responsible for preparing the analytical standards must have demonstrated experience
in performing work of a similar nature and must have a comprehensive QA program in place and
operating throughout their study operations.
Each participant laboratory (either contracted or volunteer) also must have demonstrated
experience in GC-LRMS-SIM analyses of a similar nature and must have a comprehensive QA
program in place and operating throughout their study operations.
The study report and the final draft method will be reviewed by CSRA, EPA, and the EPA work
group to ensure the QC requirements meet data quality objectives.
Cumulatively, these requirements are intended to ensure that the data produced in this study are of
appropriate and documented quality.
PCB Congener Multi-lab Validation Study Plan
2
January 2018
-------
Section 3. Study Management
The study will be managed by Adrian Hanley, the EAD Project Manager. Day-to-day management and
coordination of study activities will be performed by CSRA Study Manager, Harry McCarty, under the
supervision of the CSRA Program Manager, Lynn Walters and EAD oversight. Marion Kelly, the
Quality Assurance Coordinator for EAD, will provide QA support for EPA. Marguerite Jones, the Quality
Assurance Officer for CSRA, will provide QA oversight for CSRA. QA oversight of participant
laboratory activities will be provided by each laboratory's QA Officer (or equivalent). Each of these QA
positions is independent of the technical staff and managers who are responsible for the generation,
analysis, and use of data in this study. The organization chart below illustrates the relationship of these
parties.
Line of Authority
Line of Communication
CSRA
Technical Staff
Marion Kelly
EAD QA Coordinator
Marguerite Jones
CSRA QA Officer
Participant Laboratories
QA Officers
Participant Laboratories
Project Managers
Lynn Walters
CSRA Program Manager
Harry McCarty
CSRA Project Leader
Adrian Hanley
EAD Project Manager
Commercial Vendors of
Standards and PT Samples
Engineering and Analysis Division
Brian D'Amico, EPA Branch Chief
A commercial vendor of performance testing (PT) samples will be used to prepare real-world and
synthetic wastewater, soil/sediment, and tissue matrices for use in this study, with assistance from EAD
and CSRA as needed. CSRA will be responsible for procuring and providing oversight of the vendor.
EPA and CSRA will work together to obtain sufficient volumes of real-world wastewater, soil/sediment,
and tissue matrices from various sources and delivering them to the selected PT vendor for
homogenization, aliquoting, and shipment to the participant laboratories.
CSRA will be responsible for procuring the various PCB standards required to perform the method for
each laboratory from one or more commercial vendors. A list of the required standards is provided in
Appendix A.
PCB Congener Multi-lab Validation Study Plan
3
January 2018
-------
CSRA will be responsible for procuring and providing oversight of commercial contract laboratories that
will participate in the validation study. The number of contracted laboratories will be determined by EPA
based on factors such as cost and EPA's ability to enlist other suitable laboratories as volunteer (unpaid)
participants. In keeping with the approach described in ASTM Standard D2777 (Reference 8.4), EPA
will solicit participation from a large number of laboratories, recognizing the possibility that some
participants may drop out or otherwise fail to provide usable data. At this time, EPA is planning to
include 20 laboratories in the study if possible, far in excess of the nine laboratories recommended in the
ASTM Standard. Part of the rationale for the large number of participants is to gain additional support for
promulgation of the final method from the commercial laboratory community.
In order to comply with recent EPA policies regarding laboratory competency, CSRA will request and
evaluate information about each laboratory's certifications or accreditations relevant to the analysis of
PCB congeners in environmental matrices during the solicitation process described in Section 4.1. CSRA
recognizes that certifications or accreditations specific to analysis of individual PCB congeners may not
be offered by all accrediting bodies, and CSRA will not use the lack of certifications or accreditations to
exclude laboratories. CSRA will request similar information from any volunteer laboratories identified
by EPA. Each laboratory supporting this study must have a comprehensive QA program in place and
operating at all times during the study. CSRA will request copies of QA program documentation during
the solicitation process, as part of an assessment of laboratory capabilities. Laboratories that cannot
demonstrate competency in PCB analyses and that do not have an adequate QA program in place will not
be included as participants in the study.
All analytical results will be submitted to CSRA. As described in Sections 4.5, 4.6 and 7, CSRA will
review and evaluate all analytical data and assist EAD in drawing conclusions from the results.
Depending on the availability of resources, CSRA will either prepare a draft study report that summarizes
these results and conclusions for EAD review, or will provide data and technical assistance to aid EAD
staff in preparing such a report. As appropriate, EAD will revise the draft GC-LRMS-SIM method to
reflect study findings and add QC acceptance criteria developed from the study data.
Section 4. Technical Approach
The study will be performed in six phases. Phase 1 and 2 may occur simultaneously.
Phase 1 (Section 4.1) involves soliciting laboratories (contracted and volunteer) to participate in
the study
Phase 2 (Section 4.2) involves procuring the standards required by each laboratory, as well as the
study samples to be analyzed
Phase 3 (Section 4.3) involves the initial steps (calibration, IDCs, and MDLs) demonstrating
laboratory capability with standards and clean matrices using the draft method
Phase 4 (Section 4.3) involves using the draft method to analyze the study samples by all of the
participant laboratories
Phase 5 (Section 4.5) involves validation of all of the study results by CSRA
Phase 6 (Section 4.6) involves the development of QC acceptance criteria from the study data and
production of the final version of the method document
4.1 Phase 1 - Soliciting Laboratories
Phase 1 of the study involves identifying and soliciting up to 20 laboratories to participate in the study.
As noted earlier, some of those laboratories will be contracted by CSRA and others may participate as
volunteers. While not a hard and fast differentiation, the contracted laboratories are likely to be
commercial environmental laboratories, whereas the volunteer participants may be EPA Regional
laboratories, utility laboratories, or other organizations that are unlikely to be able to accept payment for
their participation.
PCB Congener Multi-lab Validation Study Plan
4
January 2018
-------
CSRA and EPA will develop a broad list of likely participants and contact them in advance of a formal
solicitation to determine their potential interest. Once the list of potential participants has been
established, CSRA will competitively solicit bids using government-approved procurement procedures
and an EAD-approved statement of work (SOW), or equivalent documentation that details the
requirements for sample preparation, storage, shipment, analysis, and QA/QC. The SOW also will
stipulate that the laboratory must have a comprehensive laboratory QA program in place and operating at
all times during performance under the SOW and this program must be consistent with EPA guidance for
quality systems (Guidance for Developing Quality Systems for Environmental Programs, EPA/240/R-
02/008, November 2002) and the general laboratory procedures specified in the Handbook for Analytical
Quality Control in Water and Wastewater Laboratories (EPA-600/4-79-019). CSRA will also work with
EPA to develop suitable mechanisms to engage any volunteer laboratories identified by EPA. Such
mechanisms may involve a memorandum of understanding (MOU) and/or a voluntary participation
agreement form previously developed by EPA for similar studies.
Regardless of the nature of a laboratory's participation (contracted or volunteer), the same study
requirements will apply and will be described in a study-specific statement of work and study-specific
instructions.
4.2 Phase 2 - Procuring Standards and Study Samples
Phase 2 of the study involves procuring sufficient quantities of 1) the analytical standards needed to
perform the method, and 2) the samples that will be analyzed in the study. Since the method is still in
draft form, many of the standards are not available as ready-to-use commercial products. Having each
laboratory prepare their own standards from neat materials or available stock solutions adds significant
variability to the study results that is not likely to reflect routine laboratory practice when performing the
method. Providing the same standards to each laboratory removes that aspect of variability and provides
an incentive for both contracted and volunteer laboratories to participate.
CSRA and EPA previously identified the likely commercial sources of the needed standards. CSRA will
use government-approved procurement procedures and an EPA-approved SOW (or equivalent) to obtain
sufficient volumes of the needed standards and have them shipped directly to the participant laboratories.
CSRA anticipates that these materials may need to be procured from multiple existing commercial
sources, but is also investigating the possibility of finding a single vendor who can obtain materials from
other vendors and prepare the entire suite of standards needed. The list of standards and quantities is
provided in Appendix A.
Once the sources of the standards have been identified and CSRA purchase orders are in place, CSRA
staff will work with the vendors to schedule and direct the shipments of materials to each participating
laboratory. CSRA staff will notify each laboratory of impending shipments, track each shipment from the
vendor to the laboratory, and confirm condition of the materials on receipt with each laboratory CSRA
will work with the vendors and laboratories to resolve any issues or discrepancies, and will communicate
with EPA regularly.
The focus of the study is on analysis of real-world environmental matrices, including wastewaters,
soil/sediment, biosolids, and fish tissue. A generalized list of sample types and quantities is provided in
Appendix B. Given the breadth of the matrices and samples, EPA and CSRA anticipate procuring the
services of an established commercial vendor of PT samples to prepare the study samples. EPA will work
with municipal, state, and EPA Regional contacts to obtain sufficient volumes of several real-world
wastewaters to be used in the study.
