Analytical Feasibility Support
            Document
    for the Six-Year Review of
Existing National Primary Drinking
                 lulations
            :easibility for Chemical Contaminants

-------

-------
Office of Water
Office of Ground Water and Drinking Water (4607M)
EPA815-D-02-002
www.epa.gov/safewater
March 2002
                                          Printed on Recycled Paper

-------

-------
                                                           EPA815-D-02-002
           Analytical Feasibility Support Document
                    for the Six-Year Review of
  Existing National Primary Drinking Water Regulations

           (Reassessment of Feasibility for Chemical Contaminants)
                                March 2002
                  United States Environmental Protection Agency
                               Office of Water
                    Office of Ground Water and Drinking Water
                     Standards and .Risk Management Division
                         Targeting and Analysis Branch
                     1200 Pennsylvania Avenue, NW (4607M)
                            Washington, DC 20460
 This report is issued in support of the preliminary revise/not revise decisions for the Six-Year
 Review Notice of Intent, It is intended for public comment and does not represent final agency
policy. EPA expects to issue a final version of this report with the publication of the final notice
   in 2002, reflecting corrections due to public comment on the preliminary notice  and the
                            supporting documents.

-------

-------
Acknowledgments

   The USEPA staff involved in the development and/or review of either all or portions of this
document include Wynne Miller, Ed Glick, Pat Churilla, Jeanne Campbell, Richard. Reding,
Barry Lesnik, Terry Smith, Many Allen, Judy Lebowich and Marc Parrotta. EPA staff assisted in
one of several, areas that included:

   (1) developing and/or reviewing the process used to assess whether the feasible limits have
       changed for a subset of the 68 chemical SDWA analytes tinder review;
   (2) review and comment on the preliminary results;
   (3) general QA/QC to verify that the approved analytical methods and the method detection
       limits listed for each contaminant was correct; and/or
   (4) writing and editorial review of portions of this document,

   EPA acknowledges the National Drinking Water Advisory Council (NDWAC) Drinking
Water Committee for their recommendations on how EPA should review analytical feasibility.

   EPA also appreciates the technical, support provided by The Cadmus Group, Inc, the prime
contractor for this project.  Dr. George Hallberg served, as the Cadmus project manager and
major contributions by Wendy Chou, Ashton Koo and Anne Isham are gratefully acknowledged.
Their support in the development of this document included documentation of approved
analytical methods .and their MDLs, the analyses of the Water Supply Data, the development of
the graphs and tables, a large portion of the writing  in the results section, and. general QA/QC of
the analyses and the document.                                            .  .          '

-------

-------
                                Executive Summary       '


   The Safe Drinking Water Act (SDWA). as amended in 1996, requires the Environmental
Protection Agency (EPA) to review and revise, if appropriate, existing National Primary
Drinking Water Regulations (NPDWRs), As part of the review, EPA developed a protocol,
document (entitled EPA Protocol for the Review of Existing National Primary Drinking Water
Regulations') to describe the process and strategy for regulatory review that EPA used to meet its
statutory requirement. EPA developed the protocol based on recommendations from the
National Drinking Water Advisory Council (NDWAC), through internal Agency deliberations,
and through discussions with the diverse stakeholders involved in drinking water and its
protection.  Based on the NDWAC recommendations, EPA's review included the consideration
of five key elements, as appropriate: 'health effects, analytical and treatment feasibility,
implementation-related issues, occurrence and exposure, and economic impacts. The purpose of
the analytical methods feasibility analysis was to determine whether changes in the practical
quantitation level (PQL) were possible in those instances where the Maximum Contaminant
Level is limited, or might be limited, by analytical feasibility. This document, "Analytical
Feasibility Support Document for the Six-Year Review of Existing National Primary Drinking
Water Regulations: Reassessment of Feasibility for Chemical Contaminants,1' describes the
process recommended by NDWAC and. used by EPA to address the analytical feasibility aspect
of the current (1996-2002) Six-Year Review.

    To be consistent with the accepted policy and procedures used by EPA to derive quantitation
levels for drinking water contaminants, the Six-Year Review focused on the process that has
been used by the Office of Ground Water and Drinking Water for many years. Historically,
EPA's OGWDW used two main approaches to determine.practical quantitation levels (or PQLs)
for SDWA analytes.  One approach (and the preferred approach) used data from Water Supply
(WS) Performance Evaluation (PE) studies. Although the primary use of the WS-PE data was
for EPA's laboratory certification, the data were also used. as.a secondary data source for many
years to develop PQLs when the spike concentrations were in the appropriate concentration
range. The derivation of the PQL.using WS data involved determining the concentration of an
analyte at which 75 percent of EPA'Regional and State laboratories achieved results within a
specified range around the spike value. In the absence of WS data, the other approach that EPA
used was the MDL multiplier method. In mis approach, the PQL was calculated by multiplying
.the EPA-derived MDL by a factor of 5 or 10.  The MDL multiplier method was mostly used in
the early years of rule development for NPDWRs when insufficient WS data were available.
Once sufficient WS data became.available, most of the PQLs developed using the MDL
multiplier were validated using WS data.

    For the Six-Year analytical feasibility review, EPA focused on assessing whether the
practical quantitation level has changed since promulgation for a subset of the 68 chemical
NPDWRs. EPA performed the analytical feasibility analysis for a total of 38 NPDWRs that fell
into one of two categories:

     »  First, for those contaminants where the MCL is currently limited by analytical feasibility
       (i.e., the MCL is set at the PQL) and the MCLG is'still appropriate, EPA evaluated the
       currently approved methods for those contaminants and available WS data to determine

-------
       whether it might be possible to lower the PQL and hence set an MCL that is closer to the
       MCLG.

     >  The second circumstance under which EPA re-evaluated the PQL was for contaminants
       identified under the Six-Year health effects technical review as having potential changes
       to their MCLG. Because the information for the health effects review was not completely
       available at the time the analytical methods analysis began, EPA took a broad-brush
       approach and included a number of contaminants that may not have needed a
       reassessment of their analytical feasibility,

    For each of these 38 chemical NPDWRs, the analytical feasibility reassessment included:

    (1) a methods comparison step to help identify whether the ability to detect (and therefore
       quantify) these contaminants at lower levels has increased;

    (2) a methods usage over time step to identify the analytical methods that appear to be the
       most widely used for the analysis of particular contaminants.

    (3) a Water Supply data analysis step to determine if a PQL can be recalculated (if sufficient
       WS information is available) or if there is an indication that a PQL may be lower using
       the available information.

    The results of these three steps aided in assessing whether a PQL might change for a specific
contaminant and, if so, estimating what the new PQL might be,

    The results of Six-Year analytical feasibility review concluded that the majority of the
available WS data were insufficient to actually recalculate the PQL for many of the 38
contaminants of interest. The data were considered insufficient because either the true value of
the spike concentrations used in the WS studies were above the concentration of interest and/or
the percentages of labs passing exceeded the 75 percent criterion used to calculate a PQL.
However, for many of the 38 contaminants, the available data were sufficient to indicate whether
the PQL  might change or if the current PQL is still appropriate. Of the 38 NPDWRs evaluated,  •
the available information indicates that the PQL for 23 might possibly be lower. The PQL for the
remaining 15 appears to still be appropriate.  For the 23 analytes where the WS data indicate that
a lower PQL may exist, EPA used the information about method usage over time, the MDLs for
these methods, and the 10 x MDL multiplier to estimate what the potentially lower  PQL might
be. This estimated value was used as a threshold value in the occurrence and exposure analyses
to determine whether an improvement in public health protection might be possible if EPA were
to consider gathering more definitive data to recalculate the PQL and possibly lower the MCL.
                                          IV

-------
                                 Table of Contents

Acknowledgments	,'.	  ii

Executive Summary	,	,	Hi

List of Tables	xi

List of Figures	xiii

I.   Introduction;	 1

II.  Background 	'.,.,,..	 2
    A. What is the Relationship Between SDWA Requirements and Analytical Methods?  ...... 2
    B. How Have PQLs Been Determined in the Past for SDWA Contaminants? .-	 3
      1.  How Were Water Supply Studies Conducted?	3
      2.  What Criteria Are Used to Determine a PQL?	:., .4

III.   How Did EPA identify Which Contaminants to Evaluate for a PQL Reassessment? .... 5

IV.   What Approaches Were Used to Reassess, the PQLs of Contaminants Identified by the
      Six-Year Review Process? ..,,..,..,-..,.,,.....,,	 8

V.  Results of the PQL Reassessment	11

Alachlor ...	„,..,..,,	11
    Results of the Method Comparison	.,.,.,..	 11
    Results of the Analysis of the. WS Data	.....	..	12
      a.  Method Usage Over Time	12
     •: b.  Results of the PQL Analysis	,	 . 12
    Conclusion for Alachlor	".	•..,...	14

Benzene	...,..,,	'..,.,.	14
    Results of the Method Comparison	14
    Results of the Analysis of the WS Data	•.	15
      a.  Method Usage Over Time	,.,	'..,15
      b.  Results of the PQL Analysis	16
    Conclusion for Benzene ,.„..,„.,.,.	,, 17

Benzo(a)pyrene ...	,			.,..,.,."	1.8
    Results of the Method Comparison. ,,.„..,,,.,,-,..,	18
    Results of the WS Data Analysis			.-	.18
      a.  Method Usage Over Time	,	18
      b.  Results of the PQL Analysis	,19
    Conclusion for Benzo(a)pyrene	'.	21

Beryllium ..,...,'.	'.	21"
    Results of the Method Comparison  , ^ .,.,..•;,.,,. ..„•	2.1
                                        .v

-------
   Results of the Analysis of the WS Data	22
      a.  Method Usage Over Time	.	22
      b.  Results of the PQL Analysis	,	23
   Conclusion for Beryllium	 25

Bis(2-ethylhexyl)phthalate	,,..,,	25
   Results of the Method Comparison	25
   Results of the Analysis of the WS Data	,	,	26
      a.  Method Usage Over Time	26
      b.  Results of the PQL Analysis 	,	27
   Conclusion for Bis(2-ethylhexyl)phthalate 	.	,.... 29

Cadmium	•	.„	29
   Results of the Method Comparison	;	,	29
   Results of the Analysis of the WS Data	 30
      a.  Method Usage Over Time ,	30
      b.  Results of the PQL Analysis	,. t	31
   Conclusion for Cadmium	32

Carbofuran	,	33
   Results of the Method Comparison	 33
   Results of the Analysis of the WS Data	,	33
      a.  Method Usage Over Time	, 33
      b.  Results of the PQL Analysis	34
   Conclusion for Carbofuran	.35

Carbon Tetrachloride	,	36
   Results of the Method Comparison .'.	36
   Results of the Analysis of the WS Data	 37
      a.  Method Usage Over Time	;	....,.,, 37
      b.  Results of the PQL Analysis			,	37
   Conclusion for Carbon Tetrachloride	,	38

Chlordane	,	39
   Results of the Method Comparison	 39
   Results of the Analysis of the WS Data	40
      a.  Method Usage Over Time	40
      b.  Results of the PQL Analysis	 40
   Conclusion for Chlordane	,	,	42

Chromium	42
   Results of the Method Comparison	.42
   Results of the Analysis of the WS Data	,_..	43
      a.  Method Usage Over Time	 43
      b.  Results of the PQL Analysis	:	,. 44
   Conclusion for Chromium	46

l,2-Dibromo-3-chloropropane (DBCP)	,	,		... 46
                                        VI

-------
   Results of the Method Comparison.	,....'	46
   Results of the Analysis of the WS Data	',.	..47
       a.  Method Usage Over Time ..........	-.	47
       b.  Results of the PQL Analysis	,	 47
   Conclusion for DBCP ,	,,,.,,	..;........	48

1,4-Dichlorobenzene 	,	49-
 .  Results of the Method Comparison	.........'	,.;,...49
   Results of the Analysis of the WS Data		,		... 50
       a.  Method Usage Over Tune ,	,	'	50
       b,  Results of the PQL Analysis	,	 50
   Conclusion for 1,4-Dichlorobenzene	51

1,2-Dichloroethane	»	,	,	52
   Results of the Method Comparison	,	,	52
   Results of the Analysis of the WS Data .,	,	 53
       a.  Method Usage Over Time ...,,	53
    .   b.  Results of the PQL Analysis	,	53
   Conclusion for 1,2-Dichloroethane	.,.,.,	54

l?l-Dichloroethylene	,	,	, 55
   Results of the Method Comparison	..	..,,,..,......•	 55
   .Results of the Analysis of the WS Data	56
       a.  Method Usage Over Time	,.,,.,.,..,	'.. 56
       b.  Results of the PQL Analysis	 56
 :  Conclusion for 1,1-Dichloroethylene	,	 58

Dichioromethane (methylene chloride)	58
   Results of the Method Comparison	 58
   Results of the Analysis of the WS Data	59
       a.  Method Usage Over Time	,	59
       b.  Results of the PQL Analysis	,	.60
   Conclusion for Diehloroniethane „.,,..„..„...„,	,	61

1,2-Dichloropropane	,	61
   Results of the Method Comparison	61
   Results of the Analysis of the WS Data ...........	,		62
       a.  Method Usage Over Time ,.'.............'.	,	62
       b.  Results of the PQL Analysis		.63
   Conclusion for 1,2-Dichloropropane	64

2,3,7,8-TCDD (Dioxin)	 65
   Results of the Method Comparison	65
   Results of the Analysis of the WS Data	.		... 65
   Conclusion	.65

Diquat	,	.66
   Results of the Method Comparison	66
   Results of tine Analysis of the WS Data	.67
       a.  Method Usage Over Time	67
       b.  Results of the PQL Analysis ..........	•	67
                                        VII

-------
    Conclusion for Diquat	.	68

 Ethylene Dibromide	-	69
    Results of the Method Comparison	69
    Results of the Analysis of the WS Data	 70
       a.  Method Usage Over Time	'	70
       b.  Results of the PQL Analysis	70
    Conclusion for Ethylene Dibromide	 72

 Fluoride	•	72
    Results of the Method Comparison	.	72
    Results of the Analysis of the WS Data			74
       a.  Method Usage Over Time ..,,	74
       b.  Results of Hie PQL Analysis	,	75
    Conclusion for Fluoride	 76

 Glyphosate	77
    Results of the Method Comparison	,	 77
    Results of the Analysis of the WS Data	,	78
       a.  Method Usage Over Time	.78
       b.  Results of the PQL Analysis	78
    Conclusion for Glyphosate	79

 Heptachlor	-	80
    Results of the Method Comparison	.80
    Results of the Analysis of the WS Data ..."	 81
       a.  Method Usage Over Time	 81
       b.  Results of the PQL Analysis	81
    Conclusion for Heptachlor	 83

, Heptachlor Epoxide	<	 - -"	83
    Results of the Method Comparison	 83
    Results of the Analysis of the WS Data			84
       a.  Method Usage Over Time 	84
       b.  Results of the PQL Analysis  		85
    Conclusion for Heptachlor Epoxide ...	87

 Hexachlorobenzene	8.7
    Results of the Method Comparison	.87
    Results of the Analysis of the WS Data	,		88
       a.  Method Usage Over Time	88
       b.  Results of the PQL analysis	89
    Conclusion for Hexachlorobenzene	90

 Hexachlorocyclopentadiene ....:	91
    Results of the Method Comparison		.		,	,, 91
    Results of the Analysis of the WS Data	,92
       a.  Method Usage Over Time	92
      'b.  Results of the PQL Analysis	,	92
    Conclusion for Hexachlorocyclopentadiene	94
                                        vui

-------
Mercury  	,	 94
   Results of the Method Comparison ..,.,..,,.	 94
   Results of the Analysis of the WS Data ... •		95
      • a.  Method Usage Over Time	95
       b.  Results of the PQL Analysis	:		.. 96
   Conclusion for Mercury	98

Methoxychlor	..'....	;	,98
   Results of the Method Comparison	....'.	98
   Results of the Analysis of the WS Data ;	,....._	.99
       a.  Method Usage Over Time	,	 99
       b.  Results of the PQL Analysis	.100
   Conclusion for Methoxychlor	,.,,.,	 101

Oxarnyl	;	102
   Results of the Method Comparison	,,..'	 102
   Results of the Analysis of the-WS Data	.	 103
       a.  Method Usage Over Time	103
       b.  Results of the PQL Analysis	,303
   Conclusion for Oxamyl	107

PCBs	:	-.'	107
   Results of the Method Comparison ...	107
   Results of the Analysis of the WS Data			108
       a.  Method Usage Over Time	.......,;,..„	 108
       b.  Results of the PQL Analysis	 108
   Conclusion for PCBs	.. 109

•Pentaehlorophenol	,..,..,	,, -	 110
   Results of the Method Comparison	i	......:...-	.110
   Results of the Analysis of the WS Data .	.,:.:,..			Ill
  :     a.  Method Usage Over Time ,	 Ill
       b.  Results of the PQL Analysis ,,......,..,	 112
   Conclusion for Pentaehlorophenol	,	 114

Picloram	,,........'	•«• •.	'.....,.	114
   Results of the Method Comparison .,	.,	.-....,...-... 114
   Results of the Analysis of the WS Data			'.. 115
       a.  Method Usage Over Time,	,	115
       b.  Results of the PQL Analysis	'.....	116
   Conclusion for Picloram	 117

Tetrachloroethylene	,.........,,	,	,118
   Results of the Method Comparison	,	,118
   Results of the Analysis of the WS Data			119
       a.  Method Usage Over Time	;	'	119
       b.  Results of the PQL Analysis		v.	120
   Conclusion for Tetrachloroethylene	,	, 121

Thallium . ..,	 121
   Results of the Method Comparison	121

-------
   Results of the Analysis of the WS Data	 •.,	 122
      a. Method Usage Over Time	122-
      b. Results of the PQL Analysis	-. 123
   Conclusion for Thallium	• • • •	124

Toxaphene	• • •' 125
   Results of the Method Comparison	 125
   Results of the Analysis of the WS Data			126
      a. Method Usage Over Time 	,. .•	• • •. 126
      b. Results of the PQL Analysis ...../	127
   Conclusion for Toxaphene	<	 129

1,1,1-Trichloroethane	,	 129
   Results of the Method Comparison	<	• • 129
   Results of the Analysis of the WS Data	.130
      a. Method Usage Over Time	 130
      b. Results of the PQL Analysis		..131
   Conclusion for 1,1,1-Trichloroethane  			133

1,1,2-Trichloroethane	 -	133
   Results of the Method Comparison	133
   Results of the Analysis of the WS Data		134
      a. Method Usage Over Time	—	 134
      b. Results of the PQL analysis	.,	,		135
   Conclusion for 1,1,2-Trichloroethane	»....'	136

Trichloroethylene	137
   Results of the Method Comparison	137
   Results of the Analysis of the WS Data	138
      a. Method Usage Over Time	138
      b. Results of the PQL Analysis	138
   Conclusion for Trichloroethylene	139

Vinyl Chloride	140
   Results of the Method Comparison	-... 140
   Results of the Analysis of the WS Data	,	141
      a. Method Usage Over Time	„..,..,,..141
      b. Results of the PQL Analysis	..		.141
   Conclusion for Vinyl Chloride	 143

VI.   Conclusion	143

References	,	146

Appendix A	.,..,,	147
Methods Support Document for Six-Year Review
Draft - March 2002

-------
                                    List of Tables

Table 1.   SD WA Chemical Contaminants Undergoing Analytical Methods/PQL Reassessment 7
Table 2.   Results of the Analytical Methods Comparison for Alachlor	11
Table 3.   Evaluation of Alachlor Data from WS Studies	•	13
Table 4.   Results of the Analytical Methods Comparison for Benzene	14
Table 5.   Evaluation of Benzene Data from WS Studies		16
Table 6.  . Results of the Analytical Methods Comparison for Benzo(a)pyrene	 18
Table 7.   Evaluation of PE Data for Benzo(a)pyrene from WS Studies	.	.......... 19
Table 8.   Results of the Analytical Methods Comparison for Beryllium	,'22
Table 9,   Evaluation of Beryllium Data from WS Studies	23
Table 10.  Results of-the Analytical Methods Comparison for Bis(2-ethylhexyl)phthalate .... 26
Table 11.  Evaluation of Bis(2-ethylhexyl)phthalate Data from WS Studies	 27
Table 12,  Results of the Analytical Methods Comparison for Cadmium	30
Table 13.  Evaluation of Cadmium Data from WS'Studies				... 31
Table 14.  Results of the Analytical Methods Comparison for Carbofuran	.,.,.,,,,.., 33
Table 15.  Evaluation of Carbofuran Data from WS" Studies	 34 .
Table 16.  Results of the Analytical Methods Comparison for Carbon Tetraehloride	.36
Table 17.  Evaluation of Carbon Tetraehloride Data from WS Studies .•...-	,.	38
Table 18.  Results of the Analytical Methods Comparison for Chlordane	'	39
Table 19.  Evaluation of Chlordane Data from WS Studies		..	41
Table 20.  Results of the Analytical Methods Comparison for Chromium 		43
Table 21.  Evaluation of Chromium Data from WS Studies	45
Table 22.  Results of the Analytical Methods Comparison for DBCP  	46
Table 23.  Evaluation of DBCP Data from WS Studies	,	48
Table 24.  Results of the Analytical Methods Comparison for 1,4-Dichlorobenzene.	 49
Table 25.  Evaluation of l?4-Diehlorobenzene Data from WS Studies	51
Table 26.  Results of the Analytical Methods Comparison for 1,2-Dichloroethane .......... 52
Table 27.  Evaluation of 1,2-Dichloroethane Data from WS Studies	 54
Table 28.  Results of the Analytical Methods Comparison for 1,1-Dichloroethylene	55
Table 29.  Evaluation of 1,1-Dichloroethylene Data from WS Studies	57
Table 30.  Results of the Analytical Methods Comparison for Dichloromethane 	58
Table 31.  Evaluation of Dichloromethane Data from WS Studies 	60
Table 32,  Results of the Analytical Methods Comparison for 1,2-Dichloropropane	61
Table 33.  Evaluation of 1,2-Dichloropropane Data from WS Studies	64
Table 34.  Results of the Analytical Methods Comparison for 2,3,7,8-TCDD	65
Table 35.  Results of the Analytical Methods Comparison for Diquat	.,->. 66
Table 36,  Evaluation of Diquat Data from WS Studies	".68
Table 37.  Results of the Analytical Methods Comparison for Ethylene Dibromide	69
Table 38.  Evaluation of Ethylene Dibromide Data from WS Studies	71
Table 39.  Results of the Analytical Methods Comparison for Fluoride 	72
Table 40.  Evaluation of Fluoride Data from WS Studies.	75
Table 41.  Results of the Analytical Methods Comparison for Glyphosate	77
Table 42.  Evaluation of Glyphosate Data from WS Studies	79
Table 43.  Results of the Analytical Methods Comparison for Heptachlor	80
Table 44.  Evaluation of Heptachlor Data from WS Studies  	,	82
Table 45.  Results of the Analytical Methods Comparison for Heptachlor Epoxide	84
Table 46.  Evaluation of Heptachlor Epoxide Data from WS Studies	86
Table 47.  Results of the Analytical Methods Comparison for Hexachlorobenzene 	88
Table 48.  Evaluation of PE Data for Hexachlorobenzene from WS Studies	.'	90
Table 49.  Results of the Analytical Methods Comparison for Hexachlorocyclopentadiene ... 91
Methods Support Document-for Six-Year Review
XI,
Draft -March 2002

-------
 Table 50.  Evaluation of Hexachlorocyclopentadiene Data from WS Studies	93
 Table 51.  Results of the Analytical Methods Comparison for Mercury	95
 Table 52.  Evaluation of Mercury Data from WS Studies  ,,	,	97
 Table 53.  Results of the Analytical Methods Comparison for Methoxychlor	99
 Table 54.  Evaluation of Methoxychlor Data from WS Studies   	,	100
 Table 55.  Results of the Analytical Methods Comparison for Qxamyl	102
 Table 56.  Evaluation of Oxamyl Data from WS Studies	 104
 Table 57.  Regression Results for Oxamyl	105
 Table 58.  Results of the Analytical Methods Comparison for PCBs	 108
 Table 59.  Evaluation of PCBs Data from WS Studies	109
 Table 60.  Results of the Analytical Methods Comparison for Pentachlorophenol  ......... 110
 Table 61.  Evaluation of Pentachlorophenol Data from WS Studies	113
 Table 62.  Results of the Analytical Methods Comparison for Picloram	115
 Table 63.  Evaluation of Picloram Data from WS Studies	117
 Table 64.  Results of the Analytical Methods Comparison for Tetrachloroethylene  	118
 Table 65..  Evaluation of Tetrachloroethylene Data from WS Studies		120
 Table 66.  Results of the Analytical Methods Comparison for Thallium	122
 Table 67.  Evaluation of Thallium Data from WS Studies	123
 Table 68.  Results of the Analytical Methods Comparison for Toxaphene  	126
 Table 69.  Evaluation of Toxaphene Data from WS Studies	128
 Table 70.  Results of the Analytical Methods Comparison for 1,1,1-Trichloroethane	130
 Table 71.  Evaluation of 1,1,1-Trichloroethane Data from WS Studies		132
 Table 72.  Results of the Analytical Methods Comparison for 1,1.2-Triehloroethane	134
 Table 73.  Evaluation of 1,1,2-Trichloroethane Data from WS Studies ,...	136
 Table 74.  Results of the Analytical Methods Comparison for Trichloroethylene	137
 Table 75.  Evaluation of Trichloroethylene Data from WS Studies	139
 Table 76.  Results of the Analytical Methods Comparison for Vinyl Chloride  ...	,	140
 Table 77.  Evaluation of Vinyl Chloride Data from WS Studies	142
 Table 78.  Summary of Results from the Methods Comparison and WS Analysis	144
 Table 79.  Estimated PQLs Based on Method Usage and 10 x MDL Multiplier Estimated values
           to itse in the Occurrence and Exposure. (O/E) analyses	147
Methods Support Document for Six-Year Review      xii
Draft - March 2002

-------
                                     List of Figures
 Figure 1,   Overview of the Protocol for the Revise/Not Revise
           Analytical Feasibility was Re-evaluated.  ........
 Figure 2.   Distribution of Analytical Techniques by WS Study:
 Figure 3.   Distribution of Analytical Techniques by WS Study:
 Figure 4,   Two-part Distribution of Benzene WS Data	
 Figure 5.   Distribution of Analytical Techniques by.WS Study:
 Figure 6.   PQL Evaluation from PE WS Data:.Benzo(a)pyrene
 Figure 7.   Distribution of Analytical Techniques by WS Study:
 Figure 8,   Two-part Distribution of Beryllium WS Data:	
 Figure 9.   Distribution of Analytical Techniques by WS Study:
 Figure 10.  Two-part Distribution of Bis(2-ethylhexyl)phthalate
 Figure 11.  Distribution of Analytical Techniques by WS Study:
 Figure 12,  Distribution of Analytical Techniques by WS Study:
 Figure 13.  Distribution of Analytical Techniques by WS Study:
 Figure 14.  Distribution of Analytical Techniques by WS Study:
 Figure 15.  Distribution of Analytical Techniques by WS Study:
 Figure 16.  Distribution of Analytical Techniques by WS Study:
 Figure 17.  Distribution of Analytical Techniques by WS Study:
 Figure 18.  Distribution of Analytical Techniques by WS Study:
 Figure 19..  Distribution of Analytical Techniques by WS Study:
 Figure 20.  Distribution of Analytical Techniques by WS Study:
 Figure 21.  Distribution of Analytical. Techniques by WS Study:
 Figure 22.  Distribution of Analytical Techniques by WS Study:
 Figure 23.  Distribution of Analytical Techniques by WS Study:
 Figure 24.  Distribution of Analytical Techniques by WS Study:
 Figure 25.  Distribution of Analytical Techniques by WS Study:
 Figure 26.  Distribution of Analytical Techniques by WS Study:
 Figure 27.  Distribution of Analytical Techniques by.WS Study:
 Figure 28.  Distribution of Analytical Techniques by WS Study:
 Figure 29.  Distribution of Analytical Techniques by WS Study:
 Figure 30,  Distribution of Analytical Techniques by WS Study:
 Figure 31.  Distribution of Analytical Techniques by WS Study:
 Figure 32.  Distribution of Analytical Techniques by WS Study:
 Figure 33.  PQL Evaluation of PE WS Data: Oxamy]	
 Figure 34.  Two-part Distribution of Oxamyl WS Data .......
 Figure 35.  Distribution of Analytical Techniques by WS Study:
 Figure 36.  Evaluation of PE WS Data: Pentachloropheriol
 Figure 37.  Distribution of Analytical Techniques by WS Study:
 Figure 38.  Distribution of Analytical Techniques by WS Study:
 Figure 39.  Distribution of Analytical Techniques by WS Study:
 Figure 40.  Distribution of Analytical Techniques by WS Study:
Figure 41.  Distribution of Analytical Techniques by WS Study:
Figure 42.  Distribution of Analytical Techniques, by WS Study:
Figure 43.  Distribution of Analytical Techniques by WS Study:
Figure 44.  Distribution of Analytical Techniques by WS Study:
               Decision with a Focus on Where
               	-.6
               Alachlor		'	,12
               Benzene	 15
               	17
               Benzo(a)pyrene	19
                	21
               Beryllium	,		23
               	..:.....	25
               Bis(2-ethylhexyl)phthalate  .. 27
               WS Data	29
               Cadmium	,. .31
               Carbofuran 	34
               Carbon Tetrachloride	37
               Chlordane	40
               Chromium	44
               DBCP	47
                1,4-Dichlorobenzene	'50
                1,2-DIchloroethane ...'	53
                1,1-Dichloroethylene	, 56
               Dichloromethane .....,..., 59
                1,2-Dicbloropropane ....... 63
               Diquat	 67
               Ethylene Dibromide ....... 70
               Fluoride	74
               Giyphosate	78
               Heptachlor	81
               Heptachlor Epoxi.de 	85
               Hexachiorobenzene	89
               Hexachlorocyclopentadiene  . 92
               Mercury	96
               Methoxychlor •	100
               Oxamyl	103
               	"........,...,,....106
               	..,.107
               Pentachiorophenol....,,.. 112
               	114
               Piclorani  	116
               Tetrachloroethylene ....... 119
               Thallium  	123
               Toxaphene ....,,..,...,. 127
               1,14-Trichloroethane  ..... 131
               1,1,2-Trichloroethane  ..... 135
               Trichloroethylene  	138:
               Vinyl Chloride :	141
Methods Support Document for Six-Year Review
xm
Draft - March 2002

-------

-------
                     Analytical Feasibility Support Document for the
       ' Six-Year Review of Existing National Primary Drinking Water Regulations
                  (Reassessment of Feasibility for Chemical Contaminants)

I.  Introduction

    The Safe Drinking Water Act (SDWA)., as amended in 1996, "requires the Environmental
Protection Agency (EPA) to review and revise, if appropriate, existing National Primary
Drinking Water Regulations (NPDWRs),  As part of the review, EPA developed a protocol
document (EPA Protocol for the Review of Existing National Primary Drinking Water
Regulations) that describes the process and strategy EPA used to review existing NPDWRs in
order to meet its statutory requirement  EPA developed the protocol document based 011
recommendations from the National Drinking Water Advisory Council (NDWAC), through.
internal Agency deliberations, and through discussions with the diverse stakeholders involved in
drinking water and its protection.  To more efficiently utilize limited resources, EPA performed a
series of analyses that were intended to target those NPDWRs that are the most appropriate
candidates for revision. As part of the review, and where appropriate, EPA reviewed the
following key technical elements to make decisions regarding regulatory changes: health risks
assessments; technology assessments (analytical feasibility and treatment technology); other
regulatory revisions (e.g.. monitoring and reporting): occurrence and exposure analyses; and
available economic information.  This document discusses the analytical feasibility aspect of the
current (1996-2002) Six-Year Review.