EPA plans to utilize nine wastewater matrices, submitting samples of each to all of the laboratories
participating in the study. The wastewater samples will include effluents from a publicly owned
treatment works (POTW), a substitute wastewater as specified in ASTM D 5905 - 98 (Reapproved 2013),
Standard Specification for Substitute Wastewater (Reference 8.5), and wastewaters from specific
PCB Congener Multi-lab Validation Study Plan
5
January 2018
-------
industrial discharges if they can be obtained in sufficient quantities. At least one of the wastewater matrix
types should have one of the following characteristics, such that each criterion below is represented by at
least one wastewater:
Total suspended solids (TSS) greater than 40 mg/L
Total dissolved solids (TDS) greater than 100 mg/L
Oil and grease greater than 20 mg/L
NaCl greater than 120 mg/L
CaCC>3 greater than 140 mg/L
EPA and CSRA will work with other contacts to obtain sufficient masses of soils/sediments, biosolids,
and fish tissues. All of these materials will be delivered to the selected PT vendor to be homogenized and
aliquoted into study-specific sizes, and distributed to each laboratory in accordance with the EPA-
approved PT vendor SOW. EPA plans to utilize three soil/sediment matrices, three biosolids matrices,
and three fish tissue matrices, submitting samples of each to all of the laboratories participating in the
study.
As with the standards, CSRA staff will work with the vendors to schedule and direct the shipments of
materials to each participating laboratory. CSRA staff will notify each laboratory of impending
shipments, track each shipment from the vendor to the laboratory, and confirm condition of the materials
on receipt with each laboratory. CSRA will work with the vendors and laboratories to resolve any issues
or discrepancies, and will communicate with EPA regularly.
4.3 Phase 3 - Calibration and Demonstration of Capability
Prior to analyzing any of the study samples, each laboratory will perform an initial multi-point calibration
and conduct an initial demonstration of capability for each sample matrix, as described in the sections
below.
4.3.1 Initial Calibration
Each laboratory will calibrate their instrumentation using the six standards provided by EPA for the study
and as described in the draft method. The six calibration standards cover a concentration range from 10
ng/mL in the lowest standard to 2,000 ng/mL in the highest standard (equivalent to 1 ng/L to 200 ng/L in
a one-liter aqueous sample). Each laboratory will report the relative responses (RRs) of the native
congeners, using the calculations described in the method. They will also report the response factors
(RFs) for each labeled congener, using the calculations described in the method. Each laboratory will
report the calibration linearity metric that they use (e.g., the relative standard deviation) for each congener
as well.
Twenty-three native (unlabeled) PCB congeners in the standards are calibrated by isotope dilution
quantitation, by virtue of including the 13Ci2-labeled analog of each of those congeners in the calibration
standards. An additional 14 congeners are calibrated using modified isotope dilution, by virtue of
coeluting with a congener that has a 13Ci2-labeled analog in the calibration standards. Twenty-eight more
congeners are calibrated using the response of a labeled congener in the same level of chlorination, via a
process called extracted internal standard. The remaining 144 congeners are calibrated indirectly, using
the response factor for a congener in the same level of chlorination. Each laboratory will report the
response factors (RFs) of the native congeners, using the calculations described in the method. Each
laboratory will report the calibration linearity metric that they use (e.g., the relative standard deviation)
for each congener as well.
PCB Congener Multi-lab Validation Study Plan
6
January 2018
-------
4.3.2 Initial Demonstration of Capability (IDC)
Each laboratory will perform an initial demonstration of capability (IDCs) for each of the three matrix
types in the study (aqueous, soil/sediment/biosolid, and fish tissue) using a suitable spiked reference
matrix. The spiked reference matrix is a clean matrix (void of target compounds at or above the MDL).
The reference matrix for aqueous samples is 1 L of purified or reagent water. For soil/sediment/biosolid,
the reference matrix will be clean sand, and for fish tissue, the reference matrix will be a 90/10 mixture of
clean sand and vegetable oil, which simulates the lipid content of fish tissues.
Each IDC will include an initial precision and recovery (IPR) determination and a method detection limit
(MDL) study. Although the method includes procedures for separatory funnel extraction and two forms
of solid-phase extraction of aqueous samples, each laboratory participating in the study will use the
separatory funnel extraction procedure for the IDC, because that procedure is readily available in all of
the laboratories that are likely to employ the final promulgated method.
Data from the single-laboratory study indicate that solid-phase extraction can perform as well as
separatory funnel extraction on real-world effluent samples. If a given laboratory participating in the
study also has the necessary equipment for one of the two forms of solid-phase extraction, EPA may opt
to have such laboratories also perform the IDC for aqueous using solid-phase extraction for comparative
purposes. However, EPA anticipates developing a single set of QC acceptance criteria for the method for
aqueous samples that are applicable to both types of extraction procedures.
4.3.3 IPR Determination
The IPR consists of four replicate samples of the reference matrix spiked with native congeners and
labeled compounds and carried through the entire analytical process (sample preparation and analysis).
The native congeners should be spiked around the midpoint of the calibration curve. Each laboratory will
calculate the percent (%) recovery of each native congener using Equation 1:
Eq. 1
C
Recovery = %R = x 100
where:
Cs = Measured concentration of the spiked sample aliquot
Cn = Nominal (theoretical) concentration of the spiked aliquot
The relative standard deviation (RSD) is calculated using the results of the four replicates for each native
congener using Equation 2:
Eq. 2
cn
RSD = x 100
r
avg
where:
SD = Standard deviation of Cs for the four replicates
Cavg = Average measured concentration for the four replicates
Each laboratory will perform an IPR study for each matrix and/or extraction type and will report the
results for the individual IPR samples, as well as the recoveries and RSDs for each analyte.
PCB Congener Multi-lab Validation Study Plan
7
January 2018
-------
4.3.4 MDL studies
Each laboratory will perform an MDL study using the newly promulgated MDL procedure at 40 CFR
Part 136 Appendix B, for each matrix type (Reference 8.6). Each MDL study will consist of seven
replicate reference matrix samples spiked with native congeners (MDLS) and seven replicate method
blanks (MDLb), all carried through the entire analytical process (sample preparation and analysis). The
MDL study will be conducted for all 209 congeners. The native congeners will be spiked at a
concentration near that of the lowest calibration standard in seven samples. Each laboratory will calculate
and report the MDLb and MDLS, as well as the spiking levels and the individual MDL study results for all
14 aliquots in each matrix type. If an analyte is not detected, the spiking level will be raised, and the
entire MDL study will be repeated for that analyte.
As with the IPR studies for aqueous samples, all laboratories will perform the MDL study for aqueous
samples using the separatory funnel extraction procedures described in the method, and EPA may opt to
have selected laboratories also determine MDLs using one of the solid-phase extraction procedures.
The draft method uses a 5-g sample aliquot for biosolids samples, and a 10-g aliquot for the other solid-
phase samples. In order to reduce the effort at the laboratories, each laboratory will be instructed to use
only 5 g for the MDL study for solids. CSRA and EPA anticipate that this may yield higher MDL values
than for a 10-g sample size, but that the data will still be sufficient to characterize the general sensitivity
of the method for solid matrices, since in practice, each laboratory using the final method will have to
determine their own MDL values for compliance with method requirements.
4,4 Phase 4 - Analyses of Study Samples
The focus of Phase 4 is to evaluate the GC-LRMS-SIM procedure in various real-world matrices,
including wastewaters, soils/sediments, biosolids, and fish tissues.
Wastewater Analyses
Each laboratory will receive three 1-L aliquots of each of nine wastewater samples (27 aliquots in total).
One aliquot will be prepared and analyzed unspiked. The other two aliquots will be used by each
laboratory to prepare a matrix spike/matrix spike duplicate (MS/MSD) pair. EPA and CSRA will provide
the spiking levels of the native congeners to be used for each wastewater sample type to all of the
laboratories, based on "reconnaissance" analyses conducted by SGS-AXYS Analytical, the laboratory
that developed the draft method, using aliquots of the homogenized samples provided by the study sample
vendor.
The draft method includes three extraction procedures for aqueous samples: the traditional separatory
funnel extraction available in virtually any laboratory, and two forms of solid-phase extraction that rely
on less common equipment from two or more vendors. EPA evaluated all three extraction procedures in
side-by-side testing during the single-laboratory study and found them to yield similar results.
EPA anticipates having all laboratories use separatory funnel extraction for the wastewater samples. If a
sufficient number of laboratories are identified who already possess the necessary equipment for one of
the other of the solid-phase extraction procedures, EPA will provide additional sample aliquots to those
laboratories and arrange for them to use solid-phase extraction in addition to the separatory funnel
extraction procedure. (As noted in Section 4.3, EPA currently anticipates developing a single set of QC
acceptance criteria for the method for aqueous samples that are applicable to both types of extraction
procedures.)
Assuming that 20 laboratories participate in the study either under contract to CSRA or as volunteers, the
study design will yield 180 results for unspiked wastewater samples and 360 matrix spike sample results
for each of the native and labeled congeners. Even if fewer than 20 laboratories participate, or are able to
PCB Congener Multi-lab Validation Study Plan
8
January 2018
-------
produce usable results, EPA will still have a significant body of performance data with which to judge the
method's capabilities.
Soil/Sediment andBiosolids Analyses
Each laboratory will receive three 10-g aliquots of each of three soil/sediment samples (9 aliquots in total)
and three 5-g aliquots of each of three biosolids samples (9 aliquots in total). As with the wastewater
samples, one aliquot will be prepared and analyzed unspiked. The other two aliquots will be used by each
laboratory to prepare a matrix spike/matrix spike duplicate (MS/MSD) pair. EPA and CSRA will provide
the spiking levels of the native congeners to be used for each soil/sediment sample type and each
biosolids sample type to all of the laboratories, based on "reconnaissance" analyses conducted by SGS-
AXYS Analytical, the laboratory that developed the draft method, using aliquots of the homogenized
samples provided by the study sample vendor.
Assuming that 20 laboratories participate in the study either under contract to CSRA or as volunteers, the
study design will yield 60 results for unspiked soil/sediment samples, 60 results for unspiked biosolids
samples, 120 soil/sediment matrix spike sample results, and 120 biosolids matrix spike sample results for
each of the native and labeled congeners. Even if fewer than 20 laboratories participate, or are able to
produce usable results, EPA will still have a significant body of performance data with which to judge the
method's capabilities.
Fish Tissue Analyses
Each laboratory will receive three 10-g aliquots of each of three fish tissue samples (9 aliquots in total).
As with the other matrices, one aliquot will be prepared and analyzed unspiked. The other two aliquots
will be used by each laboratory to prepare a matrix spike/matrix spike duplicate (MS/MSD) pair. EPA
and CSRA will provide the spiking levels of the native congeners to be used for each tissue sample type
to all of the laboratories, based on "reconnaissance" analyses conducted by SGS-AXYS Analytical, the
laboratory that developed the draft method, using aliquots of the homogenized samples provided by the
study sample vendor.
Assuming that 20 laboratories participate in the study either under contract to CSRA or as volunteers, the
study design will yield 60 results for unspiked tissue samples and 120 tissue matrix spike sample results
for each of the native and labeled congeners. Even if fewer than 20 laboratories participate, or are able to
produce usable results, EPA will still have a significant body of performance data with which to judge the
method's capabilities.
4,5 Phas f' Hi > "Vť "fication and Validation
All of the results from all of the laboratories participating in the study will be reviewed and validated by
CSRA relative to the study's goals. Every data submission will be checked for completeness (e.g., were
all of the samples submitted to the laboratory analyzed and results submitted?) and to determine if the
supporting documentation indicate that the laboratory followed the method and the study-specific
instructions.
Each laboratory will be required to submit the raw data (e.g., instrument printouts and copies of bench
records) that support the study results. CSRA will examine all of the raw data, perform spot checks of a
percentage of the calculations from each laboratory, and ensure that the reported results can be traced
back through all steps in the analytical process. If any issues are identified, CSRA will work with the
laboratory to clarify the situation, obtain any missing information, and document the resolution. EPA will
be advised of the status of the review efforts on a regular basis.
Because this is a method validation effort, there are no a priori quality control acceptance criteria, and
data from the study will not be excluded from consideration simply because they appear to fail some
PCB Congener Multi-lab Validation Study Plan
9
January 2018
-------
pre-conceived performance expectations. Every effort will be made to retain as many results as practical.
CSRA will flag results from samples with obvious documented failures (e.g., extracts accidentally taken
to dryness) for exclusion from use in developing method performance information and will document the
rationale for such exclusions in the project files and/or the project database. However, in the absence of
evidence of such failures, all of the results for the native and labeled congeners will be included in the
initial data set. CSRA and EPA will use statistical tests in Phase 6 of the study to determine if results for
specific laboratories, samples, or congeners may be outliers that should be removed from use in
developing QC acceptance criteria. Suspected outliers will be examined in detail by CSRA and the
laboratory before they are excluded from use in developing method performance summaries.
4.6 Phase 6 - Development of QC Acceptance Criteria
The last major phase of the study will be to develop statistically based QC acceptance criteria and
summarize method performance in real-world samples. The overall procedures used for that process are
described in Section 7 of this study plan.
Section 5. Quality Control Procedures
The GC-LRMS-SIM procedure includes many of the traditional quality control (QC) procedures found in
EPA methods for the analysis of organic contaminants. The associated QC checks are summarized in
Table 2 (following Section 8). Each laboratory is responsible for maintaining their instrumentation and
ensuring that all study samples are analyzed on a properly calibrated instrument. Therefore, if the
instrument calibrations or other instrument QC (i.e., mass spectrometer tune, mass calibration check, or
qualitative identification criteria) are outside the normal criteria (see Table 2), the laboratory will take
standard measures (e.g., cleaning the instrument, clipping column ends, or replacing column or other
instrument parts) to correct the problem before any study samples are analyzed. The laboratory is also
responsible for inspecting all study samples and standards to ensure they meet all study requirements. If
standard measures do not correct identified problems or if study schedules will be impacted due to
necessary repairs or replacement of study samples or standards, the laboratory will notify the CSRA
Project Leader to indicate the impact on study schedules, the laboratory's plans to resolve the problem(s),
and if any study samples will need to be reanalyzed.
Each laboratory will report the results from all procedure-specified QC operations, either in electronic
format, or if necessary, in hard copy. CSRA will compile the QC results in a database specific to this
project (See Section 6).
Section 6. Data Reporting an i Management
6.1 Laboratory reporting requirements
Each laboratory participating in the study will be required to (1) report summary-level electronic data and
supporting raw data, and (2) maintain their raw data for a period of five (5) years and provide them upon
request (at additional cost negotiated as necessary). Raw data will include all calibration data,
chromatograms, quantitation reports (including peak areas or heights), strip charts, spectra, bench sheets,
and laboratory notebooks showing weights, volumes, manual calculations, and other data that will allow
verification of the calculations performed and will allow the final results reported to be traced back to the
raw data.
Each laboratory also will be instructed to adhere to the following rules when reporting data:
All reports and documentation, including instrument printouts and other raw data, must be
sequentially paginated, clearly labeled with the laboratory name, and labeled to provide sufficient
identification for method blanks, calibration, interference checks, etc., necessary to link the raw
data with associated summary reports.
PCB Congener Multi-lab Validation Study Plan
10
January 2018
-------
Results from all analyses must be reported, including calibration data and any dilutions or
reanalysis performed. The laboratory also must include an explanation of any dilutions or
reanalysis performed and identify which of the analyses the lab considers to be most appropriate
for use.
Results of all measurements must be reported to three significant figures in the appropriate
reporting units (e.g., ng/L for water samples, ng/g for solid and tissue samples) to facilitate
review and evaluation
The terms "zero" and "trace" are not to be used; the term "not detected" (ND) is to be used for
each measurement for which no signal is produced or if method-specified qualitative
identification criteria are not met.
If a signal is produced, the value must be reported, even if the value is negative. If the value is
below the lowest calibration standard, a "J" flag must be applied to this value.
Results must be reported for all study samples, including QC samples.
In addition, each laboratory will be required to submit a written "narrative report" with each data package.
The narrative report will contain detailed descriptions of any difficulties encountered in the generation of
the analytical results and QC data and any attempts to resolve the difficulties. It also will contain a
detailed description of any modifications to the GC-LRMS-SIM procedure and the date that these
modifications were pre-approved by CSRA.
Finally, each laboratory will be asked to provide comments on the draft method document, focusing on
the clarity of the procedures, identifying any gaps in the descriptions of the analytical processes,
inconsistencies, etc.
6.2 CSRA Data Management and Reporting
CSRA will store all submitted data (hard copy and electronic) in master files established for this study on
CSRA's secure local area network, which is backed up nightly, and/or in hardcopy files, depending on the
source material. Cumulatively, these master files will include the following documents and records:
This study plan (including all submitted draft versions, comments, and revisions)
Documentation of the procedures used to assess the competency of laboratories participating in
this study
Documents and records associated with the solicitation and award of participant laboratories,
including the SOWs or equivalent that describe participant laboratory requirements
Documents and records associated with the procurement of standards and study samples,
including SOWs or equivalent that describe the process used to collect and produce study samples
The name, address, phone number and primary contact at the standards vendor and each
participating laboratory
Copies of all written correspondence (excluding emails) with laboratory staff, sampling
personnel, and EPA staff regarding the study
A log (or other record) that documents verbal communication with laboratory staff, sample
coordinators, sampling personnel, and EPA staff regarding study status or problems
Records concerning sample shipment and receipt
All analytical data resulting from this study
All laboratory comments on the method resulting from this study
Records of all CSRA data review assessments and statistical analyses submitted to EPA
All draft and final reports submitted to EPA pertaining to this study
PCB Congener Multi-lab Validation Study Plan
11
January 2018
-------
CSRA and EPA will develop a schedule for routine communications during the course of the study, based
on the specific activities underway at the time. For example, CSRA will communicate with the EPA
Project Manager more frequently (e.g., daily) during those periods when samples are being shipped to the
laboratories, versus less frequent communications during the periods when sample analyses are taking
place.
Section 7. Evaluation of Method Performance
EPA's overall goal is to develop method performance data for the GC-LRMS-SIM procedure. The results
of the analyses in the first four phases of this study will be evaluated using common statistical procedures
(References 8.4, 8.7, and 8.8). EPA and CSRA will use the results from the replicate samples to develop
QC criteria for initial precision and recovery (IPR) tests, ongoing precision and recovery (OPR) tests,
labeled compounds recoveries, duplicate precision, etc. A general description of the derivation of those
QC acceptance criteria is provided in Appendix C and is based on EPA's existing new method evaluation
protocol (Reference 8.9).
Finally, EPA and CSRA will develop tables of method performance data, including precision and
accuracy, as a function of analyte concentration that will provide an indication of expected performance
of the procedures under typical conditions. Such tables may be included in the revised procedure as
further evidence of its overall capabilities or limitations.
Following completion of the method performance evaluation, CSRA and EPA will prepare a formal
report on the results of the multi-laboratory validation study. EPA will submit that draft report to
appropriate levels of management review within EPA and revise the report as needed. If the study is
successful and EPA decides to move forward with a rulemaking to approve the new PCB method at
40 CFR Part 136 for use in nationwide compliance monitoring, the study report and records from the
study will be placed into the rulemaking docket.