   'The 1999-2002 Six-Year Review includes the review of 68 chemical NPDWRs promulgated
prior to the 1996 SDWA Amendments.  Because the analytical measurement feasibility may have
been the limiting factor in setting the Maximum Contaminant Level (MCL) for some of the
existing NPDWRs or because the health, effects reviews may indicate a potential change in the
MCLG. this report examines the reassessment of analytical methods capabilities including a
reassessment of whether the Practical Quantitation Levels (PQLs) may have changed since
promulgation. The PQL is generally defined as "the lowest level that can be reliably achieved
within specified limits of precision and accuracy during routine laboratory operating conditions"
(50 FR 46906, November 13, 1985).  The purpose of this support document is to:

    *   provide background information on the relationship between SDWA requirements and
       the analytical methods feasibility;
    •  'describe how PQLs have historically been determined;
    »   and. describe the process used to identify which of the 68 chemical NPDWRs under the
       1996-2002 review are subject to a further assessment with regards to analytical methods
       capabilities and a reassessment of the PQL.
Methods Support Document for Six-Year Review
Draft - March 2002

-------
II. Background

A. What is the Relationship Between SDWA Requirements and Analytical Methods?

    The SDWA [§1401(l)(C)(i); 42 U.S.C. § 300f(l)(C)(i)] states that an MCL for a national
primary drinking water regulation is set "if, in the judgment of the Administrator., it is
economically and technologically feasible to ascertain the level of such contaminant in water in
public water systems." According to SDWA, NPDWRs include "criteria and procedures to
assure a supply of drinking water which dependably complies with such maximum contaminant
levels; including accepted methods of quality control and testing procedures to insure compliance
with such levels" [§1401(1)(D); 42 U.S.C.  § 300f(l)(D)l- Except in certain circumstances, EPA
is to set the MCL as close to the Maximum Contaminant Level Goal (MCLG) as is feasible with
the best available technologies (Section 1412 (b)(4)(B))> The MCLs for several SDWA
contaminants were set due to the limits of the analytical feasibility at that time. Since the
promulgation of pre-1996 SDWA NPDWRs, newer analytical methods and updated methods for
measuring SDWA contaminants have been approved. The approval of newer analytical
techniques may have provided laboratories with the analytical capability to measure some
contaminants at lower levels. In addition, some laboratories may have improved in their ability
to measure at lower levels using the same methods that were originally promulgated.

    In considering analytical methods for use in compliance monitoring, EPA evaluates the
overall sensitivity of the techniques. In previous regulations, EPA used  two measures of
analytical capability, the Method Detection Limit (MDL) and the Practical Quantitation Level
(PQL).

>•   The MDL is a measure of method sensitivity. The MDL is defined at 40 CFR Part 136
    Appendix B as "the minimum concentration of a substance that can be reported with 99%
    confidence that the analyte concentration is greater than zero, "  MDLs can be operator,
    method, laboratory, and matrix-specific. Due to normal day-to-day and run-to-run analytical
    variability, MDLs may not be reproducible within a laboratory or between  laboratories. The
    regulatory significance of the MDL is that EPA uses the MDL to determine when a
    contaminant is deemed to be detected and it can be used to calculate a PQL for that
    contaminant.

>   In the preamble to a November 13,1985 rulemaking (50 FR 46906), the PQL-was defined as
    "the lowest concentration of an analyte that can he reliably measured within specified limits
    of precision and accuracy during routine laboratory operating conditions.  " The Agency has
    used the PQL to estimate or evaluate the irdnimum concentration at  which most laboratories
    can be expected to reliably measure a specific chemical contaminant during day-to-day
    analyses of drinking water samples. The PQL is a means of integrating information on the
    performance of the approved analytical methods into the development of a drinking water
    regulation (52 FR 25699, July 8,1987). The PQL incorporates the following (50 FR 46880,
    November 13, 1985; 52 FR 25690, July 8,1987; 54 FR 22062, May 22, 1989):

      •  quantitation,
      •   precision and bias,
      •  normal operations of a laboratory, and
Methods Support Document for Six-Year Review
Draft - March 2002

-------
       •  the fundamental need to have a sufficient number of laboratories available to conduct
          compliance monitoring analyses.

    In some cases, the quantitation limit for a particular analyte may have been the limiting factor
in the determination of the MCL for that analyte.  This could be especially true for contaminants
with MCLGs of zero. Also, there are several SDWA contaminants with non-zero MCLGs.that
have their MCL set at the PQL, •

B.  How Have PQLs Been Determined in the Past for SDWA Contaminants?

    Historically, EPA's OGWDW used two main approaches to determine a PQL for SDWA
analytes. One approach (and the preferred approach) used data from Water Supply (WS)
Performance Evaluation (PE) studies.  Although the primary use of the WS-PE data was for
EPA's laboratory certification, the data were also used as a secondary data source for many years
to develop PQLs when the spike concentrations were in the appropriate concentration range, The
derivation of the PQL using WS data involved determining the concentration of an analyte at
which 75 percent of EPA Regional and State laboratories achieved results within a specified
range around the spike value. In the absence of WS data, the other approach that EPA used was
the MDL multiplier method. In this approach, the PQL was calculated by multiplying the EPA-
derived MDL by a factor of 5 or 10, The 5 or 10 multiplier was used to account for the
variability and uncertainty mat can occur at the MDL',  The MDL multiplier method was mostly
used in the early years of rule development for NPDWRs when insufficient WS data were
available.  Once sufficient WS data became available,  most of the PQLs that were developed
using the MDL multiplier were validated using WS data,

    L  How Were Water Supply Studies Conducted?

    Water Supply Performance Evaluation (WS PE) studies were an integral part of EPA's
certification program for drinking water laboratories for over 20 years.  Historically, WS studies
were conducted semi-annually by EPA for all current and proposed drinking water contaminants.
Although the WS studies were conducted semi-annually. for certification purposes, laboratories
were only required to demonstrate acceptable performance once a year  (14L23(k)(3) and
141.24(f)(17)).  WS study samples (spike samples) were sent to all laboratories that conduct
drinking water analyses, including utility laboratories, commercial laboratories, and State and
EPA Regional laboratories.  Each WS  study included samples or sample concentrates that were
analyzed both for all SDWA analytes and for analytes that were being considered for regulation
under the SDWA.

    During these WS studies, EPA's National Exposure Research Laboratory (NERL) in
Cincinnati, Ohio., sent participating laboratories a set of stable sample concentrates in sealed
glass ampules, a data reporting form, and appropriate instructions.  Each laboratory produced the
study samples by diluting a measured quantity of the specific concentrates to volume with
reagent water. The laboratory then analyzed the samples using the .specified procedures.  The
completed reporting form was sent to EPA for evaluation,, the data were carefully reviewed
(QA/QC'ed), entered into a database, and a fully detailed report was then  returned to each
laboratory. The responsible State or EPA office contacted those laboratories mat demonstrated
potential problems.                                    '               .
Methods Support Document far Six-Year Review
Draff - March 2002

-------
    At this point in time, the WS Performance Evaluation, studies are no longer performed by
 EPA. On July 18,1996 (61 FR 37464), EPA proposed options for the externalization of the PE
 studies program (now referred to as the Proficiency Testing or PT program).  After evaluating
 public comment, in the June 12, 1997 final notice EPA stated that the Ageacv has decided (62
 FR 32112):

       ...on a program where EPA would issue standards for the operation of the program, the
       National Institute of Standards and Technology (NIST) would develop standards for
       private sector PE (PT) suppliers and "would evaluate and accredit PE suppliers, and the
       private sector would develop and manufacture PE (PT) materials and conduct PE (PT)
       studies. In addition, as part of the program, the PE (PT) providers would report the
       results of the studies to the study participants and to those organizations that have
       responsibility for administering programs supported by the studies.

    Since the last WS PE studies performed by EPA were done in the Fall of 1999, the
 externalization of the PE program  should not effect the data needed for this Six-Year Review
 process.  However, at this time the Agency has not determined how to gather data to reassess
 PQLs for subsequent reviews of NPDWRs.

    2.  What Criteria Are Used to  Determine, a PQL?

    The derivation of the PQL involves determining the concentration of an analyte at which a set
 percentage of the laboratories achieve results within a specified range of the spiked value.
 Historically, the percentage of laboratories has been set at 75 percent, while a range of
 acceptance limits around the spiked value has been used. In many cases,, EPA derived PQLs only
 from the data submitted by the EPA Regional and State laboratories that participate in the WS
 studies.

    A PQL derived from WS data in such a manner is considered a stringent target for routine
 laboratory performance because:

       •    WS samples are prepared in reagent water and therefore do not contain the matrix
           interferences that may occur in field samples,
       •    Laboratories analyze only a small number of samples for the study and are aware that
           the samples are for the purposes of performance evaluation (i.e., they are not "blind"
           samples).

   In deriving a PQL from WS study data, the Agency typically sets a fixed percentage or 2
sigma (2 standard deviation) acceptance window around the known concentration (or spike
value) of the WS samples. Then the percentage of laboratories achieving results within the
specified acceptance window (y-axis) is plotted against the known spike concentration of the
Water Supply study samples (x-axis).  While the acceptance limits for inorganics typically range
from 15 to  30 percent (40 CFR §141.23(k)(3)(ii)), the acceptance limits for organics generally
range from 20 to 50 percent (141.24(f)(17)(i) and 40 CFR §141.24(h)(19)(i)). Several SDWA
analytes have acceptance limits of 2 sigma (2 standard deviation). Linear regression or graphical
analysis is performed on the WS data to determine the concentration at which 75 percent of EPA
Regional and State laboratories achieve acceptable results.
Methods Support Document for Six-Year Review
Draft ~ March 2002

-------
HI.    How Bid EPA Identify Which Contaminants to Evaluate for a PQL Reassessment?

   For the Six-Year analytical feasibility review, EPA focused on assessing whether the
practical quantitation level (or PQL) has changed since promulgation for a subset of the 68
chemical NPDWRs. Figure 1 illustrates the, overall Six-Year protocol and the basic process used
to identify the subset of contaminants for which a PQL reassessment should be appropriate.
Using the protocol, EPA identified and performed the analytical feasibility analysis for a total of
38 NPDWRs (Table 1), which fell into one of two categories:

    >  First, for those contaminants where the MCL is currently limited by analytical feasibility
       (i.e., the MCL is set at the PQL) and the MCLG is still appropriate, EPA evaluated the
       currently approved methods for those contaminants and available WS data to determine
       whether it might be possible to lower the PQL  and hence set an MCL that is closer to the
       MCLG.

    >  The second circumstance under which EPA re-evaluated the PQL was for contaminants
       identified under the Six-Year health effects technical review as having potential changes
       to their MCLG. Because the information for the health effects review was not completely
       available at the time the analytical methods analysis began, EPA took a broad-brush
       approach and included a number of contaminants that may not have needed a
       reassessment of their analytical feasibility.                                 ; •    • .
Methods Support Document for Six-Year Review
Draft. - March 2002

-------
 Figure 1. Overview of the Protocol for the Revise/Not Revise Decision with a Focus on
 Where Analytical Feasibility was Re-evaluated.
               NPDVVRs under review
               Initial Technical Review
          Health Effects, methods and treatment
        feasibility, and other regulatory revisions.
          For methods -focused on AfCLs that
              are currently limited by PQL
               Is a health risk assessment
                  in process/planned?
                    No
                                                 Yes
                                                               risk aMtC
        Does the review suggest possible changes
             in MCLG/MCL/TT and or other
                 regulatory' revisions?
                                                 No
                                                    NPDWR remains appropriate
                                                    after data/information review
              Yes
             ..	

        In-depth Technical Analysis
   New risk assessment, methods feasibility,
treatment -effectiveness, occurrence and exposure
         and economic implications.
   For methods -focussed on determining if
  analytical feasibility an issue for NPDWRs
       with potential changes in MCLC.
                                                                                    No Revision
                                                                                     at this time
Is a significant gain in
public health protection or significant
cost savings likely to occur?
YeS
Are the data sufl
regulatory
r

icieht to support
revision ?
N° J
I
No (
1
                                                             Negligible gain in
                                                           public health protection
                                                          -   and/or cost .savings
                                                            Data gaps- determine
                                                               research needs
                                               I.  Publish FR notice with preliminary revise/not revise
                                               decisions.

                                               2.  Review Public Comments and consider revising decisions
                                               in context of new information.

                                               3,  Publish FR notice with final list of NPDWRs to be revised
                                               and planned rulemaking schedule(s).
Methods Support Document for Six-Year Revietv
                                                                              Draft - March 2002

-------
                 Table 1. SDWA Chemical Contaminants Undergoing
                       Analytical Methods/PQL Reassessment
,; '

1
2
3 .
4
5
6
7
-8
. 9 .
10
11
12
13
14
15
16 .
17
18
19
20
21
22
23
24'
25
26
27
-SDW*. Chemical Contaminant "
'* ;«r ' s
Alachlor
Benzene
Benzo(a)pyrene
Beryllium
Bis(2"Cthylhexy!)phthatate
Cadmium
Carboftiran
Carbon tetradiioride
Clilordane .
Chromium {total - Cr III and VI)
1 ,2-Dibromo-3-chloropropane
(DBCP)
L4-Dtch!orobenzcnc (para)
1,2-dichloroethanc
1 , 1 -djchloroethylene
Dichloramethane {methytene chloride)
1 ,2- Dichlorapropane :
Dioxin<2,3Js8-TCDD)
Djquat
Etiiylene dibromtde '
Fluoride
Glyphosatc
Heptachlor
Heptachlor epoxide
Hexachlorobenzene
Hexachlorocyclopentadiene
Mercury
Methoxychlor
>*MGLG *
*> (mg/t)v-
yero
zero
zero
0,004
zero
0.005
0,04
zero
zero
0.1
zero
0.075
zero
0.007
zero
zero
zero
0.02
zero
4.0
0.7
zero
zero
zero
0.05
0.002
0,04
- 'MCJU* '
- tmg/L)
0.002
0.005
0,0002
0.004
0.006
0.005
0.04
0,005
0,002
0.1
0.0002
0,075
0,005
0,007
0,005
0.005
3xlO's
• 0.02
0.00005
4,0
0,7 •
0.0004 '
0,0002
0.001
0.05
0,002
0.04
Currcnt^QL^
(mgflL)
0.002
0,005
0.0002
0,001
0.006
0.002'
0007
0.005
0.002
0.01
0.0002
0.005
0,005
0.005
0,005
0.005
3x10'*
0,004
0.00005
0,5
0.06
0.0004
0.0002
0.001
- o.oo i
0.005
0.01
Acceptance- Limit *
w
i \
+. 45 %
± 20 % or 40 %
2 Std Dcv.
±15%
2 Std Dcv
+ 20 %
+ 45%
±20% or 40%
0.45
±15%
±.40%
±20% or 40%
+ 20% or 40%
±20% or 40%
± 20%. or 40 %
±20% or 40%
2 Std Dev
2 Std Dcv
. ±,40%
± 10 %
2 Std Dev
±. 45 %
±. 45 %
2 Std Dev
2 Std Dev
± 30 %
±.45%
.Methods Support Document jar Six-Year Review
Draft- March 2002

-------

28
29
30
31
32
33
34
35
36
37
38
SDWA Chemical Contaminant
Oxamyl (Vydate)
PCBs - Polychlorinated biphenyls
(as decachlorobiphenyl)
Pcntach lorophenol
Picloram
Tctrachlorocthylene
Thallium
Toxaphcne
1,1,1 -Trichloroethane
1 , 1 ,2-Trichloroethanc
Trichloroethylenc
Vinyl chloride
MCI.G1
(rag/L)
0.2
zero
zero
0.5.
zero
0,0005
zero
0.2
0,003
zero
zero
MCI, '
(mg/L)
0.2
0.0005
0.001
0.5
0.005
0,002
0.003
0,2
O.OOS
0.005
0,002
Current PQL J<4

-------
   »•.
   >•
   »-
these approaches.  Some of the advantages and disadvantages for these PQL derivation
approaches are as follows;

(1) Analysis ofWS PE Data - Uses data from WS studies to derive a new PQL, This value is
    compared to the old PQL (i.e., the one that is currently in place).

    The advantages of the WS PE Data methods of deriving a PQL -

       Uses inter-laboratory data collected at concentrations near the MCL.
       More representative of what methods are being used for the analysis of that contaminant.
       May be the preferred approach for contaminants with MCLGs of zero,

    The disadvantages of the WS analysis method of deriving a PQL -

    *  In the past, some stakeholders have felt that the PQL may be influenced by the set of WS
       data used (i.e., using data from all laboratories as opposed to only using data from EPA
       State and Regional laboratories),    .                    .    -

    »•  Some stakeholders have felt that the laboratory performance on WS data may be skewed,
       because WS samples may be treated as special samples that are critical for laboratory
       certification.

    •»  The derivation of PQLs from WS data is  a resource- and time-intensive process,

    »•  Because the WS samples are designed to test precision and accuracy around the MCL, the
       WS data may not cover concentrations several orders of magnitude below the current
       MCL,  Hence, for some analytes, data points at lower levels may not be represented,

(2). The MDL-MultipUer Approach - Using the MDL of the currently approved method(s) for
each contaminant, the 5 or 10 multiplier method can be used to estimate the PQL. This value is
men compared to the PQL mat was derived before the 1996 SDWA Amendments,

    The advantage of the MDL multiplier approach - it is a relatively easy and clear process.

    The disadvantages of the MDL multiplier approach - ,

    >  The WS studies test laboratory performance near the MCL as opposed to the MDL.  A
       PQL derived from the MDL multiplier method may not be representative, because the
       reproducibility of a result obtained at the MDL is often not as good as that obtained near
       the MCL,           .                            .

    >  Because several, methods may be approved for the same contaminant,, it can be
       difficult to decide which MDL to select for the PQL calculation.  However, .knowledge of
       the methods that are the most widely used can be determined from the WS data since
       laboratories report which method was used to analyze spike samples,
Methods Support Document for Six-Year Review
                                                                    Draft-March 2002

-------
    Acknowledging the advantages and disadvantages of the WS analysis and the MDL
 multiplier approach, EPA used the following steps for the Six Year Review to reassess analytical
 feasibility for the 38 chemical contaminants identified:

    •   The first step is the methods comparison step. This step compared the method detection
        limits of the analytical methods which were available at the time the PQL was set to the
        method detection limits for the currently approved analytical methods. This methods
        comparison should help to identify whether the ability to detect (and therefore quantify)
        these contaminants at lower levels has increased.

        The second step is the method usage over lime. This step used, information from the last
        eight Water Supply studies (WS 34 through WS 41; 1996 to 1999) to generate a bar graph
        of the distribution of the analytical/methods used to analyze the spike samples in the WS
        studies. This analysis should give an idea of the analytical methods that appear to be the
        most widely used for the analysis of particular contaminants. Knowing which analytical
        methods are the most widely used and the MDL for these methods can aid in estimating
        where the quantisation may lie today.

    •    The third and last step is the Water Supply data analysis step.  If Water Supply data are
        sufficient, more recent WS data can be used to recalculate the PQL (using linear
        regression or graphical analyses) and determine if the quantitation level has changed.
        Data may be considered insufficient if there are not enough data points around the 75
        percent criteria to recalculate the PQL using linear regression or graphical analysis. This
        may occur if the laboratories evaluated exhibit high  passing rates (>75 percent) for all of
        the WS studies evaluated and/or no WS spike samples were below the concentration of
        the current PQL. However, even if the WS data are  insufficient to actually recalculate the
        PQL, the information may be useful to either confirm that the current PQL still appears to
        remain appropriate or it may give an indication as to whether  the PQL is likely to change
        (if the data points at concentrations close to the current PQL are available).

    Using the information from these three steps helped EPA to determine if our ability to
quantify contaminants at lower levels has increased. If there was an indication that the PQL has
or could change, then pending the results of the health effects and occurrence review, as well as
risk management considerations, these chemical contaminants may be subject to a full blown
PQL reassessment (i.e., gathering data that sufficiently covers the area around the 75 percent
laboratory passing criteria and the appropriate concentration range).

Note: Tina document will not discuss whether a full blown PQL reassessment is necessary for
specific contaminants.  Instead, that decision will be made after the integration of these results
with the health  effects and occurrence and exposure reviews and'discussed, in the Six Year Notice
of Intent.
Atctliutit Support Document for Six-Year Review
10
Draft. -. March 2002

-------
V. Results of the PQL Reassessment
Alachlor

Results of the Method Comparison            .                                    '    ,

   The approved drinking water methods for the determination of alachlor, a Phase II synthetic
organic compound (SOC), were listed in the 1991 NPDWRs (56 FR 3526). These methods ail
utilize GC or GC/MS with several extraction and/or detector variations: EPA  Methods 505,_507,
and 525,1. Since promulgation of these original methods, the Agency has eliminated EPA 525.1
from the list of approved methods, and has approved tine use of three new GC methods: EPA
Methods 508.1, 525.2, and 551.1. The three new methods are approximately 10 to 100 times
more  sensitive than the earlier methods. The. current EPA 505 is nearly equivalent m sensitivity
relative to the time of Phase II promulgation, whereas the current EPA 50? is about twice as
sensitive. Table 2 summarizes the current and previous EPA methods along with their M.DLs.
Table 2.  Results of the Analytical Methods Comparison for Alachlor (Newly Promulgated
          Methods are Indicated in Bold)
MCL = 2|ig/L Current PQL = 2 p-g/L »LA = 0.2 |*g/L Acceptance .Limit* = ± 45%
Methods Approved At Promulgation
Method
EPA 505s
E.PA5071
SPA 525*


Technique
Microextraction, GC
GC with NPD
GC/MS with LSE .


MDL

0.225
0.38
0.1


Currently Approved Methods (141,24)
Method
EPA 5052
EPA 507s
EPA S08.12
EPA S25.22
EPA 551.1s
Technique
Microextraction, GC
GC with BCD
GC with LSE, BCD
GC/MS with LSE
GC/MS with LLE,
BCD
MDL
(«?5/L)
0.223
0.14
0.009
0.069-0.11*
0.005 - 0,025*
' "Methods for the Determination of Qiganlc Compounds in Drinking \\Mer," EPA-600/4-88/039, December
1988.
2 "Methods for the Determination of Oiganic Compounds in Drinking %,ter--$upplemeirt III,"
EPA/60Q/R-95-131, August 1995,
A Regulatory DLs for organic compounds are listed at 40 CFR §141 .24(h){1 8),
'•'Acceptance" limits for organic compounds are listed at are listed at 40 CFR §141.24(h)(19)(i).
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance. . .'•...
 Methods Support Document for Six-Year Review
11
Draft - March 2002

-------
 Results of the Analysis of the WS Data

 a.  Method Usage Over Time

     Figure 2 plots the distribution of analytical techniques used by the EPA and State laboratories
 in WS 34 to 41. Methods categorized as "other" represent methods which were not specifically
 identified by participating laboratories or were-otherwise unknown. As shown in. Figure 2, EPA
 507 was used fairly consistently throughout WS 34 to 41. Use of EPA 525.1' was phased out
 after WS 36, while use of EPA 525.2 increased  significantly during the same study. EPA
 Methods 508.1,505, and "other" methods remained in use minimally throughout the study
 period.
     Figure 2. Distribution of Analytical Techniques by WS Study: Alachlor

                                         Alachior
                                                                              505
                                                                             H 507
                                                                             D 608.1
                                                                             D 525,1
                                                                             • 525.2
                                                                             • other
              WS 34   WS 35   WS 36
WS 37   WS 38

  Water Study
                                                     WS 39   WS 40   WS 41
b.  Results of the PQL Analysis

    As PE data were not available at the time of the original PQL determination, the PQL of 2
[ig/L was derived using a multiplier of 10 on the Intel-laboratory MDL (0.15 u.g/L) based on a
study conducted by the Environmental Monitoring and Support Laboratory in Cincinnati, Ohio
(54 FR 22104). Data from WS 24 to 41 were used to attempt a PQL re-evaluation. Table 3
summarizes these data, indicating the study number, the true value (i.e., the spiked value) of the
WS sample, the number of results from EPA and State laboratories, and the calculated
Methods Support Document for Six-Year Revienv
     12
Draft - March 2002

-------
percentage of laboratories whose.results successfully passed within designated acceptance limits
for alachlor (specified in 40 CFR §141.24(h)(19)(i) to be ± 45 percent).
Table 3.  Evaluation of Alachlor Data from WS Studies Using the 45% Acceptance Limits
          (in Order of Increasing Concentration)
ws#
24a
26b
29
25a
32
31
30 .
34
27
33
24b
37
35
26a
36 .
38
25b
41
39
40
Spiked "True" Value Og/L)
0.735
0,933
1,59
1.8?
2.33
2,50
3,21
3.43
3,80
. 4,27
4.53
4,87
5.27
5.66
7.34
9.52
9.80
12,9
14.8
17.7
# Results from EPA
Regional and State Labs
19
20
14
13
43
25
40
48
17
30
19
40
,27
20
50
49
13
37
40
50
% Labs Passing ± 45%
Acceptance Limit$
100
95.0
71.4
100
86.0
76.0
97.5 •'
100
88.2
86.7
100
' 85.0
96.3
95.0
100
93.9
100
100
97,5
90.0 • . • •
    The data from the available PE studies were not conducive to PQL re-evaluation, as the.
percentage of labs passing generally exceeded the standard 75 percent passing criterion xieeded to
calculate the PQL using linear regression or graphical analysis (with the exception of one study -
WS 29). However, even around the,original PQL of 2 |ig/L» the percentage of laboratories
Methods Support Document for Six-Year Review
13
Draft - March 2002

-------
passing is extremely variable and ranges from 71.4 percent in WS 29 (spike concentration = 1.59
u.g/L) to 75 percent for WS 31 (spike concentration = 2.50 u-g/L) to 100 percent for several. WS
studies at varying concentrations. Even at higher concentrations of 3.8 jxg/L (WS 27) and 4,27
Hg/L (WS 37) the laboratory passing rates dip to 88 and 87 percent, respectively. Based on this
information, EPA believes the PQL for alachlor appears to still be in the appropriate range.

Conclusion for Alachlor

    Since the promulgation of the 1991 NPDWR for alachlor and other Phase II SOCs., three new
analytical methods (EPA Methods 525.2, 508.1, and 551.1) have been approved for the
determination of alachlor in drinking water. All three new methods exhibit lower MDLs than the
original methods. Based on the distribution of method use over time (Figure 2), it appears that
EPA and State laboratories did not utilize the increased analytical sensitivity of the newer
methods, instead preferring use of EPA 507.  Meanwhile, a PQL for alachlor could not be
recalculated using the PE data from WS 24 to 41.  Nearly every study exhibited a laboratory
success rate above the 75 percent criterion needed for re-evaluating the PQL, and furthermore,
the range of true values generally exceeded the current PQL  value. Therefore, the available PE
data provide very limited evidence for revising the current PQL of 2 u-g/L, However,-based on
the available data, EPA believes the PQL for alachlor is most likely in the appropriate range.


Benzene

Results of the Method Comparison

    In July 1987, the final NPDWR for eight Phase I VOCs approved the use of EPA Methods
502.2, 503.1, 524.1, and 524.2 for the determination of benzene in drinking, water (52 FR 25690),
The currently approved methods for benzene determination are EPA, Methods 502.2 and 524,2.
Table 4 summarizes the MDLs for both the original and current approved versions of the
methods.  As compared to the original methods, the updated methods are equal in sensitivity to
the original methods.
Table 4.  Results of the Analytical Methods Comparison for Benzene
MCL«5jig/L Current PQL = 5 ug/L DL* = 0.5 fig/JU Acceptance Umit+ = ± 20% (>10u,g/L) or
± 40% (<10 pgflu)
Methods Approved At Promulgation
Method
EPA 502.2 '
EPA 503.1 '
EPA 524.1 '
EPA 524.2 '
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL*

-------
  "Methods for the Determination of Oiganic Compounds in Drinking V&ter," BPA/6QO/4-8S/039, December
  "Methods for the Determination of Oigantc Compounds in Drinking Wtfer-Suppleraent III." EPA/60G/R-95-
 131 August 1995.
 °The MDLs of the original methods for this contaminant ranged from 0.2 - 0.5 ng/L according to the July 1987
 Federal Register notice promulgating MPDWRs for the VOCs (52 PR 25690), However, the 1988 methods
 manual cited in footnote 1 lists the MDLs shown above,
 * Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
 laboratory/analyst performance.                                                     ..     .
 A Regulatory DLs for VOCs are listed at 40 CFR § 14L24(f)(17)(i).                          ,  .
 t Acceptance limits'for VOCs are listed at 40 CFR § 141.24(f)(17)(i).                               	
Resultsof the Analysis of the WS Data

a.   Method Usage Over Time

    Figure 3 summarizes the distribution of the different methods used by the EPA and State
laboratories during WS studies 34 to 41. The category of "other" contains those methods that
were either unknown or otherwise unidentified by the participating laboratories. As shown in •
Figure 3, use of EPA 524.2 during WS 34 to 41 generally increased over time while use of EPA
502.2 decreased very slightly during the same-period. Overall, usage of EPA. 524.2 remained
consistently dominant over that of EPA 502.2. Use of the original EPA Methods, 503.1 and
524.1, were not apparent during this period.

     Figure 3. Distribution of Analytical Techniques by WS Study: Benzene

                                            Benzene
      •o
      o
      f
      S
      CD
                                         m 502.2 j

                                         O 524,2 j

                                         • other i
                 WS 34   WS 35   WS 36
WS 37   WS 38

  Water Study
WS 39   WS 40    WS 41
Methods Support Document for Six-Year Review
    15
              Draft - March 2002

-------
 b.  Results of thePQL Analysis

    The original PQL of 5 u-g/L for benzene was determined by multiplying the regulatory
 detection limit of 0.5 [ig/L by a factor of 10 (52 FR 25700). To re-evaluate the PQL? multi-
 laboratory performance data from WS 24 through 41 were reviewed. Table 5 summarizes the
 results of these water studies, providing the study number, the spiked value for the WS sample,
 the number of laboratory results, and the percentage of laboratories whose reported results fell
 within the acceptance limits of ± 20 percent for true values greater than 10 u-g/L and ± 40 percent
 for true values lower than 10 u-g/L (specified at 141.24(f)( 17)(i)).


 Table 5.  Evaluation of Benzene Data from WS Studies Using the ± 20% or ± 40%
          Acceptance Limits (in Order of Increasing Concentration)
ws#
24
34
27
36
39
30
26
33 '
37
31
25
35
29
38
32
40
41
Spiked "True"
Value (ng/L)
4.32
4.94
7.09
7.49
9.39
9.51
10.3
12.0
12.5
12.6
13.5
14.0
15.3
15.3
16.5
16.7
18.7
# Results from EPA
and State Labs
57
60
38
61
43
60
59
35
48
37
37
34
34
55
65
58
41
% Labs Passing ± 20%
Acceptance Limits




'

91.5
94.3
93.8
89.2
86.5
97.1
91.2
96.4
93.8
96,6
100
% Labs Passing & 40%
Acceptance Limits
98.2
100
100
100
97.7
100











Methods Support Document for Six-Year Review
16
Draft - March 2002

-------
   The overall data for benzene, as shown in Table 5, indicate that the passing rates for
laboratories fall well above the 75 percent criterion for establishing a revised PQL value.
However, as shown in Figure 4, the PE data for benzene display different relationships with true
values concentration depending on the acceptance limits (20-or 40 percent). The data
representing laboratories passing the 40 percent acceptance limits plot a horizontal line and do
not contribute meaningfully to a PQL re-evaluation.  For the laboratories passing the 20 percent
acceptance limits, the regression line demonstrates a positive slope, although a PQL re-
evaluation is also not possible because the true values exceed the current PQL. However, the
high percentages of laboratories passing around the current PQL of 5 ng/L suggest that the PQL
for benzene could possibly be lower.
       Figure 4. Two-part Distribution of Bcnsceae WS Data

                                          Benzene
         TO
         .£

         1
         CL
                                                                      75%
                   Current PQL = 5
 * % Labs Passing ± 20% Acceptance Limits j
 » % Labs Passing ± 40% Acceptance Limits j
8     10     12

True Value (pg/L
                                                          14
                     16
                              18
20
Conclusion for Benzene

    The method comparison results show that, since the promulgation of analytical methods
under the original NPDWR for benzene, two of these methods are no longer approved for
determination of this contaminant and method sensitivity has remained about the same,
Evaluation of the quantitative PE data shows that the laboratories conducting WS analyses
exhibited very high passing rates.  Because the percentage of laboratories passing the PE testing
exceeds the 75 percent criterion, a re-evaluatioa of the PQL could not be performed using this
approach. But the high percentages of laboratories passing around the current PQL of 5 u.g/L
suggest that the PQL for benzene could possibly be lower.
Methods Support Document for Six-Year Review
17
                             Draft - March 2002

-------
 Bcnzo(a)pyrene

 Results of the Method Comparison

    With the Phase V synthetic organic compounds (57 FR 31776), three approved methods were
 listed for determination of benzo(a)pyrene in drinking water, including gas chromatography/mass
 spectrometry (GC/MS) with liquid-solid extraction (LSE; EPA 525.1), high performance liquid
 chromatography (HPLC) with liquid-liquid extraction (LLE; EPA 550), and HPLC with LSE
 (EPA 550.1).  Since this regulation was promulgated, the Agency has replaced the old GC/MS -
 LSE method with an updated version, EPA 525.2.  As shown in'Table 6,'EPA 525,2 is estimated
 to be approximately 1.5 to 4 times more sensitive than the older methods in terms of its method
 detection level (MDL).