Section 8. References
8.1 US Environmental Protection Agency. 1984. Method 608, Organochlorine Pesticides and PCBs.
http://www.epa.gov/cwa-methods/approved-cwa-test-methods-organic-compounds
8.2 US Environmental Protection Agency. 2010. Method 1668C, Chlorinated Biphenyl Congeners in
Water, Soil, Sediment, Biosolids, and Tissue by HRGC/HRMS. Office of Water, Office of
Science and Technology, Engineering and Analysis Division. EPA-820-R-10-005.
8.3 AXYS Analytical Services Ltd. 2011. AXYS Method MLA-007: Analytical Method for the
Determination of PCBs as PCB congeners, homolog totals, Aroclors or total PCBs. Rev. 13, Ver.
8, July 2015.
8.4 ASTM 2013. Method D2777-13, Standard Practice for Determination of Precision and Bias of
Applicable Methods of Committee D-19 on Water, ASTM International, West Conshohocken,
PA. http://www.astm.org
8.5 ASTM 2013. Method D5095-13. Standard Practice for the Preparation of Substitute Wastewater,
ASTM International, West Conshohocken, PA. http://www.astm.org
8.6 Federal Register. Vol. 82, No. 165. August 28, 2017. pp. 40939 - 40941.
8.7 SAS Institute Inc. 1994. SAS/STAT User's Guide, Volume 2, GLM-VARCOMP. Version 6, 4th
Edition, June 1994.
PCB Congener Multi-lab Validation Study Plan
12
January 2018
-------
8.8 Berry, D. A.; Lindgren, B. W. 1990. Statistics: Theory and Methods, pp. 286-290, 600-618.
Brooks/Cole Publishing Company. Pacific Grove, California.
8.9 US Environmental Protection Agency. 2016. Protocol for Review and Validation of New
Methods for Regulated Organic and Inorganic Analytes in Wastewater Under EPA 's Alternate
Test Procedure Program. U.S. Environmental Protection Agency. Office of Water, Engineering
and Analysis Division. EPA 821-B-16-001.
PCB Congener Multi-lab Validation Study Plan
13
January 2018
-------
Table 1. Names, Congener Numbers, and CAS Registry Numbers
for Native Chlorinated Biphenyl (CB) Congeners
CB congener1
Congener number
CAS Number
2-MoCB
PCB 1
2051-60-7
3-MoCB
PCB 2
2051-61-8
4-MoCB
PCB 3
2051-62-9
2,2'-DiCB
PCB 4
13029-08-8
2,3-DiCB
PCB 5
16605-91-7
2,3'-DiCB
PCB 6
25569-80-6
2,4-DiCB
PCB 7
33284-50-3
2,4'-DiCB2
PCB 8
34883-43-7
2,5-DiCB
PCB 9
34883-39-1
2,6-DiCB
PCB 10
33146-45-1
3,3'-DiCB
PCB 11
2050-67-1
3,4-DiCB
PCB 12
2974-92-7
3,4'-DiCB
PCB 13
2974-90-5
3,5-DiCB
PCB 14
34883-41-5
4,4'-DiCB
PCB 15
2050-68-2
2,2',3-TrCB
PCB 16
38444-78-9
2,2',4-TrCB
PCB 17
37680-66-3
2,2',5-TrCB2
PCB 18
37680-65-2
2,2',6-TrCB
PCB 19
38444-73-4
2,3,3'-TrCB
PCB 20
38444-84-7
2,3,4-TrCB
PCB 21
55702-46-0
2,3,4'-TrCB
PCB 22
38444-85-8
2,3,5-TrCB
PCB 23
55720-44-0
2,3,6-TrCB
PCB 24
55702-45-9
2,3',4-TrCB
PCB 25
55712-37-3
2,3',5-TrCB
PCB 26
38444-81-4
2,3',6-TrCB
PCB 27
38444-76-7
2,4,4'-TrCB2
PCB 28
7012-37-5
2,4,5-TrCB
PCB 29
15862-07-4
2,4,6-TrCB
PCB 30
35693-92-6
2,4',5-TrCB
PCB 31
16606-02-3
2,4',6-TrCB
PCB 32
38444-77-8
2',3,4-TrCB
PCB 33
38444-86-9
2',3,5-TrCB
PCB 34
37680-68-5
3,3',4-TrCB
PCB 35
37680-69-6
3,3',5-TrCB
PCB 36
38444-87-0
3,4,4'-TrCB
PCB 37
38444-90-5
3,4,5-TrCB
PCB 38
53555-66-1
3,4',5-TrCB
PCB 39
38444-88-1
2,2',3,3'-TeCB
PCB 40
38444-93-8
2,2',3,4-TeCB
PCB 41
52663-59-9
2,2',3,4'-TeCB
PCB 42
36559-22-5
2,2',3,5-TeCB
PCB 43
70362-46-8
2,2',3,5'-TeCB3
PCB 44
41464-39-5
2,2',3,6-TeCB
PCB 45
70362-45-7
2,2',3,6'-TeCB
PCB 46
41464-47-5
2,2',4,4'-TeCB
PCB 47
2437-79-8
PCB Congener Multi-lab Validation Study Plan
14
January 2018
-------
Table 1. Names, Congener Numbers, and CAS Registry Numbers
for Native Chlorinated Biphenyl (CB) Congeners
CB congener1
Congener number
CAS Number
2,2',4,5-TeCB
PCB 48
70362-47-9
2,2',4,5'-TeCB
PCB 49
41464-40-8
2,2',4,6-TeCB
PCB 50
62796-65-0
2,2',4,6'-TeCB
PCB 51
68194-04-7
2,2',5,5'-TeCB2
PCB 52
35693-99-3
2,2',5,6'-TeCB
PCB 53
41464-41-9
2,2',6,6'-TeCB
PCB 54
15968-05-5
2,3,3',4'-TeCB
PCB 55
74338-24-2
2,3,3',4'-TeCB
PCB 56
41464-43-1
2,3,3',5-TeCB
PCB 57
70424-67-8
2,3,3',5'-TeCB
PCB 58
41464-49-7
2,3,3',6-TeCB
PCB 59
74472-33-6
2,3,4,4'-TeCB
PCB 60
33025-41-1
2,3,4,5-TeCB
PCB 61
33284-53-6
2,3,4,6-TeCB
PCB 62
54230-22-7
2,3,4',5-TeCB
PCB 63
74472-34-7
2,3,4',6-TeCB
PCB 64
52663-58-8
2,3,5,6-TeCB
PCB 65
33284-54-7
2,3',4,4'-TeCB2
PCB 66
32598-10-0
2,3',4,5-TeCB
PCB 67
73575-53-8
2,3',4,5'-TeCB
PCB 68
73575-52-7
2,3',4,6-TeCB
PCB 69
60233-24-1
2,3',4',5-TeCB
PCB 70
32598-11-1
2,3',4',6-TeCB
PCB 71
41464-46-4
2,3',5,5'-TeCB
PCB 72
41464-42-0
2,3',5',6-TeCB
PCB 73
74338-23-1
2,4,4',5-TeCB
PCB 74
32690-93-0
2,4,4',6-TeCB
PCB 75
32598-12-2
2',3,4,5-TeCB
PCB 76
70362-48-0
3,3',4,4'-TeCB2'3
PCB 77
32598-13-3
3,3',4,5-TeCB
PCB 78
70362-49-1
3,3',4,5'-TeCB
PCB 79
41464-48-6
3,3',5,5'-TeCB
PCB 80
33284-52-5
3,4,4',5-TeCB6
PCB 81
70362-50-4
2,2',3,3',4-PeCB
PCB 82
52663-62-4
2,2',3,3',5-PeCB
PCB 83
60145-20-2
2,2',3,3',6-PeCB
PCB 84
52663-60-2
2,2',3,4,4'-PeCB
PCB 85
65510-45-4
2,2',3,4,5-PeCB
PCB 86
55312-69-1
2,2',3,4,5'-PeCB
PCB 87
38380-02-8
2,2',3,4,6-PeCB
PCB 88
55215-17-3
2,2',3,4,6'-PeCB
PCB 89
73575-57-2
2,2',3,4',5-PeCB
PCB 90
68194-07-0
2,2',3,4',6-PeCB
PCB 91
68194-05-8
2,2',3,5,5'-PeCB
PCB 92
52663-61-3
2,2',3,5,6-PeCB
PCB 93
73575-56-1
2,2',3,5,6'-PeCB
PCB 94
73575-55-0
PCB Congener Multi-lab Validation Study Plan
15
January 2018
-------
Table 1. Names, Congener Numbers, and CAS Registry Numbers
for Native Chlorinated Biphenyl (CB) Congeners
CB congener1
Congener number
CAS Number
2,2',3,5',6-PeCB
PCB 95
38379-99-6
2,2',3,6,6'-PeCB
PCB 96
73575-54-9
2,2',3',4,5-PeCB
PCB 97
41464-51-1
2,2',3',4,6-PeCB
PCB 98
60233-25-2
2,2',4,4',5-PeCB
PCB 99
38380-01-7
2,2',4,4',6-PeCB
PCB 100
39485-83-1
2,2',4,5,5'-PeCB2
PCB 101
37680-73-2
2,2',4,5,6'-PeCB
PCB 102
68194-06-9
2,2',4,5,'6-PeCB
PCB 103
60145-21-3
2,2',4,6,6'-PeCB
PCB 104
56558-16-8
2,3,3',4,4'-PeCB2'3
PCB 105
32598-14-4
2,3,3',4,5-PeCB
PCB 106
70424-69-0
2,3,3',4',5-PeCB
PCB 107
70424-68-9
2,3,3',4,5'-PeCB
PCB 108
70362-41-3
2,3,3',4,6-PeCB
PCB 109
74472-35-8
2,3,3',4',6-PeCB
PCB 110
38380-03-9
2,3,3',5,5'-PeCB
PCB 111
39635-32-0
2,3,3',5,6-PeCB
PCB 112
74472-36-9
2,3,3',5',6-PeCB
PCB 113
68194-10-5
2,3,4,4',5-PeCB6
PCB 114
74472-37-0
2,3,4,4',6-PeCB
PCB 115
74472-38-1
2,3,4,5,6-PeCB
PCB 116
18259-05-7
2,3,4',5,6-PeCB
PCB 117
68194-11-6
2,3',4,4',5-PeCB2'3
PCB 118
31508-00-6
2,3',4,4',6-PeCB
PCB 119
56558-17-9
2,3',4,5,5'-PeCB
PCB 120
68194-12-7
2,3',4,5,'6-PeCB
PCB 121
56558-18-0
2',3,3',4,5-PeCB
PCB 122
76842-07-4
2',3,4,4',5-PeCB6
PCB 123
65510-44-3
2',3,4,5,5'-PeCB
PCB 124
70424-70-3
2',3,4,5,6'-PeCB
PCB 125
74472-39-2
3,3',4,4',5-PeCB2,3
PCB 126
57465-28-8
3,3',4,5,5'-PeCB
PCB 127
39635-33-1
2,2',3,3',4,4'-HxCB3
PCB 128
38380-07-3
2,2',3,3',4,5-HxCB
PCB 129
55215-18-4
2,2',3,3',4,5'-HxCB
PCB 130
52663-66-8
2,2',3,3',4,6-HxCB
PCB 131
61798-70-7
2,2',3,3',4,6'-HxCB
PCB 132
38380-05-1
2,2',3,3',5,5'-HxCB
PCB 133
35694-04-3
2,2',3,3',5,6-HxCB
PCB 134
52704-70-8
2,2',3,3',5,6'-HxCB
PCB 135
52744-13-5
2,2',3,3',6,6'-HxCB
PCB 136
38411-22-2
2,2',3,4,4',5-HxCB
PCB 137
35694-06-5
2,2',3,4,4',5'-HxCB2
PCB 138
35065-28-2
2,2',3,4,4',6-HxCB
PCB 139
56030-56-9
2,2',3,4,4',6'-HxCB
PCB 140
59291-64-4
2,2',3,4,5,5'-HxCB
PCB 141
52712-04-6
PCB Congener Multi-lab Validation Study Plan
16
January 2018
-------
Table 1. Names, Congener Numbers, and CAS Registry Numbers
for Native Chlorinated Biphenyl (CB) Congeners
CB congener1
Congener number
CAS Number
2,2',3,4,5,6-HxCB
PCB 142
41411-61-4
2,2',3,4,5,6'-HxCB
PCB 143
68194-15-0
2,2',3,4,5',6-HxCB
PCB 144
68194-14-9
2,2',3,4,6,6'-HxCB
PCB 145
74472-40-5
2,2',3,4',5,5'-HxCB
PCB 146
51908-16-8
2,2',3,4',5,6-HxCB
PCB 147
68194-13-8
2,2',3,4',5,6'-HxCB
PCB 148
74472-41-6
2,2',3,4',5',6-HxCB
PCB 149
38380-04-0
2,2',3,4',6,6'-HxCB
PCB 150
68194-08-1
2,2',3,5,5',6-HxCB
PCB 151
52663-63-5
2,2',3,5,6,6'-HxCB
PCB 152
68194-09-2
2,2',4,4',5,5'-HxCB2
PCB 153
35065-27-1
2,2',4,4',5',6-HxCB
PCB 154
60145-22-4
2,2',4,4',6,6'-HxCB
PCB 155
33979-03-2
2,3,3',4,4',5-HxCB3
PCB 156
38380-08-4
2,3,3',4,4',5'-HxCB3
PCB 157
69782-90-7
2,3,3',4,4',6-HxCB
PCB 158
74472-42-7
2,3,3',4,5,5'-HxCB
PCB 159
39635-35-3
2,3,3',4,5,6-HxCB
PCB 160
41411-62-5
2,3,3',4,5',6-HxCB
PCB 161
74472-43-8
2,3,3',4',5,5'-HxCB
PCB 162
39635-34-2
2,3,3',4',5,6-HxCB
PCB 163
74472-44-9
2,3,3',4',5',6-HxCB
PCB 164
74472-45-0
2,3,3',5,5',6-HxCB
PCB 165
74472-46-1
2,3,4,4',5,6-HxCB
PCB 166
41411-63-6
2,3',4,4',5,5'-HxCB3
PCB 167
52663-72-6
2,3',4,4',5',6-HxCB
PCB 168
59291-65-5
3,3',4,4',5,5'-HxCB2'3
PCB 169
32774-16-6
2,2',3,3',4,4',5-HpCB2
PCB 170
35065-30-6
2,2'3,3',4,4',6-HpCB
PCB 171
52663-71-5
2,2',3,3',4,5,5'-HpCB
PCB 172
52663-74-8
2,2',3,3',4,5,6-HpCB
PCB 173
68194-16-1
2,2',3,3',4,5,6'-HpCB
PCB 174
38411-25-5
2,2',3,3',4,5',6-HpCB
PCB 175
40186-70-7
2,2',3,3',4,6,6'-HpCB
PCB 176
52663-65-7
2,2',3,3',4',5,6-HpCB
PCB 177
52663-70-4
2,2',3,3',5,5',6-HpCB
PCB 178
52663-67-9
2,2',3,3',5,6,6'-HpCB
PCB 179
52663-64-6
2,2',3,4,4',5,5'-HpCB2
PCB 180
35065-29-3
2,2',3,4,4',5,6-HpCB
PCB 181
74472-47-2
2,2',3,4,4',5,6'-HpCB
PCB 182
60145-23-5
2,2',3,4,4',5',6-HpCB
PCB 183
52663-69-1
2,2',3,4,4',6,6'-HpCB
PCB 184
74472-48-3
2,2',3,4,5,5',6-HpCB
PCB 185
52712-05-7
2,2',3,4,5,6,6'-HpCB
PCB 186
74472-49-4
2,2',3,4',5,5',6-HpCB2
PCB 187
52663-68-0
2,2',3,4',5,6,6'-HpCB
PCB 188
74487-85-7
PCB Congener Multi-lab Validation Study Plan
17
January 2018
-------
Table 1. Names, Congener Numbers, and CAS Registry Numbers
for Native Chlorinated Biphenyl (CB) Congeners
CB congener1
Congener number
CAS Number
2,3,3',4,4',5,5'-HpCB3
PCB 189
39635-31-9
2,3,3',4,4',5,6-HpCB
PCB 190
41411-64-7
2,3,3',4,4',5',6-HpCB
PCB 191
74472-50-7
2,3,3',4,5,5',6-HpCB
PCB 192
74472-51-8
2,3,3',4',5,5',6-HpCB
PCB 193
69782-91-8
2,2',3,3',4,4',5,5'-OcCB
PCB 194
35694-08-7
2,2',3,3',4,4',5,6-OcCB2
PCB 195
52663-78-2
2,2',3,3',4,4',5,6'-OcCB
PCB 196
42740-50-1
2,2',3,3',4,4',6,6'-OcCB
PCB 197
33091-17-7
2,2',3,3',4,5,5',6-OcCB
PCB 198
68194-17-2
2,2',3,3',4,5,5',6'-OcCB
PCB 199
52663-75-9
2,2',3,3',4,5,6,6'-OcCB
PCB 200
52663-73-7
2,2',3,3',4,5',6,6'-OcCB
PCB 201
40186-71-8
2,2',3,3',5,5',6,6'-OcCB
PCB 202
2136-99-4
2,2',3,4,4',5,5',6-OcCB
PCB 203
52663-76-0
2,2',3,4,4',5,6,6'-OcCB
PCB 204
74472-52-9
2,3,3',4,4',5,5',6-OcCB
PCB 205
74472-53-0
2,2',3,3',4,4',5,5',6-NoCB2
PCB 206
40186-72-9
2,2',3,3',4,4',5,6,6'-NoCB
PCB 207
52663-79-3
2,2',3,3',4,5,5',6,6'-NoCB
PCB 208
52663-77-1
DeCB2
PCB 209
2051-24-3
1. Abbreviations for chlorination levels (homologs)
MoCB
monochlorobiphenyl
HxCB
hexachlorobiphenyl
DiCB
dichlorobiphenyl
HpCB
heptachlorobiphenyl
TrCB
trichlorobiphenyl
OcCB
octachlorobiphenyl
TeCB
tetrachlorobiphenyl
NoCB
nonachlorobiphenyl
PeCB
pentachlorobiphenyl
DeCB
decachlorobiphenyl
2. National Oceanic and Atmospheric Administration (NOAA) congener of interest
3. World Health Organization (WHO) toxic congener
PCB Congener Multi-lab Validation Study Plan
18
January 2018
-------
Table 2. Routine QC Checks
QC Check
Frequency
Acceptance Criterion
Study Requirements
Initial demonstration of
capability
Once per matrix type and
extraction technique
IPR %recovery and %RSD will
be established after review of
study data. Report all results as
generated
Each laboratory will provide IPR
data for each matrix type. Study
data will be used to develop
criteria
Initial calibration
(ICAL), 6-point
minimum
Once
%RSD of the RRFs should be <
20%. Report all results as
generated.