 Table 6.  Results of the Analytical Methods Comparison for Benzo(a)pyrene (Newly
          Promulgated Methods Indicated in Bold)
MCL = 0.2ng/L Current PQL = 0.2 ng/L DL* = 0.02 ^g/L Acceptance Limit* => ± 2xSJ>.
Methods Approved At Promulgation
Method
EPA 525. 11
EPA S502
EPA 550. 12
Technique
LSE, GC/MS
LLE, HPLC
LSE, HPLC
MDL

-------
    Figure 5, Distribution of Analytical Techniques by WS Study: Benzo(a)pyrene

                                    Benzo(a)pyrene
                                                                           • 550
                                                                           ^525,1 {
                                                                           P 550,1}
                                                                           B 525.2 i
                                                                           m other I
WS34   WS35  WS36
WS 37   WS 38

 Water Study
                                                     WS 39   WS40   WS41
b. Results of the PQL Analysis

   The'original PQL was estimated at 0.2 u-g/L (57 FR 31802) based on PE data compiled from
WS studies 23,24, 26, and 27, The data used for the re-evaluation of the PQL were taken from
WS studies 24 through 41, Table 7 summarizes the results of these studies. The table provides
the WS study number, the spiked, or "true value" for the WS sample, the number of laboratory
results, and the percent of laboratories passing the WS proficiency test for benzo(a)pyrene within
the acceptance limits. The acceptance limits were calculated to be plus or minus two standard
deviations from the estimated mean recovery, as stipulated in CFR §141.24(h)(19)(i).
Table 7,   Evaluation of PE Data for Beiw0(a)pyrene from WS Studies Using 2 x S.D.
          Acceptance Limits (in Order of Increasing Concentration)
ws#
• 31
32 .
30
38 '
Spik«d ("True")Value
•foS/L)
0.202
' 0.337
0.485 •
0.53
# Results from EPA and
State Labs
14
29
1.2
38
% Labs Passing Acceptance Limits
79
93
67
84
Methods Support Document for Six-Year Review
     19
                                                     Draft - March 2002

-------
ws#
36
34
37
33
40
35
26
39
41
26
Spiked («True")Valuc
(Hg/L)
0.64
0.75
0.94
1.29
1.48
1.53
2.25
2.37
2.37
15.5
# Results from EPA and
State Labs
40
28
37
22
43
27
10
31
25
12
% Labs Passing Acceptance Limits
88
89
89
82
91
70
§3
94
84
70
    Figure 6 shows the plot of the benzo(a)pyrene data for WS 24 to 41, and the linear regression
 line.  There is no meaningful relationship in these data, in large part because the true values of
 the samples are so high that a very large percentage of the labs passed all the PE series. Further
 only one PE sample approached the current PQL (WS 31), and these results (79 percent passing)
 support the current PQL.
Methods Support Document for Six-Year Revieiv
20
Draft - March 2002

-------
        Figure 6. PQL Evaluation from PE WS Data: Benzo(a)pyrene

                                    Benzo(a)pyrene




I
& Labs Pa



rtAO/

80% ••
70% •
80% -
50%
40% •
30%
20% -
10% •
n% -









* *
* * * » *
* * 	 " * - "' 	 '•"'" 	 """ 1"""" 	 '*"-"»—""
*

75%




Current F>hl= 0.2
                           0.5
                                        1           1.5
                                        True Value (ug/L)
                                2.5
ConclusioniToriBenzo(a)pyrene

   As notedin the method comparison, a more sensitive method (EPA 525.2) has been approved
since the promulgation of benzo(a)pyrene, replacing an older and less sensitive version (EPA
525.1).  The method usage evaluation shows that in recent years, a majority of EPA and State
laboratories in the PE studies have chosen to use this more sensitive method out of all the
approved methods for benzo(a)pyrene.  The WS data do not afford a re-evaluation of the  PQL,.
and at best, support the current value. There is no clear evidence to support a change from the
current PQL of 0.2 u.g/L. The current PQL appears to be appropriate.


Beryllium

Results of the Method Comparison

   With the Phase V lOCs (57 FR 31776), EPA approved multiple'analytical methods for
determination of beryllium in drinking water, including an atomic absorption-furnace (AAF)
method (EPA 210.2), an inductively coupled plasma atomic emission spectroscopy (ICP-AES)
method (EPA Methods 200.7) and ICP-mass spectroscopy method (ICP-MS) (EPA 200,8). EPA
210.2 has since been removed from the approved list and replaced by four newer methods: EPA
200.9 (AA-platform); and three voluntary consensus standard methods, including Standard
Methods 3 3U3B (AA-fumace) and 3120B (ICP-AES) and ASTM Method D3645-93B
(AA-furnace). These methods are listed in Table 8.. The MDLs of EPA Methods 200.7 and
200.8 do not present any improved sensitivity capabilities. However, Method 200.9 is about five
Methods Support Document for Six-Year Review
21
Draft - March 2002

-------
times more sensitive than the most sensitive method approved at the time of promulgation (EPA
200.7).
Table 8.   Results of the Analytical Methods Comparison for Beryllium (Newly
          Promulgated Methods Indicated in Bold)
MCL«4ng/L Current PQL = 1 ng/L DL* «= 0.02-0.3 u-g/L Acceptance Limit* «=± 35%
Methods Approved At Promulgation
Method
EPA 200.7'
EPA 200.8'
EPA 2 10.2'



Technique
ICP-AES
ICP-MS
AA; Furnace



MDL
(u.g/L)
0.1
0.2
0.2



Currently Approved Methods (141.23)
Method
EPA 200.72
EPA200.82
EPA 200.92
3113B*
3120BJ
D3645-93B"
Technique
' 1CP
ICP-MS
AA; Platform
AA; Furnace
ICP
AA; Furnace
MDL

-------
     Figure 7. Distribution of Analytical Techniques by WS Study: Beryllium

                                         feerylltum
          60% ••
       X!
       O
       £
       0>
       CD
       £
       £
       3
          50%
40%
          30%
          20%
          10%
           0%
               WS 34   WS 35   WS 36
                             WS 37   WS 38

                              Water Study
WS39   WS40  . WS41
b.  Results of the PQL Analysis

    The current PQL (1.0 fig/L) was originally determined using PE data from WS 24 through 27
(56 FR 60949), A PQL re-evaluation was attempted using more current data spanning WS 24 to
41. Table 9 summarizes the results of these water studies, providing the study number, the
spiked value for the WS sample, the number of results from EPA and State laboratories, and the
beryllium results evaluated using an acceptance limit of ± 15 percent (as indicated at 40 CFR
Table 9.   Evaluation of Beryllium Data from WS Studies Using the 15% Acceptance
          Limits (in Order of Increasing Concentration)
ws#
25b
26a
24a
32
Spiked "True" Value (jig/L)
0.400
0,530
' 0.600
0.933
# Results from EPA and
State Labs
20
40,
35
60
% Labs Passing ± 15%
Acceptance Limits
70.0
70.0 •
71.4
73.3
Methods Support-Document for Six-Year Review
                                23
              Draft - March 2002

-------
ws#
39
35
25a
41
31
37
27
34
40
36
30
33
29
38
26b
Spiked "True" Value (ng/L)
1.20
1.33
2.00
2.58
3.27
4.26
4.67
5.33
6.60
7.70
8.47
9.07
9.76
10.1
23.1
# Results from E1PA and
State Labs
51
42
24
46
33
49
24
60
64
,61
. 49
39
25 •
63
46
% Labs Passing ±15%
Acceptance Limits
84.3
83' .3
91,7
91.3
93.9
98.0
75.0
90.0
90.6
95.1 -
91,8
84.6
80.0
90.5
84.$
    The data in Table 9 seems to indicate that a linear relationship might exist between "true"
value concentration and percentage of labs passing. Furthermore, a PQL re-evaluation appeared
possible because the percentage of labs passing fell within the 75 percent criterion for some
studies, and the true values from the water studies approximated the general range of the current
PQL. Visual evaluation of the laboratory passing percentages around the current PQL of 1 (xg/L
indicate that this value is still appropriate and unlikely to change. The graph in Figure 8
illustrates this relationship between the spike concentrations and the laboratory passing rates. As
shown in Figure 8, the recalculated PQL of 0,71 u-g/L, is slightly lower than the current PQL.
Methods Support Document for Six-Year Review
24
Draft ~ March 2002

-------
       Figure 8. Two-part Distribution of Beryllium WS Data:

                                         Beryllium






en
,£
£
to
O.
03 .
XI
_j
»









100% .

90% ,

. 80% .
70% .

60%

50% -

40% -
30% .

20% ,


10% .

0%



; • **
i +** *****

• |* ' -0- .
*;- 75%
" ,
O 1

I i

i ' • ' .
h
O
. t
•c:
;t»
j^
P

1.4, 9 14 19 . 2
True Value (ug/L)
Conclusion for
    The method comparison results indicate that EPA 200,9 is now the most sensitive method for
determination of beryllium in drinking water and its MDL has improved by five-fold over the
most sensitive MDLs achieved at the time of promulgation, EPA 200.7 was used more widely in
previous years but is less sensitive. As revealed by the results of method usage over time, EPA
200,9 is the most commonly employed method for beryllium determination in recent PE studies.
These trends seem to imply a shift in analytical capabilities for beryllium determination toward
greater sensitivity. Thus, EPA and State laboratories are likely to reach lower detection limits
today compared to the year of NPDWR promulgation. The re-evaluation .of the PQL using a
linear regression was calculated to be 0.71 jig/L, a value close to the current PQL of 1  u,g/L.
Although, it may be possible to lower the PQL slightly based on more recent PE data, the current
PQL is still appropriate.


Bis(2-ethylhexyl)phthalate

Results of the Method. .Comrjarison

    At the time of the Phase V SOC promulgation (57 FR 3 1776), EPA Methods 506 and 525.1
were the only approved methods for the analysis of bis(2-ethylhexyl)phthalate (also known as
di(2-ethylhexyl)phthalate). Since that time, EPA 506, which uses LLE or LSE GC with PID, has
remained on the approved methods list while EPA 525.1, which utilizes LSE GC/MS, has been
Methods Support Document for Six-Tear Review
25
Draft - March 2002

-------
 replaced by a more sensitive version, EPA 525.2, This increased sensitivity, however, is only
 marginal, as evidenced by the similarity of their MDLs. Table 10 lists the approved methods,
 techniques, and MDLs, during and after the time of rule promulgation.


 Table 10.  Results of the Analytical Methods Comparison for Bis(2-ethylhexyl)phthalate
           (New Methods in Bold)
MCL = 6 jig/L Current PQL = 6 u.g/L DL* = 0,6 ug/L Acceptance Limit* - ± 2 x S,D.
Methods Approved At Promulgation
Method
EPA 525.1'
EPA 5062
Technique
LSE, GC/MS
LLE or LSE,
GCw/PID
MDL
(Rg/L)
0.6 - 0.8*
2.25
Currently Approved Methods (141.24)
Method
EPA 525.2*
EPA 5063
Technique
LSE, GC/MS
LLE or LSE, GC w/
PID
MDL
(M.g/L)
0,46 - 7.9*
2.25
1 "Methods for the Determination of Oiganic Compounds in Drinking "V&ter," EPA/600/4-88/039, December
1988,
2 "Methods for the Determination of Oiganic Compounds in Drinking V&ter,' Supplement I," EPA/600/4-90/020,
July 1990.
'"Mediods for the Determination of Oiganic Compounds in Drinking Mter—Supplement III," BPA/6QO/R-95-
131, August 1995.
' Multiple method detection limit (MDL) values result from variability of reagents, instrumentation, and/or
laboratory/analyst performance.
' Regulatory DLs for oiganic compounds are listed at 40 CFR 141.24(h)(18)
* Acceptance limits for oiganic compounds are listed at 40 CFR 141 .24(h)(l 9)(i)
Results of the Analysis of the WS Data

a.  Method Usage Over Time

    The distribution of the analytical methods used during WS 34 to 4] by participating EPA and
State laboratories are illustrated in Figure 9. The category of "other" represents any alternative or
unreported methods. As shown in Table 10, use of EPA 525.1 decreased significantly,
corresponding to the introduction of EPA 525.2 during WS 36.  By WS 38, EPA 525.1 was no
longer used by any laboratories for analyzing bis(2-ethylhexyl)phthalate.  Laboratories
overwhelmingly chose to use the LSE GC/MS technique (either EPA 525.1 or its replacement
EPA 525.2) over EPA 506, which utilized the LLE or LSE GC with PID.
Methods Support Document for Six-Year Review
26
Draft - March 2002

-------
    Figure 9, Distribution of Analytical Techniques by WS Study: Bis(2-
    ethylhcxyl)phthalate

                               Bis(2-ethylhexyl)phthalate
         90%
                                                                            0525,2
                                                                              525.1
                                                                            • 506
                                                                            Bother
               WS 34   WS 35  WS 36   WS 37   WS 38   WS 39   WS 40  WS 41

                                        Water Study
b. -Results of the PQL Analysis

   PE data from WS 24 to 41 were compiled to re-evaluate the PQL. Table 11 shows the WS
number, the true value concentration, the number of participating laboratories, and the percent of
laboratories passing within acceptance limits. These limits were calculated to be two times the
standard deviation, as stipulated in 40 CFR §14L24(h)(19)(i). Note that data were unavailable
for WS 25 and 28.
Table 11.  Evaluation of Bis(2-ethylhexyl)phthalatc Data from WS Studies Using 2 x S.D.
          Acceptance Limits (in Order of Increasing Concentration)
ws#
24a
29
30
26b
Spiked "True" Value
Gig/L)-
3,18
4.58 .
6.40
7,73
# Results from EPA and
State Labs
14
8
15
13
% Labs Passing
±2 x S.D. Acceptance Limits
35.7
75
60
69.2
Methods Support Document far Six-Year Review
27
Draft - March 2002

-------
ws#
32
31
38
41
33
27
36
24b
37
34
39
40
26a
35
Spiked "True" Value
(HS/L)
9.28
. 11.7
13.7
15.3
15.8
17.3
18.3
19.1
21.3
21.3
27.7
32.4
34.2
37.7
# Results from EPA and
State Labs
27
- 12
35
25 '
23
7
39
15
33
33
• 26
38
13
20
% Labs Passing,
±2 x S.D. Acceptance Limits
92.6
91.7
91.4
84
87
85.7
92,3
93.3
93.9
93.9
100
89,5
84.6
85
    Using the data in Table 11, a linear regression was performed by plotting the percentage of
 the laboratories passing for WS 24 to 41 (excluding WS 25 and 28) against the spiked value of
 bis(2-ethylhexyl)phthalate (Figure 10). Simple visual examination of the graph shows that the
 percentage of laboratories achieving acceptable results approximates 75 percent at a
 concentration of 7.84 u-g/L.  This concentration is higher than the current PQL of 6,0 u.g/L, which
 was originally estimated using PE data from WS 23,24, and 27 (5.7 FR 60953),
A/ethods Support Document for Six-Year Review
28
                          Draft - March 2002

-------
   - Figure 10. Two-part Distribution of Bis(2-ethylhexyl)phthalate WS Data

                                 Bis{2-ethylhexyl}phthalate




CD
'55
CL
«
ss







°
90% -
80% •
70%
60%
50%
40% -
30% •

20% •

10% -
fW/.
w /O *
<
-um^uwBir-iimi-:--!.--" 	 i 	 — f 	 < 	 .MMMBWWWMWW™™-.-™-. 	 ™ 	 -...,.... 	 i 	 	 	 	 ^— 	 M 	 .^^m^^^m^^-^^-^m-^ 	 . 	
I / ^ ^ ^
V '
""" 	 "/ 	 ; 	 "" - 75o/;
/'**•? r
/ i it
/ * i
/ - id
/ i*
I
Jj '
O i . .
i i .
1 i
. o ; 	 • 	
3 5 10 15 20 25 30 35 4
                                        True Value (\igll.)
Conclusion for Bisf2-ethylhexyDphthalate                          .

   Since the time of the Phase V SOC.rule promulgation, EPA 525.2 has replaced EPA 525.1.
This new method provides a marginally greater degree of sensitivity. The EPA and State
laboratories have overwhelmingly chosen to use EPA 525,2 over any other available methods as
evidenced by the plot of method usage over time. The re-evaluation of the PQL using a linear
regression calculated from WS 24a, 26b, 30 and 32 data showed that a new PQL could be
derived. The re-evaluated PQL of 7.84 |ig/L is higher than the original. PQL of 6 (xg/L (57 FR
31802),


Cadmium

Results ,of the Method Comparison

   The 1991 NPDWRs for Phase II lOCs listed Atomic Absorption-Furnace (AAF; EPA 113.2)
and Inductively Coupled Plasma (ICP; ERA 200.7A) as approved analytical methods for the
determination of cadmium in drinking water (57 FR 31776).  Since the promulgation of the rule,
these methods have been replaced by four new or updated methods; three EPA methods (EPA
Methods 200.7,200.8, and 20Q.9), and one voluntary consensus standard method (Standard
Method 3113B).  Table 12 compares the detection limits of approved methods during and after
promulgation of the NPDWR for cadmium and shows EPA 200.9 to be the most sensitive
Methods Support Document for Six-Year Review
29
Draft - March 2002

-------
 method for detecting cadmium in drinking water, compared to all other original and current
 methods.
 Table 12.  Results of the Analytical Methods Comparison for Cadmium (Newly
           Promulgated Methods Indicated in Bold)
MCL = 5U£/L Current PQL- 2 ng/JL DL1 « 0,1-1.0 Mg/L Acceptance Limit* = ± 20%
Methods Approved At Promulgation
Method
EPA 200.7A1
EPA 2 13.2'


Technique
ICP-AES
AAF


MDL
(jig/L)
1
0.1


Currently Approved Methods (141,23)
Method
EPA 200.7*
EPA 200.8*
EPA200.92
SM 3113B*
Technique
JCP-AES
rcp-MS
AA; Platform
AA; Furnace
MDL
(fig/L)
1,0
0.5
0,05
N/A*
1 "Methods for Chemical Analysis of Wrter and Wastes (MCAWW)," EPA/600/4-79-020, March 1983.
2 "Methods for the Determination of Metals in Environmental Samples-Supplement I " ER/600/R-94/] 11, May
1994.
1 18* and 19!h editions of Standard Methods for the Examination of Wafer and Wastewater, 1992 and 1995.
American Public Health Association.
* Regulatory DLs for inorganic compounds are listed at 40 CFR § 1 4 1 .23(a)(4Xi) and depend on analytical
method.
f Acceptance limits are listed at 40 CFR §141, 23(k)(3)(ii).
• MDLs for non-EPA methods are not specified.
Results of the Analysis of the WS Data

a.  Method Usage Over Time

    Figure 11 shows the distribution of analytical techniques used by EPA and State laboratories
for WS studies 34 to 41.  The results for "other" techniques in this figure include the use of any
other technique identified by the laboratories participating in the WS study, as well as
"unknown" methods, i.e., methods for which laboratories did not report any information on the
type of method used. During WS 34 and 35, EPA 213.2 was the most widely used method for
determining cadmium in drinking water. By WS 36, its use was considerably diminished, but a
resurgence was seen in WS.38, only to be phased out in WS 39.  Subsequently, EPA Methods
200.7,200.8, and 200.9 were all used with relatively the same frequency by the participating
laboratories in the PE studies. Use of SM 3113B lessened over the period between WS 36 and
41.
Methods Support Document for Six-Year Review
30
                         Draft - March 2002

-------
     Figure 11, Distribution of Analytical Techniques by WS Study: Cadmium

                                        Cadmium
      O
      JS
      "S

      en
      XI
      ffi
100% •

 90% -

 80%

 ro% -

 60% -

 50% -

 40% -

 30%

 20% -

 10%
                      1

                WS 34   WS 35   WS 36
                             WS 37   WS 38

                               Water Study
WS 39   WS 40   WS 41
                      • 200.7

                      H 200.8

                      10200,9
                      l
                      | B213.2

                      'D3113B

                      ffl other I
b. Results of the PQL Analysis   '                          ,

   The current PQL of 2.0 |ig/L was originally set using PE data'from. WS 22 through 25 (56 FR
3549). With the availability of more current data from WS 24 to 41, a PQL re-evaluation was
attempted.  Table 13 summarizes the results of these water studies,, providing the study number,
the spiked value for the WS sample, the number of results from EPA and State laboratories, and
the cadmium results, evaluated using an acceptance limit of ± 20 percent, as designated in 40
CFR § 14L23(k)(3)(ii).       '
Table 13. Evaluation of Cadmium Data from'WS Studies Using the 20% Acceptance
          Limits (In  Order of Increasing Concentration)
ws#
38
35
29
Spiked "True" Value (fig/L)
2.12
2.80
2.80
# Results from EPA and
StateJLabs
62
39
. 34
% Labs Passing ± 20%
Acceptance Limits
85.5 '
87.2
88.2
Methods Support Document:for Six-Year Review
                                 3-1
              Draft - March 2002

-------
vvs#
32
40
26b
37
24b
31
24a
41
34
25
39
27
36
30
33
26a
Spiked "True" Value (ng/L)
4.80
6.31
9.20
. 10.2
10.4
12.8
15.4
18.2
23.0
. 27.6
28.5
29.3
34.0
39.0
49.0
53.9
# Results from EPA and
State Labs
67
65
62
51
61
35
6J
45
65
40
54
40
66
. 66
38
62
% Labs Passing A 20%
Acceptance Limits
97,4
87.7
91.9
98,0
95.1
• 91.4
90.2
95.6
• . 96.9
100
98,1
92.5
. 98.5
98.5 '
97,4
96.8
    The data in Table 13 were not adequate to perform a PQL re-evaluation. EPA's preferred
format for evaluating a PQL is to develop a regression (or graphical analysis) using the true value
concentration and the percentage of laboratories passing the performance evaluation. The PQL is
then set at a concentration in which 75 percent of those laboratories pass, In this instance,
however, the participating laboratories passed the evaluation at an average rate of 93 percent,
well above the 75 percent criterion. Also, all of the spiked "true" values were well above the
original PQL of 2 u,g/L.

Conclusion for Cadmium

    A comparison of the analytical methods approved during and after the promulgation of the
NPDWR for cadmium show that the four current methods have sensitivities similar to, or slightly
better than, those of the original methods, with EPA 200.9 being the most sensitive (with an
MDL of 0.05 fig/L). Laboratories that participated in the PE water studies chose to utilize EPA
213.2 in the WS studies prior to WS 38 but then chose to  utilize EPA Methods 200.7,200,8. and
200.9 with similar frequency.  Review and analysis of the PE data did pot result in the estimation
of a new PQL because all of the EPA and State laboratories in the WS studies evaluated
surpassed the required 75 percent criterion typically used  to determine a new PQL. In addition,
Methods Support Document for Six-Year Review
32
Draft - March 2002

-------
all of the WS spike concentrations were above the current PQL of 2>g/L,  However, laboratory
passing rates of greater .than. 85 percent at concentrations slightly above the current PQL suggest
that the PQL could be lower,


Carbofuran

Results of the Method Comparison

   At the promulgation of the NPDWRs for synthetic organic chemicals (56 FR 3552), one
analytical method (EPA 531,1) was approved for determination of carbofuran in drinking water.
Since that time, the sensitivity of EPA 531.1 has improved, as indicated by a lower MDL, One
additional method, Standard Method (SM) 6610, has been included as an approved method for
carbofuran analysis,  Both current methods use the HPLC technique and have similar MDLs.
Table 14 summarizes the approved methods, both past and present, for determination of
carbofuran,                   '
Table 14.  Results of the Analytical Methods Comparison for Carbofuran (Newly
          Promulgated Methods in Bold)
MCL==4'0|ag/JL ' Current PQjL = 7 pgflL, »L* «;0.9. jig/t- Acceptance jLinait^i 45%
Methods Approyed At Promulgation
Method
.EPA 53 1,1s

Technique
HPLC

MDL
(jig/L)
0.9

Currently Approved Methods (141.24)
Method
EPA 53 1.1*
SM 6610s
Technique
HPLC
HPLC
MDL"
G*g«-)
0.52
0.53
1 "Methods for the Determination of Oigamc Compounds in. Drinking Water," EPA/600/4-88/039, December
1988.
2 "Methods for die Determination of Qiganic Compounds in Drinking Wrter— Supplement III," EPA/600/R-95-
131, August 1995.
3 "Supplement to fee 18th edition of Standard Methods for the Examination of Witer and Wastewater." 1994,
' Regulatory DLs for oiganic compounds are listed at 40 CFR § 141 .24{h)(l 8).
f Acceptance limits for organic compounds are listed at 40 CFR § 141.24(h)(19)(i).
Results, of the Analysis of the WS Data

a.  Method Usage Over Time                                               :

   Figure 12 illustrates the methods chosen by EPA and State laboratories for carbofuran
analysis during WS PE studies 34 to 41. The category of "other" includes any unidentified or
urjreported techniques used by participating laboratories.  As shown in Figure 12, the  •
predominant method used by laboratories participating in the WS studies is EPA 531.1. The
smallest percentage of labs using this method is 91,7 percent (in WS 37), indicating wide usage
compared to SM 6610 or any other available methods.
Methods Support Document for Six-Year Review
33
Draft - March 2002

-------
      Figure 12. Distribution of Analytical Techniques by WS Study; Carbofuran

                                       Carbofuran
       I
       s
       ra
       J3
       3


VAOA

OU /O


40% -

10%
n%







"

-




















—
















_


"""""




































_

^ 	 -
—
















_

S53





































.... __



















-

	





D531.1 [
a other [



                 WS 34  WS 35
WS36   WS37   WS38   WS 39   WS40  WS 41

         Water Study
b.  Results of the PQL Analysis

    The current PQL of 7 u,g/L was derived from multiplying the interlaboratory method-
detection limit (IMDL) of 0.7 u.g/L by a factor often (54 FR 22062).  To conduct a PQL re-
evaluation, the numerical data from WS 24 to 41 were analyzed (though no data were available
for WS 28). Table 15 summarizes each WS result including the spiked (or *4true") value, the
number of participating laboratories, and the percentage of laboratories passing within the
specified acceptance limit for carbofuran (± 45 percent of the spiked value, as specified in 40
CFR§141.24(h)(19)(i)).
Table 15. Evaluation of Carbofuran Data from WS Studies Using the 45% Acceptance
          Limits (in Order of Increasing Concentration)
ws#
29 '
30
32
31
Spiked "True" Value
(WSflL)
4.00
5,78
7.67
11.3
# Results from EPA and
State Labs
10
26
36"
14
% Labs Passing
± 45% Acceptance Limits
100.
92
94
86
Methods Support Document for Six-Year Review
           34
Draft - March 2002

-------
ws#
24a
26b
34
27
25a
33
38
26a
36
35
41
24b
25b
' 37
40
39
Spiked "True" Value
' (HgflL)
15.6
17.5
18.5
20.7
24,2 . .
24.8
33.6
36.3
37.8 :
42.8
43.7 •'
44.5
48.3
48.9
55.0
74.5
# Results from EPA and
State Labs
tl
J3
. 43
8
5
33
42
. , 13
47
60
29
' 11 . . '
5
36
42
31
% Labs Passing
± 45% Acceptance Limits
100
92
91
88
80
94
95
100
98
'92
97
100
100
97
100
97
   .Re-evaluation of the PQL by a linear regression approach was not feasible using the data
shown in Table 15. The percentage of laboratories passing within the acceptance limit was well
above the 75 percent criterion historically used to calculate the. PQL, Also, very few (e,g., only
WS 29, 30, and 32) of the above WS studies provided a sample with a true value concentration
near the original PQL. Because of these data limitations, the PQL could not be re-evaluated
using the historical linear regression approach. Instead, at concentrations approaching nearly half
'of the current PQL, laboratories in WS 29 were observed to achieve a 100 percent passing rate. '
implying strong analytical capabilities at a low concentration.  This observation could have
implications for lowering the PQL.

Conclusion for Carbofuran

   Since the time of promulgation of the original methods; one new method (SM 6610) has been
added for carbofuran analysis.  The current EPA .531.1  is nearly twice as sensitive as the.previous
version, and the plot of method usage over time illustrates that EPA 531.1 was the most.
commonly used method for the determination of carbofuran.  Together, these facts imply that
analytical methods capabilities have improved over time.  The PE data from WS studies 24 to 41,
however, were not useful for a PQL re-evaluation because of the extremely high number of
Methods Support Document for Six-Year Review
35
Draft - March 2002

-------
 laboratories passing within the accepted limit for carbofuran. While the high success rate of
 laboratories at some low concentrations may present a potential argument for lowering the PQL
 from 7 |ig/L, mere are no data available to support changing the original PQL,


 Carbon Tctrachloride

 Results of the Method Comparison

    In July 1987, the final NPDWR for eight Phase I VOCs approved the use of EPA Methods
 502.1, 502.2, 524.1, and 524.2 for the determination of carbon tetrachloride in drinking water (52
 FR 25690). The current approved methods for carbon tetrachloride determination are EPA
 Methods 502.2, 524.2, and 551.1. Table 16 summarizes the MDLs for both the original and
 currently approved versions of the methods.  As shown in Table 16, the three new methods have
 greater detection sensitivity than EPA 524.1  but not when compared to EPA 502.1.


 Table 16.  Results of the Analytical Methods Comparison for Carbon Tetrachloride (New
           Methods in Bold)
MCL = 5u.g/L Current PQL = Sug/L DL* = 0.5 jig/L Acceptance Limit1' = ± 20% (>10ng/jL) «*'
± 40% (<10 ug/L) .
Methods Approved At Promulgation
Method
EPA 502.1'
EPA 502.2'
EPA 524.1'
EPA 524.2"
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL*
(WB/L)
0.003
0.01
' 0.3
0.2)
Currently Approved Methods
Method
EPA 502,2a
EPA 524.2*
EPA 551, 12

Technique
Purge and Trap GC
GC/MS
LLE/GC with BCD

MDL*

-------
Results of the Analysis of the WS Data

a.  Method Usage Over Time

    The distribution of the different methods used by the EPA and State laboratories during WS
studies 34 to 41 are shown in Figure 13. The category of "other" contains those methods that
were unknown or unidentified by the participating laboratories. As shown in Figure 13, EPA
Methods 524.2 and 502.2 are currently the preferred methods used by laboratories for
determination of carbon tetrachloride. Use of the original methods, EPA Methods 502,1 and
524.1, was not apparent over this time period; plotting data from studies prior to WS 34 might
reveal more information on the use of these methods.
    Figure 13. Distribution of Analytical Techniques by WS Study: Carbon Tetrachloride

                                  Carbon Tetrachloride
         80%
         60%
      f  50%
      S
      en
      .E  40%
      | 30%
         20%


         10%


          0%
                                              *

                                                                            H 502.2
                                                                            Bother \
               WS 34   WS 35 ' WS 36   WS 37   WS 38  WS 39   WS 40  WS 41
                                       Water Study
 b.  Results of the PQL Analysis

    The original PQL of 5 \ig/L (52 FR 25700) for carbon tetrachloride was determined from PE
 data from. WS 8 to 11. Re-evaluation of the PQL was attempted using data from, WS studies 24
 through 41.  Table 17 summarizes the results of these WS studies providing the study number,
 the spiked value for the WS sample, the number of participating laboratories, and the percent of
• laboratories passing the WS proficiency test for carbon tetrachloride within the specified
 acceptance limits for carbon tetrachloride (± 20 percent for a true value greater than 10 u-g/L, or ±
 40 percent for a true value lower than 10 u-g/L as specified at 14L24(f)(17)(i)).
Methods Support Document for Six-Year Review
                                           37
Draft - March 2002

-------
 Table 17.  Evaluation of Carbon Tetrachloride Data from WS Studies Using Either 20% or
           40% Acceptance Limits (in Order of Increasing Concentration)
\vs#
24
34
30
27
31
40
25
29
35
36
37
33
41
32
38
26
39
Spiked "True"
Value fag/L)
4.56
6.27
6.46
8.48
8.69
8.90
9.18 ,
10.4
10.8
12.6
12.7
13.4
14.2
14.5
15.6
16.7
19.2
# Results from EPA
and State Labs
56
60
59
37.
36
57
37
36
34
59
47
34
40
55
54
59
42
% Labs Passing
± 20% Acceptance Limits



s



81
88
97
87
88
100
89
94
85
S3
% Labs Passing
± 40% Acceptance Limits
96
100
98
97
100
98
95










    From Table 17, it can be concluded that the available PE data for carbon tetrachloride are
insufficient for a PQL re-evaluation.  For the WS studies evaluated, the participating labs passed
the proficiency exams at a passing rates greater than the standard 75 percent acceptance criterion
used to determine the PQL.  In addition, only one WS study (WS 24) had a concentration lower
than the current PQL. However, passing rates of greater than 96 percent at concentrations close
to the current PQL suggest that the current PQL of 5 jxg/L could be possibly be lower

Conclusion for Carbon Tetrachloride

    The method comparison results show that since the promulgation of analytical methods tinder
the NPDWR, three new methods (EPA Methods 502,2,524.2, and  551.1) have replaced the two
original analytical methods (EPA Methods 502.1 and 524.1), While EPA 551.1 is the most
sensitive of the three currently approved methods, this method is not currently used by EPA or
State laboratories according to the available WS data.  Instead, EPA 524.2, the least sensitive of
Methods Support Document for Six-Year Review
' Draft ~ March 2002

-------
the three current methods, has been the primary method of choice. The MDL of EPA 524.2
represents a mixed change in analytical methods, as it is more sensitive "than EPA 524.1 but less
sensitive than EPA 502.1.. Evaluation of the quantitative PE data showed that the majority of the
laboratories conducting WS analyses surpassed the 75 percent criterion. Because of the high
percentage of laboratories passing and a lack of spike samples at concentrations below the
current PQL, a re-evaluation of the PQL could not be performed using this approach.  However,
high passing rates at values close to the current PQL of 5 jigflL suggest that a lower PQL is
possible.        .                                                 .      '


Chlordane

Results of the Method Comparison.

   With the 1991 promulgation of the Phase II Rule for SOCs, three analytical methods were
approved for the determination of chlordane in drinking water: EPA Methods 505
(GC/rnicroextractitm), 508 (GC/ECD), and 525.1 (LSE/OC/MS) (56 FR 3526), Since the
promulgation of this rule, EPA 525.1 was removed and EPA Methods 525.2 (LSE? GC/MS) and
508.1 (GC/LSE/ECD) were added to the approved list  Table 18 lists detection limits for these
methods. Using the highest value of the range of MDLs reported, EPA Methods 508 and 508.1
are both approximately 30 times more sensitive than EPA 505.
Table 18. Results of the Analytical Methods Comparison for Chlordane (Newly
          Promulgated Methods are Indicated in Bold)
MCL = 2-u,gflL Current PQL « 2 jig/L 0k* - 0.2 u.g/L Acceptance Limit* => ± 45%
Methods Approved At Promulgation
Method
EPA SOS1
EPA 508'
EPA 525.1*

Technique
GC, microextractioii •
GC with BCD
. LSE, GC/MS

MDL
(ug/L)
0.14
0.0015
2

Currently Approved Methods (141.24)
Method
EPA 5052
- EPA 5082
EPA S08.12
EPA 525.2s
Technique
GC, microextraction
GC, BCD
LSE, GC with BCD
LSE, GC/MS •
MDL
(ug/L)
0.14
0.00041-0.0041*
0.001 - 0.004* .
0.065-0.17*
1 "Methods for the Determination of Organic Compounds in Drinking Water,'' EPA-600/4-S8/039, December
1988.
2 "Methods for the Determination of Oiganic Compounds in Drinking Wrter-Sxipplemeut III,"
EPA/60Q/R-95- 131, August 1995. . ,
A Regulatory DLs for organic compounds are listed at 40 CFR § 1 4L24{h)(l 8),
f Acceptance limits for oiganic compounds are listed at 40 CFR §I4L24(h)(I9)(i),
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance.
Methods Support Document for Six-Year Review
•39
Draft - March 2002

-------
 Results of the Analysis of the WS Data

 a.  Method Usage Over Time

     Figure 14 plots the distribution of analytical techniques used by EPA'and State laboratories in
 WS 34 to 41. The "other" techniques represent methods which were not specifically identified
 by participating laboratories or were otherwise unknown. As shown in Figure 14, the majority of
 laboratories used EPA Method 508 for.determination of chlordane in WS 34 to 41, EPA 525,1,
 which was used quite minimally in earlier WS studies, was replaced by EPA 525.2 during WS
 36. Laboratories also began to employ EPA Methods 508.1 during WS 36. For WS 34 to 41,
 EPA 505 consistently remained the second-most commonly used method after EPA 508,

      Figure 14.  Distribution of Analytical Techniques by WS Study: Chlordane

                                        Chlordane
IVW/U *
90%
80% •
•g 70% •
£
H 60%
en
| 50%
.g 40%
ra
_j
§9 30%
20%
10%
0% -





m















81
™



r-, ri




















In













in













IfL











i •



















l













n




. 	 5
• 505
ID 508
j H 508.1
i IB 525.1
I
JD 525.2

(•other


WS 34 WS 35 WS 36 WS 37 WS 38 WS 39 WS 40 WS 41
Water Study

b.  Results of the PQL Analysis

    The Agency derived the current chlordane PQL of 2 ng/L by multiplying the detection limit
by a factor often (56 FR 3552). With the availability of recent PE WS data, efforts were made to
reassess the PQL using PE data from WS 24 to 41. Table 19 summarizes the data from these WS
studies, indicating the study number, the spiked or "true" value of the WS sample, the number of
results from EPA and State laboratories, and the calculated percentage of laboratories whose
results successfully passed within federally designated acceptance limits for chlordane. These
acceptance limits are specified in 40 CFR §141.24(h)( 19)(i) to be ± 45 percent.
Methods Support Document for Six-Year Review
40
                         Draft - March 2002

-------
Table 19. Evaluation of CMordane Data from WS Studies Using the 45%  Acceptance
          Limits (in Order of Increasing Concentration)
ws#
29
24a
27
26a
33
41.
25a
39
30
37
24b
31
32
34
38
26b
40
25b
35'
36
Spiked «Tr«e" Value (jig/L)
0,833
1,32
1,84
2.70
2.76
• 2.90
3.30
3,57
4.20
4.44
4.86
5,16
.5.33-
7.26
8.20
9,60
11.8
12,6
13.6 .
16.7
# Results from EPA and
State Labs
25
40
28 .
39
36
41
22
41
49
46
40
30
56
. 52
51
39
56
22
33
52
% Labs Passing ± 45%
Acceptance Limits
88.0
87,5
96,4
94.9
88,9
95.1
95.5
95,1
93.9
93.5
85.0
83.3
83,9
90.4
98,0
97,4
94,6 •
100
93,9
94.2
    Table 19 shows that EPA and State laboratories consistently demonstrated high success rates,
surpassing the 75 percent criterion for all evaluated water studies. Therefore, a new PQL could
not be re-evaluated using this approach. Even at low spiked concentrations, including three
samples below the current PQL, well over 80 percent of participating laboratories successfully
determined results within the acceptance limits. This observation suggests the possibility that
laboratories may be capable of determining chlordane at levels below the current PQL.
Methods Support Document for Six-Year Review
41
Draft - March 2002

-------
 Conclusion for Chlordane

     The plot of method usage over time (from WS studies) indicates consistent use'of EPA
 Methods 508 and 505 over the duration of the selected studies, with laboratories favoring usage
 of EPA 508. EPA 508 represents the most sensitive currently approved method (Table 18),  It
 appears that analytical capabilities for chlordane overall have improved since the time of
 NPDWR promulgation. The available numerical WS data do not support a reassessment of the
 PQL based on the 75 percent criterion, however, because the'passing rates of laboratories always
 exceeded this value. Because such high passing rates were observed for some low spiked
 concentrations (e.g., 88 percent laboratory success for a concentration 2.5 times lower than the
 current PQL), it is possible that a lowered PQL might be appropriate.  However, the desired
 quantitative reassessment of the PQL could not be performed using the historical approach,


 Chromium

 Results of the Method Comparison

    In 1991, the Phase II rule for lOCs listed EPA Methods 218.2 (AAF) and 200.7 (ICP-AES) as
 the approved methods for determination of chromium in drinking water (57 FR 31776).  Since
 that time, EPA 218,2 has been removed from the list of approved methods and four new methods
 have been added: two EPA Methods (200.8, ICP/MS; and 200.9; AA/Platform); and two
 Standard Methods (3113B, AA/Furnace; and 3120B, ICP). The sensitivity of the continuing
 method, EPA 200.7, appears to have remained similar over time, while the newer method (EPA
 200.9) exhibits approximately ten times the sensitivity of the prior methods.  EPA 200.8 is
 comparable in sensitivity to the old EPA 218.2.  The detection limits of the Standard Methods are
 not specified. Table 20 summarizes the MDL information for all current and former approved   '
 methods.
Methods Support Document for Six-Year Review
42
                         Draft - March 2002

-------
  Table 20.  Results of the Analytical Methods Comparison for Chromium (Newly
            Promulgated Methods in Bold)
MCL~100|ig/L Current FQ&--;tO:|ig^ ' BI/>i«7u£/L • Acceptance Omitt =*««/„ •
I Methods Approved At Promulgation
Method
'EPA 200.7 A !
EPA21S.22



Technique
Inductively Coupled
Plasma
AA; Furnace



MBL
toeW •
1.0-7,0*
1.0


1
Currently Approved Methods (1 4U23)
Method
EPA 200.7J
EPA 200.83
EPA 200.93
SM3JI3B4
SM3120B4
Technique
Inductively Coupled
Plasma
1CP-MS
AA; Platform
AA; Furnace
Inducti.ve.Iy Coupled
Plasma
MDL
Ctfi/L)
• 4.0
0,9
0.1
N/A°
N/A^
SrtUios?yif rilf r1^8 At0mi Eraission Aml*sh of Drij*tag Wter," Appendix to Method 200.7,
2«ti 1 A * lu P^' Environmental .Monitonog and Support Laboratory Cincinnati, OH 45268
3 «5? H10? f°f *ihe™1™1 Analysis of Water and Wastes (MCAWW)," EPA/600/4-79-020 March 1983
1994 Determination of Metals in Environmental Samples Supplement V £k/600/R-94/i 1 1 , May
md WaS|eWater' Ame~ PuWic «««* Auoctatio-,, ^15
(MDL) VakieS ^ fr0ra VariaMi* °f rea^tS, insmunentation, and/or
onm^M^°^miC C°mpOUnds are lkted at 40 CFR §I4L23(a)(4)(I). The value may vaty depending
1 Acceptance limits are listed at 40 CFR,§ !4L23(k)(3)(ii) for inorganic compounds.
N/A = not available. MDLs are not specified for non-EPA methods
 Results of the Analysis of the WS Data.        '''•.''.

 a.  Method Usage Over Time

    The distribution of analytical methods used by participating laboratories from WS 34 to 41 w
 shown in Figure 15 J^J^ for "ote" techniques in this figure include the use of any other
 technique identified by the laboratories participating in the WS study, as well as "unknown"
 methods, i,e  methods for which laboratories did not report any information on the type of
 method used. During WS 34.and 35, EPA Methods 200.7 and 218.2 were the most widely used
 ?fvf1l^f^P^^g in the PE studies. By WS 36, EPA 218.2 was .no longer in use.
 Oveiall, EPA 200.7 remained .the most commonly used method during WS 34-41  Voluntary
 consensus standard methods SM 3120 and SM 3113B were also used in many WS studies.
Methods Support Document for Six-Year Review      43
                                                                  Draff - March 2002

-------
        Figure IS.  Distribution of Analytical Techniques by WS Study: Chromium
                                           Chromium
        £
        "o
         0)
i ww /\J
90%
80% •
70% -

60%

50%
40%

30% -
20% -

10%
0% •



















I



,
1
1




!r












1





1
is



%
j •'
'U.



i
i
li



i
i
i
fli.


R
' -i
^ ^
| , fef


1 D200".7~
H 200.8 !
• 200.9
• 2ia,2
D 608.1
H 31 138!
i
S3120B
• other !



                 WS34   WS35  WS36   WS 37   WS 38   WS 39   WS40   WS41
                                         Water Study
  b.  Results of the PQL Analysis
 FR 1549") F   h "pnf " r"°"""" """ ""tIYV'" using earlier PE data from WS 24 through 27 (56
 more recent PEdatirw^Ttnif^T^^ T6 ^ fr°m abroader ranSe «>f studies, including
 number, the spikSSS^S                                           St^
  ^?o°^^d-iherep°rted results evaluated usi«8 ^ acceptance liml of ±  U  !
 (§i4i.2J(k)(3)(n)) are summarized in TabJe 21.
Methotk Support Document for Six- Year Review
44
                                                                   Draft -

-------
 Table 21.  Evaluation of Chromium Data from WS Studies Using the 15% Acceptance
           Limits (In Order of Increasing Concentration)
ws#
34
25a
39
24b
36
26a
41
25b
32 -
37
27
31
40
26b
29
35
24a
38
33
30
Spiked "True" Value (ug/L)
11.6
15.0
23.9
25.5
37,8
5.0,2
55,5
60,0
68.1
72.9
75.3
81.6
90.9
94,6
no
119
127
148
159
200
# Results from EPA and State
Labs
65
41
50
62
65
64
47
42
67
51
35
38
65
64
33
43
62
66
35
66
% Labs Passing* 15%
Acceptance Limits
. 93.8
92.7
. • ' 94.0
90,3 •
95,4
95.3
100
95.2
95.5
98.0
94.3
92.1
96.9
93,8
87.9
90.7
88.7
95.5
91.4
93.9
    A re-evaluation of the PQL could not be performed using, the available PE data in Table 21.
To conduct a graphical PQL analysis requires laboratory success rates ranging below the 75
percent criterion, which participating laboratories consistently surpassed (achieving a passing rate
of greater than 87 percent for all the WS studies evaluated). Furthermore, the spiked
concentrations used in each water study were above the original PQL (10 u-g/L).
Methods Support Document far Six-Year Review
45
Draft -March 2002

-------
   Conclusion for Chromium
    *•            comparison results indicate that some methods approved after the promulgation
  of the Phase II Rule are more sensitive -than the original methods (one of which was
  discontinued). Currently the most sensitive method is EPA 200,9 (AA-fiunace)  However
  according to the plot of method usage over time, EPA 200.7 (JCP-AES) consistently has  *
  remained the most frequent choice for EPA and State laboratories, and the MDL of this method
  has not changed significantly.  These facts suggest no significant alteration to the analytical
  capabilities of laboratories. Using the designated ± 15 percent acceptance limit, the evaluation of
  T,?1 nJS d^a revealed tiiat these data are outside a range that would allow for a re-evaluation
  or the PQL.  I hus, the PE data continue to support the existing PQL of 1 0 p.g/L.
  l,2-Dibromo-3-chloropropane(DBCP)

  Results of the Method Comparison

     The NPDWR for DBCP, a Phase II SOQ listed EPA 504 (GC with microextraction) as the
  only approved method for determination of this compound (56 FR 3526), Since then EPA has
  replaced this method with an updated GC-microextraction method, EPA 504 1 and added a new
  technology, LLE/GC with BCD (EPA 551.1). Table 22 sumtnWs th£ approwTme&ol at
  promulgation and currently approved methods. EPA 504.1 has an MDL roughly equal to the  '
  lowest MDL of the original EPA 504, and the MDL of the other new method, EPA .551,1, is less
 Table 22. Results of the Analytical Methods Comparison for DBCP (Newly Promulgated
           Methods arc Indicated in Bold)                                       to
   MCL-0.2,ig/L     Current PQL - 0.2
                               DL* - 0,02
                    Acceptance Limit*
       Methods Approved At Promulgation
                                   Currently Approved Methods (141,24)
    Method
    Technique
                                   MDL*
                                             Method
                                              Technique
                                                                               MDL*
   EPA 504'
Microextraction, GC
                                  0.01, - 0.2
EPA 504.1*
                                                         Microextraction, GC
                                                                 0.01
                                            EPA 551. I
                                          LLE/GC with ECD
                                                                             0.14-05*
           f°r *e Determination of Oiganic Compounds in Drinking Water," EPA-600/4-88/039, December

            ^              of Oiganic Compounds in Drinking Wrier-Supplement III,"

  ^Regulatory DLs for organic compounds are listed at 40 CFR §141.24{h)(18)
   Acceptance limits for organic compounds are listed at are listed at 40 CFR 6141 2
Methods Support Document for Six- Year Revicnv
                           46
                                                                    Draft - March 2002

-------
Results of the Analysis of the WS Data

a.  Method Usage Over Time
   Figure 16 is a plot of the distribution of analytical techniques used by EPA and State
laboratories in WS 34 to 41, The, "other" techniques represent methods that were not specifically
identified by participating laboratories or were otherwise unknown. As shown in Figure 16, the
majority of laboratories used EPA 504,1 during WS studies 34 to 41. The other recently
approved method, EPA 551, was,only used minimally during WS 36 to 39. Participating
laboratories chose "other" methods approximately 10 to 15 percent of the time.
     Figure 16.  Distribution of Analytical Techniques by WS Study: DBCP

                                          DBCP
        •o
        I
        "5
        jg
        en
        .5
        "tn
        Ul
        xt
100%
90% ,
80% .
70% .
'60% .
50% -
40% .
30% .
20% .
10% .
0%


























03

—

























W"
1
ll

WS 34 WS 35













_
















••









1

-•'. • 	 	
WS36



^









ww















I
















_JBl3__
















H
•1
J— I

















I
TO



















H
WS37 WS38 WS39 WS 40 WS 41
Water Study
                                  B551
                                  j g other
b, Results of the PQL Analysis

   For DBCP, the original PQL of 0.2 jig/L was derived by multiplying the detection limit of
0,02 u.g/L by a factor often (56 FR 3551).  Recent PE data from WS 26 to 41 have enabled EPA
to attempt a reassessment of the PQL,  Table 23 summarizes the data from these WS studies
(except WS 28 and 33 which did not contain data), indicating the study number, the true value of
the WS sample, the number of results from EPA and State laboratories, and the calculated
percentage  of laboratories whose results successfully passed within the 40 percent acceptance
limits for DBCP.
Methods Support Document/or Six-Year Review
47
Draft - March 2002

-------
 Table 23.  Evaluation of DBCP Data from WS Studies Using the ± 40% Acceptance Limits
           (in Order of Increasing Concentration)
\vs#
36
32
39
37
34
38
41
40
35
27
29
26
31
30
Spiked "True" Value (fig/L)
0.196
0.233
0.246
0.286
0,363
0.429
0.451
0.527
0.589
0.653
0,980
1.13
1.78
2.65
# Results from EPA and
State Labs
50
24
36
42
45
48
34 '
50
29
18
39
35
44
24
% Labs Passing* 40%
Acceptance Lira its
92,0
95.8
91.7
88.1
88.9
97.9
100
98.0
93! 1
88.9
94.9
88,6
97.7
91 7
    Based on the data in Table 23, a revised PQL value could not be estimated because the
passing rate of participating laboratories surpassed the 75 percent criterion needed. Also, the true
values of the spiked samples in all but one study (WS 36) were greater than the existing PQL,
limiting potential conclusions regarding a lower PQL. Therefore, a regression analysis was not
performed. However, high passing rates at concentrations around the current PQL of 0.2 ug/L
are suggestive of a change in the PQL.

Conclusion for DBCP

    The MDL of EPA 504.1, a GC-microextraction technique, has not changed substantially
since the promulgation of the NPDWR for DBCP.  According to recent WS study data.
laboratories have elected to use EPA 504.1 more often than EPA 551.1. These observations
imply that detection limits for the overall contaminant have remained fairly constant over time.
Using the data compiled from WS 26 to 41, the current PQL was reassessed.  However,
laboratories in all studies surpassed the 75 percent passing rate used in a PQL re-evaluation, such
that a regression analysis was not attempted. Furthermore, the true value concentrations
exhibited in almost all of the studies were higher than the current PQL, meaning that the effects
Methods Support Document for Six-Year Review
Draft - March 2002

-------
of lower PQL could not be determined. Although, the available PE data were not suitable to
recalculate a new PQL, the high passing rates are suggestive of a change in the PQL,


1,4-DichIorobenzeiie (para-dichlorobenzene)

Results of the Method Comparison        .

   With, the promulgation of the final rule for Phase I VOCs in My 1987 (52 FR 25690), five
approved methods were listed for the determination of 1,4-dichIorobenzene (also known as para-
dichlorobenzcene) in drinking water: EPA Methods 502.1, 502.2, 503.1, 524.1, and. 524,2. Since
the promulgation of this rule, the Agency has removed EPA Methods 502.1, 503.1, and 524.1
from the list of approved 'methods. Table 24 summarizes the original and current methods and
their MDLs.
Table 24. Results of the Analytical Methods Comparison for 1,4-Bichlorobenzene
MCI* = 75 -ue/L Current PQL = S ttg/L Bl/ = 0.5 Hg/L Acceptance Limit* = ± 20% (>10 jig/L) or
* • ± 40% (<10 u-g/L)
Methods Approved At Promulgation
Method
EPA 502. 11
EPA 502.2'
EPA 503. 11
EPA 524. 1'
EPA 524.21
Technique
Purge and Trap GC
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL*
Oig/L)
ND*
0.01
' 0.006
2,0
0,03
1 "Methods for the Determination of Qiganic Compel
1988,
2 "Methods for the Determination of Oiganic Compoi
131, August 1995.
0 The MDLs of the original methods for this contami
Federal Register notice promulgating NPDWRs for tl
manual cited in footnote I lists the MDLs shown abo\
* Multiple method detection limit (MDL) values resu
laboratory/analyst performance.
* Regulatory DLs for VOCs are listed at 40 CFR § P
t Acceptance limits For organic compounds are listed
" Not determined
Currently Approved Methods (141.24)
Method
EPA 502.22
EPA 524.2Z



Technique
Purge and Trap GC
GC/MS



MDL*
Gig/L)
0,05 - 0,29
0.03 - 0.04



inds ia Drinking "ftater," EPA/600/4-88/039, December
inds in Drinking %ter--Supplement III," EPA/600/R-95-
nant ranged from 0.2 - 0-5 u-g/L according to the July 1987
w VOCs (52 FR 25690). Howevei; the 1988 methods .
'6.
t from variability of reagents, instrumentation and/or
ll,24(f)(17)
at 40 CFR §141.24(f)(17Xi)
 Methods Support. Document for Six-Year Review
49
Draft - March 2002

-------
 Results of the Analysis of the WS Data

 a.  Method Usage Over Time

     The distribution of different methods used by EPA and State laboratories during WS 34 to 41
 is charted in Figure 17. The category of "other" methods includes those methods that were
 unknown or otherwise unidentified by the participating laboratories.  During WS 34 to 41, the
 use of EPA 524.2 increased while use of EPA 502.2 diminished slightly.


        Figure 17. Distribution of Analytical Techniques by WS Study: 1,4-
        Dichlorobenzene
                                       1,4 Dichlorobenzene
           100%
                                                                                |B 502.21
                                                                                JP 524.2;
                                                                                i II other •
                  WS 34  WS 35   WS 36  • WS 37   WS 38   WS 39   WS.40   WS 41
                                           Water Study
b.  Results of the PQL Analysis
    The original PQL of 5 p,g/L for 1,4-dichlorobenzene was determined by using PE data from
Water Supply Studies 8 to 11- (50 FR 46880). For the six-year regulatory review, more recent
WS data were compiled to provide a more accurate, updated assessment of laboratory
capabilities. Hence, data  from WS studies 24 to 27 and 29 to 41 were used to attempt to re-
evaluate the PQL. Table 25 summarizes the available PE data by providing the study number,
spiked value for the WS sample, number of laboratory results, and percentage of laboratories  '
passing the proficiency test within the acceptance limits, meaning their reported results fall
within the designated acceptance limits for a particular contaminant.  The acceptance limits for
1,4-dichlorobenzene are ± 20 percent for a true value greater than 10  p,g/L, or ± 40 percent for a
true value lower than 10 u-g/L (as specified at 40 CFR  § 14L24(f)(17)(i)),
Methods Support Document for Six-Year Revie\v
50
                          Draft - March 2002

-------
Table 25. Evaluation of M-Diehlorobenzene Data *"<>*» ws Studies Using the ± 20% or ±
          40% Acceptance Limits (in Order of Increasing Concentration)
ws#
24
34
29
37
31
27
40
36
32
38
.26
33
.41
30
35
39
25
Spiked "True"
Value (ng/L)
2,50 .
5.78
6.60
7.31
9.40
9.5S
11,6
11.9
13.6
14.2
1.4.6 :
15.1
15.8
16.1
16.7
17.8
20.8
# Results from EPA
and State Labs
57
59
34
47
36
38
57
60
61
54 .
60
33
42
58 . . '
33
• 42
37
% Labs Passing ± 20%
Acceptance Limits






87.7
95.0
93.4
55.6
-93.3 .
87.9
92,9
84.5
97,0
88.1
83.8
% Labs Passing ±40%
Acceptance Limits
96,5
100
97.1
95,7
97.2
100











    Because a very large percentage of EPA and State laboratories passed the proficiency test
within the bounds of the designated acceptance limits, the PQL could not be determined .using
the historical 75 percent criterion. Therefore, a regression analysis could not be conducted.
However, the high passing rates suggest that the PQL could be lower.

Conclusion for 1,4-Dichlorpbenzene

    Since the promulgation of the NPDWR for 1,4-dichlorobenzene, the two analytical methods
approved in  1987, EPA Methods 502.2 and EPA 524,25 are still approved for use today but are
hot more sensitive. According to the distribution, of analytical methods usage over time, EPA
524.2 was more widely used than EPA 502.2 during WS 34 to 41. Upon review of the WS, data,
a high percentage of laboratories successfully passed the proficiency tests, preventing a graphical
Methods Support Document for Six-Year Review
51
Draft - March 2002

-------
estimated assessment of the PQL at the 75 percent passing rate. Although the available PE data
are insufficient to recalculate the PQL, high passing rates are suggestive of a change in the PQL.


1,2-Dichloroethane

Results of the Method Comparison

   The approved drinking water methods for the determination of 1,2-dicMorpethane, a Phase I
VOC (52 FR 25690). These methods all utilize GC or GC/MS with, several extraction and/or
detector variations: EPA Methods 502.1, 502.2, 503.1,524.1, and 524.2.  Since promulgation of
these original methods, the Agency has removed EPA Methods 502.1, 503.1, and 524.1 from the
list of approved methods, and has continued to approve the use of EPA methods 502.2 and 524.2.
Table 26 summarizes the current and previous EPA methods along with their MDLs.


Table 26.  Results of the Analytical Methods Comparison for l,2-I)ichloroethatte
MCL =* 5 ug/L Current PQL- 5 ug/L BL* = 0.5 |ig/L Acceptance taiBitf = ± 20% (>10|ig/L) or
±40%(
0.07
0,06



1 "Methods for the Determination of Oiganic Compounds in Drinking Mter," EPA/600/4-88/039, December
1988.
1 "Methods for the Determination of Oiganic Compounds in Drinking Water— Supplement III," EPA/600/R-95-
131, August 1995.
* The MDLs of the original methods for this contaminant ranged from 0.2 - 0.5 |ag/L according to the July 1 987
Federal Register notice promulgating NPDWRs for the VOCs (52 FR 25690). Howevei; the 1988 methods
manual cited in footnote 1 lists the MDLs shown above.
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance.
* Regulatory DLs for VOCs are listed at 40 CFR § I41.24(f)(17).
t Acceptance limits are listed at 40 CFR § 141.24(f)(17)(i)
Methods Support Document for Six-Year Review
52
Draft ~ March 2002

-------
 Results of the Analysis., of. the WS Data .    .       ...  .       .

 a.  Method Usage Over Time

    The types of methods used over time by-the EPA and State laboratories during WS studies 34
 to 41 are illustrated in Figure 18. The results for "other" techniques in this figure include any
 unknown or unreported methods.  Using Figure 18, it is apparent that laboratories have
 increasingly relied on EPA 524.2 over EPA 502.2 to analyze for 1,2-dichloroethane, The '
 percentage-of labs using EPA 502.2 has steadily declined while EPA 524.2 has experienced an
 increase in use over the time period between WS studies 34 and 41.

      Figure 18. Distribution of Analytical Techniques by WS Study: l,2~Dieh!oroethaMe

                                    1,2-Dichloroethane
        o
       JE
        CD
       .£
       'at

        v>
       £1
        ra
           o%
                WS 34   WS 35  WS 36
WS 37   WS 38

 Water Study
WS 39   WS 40  WS 41
b.  Results of the PQL Analysis

 .   The old PQL of 5 u-g/L (52 FR 25700) for 1,2-dichloroethane was established by using the
data from WS PE studies 8 to 11. To re-evaluate the PQL., data were taken, from WS 24 to 41
(note that data was unavailable for WS 28), This data, including the study number, the spiked or
"true" value for the WS sample, the number of laboratory results, and the percent of laboratories
passing within the accepted limits of ± 20 percent for a spiked value of > 10 ng/L and ± 40
percent for a spiked value of < 10 u-g/L, are compiled and illustrated in Table 27,
Methods Support Document for Six-Year Review
    53
              Draft - March 2002

-------
Table 27. Evaluation of 1,2-Diehloroethane Data from WS Studies Using Either 20% or
          40% Acceptance Limits (in Order of Increasing Concentration)
ws#
27
30
36
31
26
34
29
24
37
32
41
35
25
38
40
33
39
Spiked "True"
Value (|ig/L)
4.88
7.69
9.00
9.25
10.8
12.1
12.9
13.2
13.2
13.3
13.7
14.1
15.5
15.6
15.6
16.9
17.6
# Results from EPA
and State Labs
37
59
59
36
59
60
34
56
47
63
41
34
38
54
57
35
42
% Labs Passing
± 20% Acceptance Limit




95
98
79
89
81
94
93
94
90
93
88
.97
86
%. Labs Passing
* 40% Acceptance Limit
100
98
100
too













    As shown in Table 27, the percentage of laboratories passing the acceptance limit averaged
over 90 percent which is well above the 75 percent passing criterion selected to determine the
PQL. Also, the spiked (or "true") values which the laboratories received were higher than the
original PQL of 5 [ig/L (with the exception of WS 27). Therefore, recalculation of the PQL
could not be performed with these data. However, the high passing rates at concentrations close
to the current PQL suggest that the PQL could be lower,

Conclusion for 1.2-Dichloroethane

    Two new methods (EPA Methods 502.2 and 524.2) have been approved for the
determination of 1,2-dichloroethane since promulgation of analytical methods under Phase I.
One of the original methods, EPA 502.1. had the greatest detection sensitivity of the five
methods mentioned but is currently not approved for analysis. The method usage evaluation
Methods Support Document for Six-Year Review
54
Draft - March 2002

-------
 shows tihat, of the two currently approved methods, EPA'524,2 has steadily become the more
 preferred method of analysis by laboratories.  Based on an analysis of the WS data, there are not
 enough appropriate data to conduct a reassessment of the PQL.  However, high laboratory
 passing rates at concentrations close to the PQL 5 ^g/L are suggestive of a change in the PQL.