ICAL data will be provided by
each laboratory and compared to
typical calibration criteria
Calibration verification
(VER)
Initially and every 12 hrs.
Initial Calibration can be
used in place of initial VER
if samples are analyzed
within 12 hours of initial
calibration.
VER RRFs should be within
ą 20% of the mean RRFs from
the initial calibration. Report all
results as generated.
VER data will be provided by each
laboratory and compared to typical
calibration criteria
Method blank
One per batch of 20 field
samples or fewer
To be determined based on study
data. Report all results as
generated.
Perform as specified in the
procedure
Laboratory control
sample
One per preparation
batch of 20 field samples or
fewer
To be determined based on study
data. Report all results as
generated.
Perform as specified in procedure
to demonstrate performance at the
method-specified LCS
concentration. Study data will be
used to develop criteria
Matrix spike/ Matrix
spike duplicate
(MS/MSD)
One per preparation batch of
20 field samples or fewer
To be determined based on study
data. Report all results as
generated.
Not required for the method
itself, since isotope dilution
provides recovery data for every
sample.
For the study, the MS/MSD results
will be used to demonstrate
method performance.
RPD between the MS and MSD
analyses may be used to develop
acceptance criteria for duplicate
unspiked analyses.
Mass spectrometer tune
(PFTBA)
Daily, at start up
To be determined based on study
data. Report all results as
generated.
Perform as specified in procedure
Mass calibration check
(PFTBA)
Daily, at start up
Peak drift for m/z 69,219 and
502 < 0.4 (or approx. 2 mm on
the calibration printout).
Relative peak intensities (m/z
219 and 502 divided by m/z
69) between 50 - 150%)
Perform as specified in procedure
Qualitative Identification
Criteria
All peaks in all analyses
RRT within ą 3 sec of VER
calibration RRT
Ratio of 2 ions (Within 20% of
theoretical)
Perform as specified in procedure
PCB Congener Multi-lab Validation Study Plan
19
January 2018
-------
Appendix A
Standards Required for the Validation Study
PCB Congener Multi-lab Validation Study Plan
A-l
January 2018
-------
The standards in Tables A-l to A-4 will be procured as custom mixtures from one or more commercial
vendors to support this study. The standard mixes shown in Table A-5 are already commercially
available.
Table A-l. Calibration Standard Solutions
Congener number
Calibration Standards (ng/mL)
CS-1
CS-2
CS-3
CS-4
CS-5
CS-6
PCB-1
10
20
40
160
400
2,000
PCB-3
10
20
40
160
400
2,000
PCB-4
10
20
40
160
400
2,000
PCB-8
10
20
40
160
400
2,000
PCB-11
10
20
40
160
400
2,000
PCB-15
10
20
40
160
400
2,000
PCB-18
10
20
40
160
400
2,000
PCB-19
10
20
40
160
400
2,000
PCB-2 8
10
20
40
160
400
2,000
PCB-31
10
20
40
160
400
2,000
PCB-3 7
10
20
40
160
400
2,000
PCB-44
10
20
40
160
400
2,000
PCB-52
10
20
40
160
400
2,000
PCB-54
10
20
40
160
400
2,000
PCB-64
10
20
40
160
400
2,000
PCB-66
10
20
40
160
400
2,000
PCB-70
10
20
40
160
400
2,000
PCB-74
10
20
40
160
400
2,000
PCB-77
10
20
40
160
400
2,000
PCB-85
10
20
40
160
400
2,000
PCB-95
10
20
40
160
400
2,000
PCB-99
10
20
40
160
400
2,000
PCB-101
10
20
40
160
400
2,000
PCB-104
10
20
40
160
400
2,000
PCB-105
10
20
40
160
400
2,000
PCB-110
10
20
40
160
400
2,000
PCB-118
10
20
40
160
400
2,000
PCB-126
10
20
40
160
400
2,000
PCB-132
10
20
40
160
400
2,000
PCB-138
10
20
40
160
400
2,000
PCB-147
10
20
40
160
400
2,000
PCB-149
10
20
40
160
400
2,000
PCB-153
10
20
40
160
400
2,000
PCB-155
10
20
40
160
400
2,000
PCB-156
10
20
40
160
400
2,000
PCB-166
10
20
40
160
400
2,000
PCB-169
10
20
40
160
400
2,000
PCB-177
10
20
40
160
400
2,000
PCB-180
10
20
40
160
400
2,000
PCB-187
10
20
40
160
400
2,000
PCB-188
10
20
40
160
400
2,000
PCB-189
10
20
40
160
400
2,000
PCB-199
10
20
40
160
400
2,000
PCB Congener Multi-lab Validation Study Plan
A-2
January 2018
-------
Table A-l. Calibration Standard Solutions
Congener number
Calibration Standards (ng/mL)
CS-1
CS-2
CS-3
CS-4
CS-5
CS-6
PCB-202
10
20
40
160
400
2,000
PCB-205
10
20
40
160
400
2,000
PCB-206
10
20
40
160
400
2,000
PCB-208
10
20
40
160
400
2,000
PCB-209
10
20
40
160
400
2,000
13Ci2-PCB-l
400
400
400
400
400
400
13Ci2-PCB-3
400
400
400
400
400
400
13Ci2-PCB-4
400
400
400
400
400
400
13Ci2-PCB-ll
400
400
400
400
400
400
13Ci2-PCB-15
400
400
400
400
400
400
13Ci2 PCB-19
400
400
400
400
400
400
13Ci2-PCB-28
400
400
400
400
400
400
13Ci2-PCB-37
400
400
400
400
400
400
13Ci2-PCB-52
400
400
400
400
400
400
13Ci2-PCB-54
400
400
400
400
400
400
13Ci2-PCB-70
400
400
400
400
400
400
13Ci2-PCB-77
400
400
400
400
400
400
13Ci2-PCB-85
400
400
400
400
400
400
13Ci2-PCB-101
400
400
400
400
400
400
13Ci2-PCB-104
400
400
400
400
400
400
13Ci2-PCB-118
400
400
400
400
400
400
13Ci2-PCB-126
400
400
400
400
400
400
13Ci2-PCB-138
400
400
400
400
400
400
13Ci2-PCB-153
400
400
400
400
400
400
13Ci2-PCB-155
400
400
400
400
400
400
13Ci2-PCB-169
400
400
400
400
400
400
13Ci2-PCB-180
400
400
400
400
400
400
13Ci2-PCB-188
400
400
400
400
400
400
13Ci2-PCB-189
400
400
400
400
400
400
13Ci2-PCB-202
400
400
400
400
400
400
13Ci2-PCB-205
400
400
400
400
400
400
13Ci2-PCB-206
400
400
400
400
400
400
13Ci2-PCB-208
400
400
400
400
400
400
13Ci2-PCB-209
400
400
400
400
400
400
13Ci2-PCB-8
400
400
400
400
400
400
13Ci2-PCB-79
400
400
400
400
400
400
13Ci2-PCB-162
400
400
400
400
400
400
PCB Congener Multi-lab Validation Study Plan
January 2018
-------
Table A-2. Labeled Compound Standard Solution (1,250 ng/mL)
13Ci2-PCB-l
13Ci2-PCB-52
13Ci2-PCB-118
13Ci2-PCB-188
13Ci2-PCB-3
13Ci2-PCB-54
13Ci2-PCB-126
13Ci2-PCB-189
13Ci2-PCB-4
13Ci2-PCB-70
13Ci2-PCB-138
13Ci2-PCB-202
13Ci2-PCB-ll
13Ci2-PCB-77
13Ci2-PCB-153
13Ci2-PCB-205
13Ci2-PCB-15
13Ci2-PCB-85
13Ci2-PCB-155
13Ci2-PCB-206
13Ci2 PCB-19
13Ci2-PCB-101
13Ci2-PCB-169
13Ci2-PCB-208
13Ci2-PCB-28
13Ci2-PCB-104
13Ci2-PCB-180
13Ci2-PCB-209
13Ci2-PCB-37
Table A-3. Native Standard Spiking Solution (80 ng/mL)
PCB-1
PCB-52
PCB-105
PCB-169
PCB-3
PCB-54
PCB-110
PCB-177
PCB-4
PCB-64
PCB-118
PCB-180
PCB-8
PCB-66
PCB-126
PCB-187
PCB-11
PCB-70
PCB-132
PCB-188
PCB-15
PCB-74
PCB-138
PCB-189
PCB-18
PCB-77
PCB-147
PCB-199
PCB-19
PCB-85
PCB-149
PCB-202
PCB-28
PCB-95
PCB-153
PCB-205
PCB-31
PCB-99
PCB-155
PCB-206
PCB-3 7
PCB-101
PCB-156
PCB-208
PCB-44
PCB-104
PCB-166
PCB-209
Table A-4. Non-extracted Internal Standard Solution
(1,000 ng/mL)
C12-PCB-8
C12-PCB-79
C12-PCB-I62
PCB Congener Multi-lab Validation Study Plan
A-4
January 2018
-------
Table A-5. Retention Time Standards
Recommended Mixtures
#1
#2
#3
#4
#5
#6
#7
#8
#9
PCB-1
PCB-5
PCB-15
PCB-13
PCB-12
PCB-11
PCB-3 6
PCB-30
PCB-23
PCB-2
PCB-7
PCB-20
PCB-14
PCB-3 3
PCB-21
PCB-72
PCB-43
PCB-3 9
PCB-3
PCB-10
PCB-27
PCB-3 5
PCB-49
PCB-3 8
PCB-7 8
PCB-5 5
PCB-62
PCB-4
PCB-17
PCB-2 9
PCB-51
PCB-5 9
PCB-50
PCB-7 9
PCB-5 8
PCB-68
PCB-6
PCB-24
PCB-3 4
PCB-5 3
PCB-6 3
PCB-57
PCB-89
PCB-76
PCB-80
PCB-8
PCB-26
PCB-40
PCB-54
PCB-64
PCB-61
PCB-96
PCB-109
PCB-88
PCB-9
PCB-31
PCB-42
PCB-73
PCB-77
PCB-6 5
PCB-98
PCB-112
PCB-94
PCB-16
PCB-32
PCB-47
PCB-75
PCB-85
PCB-86
PCB-106
PCB-120
PCB-111
PCB-18
PCB-37
PCB-6 9
PCB-81
PCB-91
PCB-102
PCB-108
PCB-159
PCB-116
PCB-19
PCB-41
PCB-92
PCB-90
PCB-97
PCB-113
PCB-152
PCB-186
PCB-121
PCB-22
PCB-45
PCB-93
PCB-100
PCB-104
PCB-126
PCB-166
PCB-192
PCB-125
PCB-25
PCB-46
PCB-101
PCB-117
PCB-114
PCB-127
PCB-182
PCB-198
PCB-140
PCB-28
PCB-48
PCB-105
PCB-122
PCB-123
PCB-133
PCB-184
PCB-142
PCB-44
PCB-60
PCB-118
PCB-124
PCB-129
PCB-139
PCB-204
PCB-143
PCB-52
PCB-70