 1-,1-DichloroethyIene

 Results of the Method Comparison

    The analytical methods approved for the determination of 1,1 -dichloroethylene under the
 NPDWRs for Phase I VQCs include EPA Methods 502.1, 503.1, and 524.1 (52 FR 25899).
 Since the promulgation of the rule in, 1987, the Agency has added EPA 245.2, to the list of
 approved methods. The currently approved methods for 1,1 -dichloroethylene determination are
 EPA Methods 502.2 and 524.2. Table 28 summarizes the MDLs for both the original and current
 approved versions of the methods,!
Table 28. Results of the Analytical Methods Comparison for 1,1-DichIoroethylene (Newly
          Promulgated Methods Indicated in Bold)
MCL « 7 jig/L Current PQL = 5 jtg/JL ' DL* = 0,5 jxg/L Acceptance Limit'*' « ± 20% (>W |ig/L) or
±40%(<10u.g/L)
Methods Approved At Promulgation
Method
EPA 502. 1s
EPA 503. 11
EPA 524. 11
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
MDL
(tig/L)
0.003
0.2 - 0.5
0.2
Currently Approved Methods
Method
EPA 502.2s
EPA S24.21

Technique
. Purge and Trap GC
GC/MS

MDL
• fcg/L)
0.0?
0.12

1 "Methods for the Determination of Organic Compounds in Drinking Water," EPA/600/4-88/039, December
1988. . •
2 "Methods for the Determination of Ojganic Compounds in Drinking vv&ter--Supplement III," EPA/6Q0/R-95-
131, August 1995. .
* The MDLs of the original methods for this contaminant ranged from 0.2 - 0.5 jj,g/L according to the July 1987
Federal Register notice promulgating NPDWRs for the VOCs (52 FR 25690). However, the 1988 methods
manual cited in footnote 1, lists the MDLs shown above, . .
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance.
* Regulatory DLs for organic compounds are listed at 40 CFR § 141,24(f)(17).
t Acceptance limits'for VOCs are listed at 40 CFR § 14 1 ,24(f){ 1 ?)(i).
Methods Support Document for Six-Year Review
55
Draft - March 2002

-------
Results of the Analysis of the WS Data
a.  Method Usage Over Time


    Figure 19 shows the distribution of analytical techniques used by EPA and State laboratories
for WS studies 34 to 41. The results for "other" techniques in this figure include the use of any
other technique identified by the laboratories participating in the WS study, as well as
"unknown" methods, i.e., methods for which laboratories did not report any information on the
type of method used. As shown in Figure 19, EPA 502.2 was used less as EPA 524.2 was used
to most often (as shown from WS 34 to 41).
     Figure 19. Distribution of Analytical Techniques by WS Study: 1,1-
     Dichloroethylenc
        100%
                                     1,1 -Dichloroethy lene
               WS 34  WS 35   WS 36
WS 37   WS 38

     WS
WS 39   WS 40   WS 41
b. Results of the PQL Analysis


   The current PQL of 5 |ig/L was originally set using previous PE data (54 FR 22102), With
the availability of more current data from WS 24 to 41, a PQL re-evaluation was attempted.
Table 29 summarizes the results of these studies, including the study number, the spiked (or
"true") value for the sample, the number of laboratory results, and the percent of laboratories
Methods Support Document for Six-Year Review
    56
              Draft - March 2002

-------
passing the WS proficiency test for 1,1-di.chloroethylene within the acceptance limits. The
acceptance limits were calculated as ± 20 percent for a spike value of >10 u.g/L and ± 40 percent
for a spiked value of <10 |Ag/L (as specified at 40 CFR § 141.23(0(1 ?)(*))-


Table 29. Evaluation of 1,1-DicUoroethyleae Data from WS Studies Using Either 20% or
          40% Acceptance Limits (in Ordler of Increasing Concentration)
ws#
41
24
26
31
34
36
32 '
27
29
38
39
33
35
30
25
37
40
Spiked "True"
Value (iig/JL)
. 5.25
5.36
6.64
7,02
7.64
8.49
9.13
9.45
11.7
11.7
12,4
12.9 '
13.9
14.2
14,9
16.5
18.3
# Results from EPA
and State Labs
41
57
60
37 ,
60
• 60
63
37
35
56
43
33
35
45
38
48
58
% Labs Passing ± 20%
Acceptance Limits








74.3
19.6
79J '
75.8
88.6
77,8 .
94.7
83,3
86.2
% Labs Passing ± 40%
Acceptance Limits
97,6
87.7
95.0
97,3
98.3
98.3
100
89.2









    The data from the available PE studies were not conducive to PQL re-evaluation, as the
percentage of labs passing generally exceeded the standard 75 percent passing criterion needed to
estimate the PQL using the graphical approach (with the exception of two studies). However,
high laboratory passing rates for those spike concentrations just above the PQL of 5 }ig/L suggest
that the PQL could be lower.
Methods. Support Document for Six-Year Review
57
Draft - March 2002

-------
 Conclusion for Ll-Dichloroethylene


    The method comparison results indicate that EPA 502.2 is now the most sensitive method for
 determination of 1,1-dichloroethylene in drinking water.  As revealed by the results of method
 usage over time, EPA 524.2 is the most commonly employed method for 1,1-dichloroethylene
 determination in recent PE studies. Based on the evaluation of more recent quantitative PE data,
 a recalculation of the PQL is not possible. However, the high laboratory passing rates for a
 couple spike samples with concentrations slightly above the current PQL of 5 u-g/L suggest that
 the PQL could change.


 Dichloromcthane (methylene chloride)


 Results of the Method Comparison


    At the promulgation of the Phase V rule for VOCs (57 FR 31776), four analytical methods
 were approved (EPA Methods 502.1, 502.2,524.1, and 524.2) for the analysis of
 dichloromethane (also known as methylene chloride). Since that time, EPA Methods 502.1 and
 524.1 have been removed from the approved list, leaving EPA Methods 502.2 and 524.2 as the
 remaining currently approved methods; no new methods ha%?e been introduced.  Table 30
 provides descriptions of the methods and their MDLs. The MDLs of the two current methods
 remain unchanged from their values at the promulgation of the rule.


 Table 30.  Results of the Analytical Methods Comparison for Dichloromethane
MCL = 5ng/L Current PQL = 5 jig/L JW/ - 0.5 p.g/L Acceptance Limit'*' = ± 20% (>10 jig/L) or
±40%(<10>ig/L)
Methods Approved At Promulgation
Method
EPA 502. 11
EPA 502.2'
EPA 524. 11
EPA 524.21
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL
(Hg/L)
unknown
0.02
OJ3-0.25*
0.03
Currently Approved Methods (141.24)
Method
EPA 5Q2.22
EPA 524.2J


Technique
Purge and Trap GC
GC/MS


MDL
' (ng/D
0.02
,0,03 -


Afeihods Support Document for Six-Year Review
58
Draft - March 2002 ••

-------
 1 "Methods for the Determination of Ojganic Compounds in Finished Drinking Water and Raw Source Water,"
 June 1985.
 2 "Methods for the Determination of Oiganic Compounds in Drinking Mter-Supplement III," EPA/60Q/R-95-
 131, August 1995.               '    .                                  ..    •      '
 * Regulatory DLS for organic compounds are listed at 40 CFR 141.24(0(17).
 t Acceptance limits for VOCs are listed at 40 CFR 141.24(0(17)(i).          •
  Multiple method detection limit (MJ>L) values result from variability of reagents, instrumentation, and/or
                   " irmanee.	;	,„,„,	:	.• •     	
Results of the Analysis of the WS Data                                     .


a.   Method Usage Over Time .             .                            '  •  .  .


    The distribution of methods used over WS studies 34 to 41 is illustrated in Figure 20. The
designation "other" includes all unknown and unreported methods. As illustrated in Figure 20,
EPA 524.2 was the predominant method used in these WS studies. Since WS 35, EPA 524,2 has
steadily increased in its xisages experiencing only a slight decrease in WS 41. Even though EPA
502.2 features slightly better sensitivity than EPA 524.2, laboratories favored EPA 524.2, which
utilizes GC/MS,         •                                        .

   Figure 20, Distribution of Analytical Techniques by WS Study: Dichloromethane

                                    Dichtoromethane
     "O
     o
     1
     CO
                                  [•524,21
                                  JD502.2I
                                  ! m. other !
               WS34   WS35   WS36   WS 37   WS 38  WS 39  WS40   WS41

                                         Water Study
Methods Support Document for Six-year Review
59
Draft - March 2002

-------
 b.  Results of the PQL Analysis


    Table 31 summarizes the results of the data from WS 24 to 41 for the use in the PQL .re-
 evaluation. Note that data for WS 24,25,27,28, 30, and 31 are not available for this analysis,
 Table 31 includes the WS number, the spiked "true" value, number of labs that participated in the
 studies, and the percentage of those passing within the acceptance limit designated for
 dichloromethane (± 20 percent if spiked "true" value is > 10 \ig/L or ± 40 percent if the spiked
 "true" value is < 10 u-g/L.specified at 141.24(f)(17)(i)),
 Table 31.  Evaluation of Dichloromethane Data from WS Studies Using Either 20% or
           40% Acceptance Limits (in Order of Increasing Concentration)
ws#
35
40
39
32
37
36
33
38
29
26
41
34
Spiked "True"
Value (ng/L)
5.83
6.2
7.31
7.77
8.41
12.3
12.8
14.7
14.7
15.2
15.9
18.4
# Results from EPA
and State Labs
35
57
42
51
47
59
34
54
33
51
40
59
% Labs Passing
± 20% Acceptance Limits





90
94
98
97
71
85
86
% Labs Passing
± 40% Acceptance Limits
97
97
100
94
92







   The original PQL of 5 u-g/L was derived from PE data from WS 22, 23, and 26 (57 FR
60953). As shown by Table 30, the WS data are not appropriate for a PQL re-evaluation using
the linear regression approach. For most of the concentrations listed in Table 31, the percentage
of laboratories passing were well above the 75 percent criterion necessary to recalculate the PQL,
In summary, these data were not suitable for use in reassessing the current PQL using a graphical
method. However, high laboratory passing rates at concentrations close to the current PQL
suggest that the PQL may be lower.
Methods Support Document for Six-Year Review
60
Draft - March 2002

-------
Conclusion for Dichloromethane


   Since the promulgation of the Phase V rule, no new methods have been approved for the
analysis of dichloromethane, and analytical capabilities have remained essentially constant  Of
the currently approved methods, EPA .524.2 is used more frequently by laboratories for the
detection of dichloromethane, although it is not the most sensitive method available (EPA
502,2).  During WS 24 to 41, the percentage of laboratories passing was very high, limiting the
possibilities of re-evaluating the PQL using the historical 75 percent criterion.. Although, data
are insufficient to recalculate the PQL, high passing rates at values close to the current PQL
suggest that the PQL could change.                ,
1,2-Dichlpropropane


Results of the Method Comparison


    The final NPDWR for 1,2-dichloropropane, a Phase IIVOG, has four approved methods:
EPA Methods 502.1, 502.2, 524.1, and 524.2,  Since that time, EPA Methods 502.1 and 524.1
have been removed from the approved list, leaving EPA Methods 502.2 and 524.2 as the
remaining currently approved methods; no new methods have been introduced, Table 32
provides descriptions of the methods and their MDLs. The MDLs of the two current methods
remain essentially unchanged from their values at the promulgation of the rule.
Table 32. Results of the Analytical Methods Comparison for 1,2-Dichloropropane (Newly
          Promulgated Methods in Bold)
MCI, <= 5 us/I, Current PQL = 5 pg/L 0I/ = 0.5 ng/L Acceptance 'Limits1' = ± 20% (>10 \ig/L) or '
- . - ' - ± 40% <<10. jig/L)
Methods Approved At Promulgation
Method
EPA 502. 11
EPA 502,2 l
EPA 524.1'
EPA 524.2'
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL
(Hg*L)
N.D.
0.01
0.2
0.02 - 0.04
Currently Approved Methods (141.24)
Method
EPA 502.2-
EPA 524:22


Technique
Purge and Trap GC
GC/MS '


.. MDL*
(ug/L)
0.02-0.03
0.02 - 0.04


Methods Support Document for Six-Year Review
61
Draft - March 2002

-------
  1 "Methods for the Determination of Ojganic Compounds in Drinking %ter," BPA-6Q0/4-88/Q39, December
  1988.
  2 "Methods for the Determination of Oiganic Compounds in Drinking Water—Supplement III," EPA/600/R-95-
  131, August 1995.
  * Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
  •laboratory/analyst performance.
  N.D. = Not determined
  » Regulatory DLs for oiganic compounds are listed at 40 CFR § 141,24(f)(17).
  t Acceptance limits for VOCs are listed at 40 CFR. § 141.24(f)(17)0),	
Results of the Analysis of the WS Data  ,
a.  Method Usage Over Time
    For PE WS studies 34 to 41, the distribution of methods used by EPA and State laboratories
is illustrated by Figure 21. The category of "other" contains those methods that were unknown or
unidentified by participating laboratories.  As Figure 21 shows, EPA 524,2 was the preferred
method for laboratories participating in WS 34 to 41.  For the most part, the use of EPA 502.2
decreased over time.
Methods Support Document for Six-Year Revie\v
62
Draft - March 2002

-------
     Figure 21. Distribution of Analytical Techniques by WS Study: 1,2-
     Dichloropropane
                                    1 ,2-DichIoropropane
         100%
      TS
      o

      tu
      S
      O>

      3
      o
                WS34   WS35   WS36
WS 37   WS 38

  Water Stydy
            W839   WS40   WS41
b.  Results of the PQL Analysis
    The original PQL of 5 ^g/L (56 FR 3526) for ] ,2-dichloroptopane was determined by using
PE data from WS 18. A re-evaluation of the PQL. was attempted using more recent data from
WS studies 29 through 41, Table 33 summarizes the results of these WS studies, including the
study number, the true value concentration of the spiked sample, the number of laboratory results
returned, and the percentage of laboratories passing the proficiency test within acceptance limits
of ± 20 percent for.a true value greater than 10 |ig/L, or ± 40 percent for a true value lower than
10 \ig/L (as specified in 141.24(i)(l 7)).
Methods Support Document for Six-Year Review
63
                             Draft - March 2002

-------
 Table 33.  Evaluation of 1,2-Diehloropropanc Data from WS Studies Using the ± 20% or ±
           40% Acceptance Limits (in Order of Increasing Concentration)
ws#
32
35
30
39
34
37
33
41
29
36
38
40
Spiked "True"
Value ftig/L)
6.46
9.00
10.9
12.2
12.3
14.2
14.3
15.4
15.8
16.4
18.3
19.0
# Results from EPA
and State Labs
61
34
58
43
59
47
32
40
33
59
55
57
% Labs Passing ±20%
Acceptance Limits


98,3
93.0
96.6
91.5
93.8
97.5
87,9
93.2
94.5
94,7
% Labs Passing ± 40%
Acceptance Limits
TOO
100










    For this contaminant, the participating laboratories studies achieved success rates far greater
than the 75 percent criterion typically used to estimate the PQL.  Also, all of the true value
concentrations used in the available studies exceeded the current PQL. Thus, the PQL could not
be re-evaluated using the regression method.  However, high passing rates of 100 percent for WS
32 with a spike concentration of 6.46 (ig/L is very close to the current PQL, This suggests that
the PQL  could be lower.
Conclusion for 1.2-Dichloropropane
    The method comparison results show that no new methods have been approved since the
promulgation of the NPDWR for 1,2-dichloropropane. Evaluation of the quantitative PE data
showed that laboratories conducting WS analyses had surpassed the 75 percent criterion,
Because the data featured a large percentage of laboratories passing, and veiy high true value
concentrations, a re-evaluation of the PQL could not be performed using the typical graphical
estimation approach.  Thus, the available WS data are insufficient to recalculate the PQL for 1,2-
dichloropropane. However, high laboratory passing rates for WS studies with concentrations
close to the current PQL are suggestive of a change in the PQL,
Methods Support Document for Six-Year Review
64
Draft - March 2002

-------
2,3,7,8-TCDD (Dioxin)
Results of the Method Comparison
   2,3,7,8-TCDD, commonly known as dioxin, was listed with the Phase V SOCs (57 FR
31776), Dioxin has not had any new methods approved for analysis since the promulgation of
the NPDWRs'for Ms contaminant However, the MDL for EPA 1613 has changed from 5 pg/L
(picogram = 10"12 gram) to 1 pg/L, indicating a slight increase in sensitivity. Table 34 shows the
specifications of Method 1613m both past and current periods.
Table 34, Results of the Analytical Methods Comparison for 2,3,7,8-TCDD
MCL=30pg/t Current PQL ==30 pg/L PL1 = 5 pg/L Acceptance Limit* = ± 2*8,0.
Methods Approved At Promulgation
Method
EPA 1613'
Technique
GC/MS
. MDL
5 :
Currently Approved Methods (141.24) '
Method
EPA 16 13'2
MDL
Technique
GC/MS i
! Method 1613, Revision A, USEPA, April 1990.
2 Method 1613, "Tetra-through Octa-chlorinated Dioxins and Fwran$ bv Isotope-Dilution HRGC/HRMS,"
EPA/821-B-94-005, October 1994.
' Regulatory DLs- for oiganic compounds are listed at 40 CFR 141. 24(h)(l 8).
f Acceptance limits for organic compounds are listed at 40 CFR i4l.24(h)(l9)(i).
Results of .the Analysis..of the. WS Q.ata
   WS data were not available for dioxin; hence, no analyses on these data could be performed.
Conclusion
   No new methods have been approved for the analysis of dioxin since the promulgation of the
NPDWRs. However, the sensitivity of Method 1613 has slightly increased from 5 to 1 pg/L
since that time. A re-evaluation of the PQL could not be determined because no PE data were
available for examination. Therefore, the current PQL is likely to remain unchanged.
Methods Support Document for Six-Year Review
65
Draft - March 2002

-------
 Diquat
 Results of the Method Comparison
    Diquat became a regulated SDWA contaminant with the promulgation of the July 1992 Phase
V rule for SOCs. Table 35 compares the approved methods at promulgation with currently
approved methods.  At the time of the Phase V regulation only one method, EPA 549, was
approved for determination of diquat in drinking water (57 FR 31776), In August 1992, EPA
549 was replaced by an updated method, EPA 549.1  (''Methods for the Determination of Organic
Compounds in Drinking Water—Supplement 2"), Because EPA 549.1 received approval so soon
after the promulgation of the Phase V rule, it is listed in the column of methods approved at
promulgation in Table 35. In 1999 EPA approved EPA 549.2 (64 FR 67450) and discontinued
EPA 549.1. The MDL of EPA Method 549.'2 is approximately five times more sensitive than the
original version of the method (EPA 549).
Table 35.  Results of the Analytical Methods Comparison for Diquat (Newly Promulgated
           Methods Are Indicated in Bold)
   MCL = 20
Current PQL .== 4 p,g/L
DL* = 0.4 u-g/L
Acceptance Limir = ± 2*S.D.
Methods Approved At Promulgation
Method
EPA 549'
EPA 549.1-
Technique
LSE. HPLC with UV
LSE, HPLC with UV
MDL

-------
 Results of the Analysis of the WSLPata


 a.  Method Usage Over Time


    The distribution of analytical techniques used by the EPA and. State laboratories in WS 34 to
 41 is shown in Figure 22, The "other" techniques represent methods which were not specifically
 identified by participating laboratories or were otherwise unknown, EPA has consistently
 approved a single method for diquat at any given time, EPA Method 549 was predominantly used
 prior to WS 36 and. EPA 549.1 was used predominantly after WS 36.


     Figure 22. Distribution of Analytical Techniques by WS Study: Diquat

                                          Diquat
f \J\Jt /G "
90% -

80%

•g "70% -
| 60% •
O)
| 50%-
D
j§ 40%
Jj
SS 30%
20%
10% •
n% ,.















TM

f

K
*»
1
W
m
1
I
m
^
&
si
SB

1


»
*
f '['!!»













,
I
&
||
Sj
1
1
tej
Pj
'<&?.
I
&
f
1
s
1
11













B
—























1
—

























1
—

























•

























































.






























                                                                             |S549  |
                                                                             JD549.1J
                                                                             • other!
                WS 34   WS 35   WS 36   WS 37   WS 38  WS 39  ' WS 40   WS 41

                                         Water Study
b. Results of the PQL Analysis


   The original PQL for diquat (4;|ig/L) was derived by using PE data from WS 23 to 27 (56 FR_
60949). A reassessment of the PQL was attempted after compiling additional data from WS 24
to 41, However, for several of these available water studies, the number of participating
laboratories were too few (under ten) to include for consideration in the analysis. After omitting
these studies, the remaining usable PE data were summarized, including the study number, the
true value of the WS sample, the number of results from EPA and State laboratories, and the
calculated percentage of laboratories whose results successfully passed within, federally
Methods Support Document for Six-Year Review
67
Draft-March 2002

-------
designated acceptance limits (Table 35).  The acceptance limits for diquat are specified in 40
CFR §141.24(h)(19)(i) to be twice the standard deviation, or ± 2*S.D.,. from the value *'xn (where
x = aT+b and T represents the true value).


    Table 36 reveals that laboratory success rates ranged between 78,6 and 100 percent. Thus,
use of the 75 percent criterion for estimation of the PQL by the graphical method would be
meaningless. The dataset is also incompatible with the goals of a PQL reassessment-particularly
a potential lowering of the PQL-because the concentrations of all the spiked samples exceeded
the current PQL. Also, the concentration of the spike samples was well above the PQL of 5
Table 36. Evaluation of Diquat Data from WS Studies Using the ± 2* S.D. Acceptance
          Limits (in Order of Increasing Concentration)
\vs#
37
36
40
38
32
34
39
35
41
Spiked "IVue" Value
(WS/L)
8.41
14,7
14.8
23.7
2S.2
29.2
32.2
37.4
44.0
# Results from EPA and
State Labs
21
28
27
22
11
17
21
15
,17
% Labs Passing ± 2* S.0.
Acceptance Limits
90,5
78.6
96.3
95,5 '
100 •
94.1
95.2
86.7 .
100
Conclusion for Diquat
   As shown by the results of the method comparison, few methods have been approved by EPA
for the determination of diquat.  The analytical technology supporting the determination of diquat
has remained essentially unchanged over time. However, analytical capabilities have improved.
The original method, EPA 549, was much less sensitive than today's method, EPA 549.2,
indicating significant improvement. The plot of method usage over time reveals that the
techniques predominantly used by participating laboratories were EPA Methods 549 (prior to WS
36) and 549.1 (subsequent to WS 36). Although the method comparison indicates increased
analytical capabilities, the available WS data did not provide information suitable for a PQL
Methods Support Document for Six-Year Review
68
Draft - March 2002

-------
reassessment. The percentage of laboratories passing within the acceptance window was
generally well above the 75 percent criterion for use in the linear regression approach.
Ethylene Dibromide


Results of the Method Comparison •.'.••


   With the Phase II SOCs .(56 FR 3526), EPA 504, was the only approved method listed for
analysis of ethylene dibromlde (EDB) in drinking water. Since this regulation was promulgated,
the Agency has approved a new revision of GC with microextraction '(EPA 504.1) and approved
an additional analytical method (EPA 551.1, GC with. LLE and BCD). As shown, in Table 37, the
MDLs of EPA Methods 504 and 504.1 are essentially equivalent, but EPA 551.1 offers a slightly
greater level of sensitivity.
Table 37. Results of the Analytical Methods Comparison for Ethylene Dibromide (New
          Methods in Bold)
MCL = OJS;jig/L Current PQL = 0.05 jig/L • DLA = 0.01 ng/L Acceptance Limitf = ± 40%
Methods Approved At Promulgation
Method
EPA 5041

Technique
Microextraction, GC

MDL
d»8*-)
0.01

Currently Approved Methods (141.24)
Method
EPA 504. 12
EPA 5S1.1»
Technique
MicroexiJ'action, GC
LLE/GCw/ECD
MDL
(l»g/L) . '
0,01
0.008
! "Methods for the Determination of Giganic Compounds in Drinking V&ter," EPA/600/4-88/039, December
1988.
z "Methods for the Determination of Oiganic Compounds in Drinking \teter-Suppleraent III," EPA/6QO/R-9S-
131, August 1995. , ' '
A Regulatory DLs for oigatric compounds are listed at 40 CFR § 1 41 .24(h)(1 8).
f Acceptance limits are listed at 40 CPR§14L24(h)(I9)(i}. .
Methods Support Document for Six-Year Review
69
Draft - March 2002

-------
 Results of the Analysis of the WS Data


 a.  Method Usage Over Time


    Figure 23 illustrates the distribution of the analytical techniques used by the EPA and State
 laboratories in WS studies 34 to 41. 'The "other" designation includes methods for which
 laboratories did not report any information on the method used or non-EPA methods. The use of
 EPA 504.1 greatly exceeds the use of any other method.  This distribution is consistent through
 the duration of the WS studies shown in Figure 23.

      Figure 23. Distribution of Analytical Techniques by WS Study: Ethylene Dibromide

                                    Ethylene Dibromide
        g*
        in
        J2
        TO
90%
80% •
70%
60%

40%
30%
20%
10%
0% -





-











-











i







	 ,















1





—











•





1







"**



™

















IP

—



i;





— — >











Mi i




ID 504.1
i
j • 524.2
pother
                 WS34   WS35  WS36   WS37  WS 38   WS39   WS40   WS41

                                         Water Study
b.  Results of the PQL Analysis


    The current PQL, 0.05 p,g/L, was derived from a multiplier of 5 from the MDL of 6.01 u.g/L
(56 FR 3552). The data used for PQL re-evaluation were taken from WS studies 24 to 41. Table
38 summarizes the results of the WS studies, providing the study number, the true concentration
of the spiked sample, the number of laboratories participating, and the percentage of laboratories
passing the WS study (evaluated using acceptance limit of ± 40 percent for EDB as cited in 40
CFR§141.24(h)(19)(i)).
Methods Support Document for Six-Year Review
70
Draft - March 2002

-------
    Using the 75 percent criterion, it appears from the numerical data that EPA and regional
laboratories are able to achieve acceptable .results within the ± 40 percent acceptance window at
concentrations from 0.14 to 2.3 u.g/L.
Table 38. Evaluation of Ethylene Dibromide Data from WS Studies Using the ± 40%
         • Acceptance Limits (in Order of Increasing Concentration)
ws#
37
33
39
36
38
41
34
26. .
24
35
31
40
29
25
30 ..
27
32
Spiked ("True") Value
(M5/L)
0.138
0,143
0,227
Q.2S3
0,336
0,344
0,406
0.434
0.480
,0.609
0.637
0.638
0,850
0.944
1.39
1.45
2.29
# Results from EPA and
State Labs
42
24
'36
51
. - 48
34
45 ,
35 .
32
30
. 25
.50
25
18
39
17
44
% Labs Passing
Acceptance Limits
81,
75
94
94
96
100
87
91
100
90
84
98
84
JOO
92
88
. 98
 •  The percentages of the acceptable results for laboratories were not plotted, as these values all
exceeded the 75 percent criterion and therefore could not contribute meaningful information
toward the re-evaluation of a PQL using the linear regression approach.  In addition, none of the
WS studies evaluated had spike concentrations below the current PQL of 0.05 u.g/L. However.
WS studies at concentrations slightly above the PQL (WS 37 and 33) had laboratory passing
Methods Support Document for Six-Year Review
71
                          Draft - March 2002

-------
rates of 81 and 75 percent, respectively. This would indicate that the current PQL is probably
appropriate and unlikely to change.


Conclusion for Ethvlene Dibromide
   The method comparison results indicate that the sensitivity of the available methods has not
improved significantly since the promulgation of NPDWRs for BOB. Evaluation of more recent
WS data provides no evidence that would support a change from the current PQL of 0.05
Fluoride
Results of the Method Comparison


   Fluoride, a Phase II IOC, is unique among SDWA contaminants because it is often added to
drinking water to provide well-known health benefits.  Because low concentrations of fluoride
are often added to public water supplies to protect dental health, EPA has not published a
detection limit for this contaminant. Since the Agency's promulgation of the MCL for fluoride
(April 1986,51 FR 11397), this contaminant has gained several additional approved methods,
mostly developed by voluntary consensus standard otganizations. Currently, the only EPA-
approved method for fluoride determination is EPA 300,0, an ion chromatography method with
greater sensitivity than the EPA methods approved at the~time of promulgation. Table 39
summarizes original and current methods, and their individual detection limits. The voluntary
consensus standard method MDLs are not listed in this table because non-EPA methods are not
required to  document detection limits.
Table 39. Results of the Analytical Methods Comparison for Fluoride
MCL = 4 mg/L Current PQL = 0.5 mg/L DL = N/A* Acceptance Limit* = ± 3t 0% ,
Methods Approved At Promulgation
Method
EPA 340.1'
EPA 340.2'
EPA 340.3'
Technique
Colorimetric SPADNS, with
Bellack Distillation
Potentiometric, with
ion-selective electrode (ISE)
Automated Alizarin, with
distillation (complexone)
MDL
(mg/L)
0. 1- 1,4*
NA*
0,05 - 1.5*
Currently Approved Methods (141.23)
Method
EPA 300.0*


Technique
Ion. chromatography


MDL
(mg/L)
0.01


Methods Support Document for Six-Year Review
72
Draft -March 2002

-------
D1179-72AZ
D1179-872B2
414A; C3
414B3
4I3E3

129-7 1W4
3 80-75 WE5
Colorimetric SPADNS, with
distillation
Potentiometric, with ISE •
Colorimetric SPADNS, with
distillation
Potentiometric, with ISE
Automated Alizarin,
w/distillation

Automated Alizarin,
w/disti Hation
Automated electrode
N/A*
N/A*
.N/A'
N/A' .
N/A*

WA-
N/A'
1)1179-931^
D4327-918
4500F-B,D8
4SOOF-C8
4500F-E8
4110B8
129-7 I W
38G-75WE5
Manual electrode .
Ion chromatography
Manual distillation;
colorimetric
SPADNS
Manual electrode
Automated Alizarin
Ion chromatography
Automated Alizarin
Automated
electrode
N/A*
N/A'
N/A*
N/A"
N/A '
N/A'
N/A*
N/A*
 1 "Methods of Chemical Analysis of Water and Wastes (MCAWW)," EPA/600/4-79-020, EPA Environmental
 Monitoring Laboratory, Cincinnati, Ohio 45268. March 1983.
 2 Annual Book of ASTMStandards, part. 31, Water.  American Society for Testing and Materials, 1961 Race
 Street Philadelphia, PA  19103.
 3 Standard Methods for the Examination of Water and Wastewater, 16* edition, American Public Health
 Association, American Water Works Association, Water Pollution. Control Federation, 1985,

 4 "Fluoride in Water and Wastewater, Industrial Method 129-71 W>" Technicon Industrial Systems, Tarrytown,
 New York 10591.  December 1972.

 * "Fluoride in Water and Wastewater," Technicon Industrial Systems, Tanytown, New York 10591,  February
 1976,      '      '                                                     .           "         •
 6 "Methods for the Determination of Metals in Environmental Samples Supplement 1," ER/600/R-94/111, May
 1994,

 7 Annual Book ofASTM -Standards^ Vol.  11.01. - American Society for Testing and Materials, 1961 Race Street.
 Philadelphia, PA 19103.            .                                .
 8 Standard Methods for the Examination of Water and Wastewater., 19* edition, American Public Health
 Association, American Water Works Association, Water Pollution Control Federation,
 * N/A: not available.  See above text for explanation.

 * Acceptance limits for inoiganic compounds are listed at 40 CFR § 141,23{k)(3)(il).

 * Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
 laboratory/analyst performance.	;	       ;	
    As shown in Table 39, multiple versions of'methods exist for some technologies such as ion
chromatography, automated electrode, automated alizarin, and manual electrode.
Methods Support Document for Six-Year Review
73
Draft - March 2002

-------
Results of the Analysis of the WS Data
a.  Method Usage Over Time


    The distribution of the analytical techniques used by the EPA and State laboratories in WS 34
to 41 is shown in Figure 24. The results for "other" techniques include the use of other
techniques identified by the laboratories participating in the WS study, as well as "unknown"
methods, i.e., methods for which laboratories did not report any information on the type of
method used.
    Figure 24.  Distribution of Analytical Techniques by WS Study: Fluoride

                                          Fluoride
•g
£
XI
ra
60%


50%


40%


30% -


20% -


10%
                     i
                              L
                                                                        |«300,0        ;
                                                                        {•340,1
                                                                        |S340,2        ;
                                                                        IB 340.3        •
                                                                        [H38Q-75WE    ;
                                                                        ID Manual Bee,*  !
                                                                        IB Auto, Alizarin**.
              WS34  WS35   WS36  WS 37  WS 38  WS 39  WS40  WS41

                                     Water Study
» Manual electrode combines the methods D1179-93B (ASTM) and 4500F-C (SM).
** Automated Alizarin combines the methods 4500F-E (SM) and 129-71W fBchnicon),
    Fluoride determination has involved the use of a wide variety of analytical methods. In
earlier years, the most popular method was EPA 340.2, an ion selective electrode method, but its
usage dropped significantly after WS 36. Since WS 36, method usage by laboratories
participating in the water supply studies has been dominated by a non-EPA manual electrode
method.  Laboratories have also increasingly favored the use of EPA 300.0, though to a lesser
extent relative to the manual electrode methods (D1179-93B, 4500F-C).  A small fraction of
Methods Support Document for Six-Year Review
                                       74
                                                              Draft - March .2602

-------
participating laboratories have employed the voluntary consensus standard method 380-75WE
(automated electrode) throughout WS 34 to 41.                          ,
b.  Results of the PQL Analysis
    The current PQL (500 ng/L or 0,5 mg/L) was originally determined from, older PE water
supply study data (WS 8 to 12, see USEPA, "Monitoring for Fluoride in Drinking Water:
Revised." March 1986). The PQL was re-evaluated using a broader range of PE data including
more recent studies (WS 24 to 41). Table 40 summarizes the results of these studies, providing
the study number, the spiked or "true" value for the WS sample, the number of results from EPA
and State laboratories, and the actual responses returned by laboratories. For fluoride, EPA
stipulates acceptance limits of ± 10 percent (§141,23(k)(3)(ii)).
Table 40, Evaluation of Fluoride Data from WS Studies Using the 10% Acceptance Limits
          (in Order of Increasing Concentration)
ws#
29
34
26a
24b
40
25b
24a
37
32
• 25a
39
26b
35
27
38
Spiked "True" Value (mg/L)
0.33 -
1.10
1,25
1.30
1.40
1.50
1,72
1.80
2.00 '
2.50
2.90
3.41
3.80
4.35
4,70
# Results from EPA awd
State Labs
33
65 '
63
62
66
40
62
55
67
40
54 ,
64
43 '
37
65
% Labs Passing ±10% .
Acceptance Limits
84,8
93,8
93,7
96.8
60.6
87,5
93.5
87.3 '
92,5
87.5
92,6
8.9.1 '
97.7
86,5
98.5
Methods Support Document for Six-Year Review
75
Draft - March 2002

-------
vvs#
31
41
33
36
30
Spiked "True" Value (mg/L)
5.70
6.20
6.60
7.20
7.90
# Results from EPA and
State JLabs
33
46
35
61
61
% Labs Passing ± 10%
Acceptance fjm its
97.0
95,3
94.3
95.1
93.4
    The WS data indicate that EPA and State laboratories performed with high success rates In
WS studies involving fluoride determination. On average, 91 percent of participating
laboratories achieved results within the ± 10 percent acceptance window. The range of true value
concentrations contained many spikes above the current PQL (0.5 mg/L), preventing a thorough
analysis of laboratory capabilities at or below the PQL. For WS 29, the only study where the true
value was below the PQL, 85  percent of laboratories successfully passed within the specified
acceptance limits, well above  the 75 percent criterion. Thus, it might be possible for laboratories
to pass the PE studies at even  lower concentrations: however, such a conclusion is uncertain in
the absence of additional data below the PQL value.
Conclusion for Fluoride
    The MDL of the only currently approved EPA method, EPA 300,0, is much lower than those
of previous EPA methods and use of this method has increased gradually over time.  However,
EPA 300.0 still accounts for less than a-quarter of method usage for fluoride and thus cannot be
representative of overall laboratory analytical capabilities. Based on the graph of method usage
over time (Figure 24), the method most commonly used in recent years is a manual electrode
method, whose detection limit is not specified. Hence, the combined results of the method
comparison and method usage over time suggest that analytical capabilities have improved for
only a subset of EPA and State laboratories.


    Based on the evaluation of quantitative WS data, the current PQL of 0,5 mg/L is still
supportable and appropriate, although an even lower value might be attainable based on the high
success rates of laboratories in water studies. However, this hypothesis would require further
analysis using WS data at spiked concentrations below 0.5 mgYL, and such data are currently not
available.
Methods Support Document for Six-Year Review
76
Draft - March 2002

-------
Glyphosate
Results of the Method Comparison-
    In 1992, the Agency listed EPA 547 as the only approved method for determination of
glyphosate in drinking water, according to the Phase V rule for SOCs (57 PR 31776). Since that
time, EPA has added Standard Method (SM) 6651, a voluntary consensus standard method, to
the approved list. The MDL of SM .6651 is not specified, but it is reasonable to expect a similar
detection limit compared to EPA 547 due to the similarity in determinative technique. Because
the MDL of EPA 547 has not changed over time and no additional EPA methods were approved,
the analytical capabilities for determination of glyphosate have remained constant since the
approval of EPA 547.
Table 41, Results of the Analytical Methods Comparison for Glyphosate (Newly
          Promulgated Methods are Indicated in Bold)
MGL = 0,7«ag/L Current PQL = 0.06 mg/L DLA - 0,006 mg/L Acceptance Lim»tf = ± 2*S,D.
Methods Approved At Promulgation
Method
EPA 54?1

Technique
HPLC, post-column
derivatization,
fluorescence detection

MDL
(mg/L)
0.006

Currently Approved Methods (141.24)
Method
EPA 547*
SM 66512 .
Technique
HPLC, post-column
derivatization, fluorescence
detection
Liquid chromatography, post-
column fluorescence
MDL
- (wg/L)
0.006
N/A°
! "Methods for the Determination" of Oiganic Compounds in Drinking Witer— Supplement I," EPA/600/4-90/020,
July 1990.
2 Standard Methods for the Examination of Water and Wastewater. 1.8* edition. America!* Public Health
Association, 1015 Fifteenth Street NW, Washington, DC 20005,
A Regulatory DLs for ojganic compounds are listed at 40 CFR §14L24(h)(J8). "
* Acceptance limits for otganic compounds are listed at are listed at 40 CFR § 141 .24(h)(19)(i).
0 N/A = not available. MDLs are not necessarily specified for noa-EM methods.
Methods Support Document for Six-Year Review
77
Draft - March 2002

-------
Results of the Analysis of the WS Data


a.  Method Usage Over Time


    Figure 25 is a plot of the distribution of analytical techniques used by the EPA and State
laboratories during WS 34 to 41. As shown in Figure 25, the majority of the participating
laboratories utilized EPA  547 for the determination of glyphosate. A small fraction of
laboratories employed the voluntary consensus standard method, SM 6651 and a minority of
laboratories used "other" methods or methods that were not specified.
      Figure 25. Distribution of Analytical Techniques by WS Study: Glyphosate

                                        Glyphosate
1UU/O '
90% •
80% •
70% •
60% •
50% -
40%
30%
20% -
10%
n% .










-


















H
[—


















a
















J
1
__


















fH

















1



















m
P


















I























j Q 547
136651
{
E5!?£L




                 WS 34   WS 35   WS 36  > WS 37  WS 38   WS 39   WS 40   WS 41

                                        Water Study
b. Results of the PQL Analysis
   Glyphosate currently has a PQL of 0,06 mg/L (or 60 p.g/L) which was derived from previous
WS results from EPA and State laboratories (WS 24 to 27,56 FR 60949). In light of the
availability of more recent PE WS data from WS studies 32 to 41, efforts were made to reassess
the PQL. Table 42 summarizes the data from these water studies, indicating the study number,,
the true value of the WS sample, the number of results from laboratories, and the calculated
percentage of laboratories whose results successfully passed within the designated acceptance
limits for glyphosate (± 2*S.D.  as specified at 141.24(h)(19)(i).
Methods Support Document for Six-Year Review
78
Draft -March 2002

-------
    Table 42 shows that the laboratories in these water studies were able to achieve results within
the acceptance window with 73 to 100 percent passing rates. Using the 75 percent criterion for
estimation of the PQL will not be meaningful for this dataset because of the high success rate and
the lack of spike values at concentrations around the current PQL of 60 ug/L,


Table 42. Evaluation of Glyphosate Data from WS Studies Using the ± 2* S.D. Acceptance
          Limits (in Order 0f Increasing Concentration)
ws#
33
40
38
34
32
36
41
39
35
37
Spiked "True" Value (fig/L)
308
375 •
4)0
.438
447
528
560
620
665
780
# Results from EPA and
State Labs
11
30
' 25
22
15
27
17
22 .
15
22
% Labs Passing ±'2*S.D.
Acceptance Limits
72.7
96.7
88,0
90.9
86.7
96,3,
100
95.5
86.7
77.3
Conclusion for Glvphosate
    Currently, EPA 547 is the sole approved EPA method for determination of glyphosate and its
MDL has hot changed over time (Table 41).  A voluntary consensus standard method (SM 6651)
with similar determinative technology is also approved for use, but was used minimally over the
duration of the selected studies according to PE records (Figure 25). The MDL of SM 6651 is
not specified but in all likelihood, resembles the MDL of EPA 547,  Therefore, the analytical
methods capabilities for glyphosate determination have stayed constant for most laboratories.
The available numerical WS data do not support a reassessment of the PQL based on the 75
percent criterion, because the passing rates of laboratories typically exceeded this value. In
addition, the spike concentration of the WS samples were much higher than the current PQL. -
Methods Support Document for Six-Year Review
79
-Draft - March 2002

-------
 Ileptachlor
Results of the Method Comparison
    The approved methods for the analysis of heptacblor, a Phase II SOC, in drinking water were
listed in the NPDWRs (56 FR 3526),  These original methods included gas chromatpgraphy with
several extraction and/or detector variations (EPA Methods 505, 508, and 525.1).  Since
promulgation of the Phase II rale, EPA has removed one older method (EPA 525.1) and
approved the use of three new analytical methods (EPA Methods 508.1» 525.2, and 551.1).- As
shown in Table 43, EPA Method 505 has experienced an estimated 10-fold increase in
sensitivity. EPA Method 508.1 is nearly 7 times more sensitive than it was at the time of
NPDWR promulgation, and represents the most sensitive of the new methods, with a current
MDL of approximately 0.0015 u-g/L.  The MDLs of the newly approved, methods, EM Methods
508.1, 525.2, and 551.1, range from 0,5 to 15 times the MDLs of the most sensitive method at
the time of the NPDWR promulgation.
Table 43. Results of the Analytical Methods Comparison for Heptachlor (Newly
          Promulgated Methods are Indicated in Bold)
MCL = 0.4 ng/L Current PQL = 0.4 jig/L PLA = 0.04 ng/L Acceptance Limit* - * 45%
Methods Approved At Promulgation
Method
EPA 505'
EPA 508'
EPA 525. 11


Technique
GC, microextraction
GC, BCD
GC/MS, LSE


MDL
0*8/1-)
0.03
0.01
0.04


Currently Approved Methods (141.24)
Method
EPA 505*
EPA 5Q82
EPA 508.12
EPA S25.22
EPA 551.1*
Technique
GC, raicroextraction
GC, BCD
GC, LSE, BCD
GC/MS, LSE
GC/MS, LLE, BCD
MDL

-------
Results of the Anal Y.SJS., of the WS Data


a.  Method Usage Over Time                                '


    Figure 26 is a plot of the distribution of analytical techniques used by the EPA and State
laboratories in WS 34 to 41. The "other1* techniques represent methods which were not
specifically identified by participating laboratories or were otherwise unknown. As shown in  .
Figure 26, the majority of laboratories used EPA 508 for determination of heptachlor in WS
studies 34 to 41. EPA 525,1, which was used in earlier WS studies, was replaced by EPA 525,2
after WS 36, Other than this shift, there has been little overall change in method usage over time
for this contaminant.
     Figure 26. Distribution of Analytical Techniques by WS Study: Heptachlor

                                        Heptachlor
         70%
         60%
      TS
      O
      £
      (A

      5
50% -


40%


30% -


20% -


10% -


 0% J
                                      ! 0525.2!
                                      SB 525,1 !
                                      (•508
                                      1P505  r
                                      10508.1!
                                      -;        i
                                      Ifl other •;
               WS 34   WS 35  WS 36
WS 37   WS 38

  Water Study
                                             WS 39   WS 40   WS 41
b. Results of the PQL Analysis
   The original PQL for heptachlor (0,4 u-g/L) was derived via multiplication of the IMDL by a
factor of 10 (56 FR 3552).  Witt the availability of recent PE WS data, efforts were made to
reassess the PQL using data from WS studies 24 to 41. Table 44 summarizes the data from'these
WS studies, indicating the study number, the true value of the WS sample, the number of results
from EPA and State laboratories, and the calculated percentage of laboratories whose results
Methods Support Document for Six-Year Review
                              Draft ~ March 2002

-------
 successfully passed within federally designated acceptance limits for heptachlor.  These
 acceptance limits are specified in 40 CFR §I41.24(h)(19)(i) to be ± 45 percent from the true
 concentration of the spike sample.


    The numerical data in Table 44 demonstrate that the laboratories in these WS studies are able
 to achieve results within the ± 45 percent acceptance window with 82 to 100 percent passing
 rates.  Using the 75 percent criterion for estimation of the PQL, the dataset are not adequate to re-
 evaluate the PQL. Further, the majority of the spike samples had concentrations in excess of the
 current PQL of 0.4 u.g/L.  Nevertheless, the high passing rates of laboratories at concentrations
 around the PQL (e.g., 100 percent passing for a true value equal to 0.11 u.g/L) support a possible
. reconsideration of the current PQL.
 Table 44.  Evaluation of Heptachlor Data from WS Studies Using the 45% Acceptance
           Limits (in Order of Increasing Concentration)
ws#
25b
24a
29
32
37
27
39
36
41
34
38
30
25a
31b
33
31a
26
Spiked "True" Value (fig/L)
0.113
0.263
0.370
0.443
0.563
0.642
0.667
0.751
0.83
0.914
1.20
1.38
1.42
1.44
1.73
1.92
2.27
# Results from EPA and
State Labs
18
33
24
44
42
24
36
51
34 ''
45
48
47
19
25
24
27
31
% Labs Passing
± 45% Acceptance Limits
100
94
95.8
94,6
95.6
100
97,6
94.5
93
90.2
94
93.6
100
84.6
90.6
81.5
90.3
 Methods Support Document for Six-Year Review
82
Draft-March 2002

-------
ws#
40
35
24b
Spiked "True" Value (jig/L)
233
2,54
3.15 "'
# Results from EPA and
State Labs
50 '
. 30 . .
33
% Labs Passing
'As 45% Acceptance Limits
• 90.7
90.9
93.9
Conclusionfor Heptachlojr


    The plot of the method usage over time (from WS studies) indicates consistent use of two
methods, EPA Methods 508 and 505 over the duration of the selected studies, with laboratories
slightly favoring usage of EPA 508. EPA 508 represents the most sensitive currently approved
method, according to Table 43. The overall level of analytical sensitivity for this contaminant
has improved, approximately seven-fold from the time of NPDWR promulgation.  The available
numerical WS data do not support a reassessment of the PQL based on the 75 percent criterion,
because the passing rates of laboratories always exceeded this value.  Because such high passing
rates were observed for some low spiked concentrations (e.g., 100 percent laboratory success for
a concentration three times lower than the current PQL), it is possible that a lowered PQL might
be appropriate. However, the desired quantitative reassessment of the PQL could not be
performed using the historical approach.
Heptachlor Epoxide
Results of the Method Comparison
   . The approved methods designated in the Phase II rule for heptachlor epoxide (56 FR 3526)
included EPA Methods 505, 508, and 525,1. These methods are variations of GC methods xvith
different extraction and/or detection techniques as shown in Table 45. The most sensitive of the
current methods is EPA 508,1 (GC with BCD), with an MDL of 0.001 (ig/L. This method was
introduced subsequent to the promulgation of the Phase II rale, along with two additional
methods. EPA Methods 525,2 and 551.1.
Methods Support Document for Six-Year Review
83
Draft - March 2002

-------
Table 45. Results of the Analytical Methods Comparison for Heptachlor Epoxide (New-
          Methods Indicated in Bold)
MCL = 0.2[ig/L Current PQL = 0.2 uf/L »LA = 0.02 |ig/L Acceptance Liwitt = ±45%
Methods Approved At Promulgation
Method
EPA 505'
EPA 508'
EPA 525.1'


Technique
GC, microextractton
GC, BCD
GC/MS, LSE


MDL
0.004
0.00015
0.2


Currently Approved Methods (141.24)
Method
EPA 505s
EPA 5082
EPA S08.lz
EPA 525.2*
EPA 551. 12
Technique
GC, microextraction
GC, BCD
GC, LSE, BCD
GC/MS, LSE
GC/MS, LLB, BCD
MBL
0.004
0,0059
0.001
0.048-0.13*
0.002
1 "Methods for the Determination of Oiganic Compounds in Drinking V&ter," EPA-600/4-88/039, December
1988.
2 "Methods for the Determination of Organic Compounds in Drinking Water— Supplement 111,"
EPA/600/R-95-131, August 1995. ,
A Regulatory DLs for oiganic compounds are listed at 40 CFR §141,24(h)(18).
t Acceptance limits for organic compounds are listed at are listed at 40 CFR 141.24(h)(19)(i).
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance.
Results of the Analysis of .the WS Data
a.  Method Usage Over Time
   Figure 27 illustrates the distribution of the analytical techniques used by the EPA and State
laboratories in WS 34 to 41. Analytical methods which were not reported by laboratories or were
otherwise unknown were grouped into the category of "other."  As illustrated by Figure 27, EPA
508 represents the most commonly used method for heptachlor epoxide WS analyses. EPA
Methods 505 and 525.2 were used to a lesser degree; together, these methods account for 40 to
50 percent of the techniques used during the indicated time frame.
Hfethods Support Document for Six-Year Review
84
Draft - March 2002

-------
    Figure 27. Distribution of Analytical Techniques by WS Study: Heptachlor Epoxide

                                   Heptaehlor Epoxide
         70%
                                                                               505
                                                                               525.2,
                                                                               508  I
                                                                               508.1- i
                                                                             n other
               WS34  WS35   WS36  WS 37  WS 38   WS 39  WS40  WS41
                                        Water Study
b. ; Results of the PQL Analysis     •


  . The PQL for heptachlor epoxide is currently. 0.2 u-g/L. EPA obtained this value by
multiplying the estimated IMDL by a factor of 10 (56 FR 3552). The Agency wished to re-
evaluate the PQL using available WS data from WS studies 24 to 41 (although no data were
available for WS 28). Table 46 summarizes these data, indicating the study number, the true
value of the sample, the number of non-zero responses from laboratories, and the calculated
passing rates of laboratories, who satisfied the specified acceptance limits for heptachlor epoxide.
These limits are designated as ± 45 percent (40 CFR § 141.24(h)(19)(i)). .
Methods, Support Document for Six-Year Review
85
Draft - March 2002

-------
 Table 46. Evaluation of Heptachlor Epoxide Data from WS Studies Using the 45%
           Acceptance Limits (in Order of Increasing Concentration)
ws#
24a
26a
29
39
32
37
27
34
41
33
38
25a
30
25b
40
35
24b
26b
31
Spiked "True" Value 
-------
Conclusion for Heptaehlor Epoxlde


   According to the method comparison results, the introduction of three new analytical
methods since the promulgation of the Phase II rule has not presented any significant
improvements in overall method sensitivity.  The most-sensitive method (EPA 508) at the time of
NPDWR promulgation is no longer the most sensitive method today, due to its MDL change, as
evident in Table 45.  Based on the analysis of method usage over time, EPA 508 is more widely
•used than the current most sensitive method, EPA 508,1.-  In light of the usage patterns for
heptachlor epoxide, analytical method capabilities for the overall contaminant are probably
similar to those of the original methods. Although the WS data did not provide enough
information to perform a regression analysis, the passing rates at concentrations around the
current PQL of 0.2 jxg/L suggest that the PQL could be lower.
Hexachlorobenzene
Results of the Method Comparison
   Hexachlorobenzene. a Phase V SOC, had three approved methods for drinking water analysis
at the time NPDWRs were promulgated (57 FR 31776). All three original methods consisted of
GC with extraction and/or detector variations (EPA 505, microextraction; EPA 508, BCD; and   .
EPA 525.1, LSE), Since the Phase V rule was promulgated, the Agency has" retained two of the
old methods, removed EPA 525.1, and approved three additional analytical methods, EPA
Methods 508.1, 525.2, and 551.1 (LSE, BCD; LSE; and LLE, BCD, respectively). As shown in
Table 47, the most sensitive method of both the old and new categories is EPA 508, with an
MDL of 0.0077 |ig/L.
Methods Support: Document for Six-Year-Review
87
Draft - March 2002

-------
Table 47. Results of the Analytical Methods Comparison for Hexachlorobenzene (New
          Methods in Bold)
MCL - 1 ng/L Cu rrent PQJL = I ng/JL BL* = 0.1 ji g/JL. Acceptance Limit* = ± 2 x S.D,
Methods Approved At Promulgation
Method
EPA 505'
EPA 508'
EPA 525. 11


Technique
GC, microextraction
GC, BCD
GC/MS, LSE


MDL

0,002
0.0077
0,2


Currently Approved Methods (141.24)
Method
EPA,505Z
EPA SOS*
EPA 508.1*
EPA 525,2*
EPA 551.1*
Technique
GC, raicroextractton
GC, BCD
GC, LSE, BCD
GC/MS, LSE
GC/MS, LLE, ECD
MBL
frg/L)
0.002
0.0077
0,001
0,001
' 0.001
1 "Methods for the Determination of Oiganic Compounds in Drinking Water," EPA-600/4-88/039, December
1988.
2 "Methods for the Determination of Oiganic Compounds in Drinking Water—Supplement 111,"
EPA/600/R-95- 131, August 1995,
* Regulatory DLs for onjanic compounds are listed at 40 CFR §141.24(h)(18).
f Acceptance limits are listed at 40 CFR §141.24(h)(19)(i). Substandard deviation.
Results of the Analysis of the WS Data
a.  Method Usage Over Time
   The distribution of the analytical techniques used by EPA and State laboratories in PE WS
studies 34 to 41 is illustrated by Figure 28. Methods designated as "other" include methods for
which laboratories did not specify the method used or were otherwise unknown. As shown by
Figure 28, laboratories responding to these PE WS studies, mainly determined hexachlorobenzene
using EPA 508, with the exception of one study (WS 39) where use of EPA 525.2 was more
prevalent. Use of EPA 525.2 began to appear in WS 36 and has increased slightly in recent
years. The analytical methods used most recently (i.e., WS 41) for determination of
hexachlorobenzene are divided between the several available current methods. Use of EPA
508.1 was quite small relative to the other methods used.
Methods Support Document for Six-Year Review?.
                                                                   •Draft - March 3002

-------
     Figure 28, Distribution of Analytical Techniques-by WS Sjudy: Ilexachlorobenzenc

                                   Hexachlorobenzene
          60%
                WS34  WS35  WS36   WS 37  WS38   WS39   WS40  WS41
                                        Water Study
b,  Results of the PQL analysis
    The original PQL of hexachlorobenzene (1 u-g/L> 56 FR 3552) was dete.rmin.ed using PE data
from WS 27, The data used for the re-evaluation of the PQL were taken from WS 27 through 41
(minus 28),  The results of these WS studies are summarized in Table 48, which provides the
study number, tine spiked value for the WS sample, the number of results from EPA and State
laboratories, arid the results evaluated using acceptance limits of ± 2 x S.D (as specified at
14L24(h)(19)(i)).


    It appears from the numerical data that EPA Regional and State laboratories are able to
achieve successful results within the ± 2 x S.D. acceptance limits at rates well above the 75
percent criterion typically tised for determining a PQL, Thus, a regression analysis using the
available PE data was not attempted due to the nature of these data.  On average, the success
rates of the WS laboratories were about 92 percent over spike concentrations ranging from 0.4 to
3.6 ug/L, The high passing rates at concentrations around the PQL of 1 M-g/L suggest that the
PQL could be lower.
Methods Support Document for Six-Year Review
89
Draft - March 2002

-------
 Table 48.  Evaluation of JPE Data for Hexachlorobenzene from WS Studies Using the 2 x
           S.D. Acceptance Limits (listed in order of increasing concentration)
ws#
29
27
38
35
30
37
36
32
41
33
39
31
40
34
Spiked ("True") Value (p.g/L)
0.417
0.483
0,538
0.635
0.667
0.806
0.847
0.857
1.03
1.32
1.68
2.4
2.9
3.57
if Results from EPA and
State Labs
18
15 .
45
31
38'
42
50
44
38
29
37
23
48
46
% Labs Passing ± 2 x S.B.
Acceptance Limits
83
100
93
. 90
95
95
88
89
97
93
97
87
92
89
Conclusion _for Hexachlorobenzene
    Together, the results from the method comparison and the method usage over time show that
the most sensitive method, EPA 508, has been consistently used since the promulgation of the
NPDWRs and is one of the more frequently used methods for determination of
hexachlorobenzene. Thus, the MDL for the overall contaminant appears to be unchanged since
the original methods were promulgated. Other less sensitive methods that are also currently in
use include EPA Methods 505 and 525.2.. Examination of the quantitative PE WS data reveal
that the percentage of laboratories successfully passing the proficiency exams is too high to
provide insight toward a re-evaluated PQL.  However, the high laboratory passing rates at
concentrations around the current PQL are suggestive of a change in the PQL,
Methods Support Document fat- Six-Year Review
90
Draft - March 2002

-------
HexacWorocyclopentadiene


Results of the Method Comparison


   With the promulgation of the NPDWRs for Phase V SOCs, the approved methodology for
determination of hexachlorocyclopentadiene included two EPA gas chronaatography methods,
EPA 505 and 525.1. Since that time, EPA 525.1 was discontinued and four new methods were
added (EPA Methods 525.2, 508.1, 551.1, and 508). As indicated in Table 49, the MDLs for the
newer methods (EPA Methods 508,1 and 551.1) are lower than the MDLs of both original
methods, while the MDLs of EPA Methods 505 and 525.2 are comparable to the MDL of the
original EPA 505.                                                               .  •
Table 49.  Results of the Analytical Methods Comparison for Hexachlorocyclopentadiene
MCL = 50ng/L Current BQL = I jig/L " DL* = 0.1 pg/L Acceptance Limit1" = ±2*S.D.
Methods Approved At Promulgation
Method
EPA 505s
EPA 525. 11



Technique
Microextraction, GC
LSE, GC/MS



0^)
0.1
0,03



Currently Approved Methods (141,24)
Method
EPA 5052
EPA 52S.21
EPA 508.12
EPA 551.1*
EPA5082
Technique
Microextraction, GC
LSE, GC/MS
LSE, GC with BCD
LLE, GC with BCD
GC with BCD
MDL
0.13
0.1
0.004
0.018
N/A*
1 "Methods for the Determination of Cnganic Compounds in Drinking Mter," EPA-6QQ/4-88/039, December
1988. . - ' :
2 "Methods for the Determination of Qzganie Compounds in Drinking \Mtter— Supplement HI,"
EPA/60Q/R-95-131, August 1995.
A Regulatory DLs for ojganic compounds are listed at 40 CFR §141 .24(h)( 18).
f Acceptance limits for organic compounds are listed at 40 CFR §141.24(h)(19)(i).
*N/A = not available. • - •
Methods Support Document for Six-Year Review
91
Draft - March 2002

-------
Results of the Analysis of the WS Data
a.  Method Usage Over Time


   Figure 29 illustrates the distribution of the analytical techniques used by EPA and State
laboratories in WS studies 34 to 41. The results for "other" techniques in this figure include the
use of any other technique identified by the laboratories participating in the WS study, as well as
methods for which laboratories did not report any information on the type of method used. The
distribution of methods used by EPA and State laboratories has been fairly well-mixed.  During
WS 34 and 35, EPA 505 was the predominant choice for determination of
hexachlorocyclopentadiene. From WS 36 to 41, however, use of EPA Methods 508 and 525.2
began to eclipse that of EPA 505. The remaining methods were used minimally throughout WS
34 to 41.


     Figure 29. Distribution of Analytical Techniques by WS Study:
     Hexachlorocyclopentadiene

                                 Hexachlorocyclopentadiene
       o>
       "55

       in
       JO
       n
100% -

 90%

 80%

 70%

 60% •

 50%

 40% •

 30% -

 20%

 10%

  0%
                                                              4
                                                              rf
                                                              &
                                                              /*<
                                                                             IB'SQS"
                      JD5D8.1!
                      j IS 525,11
                      JH625.2J
                      |D other |
                 WS 34  WS 35   WS 36
                             WS 37  ' WS 38

                               Water Study
WS 39   WS 40   WS 41
b.  Results of the PQL Analysis


   The current PQL of 1.0 jxg/L was originally set from PR data from WS 23 through 27 (57 FR
31801). The PQL re-evaluation used data from WS 24 to 41 (with the exception of WS 28 which
had no available data). Table 50 summarizes the results of these studies, providing the study
Methods Support Document for Six-Year Review
                                92
             Draft -March 2002

-------
number the spiked value for the WS sample, the Dumber of results from EPA Regional and State
laboratories, and the passing rates of these laboratories when evaluated using an acceptance limit
of ± twice the standard deviation (as-specified at 141.24(h)(19)(i))-                        •
Table 50, Evaluation of Mexachiorocyclopentadiene Data from WS Studies Using the ± 2*
          SJX Acceptance Limits (in Order of Increasing Concentration)
ws#
25b
29
26b
24b
27
32
31
40
38
30
35
25a
41
34
26a
37
33
39
24a
36
Spiked "True" Value (ng/JL)
0,267
0.313
0,367 .
0.736
' 0.774
0,823
1,11
•1.22
1.47
1,72
1.84
1.87
1.93
2,14
'2.47
, 2.49
' 2.92
3.26
4.42
4.71
# Results from EFA and
State Labs
11
14
21
21
14
41
16
47
44
'35
27
12
35
43
21
40
26
.35
.' 21
47
% Labs Passing ± 2* S.D.
Acceptance Limits
54.5
92.9
95.2
95.2
100
87,8
93.8
93,6
93.2
100
, 92.6
83.3
97.1
95.3
81.0.
95.0
96.2
94.3
90.5
93.6 "
 Methods Support Document for Six-Year Review
93
Draft - March 2002

-------
     As shown in Table 50, laboratories exhibited passing rates over 75 percent in all WS studies
 with the exception of one (WS 25b),  Because the passing rates for laboratories determining
 hexachloropentadiene were well above 75 percent, a re-evaluation of the PQL was not feasible.
 However, the demonstrated success of laboratories at concentrations well below the existing PQL
 of 1 ug/L (e.g., WS 27, 29' and 32) suggests that a lower PQL may be possible.


 Conclusion for Hexachlorocyclopentadiene


     The method comparison results indicate mat method sensitivities have either remained
 similar or improved slightly (e.g., for EPA Methods 508.1 and 551.1).  A review of method
 usage over time shows that EPA Methods 505,525.2, and 508 were the most commonly used
 methods in recent WS studies. Because the more sensitive methods were not used with great
 frequency, it would appear that method capabilities for hexachlorocyclopentadiene have
 remained more or less unchanged over time. Although it was not possible to recalculate the
 PQL, high laboratory passing rates at concentrations below the current PQL suggest that a lower
 PQL may be feasible.
 Mercury
 Results of the Method Comparison
                                                 i

    The analytical methods approved for the determination of mercury under the NPDWRs for
 Phase II lOCs include EPA Methods 245.1 and 245.2 (56 FR 3526).  Since the time of
 promulgation, EPA Method 200.8 has been approved. The currently approved methods for
 mercury determination include EPA Methods 200.8,245.1, and 245.2. Table 51 summarizes the
 MDLs for both the original and current approved versions of the methods. As shown in Table
 51, the MDLs for current methods are equal in sensitivity to past methods.
Methods Support Document for Six-Year Review
94
Draft - March 2002

-------
Table 51, Results of the Analytical Methods Comparison for Mercury (Newly Promulgated
          Methods Indicated in Bold)
M,CL = 2|ig7L Current PQ.L - 0','5 }ig/L • »!/ = 0.2 -fig JL Acceptance Ltai if -±30% , •
Methods Approved At Promulgation
Method
EPA 245, 11
EPA 245 .21

Technique
Manual, Cold Vapor
Automated, Cold Vapor

1 "Methods for Chemical Analysis of VMate
2 "Methods for the Determination of Meta
1994.
" Regulatory DJLs for inorganic compound,
method.
f Acceptance limits are listed at 40 CFR §"
MDL :
(fig/L)
0.2
0,2

Currently Approved Methods
Method
EPA 200.8*
EPA 245. 11
BPA245.21
Technique
1CP-MS
Manual, Cold Vapor
Automated, Cold Vapor
MDL*
(M8/D
0,2 .
0,2
0.2
r and Wastes," EPA/600/4-79/020, March 1983,
s in Environmental Samples-Supplement L" EE\/6QQ/R-94/l 1 1, May
5 are listed at 40 CFR §14!.23(a)(4)(i) and depend on analytical
(41.23(k)(3)(ii) for inorganic compounds.
Re.su Its of the Analysis of the WS.. Data
a.  Method Usage Over Time
    Figure 30 shows the distribution of analytical techniques used by EPA and State laboratories
for WS studies 34 tp 41.  The results for "other" techniques in this figure include the use of any
other technique identified by the laboratories participating in the WS study, as well as
"unknown" methods, (i.e.. methods for which laboratories did not report any information on the
type of method used).


    As shown in Figure 30, EPA Method 245.2 (automated cold-vapor) was the most widely used
method in WS 34 and 35. From WS 36 to WS 41, EPA Methods 245.1 (manual cold vapor) and
245.2 were most commonly used.
Methods Support Document for Six-Year Review
95
Draft - March 2002

-------
      Figure 30. Distribution of Analytical Techniques by WS Study: Mercury

                                          Mercury
100%
90%
80%
0 70%
i 60%
D)
I 50%
3
Ja 40%
.3
5? 30%
20%
10%
0% -














n
WS34











Fl



r\m
~~






~l 1
I





n

T






i •
'i

::
IT


n
~H







I H






41
1
	 -;• 	 ~— : 	 1 — "~ 	 	 — r™ — 	 -"T"""™ — "-™1 	 r™™ 	 '"-"•" •"™"M!
WS35 WS36- WS37 WS38 WS 39 WS40 WS41



i B 245,1
• 245.2
D 200.8
D3112B
• other



Water Study
b.  Results of the PQL Analysis


    Tlie current PQL of 0.5 jxg/L was originally set using previous PE data (56 FR 3549), With
the availability of more current data from WS 24 to 41, a PQL re-evaluation was attempted.
Table 52 summarizes the results of these water studies, providing the study number, the spiked
value for the WS sample, the number of results from EPA and State laboratories, and the
mercury results evaluated using an acceptance limit of ± 30 percent, as designated in 40 CFR §
Methods Support Document for Six-Year Revie\v
96
Draft - March 2002

-------
 Table 52.  Evaluation of Mercury Data from WS Studies tJsing the 30% Acceptance Limits
           (in Order of Increasing Concentration)
ws#
29
25a
35
31
27
40
33
24b
26b
36
30
39
25b
26a
34
24a •
41
32 •
38
37
Spiked "True" Value fttg/L)
0,506
0.720
Q.S97
0.908
1,29
1.50
1.77
2,16.
2.47
3.00
' 3.46
3.80
432
4.56
5.09
5.76
5.82
-6.23
6.39
8.16
# Results from EPA and
State Labs
31
37
37
31
33 . '
60
34
59
61
61
62
47
.37
61
61
59
44
64
60
47
% Labs Passing ± 30%
Acceptance Limits
87.1
70,3
94.6
87.1
97,0
93.3
94.1
94.9
93.4
95.1
1 00.0
83,0
91.9
96.7
100.0
96.9
100.0
96.9
95.0
91.5
    The data from the available WS studies were not conducive to recalculation of the PQL
because the percentage of labs passing (with the exception of one study) generally exceeded the
standard 75 percent passing criterion needed to evaluate the PQL using the either a linear
regression or graphical approach. In addition, the majority of the true values were above the
original PQL of 0.5 u.g/L. However, at values slightly above the current PQL (e.g.,  WS 29 and
Methods Support Document for Six-Year Review
97
Draft - March 2002

-------
25a) the passing rates of 87 and 70 percent would indicate that the current PQL is most likely in
the appropriate range.
Conclusion for Mercury
    Since the promulgation of the NPDWR for mercury, EPA 200,8 has been added to the
original two analytical methods approved for the measurement of mercury in drinking water
(EPA 245.1 and 245.2). The three methods currently approved for mercury are all equal in
sensitivity with an MDL of 0.2 u-g/L. According'to the distribution of analytical methods usage
over time, EPA 245.2 was more widely used than EPA 245.1 during WS 34 to 35.  From WS 36
to 41, these two methods were utilized with approximately the same frequency. Upon review of
the WS data, a high percentage of laboratories successfully passing the WS studies prevented a
recalculation of the PQL at the 75 percent passing rate, Thus, the available PE data provided
little evidence for a lower PQL using this approach. Observation of laboratory passing rate's at
concentrations slightly above the current PQL of 0.5 u.g/L suggests that this PQL is still in all
likelihood appropriate.
Methoxychlor
Results of the Method Comparison
    At the promulgation of the NPDWRs for methoxychlor, three methods were approved for the
determination of mis compound: EPA Methods 505, 508, and 525.1 (56 FR 3552). Since that
time, use of EPA 525.1 has been discontinued and additional methods have been approved,
including EPA Methods 508.1, 525.2, and 551.1. All approved methods utilize GC in various
forms. As shown in Table 53, the MDL of EPA 505 has changed little over time, whereas the
MDL of EPA 508 has improved in sensitivity. The MDLs of the newer methods (in bold) range
from 0.008 to 0.1 |ig/L. The most sensitive method currently available (EPA 551) has about 125
times the sensitivity of the original EPA 505.
Methods Support Document for Six-Year Review
98
Draft - March 2002

-------
Table 53.  Results of the Analytical Methods Comparison for Methoxychlor (Newly
          Promulgated Methods in Bold)
MOL = 40iAg/JL Current PQL = 10-ng/JL »I/ = 0.1 ngflU Acceptance Limit* =• ± 45%
Methods Approved At Promulgation
Method
EPA 505!
EPA 508'
EPA 525. 11


Technique
Microextraction, GC
GC with BCD
GC/MS with LSE


MDJL
ow
1.0
0,05
40


Currently Approval Methods (141.24)
Method
EPA 5052
EPA 508 2
EPA. 508.1s
EPA 525.2 2
EPA 551.1*
Technique
Microextraction,
GC
GC with BCD
GC with LSE; BCD
GC/MS with LSE
GC with LLE, BCD
MDL
(l-ig/L)
0.96
0,022
0.015
0,1
• 0,008
1 "Methods for the Determination of Organic Compounds in Drinking U&ter," (EPA/600/4-88/039), December
1988.
2 "Methods for the Determination -of Oiganic Compounds in Drinking Mter— Supplement III," EPA/600/R-95-
13.1, August 1995,
* Regulatory DLs for ojganic compounds are listed at 40 CFR § 141.24{h)(18),
f Acceptance limits for oiganic compounds are listed at 40 CFR § 141.24(li)(I9)(i).
Results. Of the Analysis of the WS Data
a. Method Usage Over Time


  • The distribution of methods used by EPA and State laboratories participating in WS 34 to 41
is depicted in Figure 31. The category of "other" includes any unidentified techniques used by
participating laboratories.  As Figure 31 shows, EPA 508 was used most widely by participating
laboratories during WS 34 to 41, As EPA 525.1 was phased out in WS 36, use of newer
methods, EPA Methods 525.2 and 5084» began. EPA 505 and other unidentified methods were
used intermittently throughout WS 34 to 41.
Methods Support Documentfo'r Six-Year Review
99
Draff - March 2002

-------
     Figure 31. Distribution of Analytical Techniques by WS Study: Methoxychior

                                        Methoxychfor
        o>

        en
        c
        at
100%

 90%

 80%

 70%

 60%

 50%

 40%

 30% •

 20%

 10%

  0% -I
                     I
    I
1
                                                                              [is 505 T
                                      |® 508.1 j
                                      JD 525.11
                                      ID 525.21
                                                                              [•other
                                                                                     !
                 WS 34   WS 35  WS 35
WS 37   WS 38  WS 39

  Water Study
              WS 40   WS 41
b.  Results of the PQL Analysis
    The PQL for methoxychlor was originally proposed at 0,001 mg/L, but was finalized in 1991
at 0.01 mg/L (or 10 p.g/L). The method for deriving this value was multiplication of the
detection limit by a factor often (56 FR 3551).  Recently, with the availability of more recent
laboratory performance d.ata from. WS 24 to 41, a reassessment of the existing PQL was
conducted. Table 54 summarizes each WS result including the spiked (or "true"'} concentration
in the test sample, the number of participating laboratories, and the calculated percentage of
laboratories successfully passing within the specified ± 45 percent acceptance limit for
methoxychlor (141.24(h)(19)(i)).
Table 54. Evaluation of Methoxychlor Data from WS Studies Using the 4:5% Acceptance
          Limits (in Order of Increasing Concentration)
ws#
2<5b
25a
29
Spiked "True" Value (jig/L)
2.18
3.17
5.21
# Results from EPA and
State Labs
59
37
33
% Labs Passing ±45%
Acceptance Limits
98,3
94.6
97.0
Methods Support Document for Six-Year Review
   100
                   Draft.'- March 2002

-------
ws#
24b
31
34
27
32
37
41
36
30
38
33
40
25b
39
35
24a
26a
Spiked "True" Value Oig/L)
5,37
12.9
J4.2
16.6
17.4
18.5
26.8
28.9
34,2
34.8
42.3
42.8
48,8
53.8
62.6
73.2
92.8
# Results from EPA and
State Labs
61 " .
•32
52
39 ' '
59
45
. 43
56
54
51
35
54
37
43
30
62
59
%La'bs Passing ±45%
Acceptance Limits
93.4
.81.3
92.3
89,7
88.1
. 93.3
90,7
92,9
88.9
88,2
97.1
96.3
94.6
93,0 .
96.7
91.9
93,2
    Table 54 reveals that the percentage of passing laboratories was well above the 75 percent
criterion, and thus, the PQL could not be re-evaluated using the regression technique. Table 53
also shows that even at very low concentrations (e.g., one-fifth of the existing PQL), a large
percentage of EPA and State laboratories was able to pass the WS study, suggesting that a lower
PQL may be feasible.  .       •                              •    •


Conclusion for Methoxychior


    EPA Methods 505, 508, and 525.1 were originally approved for the determination of
methoxychlor with the promulgation of the NPDWRs for Phase II SOCs in 1991. Since then,
EPA 525.1 has been removed while EPA Methods 508,1, 525.2, and 551.1 have been added to
the approved list,  A review of the method usage over time shows that laboratories participating
in WS 34 to 41 utilized EPA 508 more frequently than all other approved methods, although this
margin of difference began to decrease in more recent WS studies. A revised PQL for
Methods Siipport Document for Six-Year Review
101
Draft - March 2002

-------
methoxychlor could not be determined due to the large percentage of laboratories surpassing the
75 percent criterion. However, the data do reveal a possible basis for lowering the PQL based on
the high success rates of laboratories at concentrations well below the current .PQL.
Oxamyl
Results of the Method Comparison
    Oxamyl is one of several SOCs first regulated under the Phase V Rule (57 FR 31,776). Under
this rule, EPA 531.1 (a HPLC method) was approved for determination of oxamyl. More
recently, the Agency has approved an additional analytical method, Standard Method (SM) 6610
(HPLC followed by post-column reaction and fluorescence detection), while retaining the use of
EPA 531.1. The MDL for SM 6610 is not specified, as methods published by organizations
outside the Agency are not required to calculate an MDL.  The MDL for EPA 53LI has not
changed since the 1988 Phase V Rule promulgation, signifying no change in analytical sensitivity
for this contaminant.
Table 55. Results of the Analytical Methods Comparison for Oxamyl
MCL <= 200 p.g/L Current PQL = 20 u.g/L DI/ = 2p,g/L Acceptance Limit1" = ± 2*S.D,
Methods Approved At Promulgation
Method
EPA53L1'

Technique
HPLC

MDL
2.0

Currently Approved Methods
Method
EPA 531, la
SM 6610J
Technique
HPLC
HPLC-post column reaction/
fluorescence detection
MDL
(Hg/L)
2.0
NA*
' "Methods for the Determination of Oiganic Compounds in Drinking Water," EPA-600/4-88/039, December
1988.
2 "Methods for the Determination of Otganic Compounds in Drinking Water—Supplement III,"
EPA/600/R-95- 131, August 1995."
J Standard Method 6610. Supplement to the 18th edition of Standard Methods for the Examination of Wter and
Wastewater, 1994, American Public Health Association, 1015 Fifteenth Street NW, Washington, D.C. 20005.
* Regulatory DLs for inoiganic compounds are listed at 40 CFR §141.24(h)(18).
f Acceptance limits are listed at 40 CFR § 141 ,24(h)(19)(i).
* MDLs not provided for voluntary consensus standard methods.
Methods Support Document for Six-Year Review      102
Draft -March 2002

-------
Results of the Analysis pftheWS Data
 a.  Method. Usage Over Time
    The distribution of the methods used by EPA Regional and State laboratories in WS studies
 34 to 41 is shown in Figure 32, The results for "other" techniques in this figure include methods
 for which laboratories did not report any information on the type of method used or reported
 codes that could not be identified.  As shown in Figure 32, EPA 531,1 was the most widely used
 analytical method for determination of oxamyl during WS 34 to  41. By contrast, the more
 recently approved SM 6610 was only used minimally during WS 40.

      Figure 32. Distribution of Analytical Techniques by WS' Study: Oxamyl

                                         Oxamyl
100% -
90% -
80%--
"g 70% -
£
1 60% -
ra
•f 50%
D
J§ 40%
3
SS 30%
20% -
10% -
0% -






























ESI.











m











m











	 EH 	























•H















( D531.il
JH6310 j
j Bother I


WS34   WS35   WS36   WS 37  WS38

                        Water Study
                                                      WS39   WS40  WS41
b. Results of the PQL Analysis
   The current PQL "of 20 u-g/L was originally determined using a ten times MDL multiplier (56
FR 30370). In re-evaluating the PQL, a broader range of PE data from WS 24 to 41 were
analyzed. Table 56 summarizes the results of these water studies, providing the study number,
the spiked (or "true") value for the WS sample, the number of results from EPA and State-
laboratories, and the percentage of laboratories that successfully passed the test using an
acceptance limit of + 2*S.D, (specified at 141 ,24(h)(19)(i)).
Methods Support Document for Six-Year Review
                         103
Draft - March 2002

-------
Table 56.  Evaluation of Oxamyl Data from WS Studies Using the ± 2* S.D. Acceptance
           Limits (in Order of Increasing Concentration)
\vs#
29
31
30
27
24a
32
25a
34
26a
33
24b
41
36
40
37
26b
35
25b
38
39
Spiked "True" Value 
-------
   The use of regression techniques assumes that the data-from the various WS studies form a
single continuous data set. In reality, the.study results do not form a continuous data set, but
represent results from samples spiked at specific discrete concentrations of oxamyl and analyzed
a few at a time over an extended period. However, through the use of a linear regression, the
data can be used to create a model (the regression line) that may be useful in predicting accuracy
and precision as a function of the concentration of the samples. The regression determines the
linear relationship that best fits the observed results, in effect smoothing the curve and ensuring
that there is a unique concentration that corresponds to any percentage of acceptable laboratories.


   Calculating the regression equations also provides the correlation coefficient (r) for the
regression, which is a measure of the degree to which the actual data fit the linear model
represented by the regression line. An r value of one would indicate a perfect fit with a positive
slope of the data to the model. A value, p, can also be calculated for the regression that indicates
the probability of concluding the null hypothesis (in mis case that the spiked value concentration
is linearly correlated with the percentage of labs achieving acceptable results) is false, when in
fact the null hypothesis is true, for the given data set. In statistical terms, p indicates the
probability of a Type I error. The results for the regression equation are summarized in Table 57.
Table 57. Regression Results for Oxamyl
Regression Term
m
; b
r
P
±2*$J>. Acceptance Limits
0,0125
0.223
0.819 '
0.005
    The r value of 0.82 indicates that the data fit the linear model fairly well. The p value of
 0.005 indicates .that the spiked value and the percentage acceptable are linearly correlated with
 more than 99 percent confidence.


    Figure 33 shows the graphical results of the regression.  The observed data were plotted
 against the WS spiked value concentration and the results predicted from the linear regression
 line were superimposed.  The observed success rate for the EPA and State laboratories is
 noticeably lower than the predicted success rate in the region, below 20 [ig/L.  While attempts
 might be made to model the observed results using a second order (non-linear) regression, the
 Agency does not believe that there is a scientifically valid reason to do so.  Figure 34 also shows
 that the percentage of laboratories achieving acceptable results reaches 75 percent at a
 concentration of 39.5 u-g/L which, by the graphical method, indicates the value of the re-
 evaluated PQL (RPQL).  This concentration is higher than the existing PQL of 20 jig/L.    ,
Methods Support Document for Six-Year Review
105
Draft ~ March 2002

-------
     The oxamyl PE data illustrate the expected outcome from PE testing: a lower percentage of
 labs passing at the lowest true values, with an increasing percentage of labs passing as true values
 increase, due to increasing lab analytical capabilities at higher concentrations. As true values
 increase further, the percentage of labs passing approaches a stable plateau beyond which lab
 analytical capabilities do not improve.  This results in a two-part distribution: the lower true
 values (<30 u.g/L) are characterized by a line of steeper slope, while higher true values (>30
 u.g/L), are characterized by a line that is nearly flat (see Figure 33). Using this graph to visually
 estimate a re-evaluated PQL, the original choice of 20 u-g/L appears to be appropriate.
          Figure 33. PQL Evaluation of PE WS Data: Oxamyl

                                              Oxamyl

              100%
                         10
20
                                        30      40     SO
                                          True Value (ug/L)
                              60
70
                                             80
Methods Support Document for Six-Year Review
           106
                                      Draft -March 2002 •

-------
        Figure 34, Two-part Distribution of Oxamyl WS Data

                                           Oxamyi



os
£
•*c5 :
in
to
O.
tn
a
3



I UU /O ~^~™™~~™^^™~~™™~^~~r~ 	 j
90% J .
80% -
'
70% j
60%. •' /
50%,. /
40% \ /
30% - /
* / '*
20% , /
10% -I S
n% J 	 » 	 „... 	 _„ 	 	
»• 	 .-*
, — - ' ~~" **~ - I
/ i
/
'. *
.


current PQL" 20
                        10      20      30     40     50-
                                        True Value 
-------
 508A for determination of DCBP, a Phase IISOC (56 FR 3526), in drinking water. This
 analytical method utilizes gas chrornatography with perchlorination. Since promulgation of the
 Phase II rule, EPA has not changed the status of EPA 508 A and its MDL remains at 0,1 'jig/L
 (Table 58).            '       .                                       ,
 Table 58. Results of the Analytical Methods Comparison for PCBs (Newly Promulgated
          Methods are Indicated in Bold)
MCL»0.5jig/L Current PQL = 0.5 [ig/L DLA=0.1 pg/L Acceptance Limit* « A 300%
Methods Approved At Promulgation
Method
EPA 508A1
Technique
Perchlorination with GC
MDL
(Hg/L)
0,1
Currently Approved Methods
Method
EPA 508A1
Technique
Perchlorination with GC
MDL
Gigfc)
0.1
1 "Methods for the Determination of Oiganic Compounds in Drinking Water," EPA-600/4-88/039, December
19SS.
A Regulatory DLs for organic compounds are listed at 40 CFR §14L24(h)(18),
f Acceptance limits for organic compounds are listed at are listed at 40 CFR §141.240)){19)(i).
Results of the Analysis of the WS Data
a.  Method Usage Over Time
    Because only one method, EPA 508A, was approved for determination of PCBs (as DCBP)
over the duration of WS 34 to 41. its usage was not plotted.
b. Results of the PQL Analysis
   The original PQL for PCBs was proposed at five times the MDL of 0.1 |xg/L. For the final
rule, EPA compared this value with multilaboratory performance data from WS studies 22 to 24,
and found the PE data to support the proposed value, 0.5 u.g/L (56 FR 3552), For the six-year
regulator}' review, new efforts have been made to reassess the PQL using more recent PE WS
data from WS 31 to 41. These data are summarized in Table 59, which indicates the study
number, the true value of the WS sample, the number of results returned  by EPA and State
laboratories, and the calculated percentage of laboratories whose results successfully passed
within designated acceptance limits for PCBs (±100 percent, as specified at 40 CFR
Methods Support Document for Six-Year Review
108
Draft - March 2002

-------
Table 59. Evaluation of PCBs Data from WS Studies Using the ± 100% Acceptance Limits
          (in Order of Increasing Concentration)
ws#
31
37
35
39
38
33
32 .
34
36
40
41
Spiked "True" Value (|ig/L)
0,445
0,527
0.596
' 0.667
0,733
0.807
0,959
. 1.08
1.13 ,
1.23
1.80
# Results from EPA and
State Labs
13
27
15
26
27
14
22
26
30
27
21
% Labs Passing ± 100%
Acceptance Limits
69,2
92.6
80.0
96.2
92.6
92.9
95,5
96.2
93.3
100
100
    A PQL is historically derived from a concentration at which 75 percent of the participating
laboratories pass, or report concentrations that fall within the specified acceptance limits.
However, the data in Table 59 indicate that laboratories exceeded the required 75 percent
criterion in almost all studies (with the exception of WS 31). Because of the high laboratory
passing rates and a lack of sufficient spike concentrations below 0.5 u-g/L, the PQL for PCBs
could not be re-evaluated graphically with these data. However, the two of the three lowest spike
'concentrations close to the current PQL, WS 31  (0,445 u,g/L) and WS 35 (0.596 |j.g/L), had
passing rates of 69 and 80 percent, respectively.  These passing rates suggest that the current PQL
of 0.5 jag/L is unlikely to change.         -


Conclusion for PCBs


 '   Since the promulgation of the NPDWR for PCBss EPA 508A has been the only method
approved for PCB determination. As the current PQL was derived from the MDL multiplier
method rather than multi-laboratory performance data, a PQL re-evaluation was attempted using
current PE data from WS 31 to 41. The high percentage of laboratories passing the PE testing   .
within the designated acceptance limits and the relatively high spike concentrations prevented a
conclusive re-evaluation of the PQL using the historical approach. However, spike
concentrations close to the- current PQL had laboratory passing rates close to 75 percent. This
suggests that the current PQL is appropriate and unlikely to change.
Methods Support Document for Six-Year Review
109
Draft - March 2002

-------
 Pcntaehlorophcnol
 Results of the Method Comparison
    With the promulgation of NPDWRs for Phase II SOCs, two methods were approved for the
 determination of pentaehlorophenol in drinking water: EPA Methods 515,1 and 525,1 (56 FR
 3526). Since promulgation of this rule, EPA removed EPA 525.1 and approved the use of five
 new or updated methods: EPA Methods 515.2, 515.3, 555, 525.2, and ASTM Method D5317-93
 (GC with BCD), a voluntary consensus method. The MDLs of these methods are indicated in
 Table 60.


 Table 60. Results of the Analytical Methods Comparison for Pentaehlorophenol (Newly
          Promulgated Methods Are Indicated in Bold)
MCL =* 1 |ig/JL Current PQL=ljig/L DLA = 0.04 |ig/L Acceptance Limit* *ȣ 50%
Methods Approved At Promulgation
Method
EPA 515.11
EPA 525. 11




Technique
GC with BCD
LSE, GCMS




MDL
(Hg/L)
0.076
0.3-3.0*




Currently Approved Methods (141.24)
Method
EPA 5 15. 12
EPA 52S.22
EPA 515.2"'
EPA 555*
EPA 515.3"
D5317-935
Technique
GC with BCD
LSE GC/MS
LSE, GC with BCD
LLE, derivatization and
GC with. EdD
LLE, derivatization and
GC with BCD
GC with BCD
MDL
fcg/L)
0,032
0.72 -0.1*
0,16
0.15- 1.6*
0,021-0.085*
•N/A*
Methods Support Document for Six-Year Review
110
Draft - March 2002

-------
 1 "Methods for the Determination of Qiganic Compounds in Drinking Mter," EPA-600/4-88/G39, December
 1988,
 1 "Methods for the Determination of Oigatuc Compounds in Drinking U&ter-Supplement III,"
 EPA/600/R-95-131, August 1995.
 3 "Methods for the Determination of Otganic Compounds in Drinking Mter-Suppieraent II," EPA/600/R-
 92/129, August 1992.
 4 "Determination of Chlorinated Acids in Drinking \teter by Liquid-liquid Extraction, Derivatizatton and Gas
 Chromatography with Electron Capture Detection," Revision 1,0, EBi/815/B-99/001, July 1996.
 5 Annual Book ofASTM Standards, Vol.  11.01,  American Society for Testing and Materials, 1961 Race Street,
 Philadelphia, PA 19103.                .    •             :
 A Regulatory DLs for ojganic compounds are listed at 40 CFR §141,24(h)(18).
 f Acceptance limits for organic compounds are listed at are listed at 40 CFR §141.24(h)(19)(i).
 * Multiple MDL values result from variability of reagents, instrumentation and/or laboratory performance.
 0 N/A - not available. MDLs for voluntary consensus standard methods are not specified.	
    As shown in Table 60, most of the current EPA methods for pentachlorophenol do not
display a significant increase in analytical sensitivities as compared to the methods approved at
the time  of the Phase II rule promulgation.  The exception is EPA 515.1 „ which has
approximately twice the sensitivity of the prior version of the same method approved at the time
of promulgation.


Results .ofthe^Analysis, of the. W.S .Data                     -       -     '


a.   Method Usage Over Time                      .


    The distribution of analytical methods used by the EPA and State laboratories in WS 34 to 41
is .plotted in Figure 35. The "other" techniques represent methods which were not specifically
identified by participating laboratories or were otherwise unknown. As shown in Figure 35,
despite the addition of several new analytical methods, the majority of laboratories still favored
the use of EPA 515.1 for determination of pentachlorophenol during WS 34 to 41. The more
recently  approved EPA Methods 515.2 and 555 were used much less frequently, by comparison,
EPA 525.2 was only used intermittently during the study period.
Methods Support Document for Six-Year Review
111
Draft - March 2002

-------
     Figure 35. Distribution of Analytical Techniques by WS Study: Pentaehlorophenol

                                     Pentachlorophenol
       "g
       S
       D)


       I
       in
100%

 90% -

 80%

 70% •

 60%

 50%

 40%

 30%

 20%

 10% •

  0%
                 ,
                                     jdsi5.Tj
                                      • 515.2 i
                                      Ol 525.21
                                      B555  j
                                      S other i
                WS 34   WS 35   WS 36
WS 37   WS 38

  Water Study
                                             WS39  WS40   WS41
b.  Results of the PQL Analysis
   EPA determined the current PQL of 1 u.g/L (0.001 mg/L) for pentachlorophenol using earlier
water supply data (WS 22 to 25, 56 FR 3552).  A re-evaluation of the PQL was attempted using
more recent PE data from WS 24 to 41.  Table 61 summarizes the data from these WS studies
(with the exception of WS 25, 27, 28, and 30), indicating the study number, the true value of the
WS sample, the number of results from EPA and State laboratories, and the percentage of
laboratories whose results successfully passed the acceptance limits of ± 50 percent for
pentachlorophenol (40 CFR §141.24(h)(19)(i)).
Methods Support Document for Six-Year Review
   112
                                                          Draft - March 2002

-------
Table 61.  Evaluation of Pentachlorophenol Data from WS Studies Using the ± 50%
           Acceptance Limits (in Order of Increasing Concentration)
ws#
24a
33
. 26a
37
29
35
32
3J
36
38
24b
40
34
41
26b
39
Spiked "True" Value (fig/L)
0,924
2.72
3,75 ,
6.59
6.73
8.91
10.7
11.4
14.6
14.7
16.2
22.3
22.6
34.6
38,5
43.7
# Results from EPA and
State Labs
13
24
16,
40
11
35
35
16
46
48
13
47
45
• 36
16
41
h«-tTTr..^...,.,.;;;;;;:.;,;,;,^
% Labs Passing ± 50%
Acceptance Limits
84.6
83,3
87.5
75.0
72.7
74.3
85.7
62.5
87.0
87.5
76,9
. ' 95.7
66.7
91,7
75.0
'85,4
    Visual examination of the data show that with the exception of three WS studies (WS 29, 35
and 34), the laboratory passing rates exceeded the 75 percent criterion and only one WS study
had a spike concentration below the current PQL of 1 u.g/L. Because many of the laboratories
achieved high success rates during PE testing for pentachlorophenol, the resulting regression line,
shown in Figure 36, failed to intersect the 75 percent line that would have indicated the new
estimated PQL. Furthermore, there were very few water studies with spiked samples below the
current PQL.  Thus, a new PQL could not be derived using this procedure.
Methods Support Document for Six-Year Review
113
Draft -March 2002

-------
     Figure 36. Evaluation of PE WS Data: PentacMorophenol

                                     Pentaohlorophenol
       en
       To
       
       ra
       CL
       to
       JO
       2
100% -

 90% •'

 80% •

 70%

 60% •

 50%

 40% •

 30%

 20% •

 10%

  0% •
                                                                             75%
                 Current PQL = 1
                             10      15     20     25

                                         True Value (ug/L)
                                                30
35
                                                                        40
45
Conclusion for Pentachlorophenol
   The plot of the method usage over time for WS 34 to 41 indicates laboratories consistently
used EPA 515.1 more frequently than other methods.  The detection limit for EPA 515.1 has, not
changed over time, as shown by the results of the method comparison (Table 60). The available
WS data did not provide a basis for lowering the PQL for two reasons: the passing rates of
laboratories were generally greater than the 75 percent criterion for PQL estimation and the true
value concentrations typically exceeded the current PQL.
Picloram
Results .of the Method Comparison
   Picloram became a regulated SDWA contaminant in 1992 with the promulgation of
NPDWRs for Phase V SOCs (57 PR 31776),  At that time, the sole method approved for
determination of picloram in drinking water was EPA 515.1. Since promulgation of the Phase V
rule, EPA has added two methods to the approved list: EPA Methods 515,2 and 555. Table 62
Methods Support Document for Six-Year Review
                                114
  Draft-March. 2002

-------
summarizes MDL information for the current and former approved methods for picloram, EPA
555 is about five times less sensitive compared to earlier methods.
Table 62. Results of the Analytical Methods Comparison for Picloram (Newly
          Promulgated Methods Are Indicated in Bold)
M.CJL = '500 jig/L Current PQL ~ 1 ng/L DLA ~ 0,1 |ig/L Acceptance Limit* = 2*S.D.
Methods Approved At Promulgation
Method
EPA 515.1'


Technique
GC with BCD


MDL
0.)