PCB-119
PCB-130
PCB-137
PCB-145
PCB-148
PCB-56
PCB-83
PCB-128
PCB-154
PCB-156
PCB-161
PCB-150
PCB-66
PCB-84
PCB-134
PCB-163
PCB-167
PCB-169
PCB-155
PCB-67
PCB-95
PCB-136
PCB-165
PCB-176
PCB-181
PCB-160
PCB-71
PCB-103
PCB-144
PCB-175
PCB-185
PCB-162
PCB-74
PCB-107
PCB-151
PCB-200
PCB-189
PCB-168
PCB-82
PCB-115
PCB-157
PCB-201
PCB-188
PCB-87
PCB-131
PCB-158
PCB-202
PCB-99
PCB-132
PCB-190
PCB-110
PCB-135
PCB-191
PCB-138
PCB-141
PCB-207
PCB-146
PCB-149
PCB-208
PCB-147
PCB-164
PCB-209
PCB-153
PCB-170
PCB-173
PCB-171
PCB-174
PCB-172
PCB-177
PCB-178
PCB-179
PCB-183
PCB-180
PCB-193
PCB-187
PCB-196
PCB-194
PCB-197
PCB-195
PCB-205
PCB-199
PCB-203
PCB-206
PCB Congener Multi-lab Validation Study Plan
January 2018
-------
Appendix B
Sample Types for the Validation Study
Sample Matrix Types to be Obtained for the Study
Matrix Type
# Matrices
Approximate Amount
# Aliquots of each
Aliquot size
Wastewater
9
120 liters each
120*
1-liter
Soil/sediment
3
800 grams each
80
10-grams
Biosolids
3
400 grams each
80
5-grams
Fish tissue
3
800 grams each
80
10-grams
Total # Samples
1,800
* The 120 wastewater aliquots will provide 80 aliquots for the separately funnel extraction analyses
(including 20 backup samples in the event of breakage or laboratory accidents) and an additional 40 aliquots
to support possible solid-phase extraction analyses by a subset of the study participants.
PCB Congener Multi-lab Validation Study Plan
B-l
January 2018
-------
Appendix C
Procedures for Derivation of QC Acceptance Criteria from the Validation Study Results
PCB Congener Multi-lab Validation Study Plan
C-l
January 2018
-------
The information below has been excerpted from Reference 8.9 and all citations to section numbers and
references apply to the original document, not this study plan. Not all of the calculations in the document
may apply to this specific method study. The calculations used for this study will be adjusted for the
actual number of laboratories involved.
Quality Control Acceptance Criteria Development for New Methods
Method Detection Limits and Minimum Levels
Each laboratory participating in the validation study must perform an MDL study as described in Section
3.1.1. The organization responsible for developing the new method must establish an MDL for the
method, using a pooled MDL from the at least six laboratories. A pooled MDL is calculated from m
individual laboratory MDLs by computing the square root of the mean of the squares of the individual
MDLs and multiplying the result by a ratio of ^-values to adjust for the increased degrees of freedom.
Note: The MDL values used in this calculation are those determined in each of the six or more
laboratories. If one laboratory reports an MDLS (from spiked samples), that value is used in
conjunction with the MDL values from the other laboratories, including any values reported as
MDLb (from blanks).
MDLpooled
where:
,di (MDL^1
V Ł(0.99,^)
)\d
\ c(0.99,d2) J
, fMDLLab m\
Ul \ t(0.99,dm) J
1 ^2 1
x Ł(0.99,[d1+d2+---dm]
m = The number of laboratories, and
di = The number of replicates used by Lab i to derive the MDL.
In the case of 9 laboratories with 7 replicates per laboratory, the equation simplifies to:
MDLpooled
M
MDL\abi + MDL\ab2 +ŚŚŚ MDL2Lab9 2.41
9 X 3.14
The organization responsible for developing the method must also use this MDL to develop an ML.
Procedures for determining the ML are given in Section 3.1.1.
Calibration Linearity
The instrument or analytical system is then calibrated with six standards specified in the method to
calculate an initial RSD for the response factor
The RSD and the RSD limit for the CF, RF, or RR is determined as follows:
1. Calculate the mean and standard deviation of the CFs, RFs, or RRs for each laboratory.
Łf=i Factor^
Mean Factor = Factor =
n
PCB Congener Multi-lab Validation Study Plan C-2 January 2018
-------
Y,f=1(Factori Factor)2
where:
Factor = The "Factor" terms are replaced by the CF, RF or RR terms, based on the quantitation
approach described in the method in question, and
n = The number of calibration points used in each laboratory.
2. Calculate the relative standard deviation of the CFs, RFs, or RRs of each laboratory and analyte as:
RSDi = 100 x Sl
Factort
where s and Factor are the standard deviation and mean of the CFs, RFs, or RRs for laboratory /.
3. Calculate the pooled RSD of the CFs, RFs, or RRs for each analyte from all laboratories. The pooled
RSD is calculated as the square root of the mean of the squares of the sample RSDs from each
individual laboratory. For example, for nine laboratories, the pooled RSD is calculated as:
RSD.
pooled
M
RSDl + RSD^ + RSDl
4. Calculate RSDmax as the smaller of 35% and:
RSDmax k(RSDp0oled)
where:
k = The square root of the 95th percentile of an F distribution with n- 1 degrees of freedom in the
numerator and m(n - 1) degrees of freedom in the denominator,
m = The number of laboratories, and
n = The number of calibration points.
For nine laboratories using a five-point calibration (m = 9,n = 5), the value of k is 1.6. The maximum
allowable specification for RSDmax is 35%.
Calibration Verification
As noted in Section 2.2., acceptance limits for calibration verifications can be determined in three
different ways, each of which is described below.
The calibration verification criterion may be specified as a maximum relative distance between the mean
CF, RF, or RR obtained by a future laboratory's initial calibration (Factor) and the CF, RF or RR
obtained from its calibration verification standard (FactorVER). The maximum allowable deviation is based
on the pooled relative standard deviation (RSDpooied) calculated in Section 3.2.2.
1. Determine kVER by multiplying the 97.5th percentile of a Student's t distribution with (m[n-l])
degrees of freedom times the square root of (1+1 In), where there are n points in the calibration and m
laboratories:
kvER t
1+ -
n
PCB Congener Multi-lab Validation Study Plan
C-3
January 2018
-------
For a five-point calibration, the Student's t value is 2.0, resulting in combined multipliers of 2.4 for a
three-point calibration, and 2.2 for a five-point calibration.
2. The calibration verification criterion for the new method would then be stated as the maximum
percent difference as follows:
where:
Factor = The "Factor" terms are replaced by the CF, RF or RR terms, based on the quantitation
approach described in the method in question, and
For example, if the calibration verification criterion, calculated as Icver RSDpooled, equals 17%, then the
difference between the Factor from the initial calibration and the FactorVER from the calibration
verification sample must be less than or equal to 17% of the Factor.
When using either the concentration or the recovery approach, the calculations are very similar to
those used for the "factor" limits shown above:
3. Calculate the upper and lower QC acceptance criteria for the known concentration of the analyte in
the calibration verification standard, using the lower and upper percentages calculated in Step 2
Lower limit = (Lower Percentage in Step 2) x (Known Concentration in Standard)
Upper limit = (Upper Percentage in Step 2) x (Known Concentration in Standard)
Alternatively, calibration verification criteria may be specified as the range of acceptable recoveries
calculated for the analytes in the calibration verification standard, using the lower and upper percentages
calculated in Step 2 above to create a window around 100% recovery.
Initial and Ongoing Precision and Recovery
For the IPR and OPR tests, QC acceptance criteria are calculated using the mean percent recovery and the
standard deviation of recovery from the IPR tests of four aliquots of the reference matrix and the OPR test
of one aliquot of the reference matrix (for a total of five samples) in nine laboratories. The QC
acceptance criteria are developed using the following steps:
1. Calculate the mean percent recovery (X) for each analyte, based on all data points from all
laboratories, the between-laboratory standard deviation (sb) of the mean results for each of the six or
more laboratories (standard deviation of the nine laboratory means X1 + X2 + ... Xg). and the pooled
within-laboratory standard deviation (sw). The value of s is calculated as the square root of the mean
of all within-laboratory variances. For example, for nine laboratories:
Factor>
Factor
Percent Difference = 100 x
pooled
Factor
above:
where:
Xj = The mean percent recovery for the jth laboratory
m = The number of laboratories, and
X = The overall mean of the percent recoveries from all laboratories
PCB Congener Multi-lab Validation Study Plan
C-4
January 2018
-------
S1 + S2 + S9
M
9
Note: CSRA will provide direction to the participating laboratories to ensure they are spiking IPR
and OPR samples at the same concentration.
2. QC acceptance criteria for IPR recovery - Calculate the combined standard deviation for
interlaboratory variability and estimation of the mean (sc) as:
3. Calculate the QC acceptance criteria for recovery in the IPR test by constructing a ą 2.3 sc window
around the mean percent recovery X, where 2.3 is the 97.5th percentile Student's t value for 10
degrees of freedom (an estimated degrees of freedom based on the variance ratios observed with EPA
Method 1625):
If more than 9 laboratories are used, the degrees of freedom for t will increase, but a complete
calculation is beyond the scope of this document. An approximation of degrees of freedom equal to
the number of laboratories will serve for most situations.
4. QC acceptance criterion for IPR precision - The maximum acceptable RSD for the four IPR aliquots
is approximated by a 95% upper confidence limit around the observed RSD of the results from all of
the laboratories. The RSDipr (computed as sw divided by X) is multiplied by the square root of a 95th
percentile /Ś' value with 3 degrees of freedom in the numerator and m (n - 1) degrees of freedom in the
denominator, where m = the number of laboratories, and n is the number of data points per laboratory.
For example, the resulting multiplier on the RSD for nine laboratories and five data points per
laboratory will then be 1.7, and the QC acceptance criterion for precision in the IPR test is calculated
as follows:
where:
m = the number of laboratories, and
n = the number of data points per laboratory.
For 9 laboratories and 5 data points per laboratory, the calculation becomes:
Lower limit (%) = X 2.3sc
Upper limit (%) = X + 2.3sc
Maximum RSDipr = (1.7)xRSDipr
5. QC acceptance criteria for OPR recovery - Calculate the combined standard deviation for
interlaboratory variability and estimation of the mean (sc) as:
PCB Congener Multi-lab Validation Study Plan
C-5
January 2018
-------
where:
m = the number of laboratories, and
n = the number of data points per laboratory.
For 9 laboratories and 5 data points per laboratory,
6. Calculate the QC acceptance criteria for recovery in the OPR test by constructing a ą 2.1 sc window
around the mean percent recovery X, where 2.1 is the 97.5th percentile Student's t value for 19
degrees of freedom (an estimated degrees of freedom based on the variance ratios observed with EPA
Method 1625):
Lower limit (%) = X 2.1 sc
Upper limit (%) = X + 2.1 sc
If more than 9 laboratories are used, the degrees of freedom for t will increase, but a complete
calculation is beyond the scope of this document. An approximation of degrees of freedom equal to
the number of laboratories will serve for most situations.
Matrix Spike and Matrix Spike Duplicate
Results of the MS/MSD analyses performed in the validation study are used to develop the MS/MSD QC
acceptance criteria. Calculate the MS/MSD performance criteria as follows:
1. Calculate the mean and sample standard deviation of the recoveries of each MS/MSD pair, and then
compute the overall mean recovery X, the between-laboratory standard deviation (sb) of the mean
results for each of the nine laboratories, and the pooled within-laboratory standard deviation (s) for
each target analyte using the MS and MSD analyses.
- *>2
where:
Xj = The mean percent recovery for the jth laboratory
m = The number of laboratories, and
X = The overall mean of the percent recoveries from all laboratories
In order to allow for interlaboratory variability, calculate the combined standard deviation (sc) for
interlaboratory variability and estimation of the mean as:
sr
1 Ś -
771
~h S^
+ 2 sw
where:
m = the number of laboratories.
For nine labs, this becomes:
sr =
PCB Congener Multi-lab Validation Study Plan
C-6
January 2018
-------
2. QC acceptance criteria for MS/MSD recovery - Calculate the QC acceptance criteria for recovery in
the MS/MSD test by constructing a ą 2.2sc window around the mean percent recovery (X) using the
combined standard deviation. This factor comes from a t value for an estimated 11 degrees of
freedom (based on this experimental design and variance ratios observed in Method 1625):
If more than 9 laboratories are used, the degrees of freedom for t will increase, but a complete
calculation is beyond the scope of this document. An approximation of degrees of freedom equal to
the number of laboratories plus 2 will serve for most situations.
Note: For highly variable methods, it is possible that the lower limit for recovery for both the IPR
and OPR analyses will be a negative number. In these instances, the data should either be
log-transformed and the recovery window recalculated, or the lower limit established as
"detected," as was done with some of the methods in 40 CFR Part 136, Appendix A.
3. QC acceptance criteria for MS/MSD relative percent difference (RPD) - To evaluate a 95% upper
confidence limit for precision, the RSD (computed using the pooled within-laboratory standard
deviation s of the MS/MSD samples, divided by X, is multiplied by the square root of the 95th
percentile F value with 1 degrees of freedom in the numerator and m degrees of freedom in the
denominator multiplied by the square root of 2 (i. e., V2), where m is the number laboratories. The
resulting multiplier on the RSD for 3 laboratories will then be 3.2. The QC acceptance criterion for
precision in the MS/MSD test (RPDmax) is calculated as follows:
Absolute and Relative Retention Time
Establishing QC acceptance criteria for RT and RRT precision is problematic when multiple laboratories
are involved because laboratories have a tendency to establish the chromatographic conditions that suit
their needs. Calculating mean RTs and RRTs based on different operating conditions will result in the
establishment of erroneously wide windows. Therefore, it is advised that the organization developing the
method specify to the participating laboratories the chromatographic conditions and columns to be used.
Any future laboratories operating under different conditions will need to develop new acceptance criteria
for RT and RRT precision.
Determine the mean retention time, RT (and/or the mean relative retention time RRT) and the standard
deviation (s) of the RT and/or RRT for each analyte and standard. Determine the upper and lower
retention time (or relative retention time) limits as follows:
where:
t = The 97.5th percentile of a t distribution with n - 1 degrees of freedom, and
n = The number of retention time or relative retention time values used.
The relative retention time upper and lower limits are determined by replacing RTwith RRT in the
equations above.
PCB Congener Multi-lab Validation Study Plan C-7 January 2018
Lower limit (%) = X 2.2sc
Upper limit (%) = X + 2.2sc
RPDmav = 3.2 RSD
max
Lower limit
-------
Blanks
Establish the QC acceptance criteria for blanks. The historical requirement has been that the
concentration of an analyte in a blank must be below the ML or below one-third (1/3) the regulatory
compliance level, whichever is higher. However, other limits (including those below the ML) may be
used for a specific method. In instances where the level of the blank is close to the regulatory compliance
level or the level at which measurements are to be made, it may be necessary to require multiple blank
measurements and establish the QC acceptance criteria based on the mean of the blank measurements
plus two standard deviations of the blank measurements.
Labeled Compound Recovery
The labeled compound recoveries from all of the samples analyzed in the validation study can be used to
develop the labeled compound QC acceptance criteria. Calculate the labeled compound performance
criteria as follows:
1. Calculate the mean and sample standard deviation of the recoveries of each labeled compound, and
then compute the overall mean recovery X, the between-laboratory standard deviation (sb) of the
mean results for each of the nine laboratories, and the pooled within-laboratory standard deviation
(s) for each labeled compound.
-n 2
Sb
" *)
M
m Ś
where:
Xj = The mean percent recovery for the jth laboratory
m = The number of laboratories, and
X = The overall mean of the percent recoveries from all laboratories
In order to allow for interlaboratory variability, calculate the combined standard deviation (sc) for
interlaboratory variability and estimation of the mean as:
sr =
where:
m = the number of laboratories.
1 + ^)sť2 + is"
For nine labs, this becomes:
sr
\
10\
9 )S
1+ [\
2. QC acceptance criteria for labeled compound recovery - Calculate the QC acceptance criteria for
recovery the labeled compounds by constructing a ą 2.2sc window around the mean percent recovery
(X) using the combined standard deviation. This factor comes from a t value for an estimated 11
degrees of freedom (based on this experimental design and variance ratios observed in Method 1625):
Lower limit (%) = X 2.2sc
Upper limit (%) = X + 2.2sc
If more than 9 laboratories are used, the degrees of freedom for t will increase, but a complete
calculation is beyond the scope of this document. An approximation of degrees of freedom equal to
the number of laboratories plus 2 will serve for most situations.
PCB Congener Multi-lab Validation Study Plan
C-8
January 2018
------- |