Currently Approved Methods
Method
EPA 515 J2
EPA 555-
EPA 515,2'
Technique
GC with BCD
HPLC with a Photodiode
Array Ultraviolet Detector
LSE, GC with BCD

-------
     Figure 37. Distribution of Analytical Techniques by WS Study: Pieloram
                                          Picloram
       •O
       O


       1
       CO


       I
       U)

       2
       g?
100%

 90%

 80% -

 70%

 60%

 50%

 40% •

 30% •

 20%

 10% •

  0%
                          1
                            fs~j

                            Ifc
                                   |S515.1|
                                   [0515.21
                                   [•'555  [
                                   [Bother;
                 WS34   WS35  WS36   WS 37   WS 38  WS 39   WS40   WS41

                                         Water Study
b.  ResuIts of the PQL Analysis


    The original PQL for picloram of 1.0 u.g/L was derived by multiplying the detection limit
(DL) by a factor of 10 (57 FR 31776).  With the availability of more recent PE data, a
reassessment of the PQL was attempted.  Table 63 summarizes the data from WS 24 to 41 (with
the exception of WS 25, 27,28, and 29, which lacked data), indicating the study number,, the true
value of the WS sample, the number of results front EPA and State laboratories, and the
percentage of laboratories whose results successfully passed within the acceptance limits for
picloram.  These limits (±2*S.D.) are not fixed but essentially represent a function of the true
value (40 CFR §141.24(h)(19)(i)).
Methods Support Document for Six-Year Review
116
                                                          Draft - March 2002

-------
Table 63.  Evaluation of Pieloram Data from WS Studies Using the ± 2* S.JX Acceptance
          Limits (in Order of Increasing Concentration)
ws#
26b
24b
32
34
33
24a
30
•37
31
26a
36
40
38
41
35
39
Spiked "True" Value (jigflU)
1,33 '
2.63
10.6
13.2
17.4
17,5
22.4
23.3
26,7
31.2
42.2
44,0
56,4
62.1
' 62,5
74,9
# Results from EPA and
• State Labs
11
12
32
' 43
24
* 11
26
38
12
12
45
43
43
. - • 35
24
37
% Labs Passing ± 2* S.D.
Acceptance Limits
63,6
91.7
93.8
93.0
100
100,
88.5
100 • •
91.7
91.7
100
95,3
93.0
97.1
79,2 .
100
    As.shown by Table 63, participating EPA and State laboratories in every water study but one
 (WS 26b) achieved a passing rate above the 75 percent criterion for determination of the PQL.
 Therefore, estimation of the PQL is likely to not be meaningful for this'dataset. In addition, the
 entire range of true values for this contaminant exceeded the current PQL3 some by nearly two
 orders of magnitude, preventing any assertions on the appropriateness of a lower PQL.


 Conclusion for Plcloram


    The available WS data do not support a reassessment of the PQL based on the 75 percent
 criterion because the passing rates of laboratories almost always exceeded this value. Therefore,
 a quantitative reassessment of the PQL could not be performed using the graphical approach.
 The current PQL of 1 o>g/L appeai-s to still be appropriate.
 Methods Support Document for Six-Year Review
1.17
Draft - March 2002

-------
 Tctrachloroethylene
 Results of the Method Comparison
    The final January 1991 NPDWR for Phase II VOCs (56 PR 3526) approved several analytical
 methods for tetrachloroethylene. These included EPA Methods 502,1, 502.2, 503.1, 524,1, and
 524.2.  Since this regulation was promulgated., the Agency retained EPA Methods 502.2 and
 524.2 for determination of tetrachloroethylene and introduced a new GC variation, EPA Method
 551.1.  Table 64 summarizes the MDLs for both the original and currently approved versions of
 the methods. As shown in Table 64, EPA Methods 502.2 and 551,1 have greater detection
 sensitivity than EPA 524.2.
 Table 64.  Results of the Analytical Methods Comparison for Tetrachloroethylene (Newly
           Promulgated Methods in Bold)
MCL « 5 u.g/L Current PQL - 5 ng/L DL* - 0.5 ng/L Acceptance Limits* * A 20% (>10 |ig/L) or
. ±40%(
-------
Results of the Analysis of the WS Data


a.  Method Usage Over Time          .                                     .


   The distribution of the different methods used.by EPA and State laboratories during WS
studies 34 to 41 are shown in Figure 38. The category of "other" contains those methods that
were unknown or unidentified by the participating laboratories. During WS 34 to 41, EPA 524.2
was the favored method for determining tetrachloroethylene 'among participating laboratories, .
Use of EPA 524.2 generally increased slightly over time while usage of EPA 502,2 declined. No
use of EPA 551.1 was observed during these studies.
     Figure 38, Distribution of Analytical Techniques by WS Study:
     Tetrachloroethylene

                                    Tetrachloroethyletie
         100%
      •o
      a

      i
      0)

      I
      JSt
      TO
                WS 34   WS 35   WS 36
.WS37   WS38   WS39

  Water Study
WS40   WS41
Methods Support Document far Six-Year Review      119
                              Oraft - March 2002

-------
 b.  Results of the PQL Analysis
    The Agency set the original PQL at 5 p,g/L (52 FR 25700 and 56 FR 3526) for all VOCs
 except vinyl chloride. More recent data from WS 27 through 41 (except WS 28 which had no
 available data) were used to re-evaluate the PQL for tetrachloroethylene. Table 65 summarizes
 the results of these WS studies providing the study number, the spiked value for the WS sample.,
 the number of laboratory results reported, and the percent of laboratories passing the WS
 proficiency test for tetrachloroethylene within the acceptance limits of ± 20 percent for a true
 value greater thanlO u.g/L, or ± 40 percent for a true value less than 10 u-g/L (specified at
 Table 65.  Evaluation of Tetrachloroethylene Data from WS Studies Using the ± 20% or ±
           40% Acceptance Limits (in Order of Increasing Concentration)
\vs#
32
39
27
30
37
41
31
35
33
38
40
29
34
36
Spiked "True"
Value (|ig/L)
7.43
7.60
7.76
9.00
9.60
11.5
11.6
11.6
12.9
14.1
14.7
15.6
16.5
18.5
# Results from EPA
and State Labs
63
45
35
60
48
41
33
36
34
56
58
34
60
61
% Labs Passing ± 20%
Acceptance Limits


-


92,7
97.0
91,7-
91.2
94,6
9.1.4
85,3
96.7
91,8
% Labs Passing ± 40%
Acceptance Limits
100
97.S
• 100
100
100









Methods Support Document for Six-Year Review
120
Draft - March 2002

-------
    The data for tetrachloroethylene could not be used to re-evaluate the PQL because State and
EPA laboratories, on average, passed the performance evaluation at a rate over'90 percent.  Thus,
laboratories surpassed the standard 75 percent acceptance criterion typically1 used to determine
the PQL, IE addition, the true value concentrations observed in the available WS data were all
greater than the current PQL of 5 u-g/L. At concentrations close to the current PQL (7.43, 7,60
and 7.76 u-g/L), the passing rates were 100, 98 and 100 percent, respectively. This data suggest
that the current PQL of 5 u-g/L could be lower.
Conclusion-for Tetrachloroethylene
    The method comparison results show that since the promulgation-of analytical methods for
tetrachloroethylene under the NPDWR, one method was retained (EPA 524.2) and two methods
were added (EPA Methods 502,2 and 551,1). The most commonly used method in recent WS
'studies has been EPA 524,2, which is also the least sensitive method of the past and present
methods.  The MDL and method usage information together imply that observable analytical
sensitivities for this contaminant have not improved since the promulgation of the Phase I rule,
This conclusion is further supported by the observation that EPA 551.1, the most sensitive of the
three currently approved methods, appears from the WS data to not be employed by EPA or State'
laboratories.  Evaluation of the quantitative PE data showed that, the majority of the laboratories
conducting WS analyses had surpassed the 75 percent criterion.  The high percentage of
laboratories passing and high true value concentrations apparent hi the WS data prevented a re-
evaluation of the PQL using the  graphical approach.  However, the high laboratory passing rates
at concentrations slightly above the current PQL may suggest that the PQL could be lower
Thallium
Results of the Method Comparison
    With the Phase V lOCs (57 FR 31776), Furnace Atomic Absorption (Furnace AA; EPA
279.2)s Inductively Coupled Plasma (ICP)/MS (EPA 200.8), and Platform Furnace AA  •
Spectrometry (EPA 200,9) were the approved methods listed for analysis of thallium in drinking
water.  Since this regulation was 'promulgated, the only change in approved analytical methods
made by the Agency was the removal of EPA 279.2 from the list of approved analytical methods.
MDLs for EPA Methods 200.8 and 200.9 have not changed, as indicated in Table 66. This table
also shows that the discontinued EPA 279.2 was less sensitive than EPA Methods 200,8 and
200.9, meaning that the collective methods approved since promulgation are, on average, more
sensitive than the methods approved at the time of promulgation, EPA 200.8 is approximately
twice as sensitive than the other currently approved method., EPA 200.9,
Methods Support Document for Sis-Year Review
121
Draft - March 2002

-------
Table 66. Results of the Analytical Methods Comparison for Thallium
MCL = 2u.g/JL Current PQL = 2 ug/L . 1>I/ = 0.3 - 0,7 pg/L Acceptance Limit* = ± 30%
Methods Approved At Promulgation
Method
EPA 279.2'
EPA 200.8'
EPA 200.9'
Technique
Furnace AA
ICP/MS (Platform)
Furnace AA
Spectrometry (Platform)
MDJL

-------
     Figure 39. Distribution of Analytical Techniques by WS Study: Thallium

                                        Thallium
         90% -
         80%
               WS 34   WS 35   WS 36
WS 37   WS 38

  Water Study
WS 39   WS 40   WS 41
b. Results of the PQL Analysis    '


   The current PQL (2 u-g/L) was originally set using PE data from WS 24 through 27, For the
PQL re-evaluation, data were taken from WS 24 to 41 (57 FR 31801). Table 67 summarizes the
results of these water studies, providing the study number, the spiked value for the WS sample,
the number of results from EPA and State laboratories, and .the results evaluated using
acceptance limits of ±3.0 percent (specified at 141.23(k)(3)(ii)).              •
Table 67.  Evaluation of Thallium Data from WS Studies Using the 30% Acceptance Limits
          (In Order of Increasing Concentration)'
ws#
31
24a
37
Spiked "true" Value
(Hg/L)
1.44
2.00
2.38 . •
# Results from EPA and State
Labs
2?
30
43
% Labs Passing ± 30%
Acceptance Limit
82
80
91
Methods Support Document for Six-Tear Review
   123
             Draft - .March 2002

-------
\vs#
32
25b
41
26
36
30
39
34
35
38
33
29
40
24b
27
25a
Spiked "true" Value
(Hg/L)
2,56
3.00
3.50
4.00
4.50
5.30
5.60
6.19
8.00
8.91
9.56
9.74
10.0
18.0
26.9
36.0
n Results from EPA and State
Labs
60
21
44
37
61
48
47
58
41
59
32
21
59
35
23
26
% Labs Passing A 30%
Acceptance Limit
88
81
96
87
98
92
98
97
93
95
91
95
98
89
91
96
    Using the 75 percent criterion, EPA and regional laboratories were able to achieve acceptable
results within the ±30 percent acceptance window over the entire range of tested concentrations,
i.e., 1.4 to 36 u.g/L. Upon examination of the data, the current PQL of 2 \ig/L appears to be
easily supportable, as over 75 percent of laboratories successfully passed within the acceptance
limits when tested at those concentration.  It might even be possible for the labs to pass at a
slightly lower concentration; however, a conclusion is uncertain in the absence of additional data
below the PQL, concentrations.
Conclusion for Thallium
   The method comparison results indicate that use of the least sensitive method was
discontinued since promulgation of NPDWRs.  In addition, the MDLs of all the possible methods
do not differ much between themselves (only by about a factor of two) with the ICP/MS method
having greater sensitivity. The method usage over time shows that EPA 200.9, the less sensitive
method, was the preferred choice for EPA and State laboratories.  Based on the evaluation of
Methods Support Document for Six-Year Review
124
Draft - March 2002

-------
more recent quantitative PE data, the current PQL of 2ug/L using a ± 30 percent acceptance
limit appears to.still be supportable and appropriate. Data.do suggest that it may possible for the
labs to pass at a slightly lower concentration-
Toxaphene
Results of the Method Comparison
    At the promulgation of the NPDWRs for Phase II SOCs, three, GC methods, EPA 505 (GC
with inicroextractionX EPA 508 (GC with BCD), and EPA 525.1 (GC/MS with LSE)S were
approved for the determination of toxaphene in drinking water (56 FR 3552),  Since that time,
EPA has approved two additional GC and GC/MS methods, respectively: EPA Methods 508,1
and 525.2 (both featuring LSE extraction). Table 68 provides a -summary of MDLs for the
approved methods. As shown in Table 68, EPA 508 does not specify the MDL for toxaphene.
The MDL for EPA 525.2 is comparable to that of EPA 505, The MDL of EPA 505 has not  -
changed over time, 1 |ig/L. The MDLs of the newer methods range from about 0.03  to 1 ,7
The most sensitive method currently available, EPA 508.1, has about thirty times the sensitivity
of EPA 505,
 Methods Support Document for Six-Year Review
125
Draft - March 2002

-------
 Table 68. Results of the Analytical Methods Comparison for Toxaphene (Newly
           Promulgated Methods in Bold)
MCL=*3ug/L Current PQL = 3 u-g/L PJL' = l^g/L Acceptance Limit* = ± 45%
Methods Approved At Promulgation
Method
EPA 5051
EPA 508'
EPA 525.1 '

Technique
Microextraction, GC
GC with BCD
GC/MS with LSE

MDL
(Hg/L)
I
No data
7.8-15*

Currently Approved Methods (141,24)
Method
EPA5052
EPA 508 2
EPA 508.1*
EPA 525.2 2
Technique
Microextraction, GC
GC with BCD
GC with LSE, ECD .
GC/MS with LSE
MDL*

1
No data
0.029 '
1.0- 1,7*
1 "Methods for the Determination of Oiganic Compounds in Drinking V&ter," EPA/600/4-88/039, December
1988.
2 "Methods for the Determination of Oiganic Compounds in Drinking Water-Supplement HI," EPA/6QQ/R-95-
131, August 1995.
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation, and/or
laboratory/analyst performance.
•Regulatory DLs forotganic compounds are listed at 40 CFR § I41,24(h)(18),
1 Acceptance limits for organic compounds are listed at 40 CFR § 14L24(h)(19)(i).
 Results of the Analysis of the WS Data
a.  Method Usage Over Time
    Figure 40 illustrates the methods chosen by EPA and State laboratories for toxaphene
analysis during WS PE studies 34 to 41. The category of "other" includes any unidentified or
unreported techniques used by participating laboratories. As shown in Figure 40, the
predominant method used by laboratories participating in the WS studies was EPA 508, followed
by EPA 505.  Other methods such as EPA Methods 525.1, 525.2, and other unidentified methods
were used intermittently throughout WS 34 to 41. EPA 508.1 was not used by laboratories
according to the PE data.
Methods Support Document for Six-Year Review
126
                         Draft - March 2002

-------
    Figure 40, Distribution of Analytical Techniques by WS Study; Toxaphene

                                       Toxaphene
        100% -

         90% -

         80% -

         70%
 •o
 o

 I   60%

 f   50%
 3
' J8   40%

 SS   30%

     20% •

     10%

      0%
                          J3.

               WS 34   WS 35  . WS 36  WS 37   WS 38   WS 39   WS 40   WS 41

                                        Water Study
b. Results of the PQL Analysis *


   The current PQL of 3 'jig/L was derived by multiplying the IMDL by factor of five (56 FR
3526), With the availability of more recent WS data, EPA reviewed the data from WS studies 24
to 41 to attempt a PQL re-evaluation. Table 6.9 summarizes each WS result including the spiked
(or "true") value, the number of participating laboratories,, and the percentage of laboratories
passing within  the specified acceptance limit for toxaphene (± 45 percent as designated in 40.
CFR§14L24(h)(19)(i)),   •
Methods Support Document for Six-Year Review
                                     127
Draft- March 2002

-------
 Table 69. Evaluation of Toxaphene Data from WS Studies Using the 45% Acceptance
           Limits (in Order of Increasing Concentration)
\vs#
25b
24b
30
31
39
26b
32
25a
34
27
41
24a
29
37
33
26a
38
36
40
35
Spiked "True" Value (ng/L)
1.41
2,33
2.80
3.31
3.65
3.68
3.71
4.22
5.37
6.39
6.90
7.58
7.60
8.83
9.23
10.8
12.7
14.7
16.5 .
18.3
# Results from EPA and
State Labs
35
59
54
31
41
58
56
35
54
34
39
60
33
46
33
58
50
• 52
56 -
34
% Labs Passing ± 45%
Acceptance Limits
97.1
91.5
• 92.6
93.5
92.7
94.8
92.9
94.3
94,4
91.2
92.3
91.7
97.0
87.0
90.9
96.6
92.0
90.4
' 94,6
• 97,1
    EPA prefers to evaluate the PQL as the concentration at which 75 percent'of laboratories are
able to pass the proficiency exam within the 45 percent acceptance limits. In the case of
toxaphene, however, this approach could not be used, since the percentage of laboratories
passing within acceptance limits was well above the 75 percent criterion.  Also, very few WS
studies (e.g., 24b, 25b and 30) involved true value concentrations near the current PQL; all other
studies involved spiked values above 3 ug/L.  Because of these data limitations, the PQL could
Methods Support Document for Six-Year Review
128
                          Draft - March 2002

-------
not be re-evaluated using the historical linear regression approach.  However,, the observed high
success rates of laboratories suggests that the PQL maybe lowered. Since 97 percent of .
participating laboratories, on average, were able to determine concentrations at about half of the
existing PQL, this success rate suggests that a lower PQL probably would not challenge most
laboratories'analytical capabilities.


Conclusion for Toxaphene,


   The 1991 NPDWR for toxaphene approved the use of EPA Methods 505, 508, and 525.1 for
toxaphene determination in drinking water.  EPA Methods 508.1 and 525.2, using similar GC
technologies, were approved more recently.  The MDL of EPA 508,1 was the most sensitive
compared to all other methods; however, data on method usage over time revealed that this
method was not employed by laboratories. Instead, laboratories participating in the PE studies
used EPA Method 508 with .the greatest frequency.  Unfortunately, no data are available
regarding the MDL of this method.  A PQL re-evaluation could not be completed due to the high
success rates (well above the 75 percent criterion) in each water study as well as the. high spiked.
concentrations in all but three water studies, • The high success rates of laboratories during PE
testing at concentrations well below the current PQL leads to the inference that a lower PQL
would not greatly lessen laboratory performance.
 1,1,1-Trichloroethane


 Results of the Method Comparison  .


    The NPDWRs for 1,1,1 -triehloroethane approved certain analytical methods for the
 determination of this contaminant (52 FR 25690) and seven other Phase I VOCs. These methods
 included EPA Methods 502.1, 502.2, 503.1, 542.1, and 524.2, Since this regulation was
 promulgated, the Agency retained EPA Methods 502.2 and 524.2 for determination, of 1,1,1-
 trichloroethane and introduced a new GC variation, EPA 551.1, LLE/GC with BCD. Table 70
 summarizes the MDLs for both the original and current methods.
 Methods Support Document for Six-Year Review
129
Draft - March 2002

-------
  Table 70. Results of the Analytical Methods Comparison for 1,14-TttchIoroetiuine (Newly
            Promulgated Methods in Bold)
2QOug/L  Current PQL = 5 |ig/L   M/ = 0.5
                                                        Acceptance .Limit*— ± 20% (>10 ng/JL) or
Methods Approved At Promulgation
Method
EPA 502. 11
EPA 502.21
EPA 524. 11
EPA 524.2'
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL°
'0*gft.)
0.003
0.03
0.3
0.08
Currently Approved Methods (141,24)
Method
EPA 502.22
EPA S24.22
EPA 551.1s •

Technique
Purge and Trap GC
GC/MS
LLE/GC with BCD

MDL
(WJ/i<>
0.0 i- 0.03*
0.04 - 0,08*
0.005

  1 "Methods for the Determination of Oiganic Compounds in Drinking Water," EPA-600/4-88/039, December
  1988.

  2 "Methods for the Determination of Oiganic Compounds in Drinking Mter-Supploraent IH " EPA/600/R-95-
  131, August 1995.

  c The MDLs of the original methods for this contaminant ranged from 0,2 - 0.5 ng/L according to the July 1987
  Federal Register notice promulgating NPDWRs for the VOCs (52 FR 25690). However, the 1988 methods
  manual cited in footnote 1 lists the MDLs shown above,

  * Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
  laboratory/analyst performance.

   Regulatory DLs for organic compounds are listed at 40 CFR § 141,24(f)(I 7)(i).

  tAcceptance limits are listed at 40 CFR § 141.24(f)( 17)(i).
 Results of the Analysis of the WS Data


 a.  Method Usage Over Time


    The distribution of the different methods used by the EPA and State laboratories during WS
 studies 34 to 41 are shown in Figure 41. The category of "other" contains those methods that
 were unknown or unidentified by the participating laboratories. As shown in Figure 41, the
 increase in EPA and State laboratory usage of EPA 524.2 was accompanied by a gradual
 decrease in use of EPA 502.2 during WS 34 to 41.  Although EPA 551.1 was approved during
 this time, laboratories participating in WS studies did not report usage of this method.
Methods Support Document for Six-Year Review
                                   130
                                                                        Draft - March 2002

-------
     Figure 41. Distribution of Analytical Techniques by WS Study:
     Triehloroethane

                                    1,1,1-Trlchf oroethane
       O
      £
      "S
      5
       01
      "5>

       
-------
Table 71. Evaluation of 1,1,1-Trichloroethanc Data from WS Studies Using the ± 20% or ±
          40% Acceptance Limits (in Order of Increasing Concentration)
vvs#
24
34
30
40
27
35
29
32
37
39
25
41
31
26
36
33
38
Spiked "True"
Value (ptg/L)
3.21
5.73
7.13
7.20
7.38
8.78
8.80
10.1
10.3
11.2
11.3
12.6
13.0
13.6
14.5
14.6
17.2
# Results from EPA and
State Labs
57
59
59
57
37
34
35
62
47
43
37
41
36
59
59
33
54
% Labs Passing ± 20%
Acceptance Limits







95,2
87,2
90.7
83,8
100
100
86,4
100
87.9
96.3
% Labs Passing ± 40%
Acceptance Limits
98,2
100
98.3
98.2
97.3
100
100










   The 1,1,1-trichloroethane data from Table 71 are insufficient for a PQL re-evaluation using
the graphical or linear regression approach.  The high laboratory passing rates do not permit
evaluation of the PQL using the 75 percent criterion. In addition, only one spike concentration
(WS 24) was below the current PQL of 5 u-g/L. However, the laboratory passing rates at spike
concentrations around the current PQL exceeded 98 percent. This information suggests that a
lower PQL could be possible.
Methods Support Document for Six-Year Review
132
Draft - March 2002

-------
Coiicliisionfgr.Ll.l-Trkhloroethane
   The method comparison results show that since the promulgation of analytical methods under
the NPDWR, EPA Methods 502 2 and 524,2 have remained in use whereas EPA 551.1 was more
recently approved.  While EPA 551.1 is the most sensitive of the three currently approved
methods, this method is not currently used by EPA or State laboratories according to the
available WS data.  However, all current methods are more sensitive than the methods approved.
at promulgation, as shown in Table 70, Evaluation of the quantitative PE data showed that the
majority of the laboratories conducting WS analyses were able to surpass the 75 percent criterion
needed to evaluate the PQL. Because of the high percentage of laboratories passing and an
insufficient number of spike concentrations below the 5 U-g/L, are-evaluation of the. PQL could
not be performed using the graphical approach. However, the available data do suggest that the
PQL could be lower.
14,2-Trichloroethane
Results of the ...Method .Comparison
    As determined by the Phase I rules for volatile organic compounds (57 FR 31776), the
approved methods listed for analysis of 1,1,2-trichIoroethane in drinking water were purge and.
trap GC and GO/MS. Since this regulation was promulgated, the detection capability of EPA
Method 524.2 have become slightly more sensitive. One new analytical method, LLE/GC with
ECD (EPA 551.1), has been approved by the EPA since the promulgation of the original
methods. As shown in Table 72,' all three currently approved methods have comparable MDLs..
However, EPA 502.1, which is no longer approved for analysis, has the greatest sensitivity level
of all the approved methods, past and present   '     -    .
Methods Support Document for Six-Year Review
133
Draft ~ March 2002

-------
Table 72.  Results of the Analytical Methods Comparison for 1,1,2-Trichloroethane (New
           Methods in Bold)
MCL « 5
                Current PQL = 5 ptg/L    0.L* = 0,5 u.g/1
Acceptance Limit1" = ± 20% (> JO u.g/L) or
               ± 40% (
0.007
ND'
unknown
0.10
Currently Approved Methods (141.24)
Method
EPA 502.22
EPA 524.2*
EPA S5I.12

Technique
Purge and Trap GC
GC/MS
LLE/GC with BCD
•
MDL
(Hg/L)
0.03 '
0,01 - 0.03*
0.012-0.017*

  1 "Methods for the Determination of Oiganic Compounds in Drinking V&teiy" EPA/600/4-88/039, December
  1988.

  2 "Methods for the Determination of Oiganic Compounds in Drinking \\&ter—Supplement 111," EPA/6QO/R-95-
  131, August 1995.

  *The MDLs of the original methods for this contaminant ranged from 0,2 - 0.5 ug/L according to the July 1987
  Federal Register notice promulgating NPDWRs for the VOCs (52 PR 25690). However, the 1988 methods
  manual cited in footnote 1 lists the MDLs shown above.

  * Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
  laboratory/analyst performance.

  A Regulatory DLs for organic compounds are listed at 40 CFR §141,24(f)(17)(i).
  * Acceptance limits are listed at 40 CFR § 141,24(f)( J 7)(i).
  * Not determined
Results of the Analysis of the WS Data
a.  Method Usage Over Time
    Figure 42 shows the distribution of methods.used by the EPA and State laboratories during
WS 34 to 41.  The category of "other" includes any unknown or unreported techniques used by
the participating laboratories. As shown in Figure 42, the predominant methods used since the
time of promulgation are EPA Methods 524.2 and 502.2. Although EPA approved a new method
(EPA 551.1), none of the laboratories that responded used it for analysis.
Methods Support Document for Six-Year Review
                                           134
                 Draft - March 2002

-------
     Figure 42, Distribution of Analytical Techniques by WS'Study: 1,1*2-
     Trichloroethane

                                  1,1,2-Trichloroethane
o
I
      |
      J3
      «J '
80%


70%


60%


50%


40%


30%


20%


10%


 0%
1
                 —i

                in
                i
                                                              £
^
V
fn
                                                                        m.
                                                                             SS24.2
                                                              0502,2


                                                              • other
               WS 34  WS 35
                        WS 36   WS 37  WS 38   WS 39   WS 40  WS 41

                                 Water Study
b. Results' of the PQL analysis                                           '


   The original PQL for Ll>2-trichl0roethane was estimated from PE data (specifically WS 20,
23, and 26) to be 5 iig/L (55 FR 30414), The PQL re-evaluation-data were taken from WS 24
through 41 (note that data were not available for WS 24,25, 27 to 29, and 31). Table 73
summarizes the results of these studies, including the study number, the spiked (or "true") value
for the sample, the number of laboratory results, and the percent of laboratories passing the WS
proficiency test for 1,1,2-trieMoroethane within the specified acceptance limits (± 20 percent for
a spiked value of >10 u-g/L and ± 40 percent for a spiked value of <10 ug/L, as stipulated in 40
CFR§141,23(i)(17)(i)).                           '
Methods Support Document for Six-Year Review
                                     135
                                                            Draft -March 2002

-------
 Table 73.  Evaluation of 1,1,2-Trichloroethane Data from WS Studies Using Either 20% or
           40% Acceptance Limits (in Order of Increasing Concentration)
ws#
36
34
37
30
39
35
32
41
33
38
40
26
Spiked "True"
Value (ng/L)
6,46
8.50
10.7
11.5
12.3
12.8
13.2
13.3
15.7
16.3
17.2
26.9
# Results from EPA
and State Labs
59
59
- 47
50
42
35
51
40
34
54
57 '
54
% Labs Passing
± 20% Acceptance Limits


89.4
94.0
97.6
90.9
90.6
95.1
91.4
98.1
98.2
94.4
% Labs Passing
•± 40% Acceptance Limits
98.3
100










    The standard approach to develop or evaluate a PQL could not be performed with 1,1,2-
trichloroethane because the high laboratory passing percentage data (as shown in Table 73) does
not permit the plotting of a useful linear regression line. Therefore, the re-evaluation of the PQL
could not be done using these WS data. However,, high laboratory passing rates at concentrations
slightly above the current PQL of 5 u.g/L suggest that a lower PQL may be possible.


Conclusion for U .2-Trichloroethane


    The method comparison shows that since the promulgation of the original analytical methods
for 1,1,2-trichloroethane, one new method (EPA 551.1) has been approved for contaminant
analysis.  However, laboratories who responded to the WS studies primarily chose to use EPA
Methods 524.2 and 502.2 rather than EPA  551,1, as shown by the plot of method usage over
time. The MDL for 524.2 has decreased slightly overtime, indicating slightly greater sensitivity.
While EPA 502.1 had a maximum MDL of 0.007 u,g/L, the lowest of any methods, it is currently
not approved for 1,1,2-trichloroethane analyses. The evaluation of the WS data show that the
majority of participating labs surpassed the 75 percent EPA criterion.  Hence, because of the high
percentage of labs passing and the lack of spike concentrations below the current PQL, a re-
evaluation of the PQL could not be performed using linear regression or graphical analysis.
Methods Support Document for Six-Year Review
136
Draft - March 2002

-------
However, the high passing rates at concentrations slightly above 5 fig/L are suggestive of a
potential change in the current PQL.
 Trichloroethylene


 Results of the Method Comparison


    The analytical methods approved for the determination of trichloroethylene under the
 NPDWRs for Phase I VOCs include EPA Methods 502.L 502.2, 524.1, and 524.2 (52 FR
 25899), Since the promulgation of the rule in 1987, the Agency has added EPA 551.1, a GC
 method with liquid-liquid extraction (LLE) and electron capture detector, to the list of approved
 methods.  The currently approved methods for trichloroethylene determination are EPA Methods
 502.2, 524.2, and 551.1. table 74 summarizes the MDLs for both the original and current
, approved versions of the methods. As shown in the Table the current methods are more sensitive
 than the past methods,


 Table 74, Results of the Analytical Methods Comparison for TrichloroethyJene (Newly
          Promulgated Methods in Bold)
MCL = 5itg/L Current PQL = 5 jig/L BL* ~ 0,5 fig/L Acceptance t,hnitf = ± 20% (>10 -p.g/L) or
± 40% (<10 |ig/L)
Methods Approved At Promulgation
Method
EPA 502, 11
EPA 502.21
EPA 503.1*
EPA 524'. I1
EPA 524.2'
Technique
Purge and Trap GC
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL*
(ng/L)
0.001
0.06
0.01
0,4
0.02
Currently Approved Methods (141.24)
Method
EPA 502.2*
EPA524.22 -•
EPA 551 .lz


Technique
Purge and Trap GC
GC/MS
LLE/GC with ECD .


MDL
Cng/L) •
0.01 - 0,4*
0.02-0,19*
0.002


! "Methods for the Determination of Oiganic Compounds in Drinking Witer," EPA/600/4-88/039, December
1988.
2 "Methods for the Determination of Oiganic Compounds in Drinking Mter— Supplement HI," EPA/6QO/R-95-
131, August 1995. . ' . .
* The MDLs of the original methods for this contaminant ranged from 0.2 - 0.5 ng/L according to the July 1987
Federal Register notice promulgating NPDWRs for the VOCs (52 FR 25690). However; the 1988 methods
manual cited in footnote 1 lists the MDLs shown above.
* Multiple method detection limit (MDL) values result from variability .of reagents, instrumentation and/or
laboratory/analyst performance.
* Regulatory DLs for otganic compounds are listed at 40 CFR § 14 1 ,24(f)( 1 7)(i).
t Acceptance limits for VOCs are listed at 40 CFR § 141,24{f)(17)(i).
 Methods Support Document:for Six-Year Review
137
Draft - March 2002

-------
 Results of the Analysis of the WS Data

 a.  Method Usage Over Tune

     Figure 43 shows the distribution of the different methods used by the EPA and State
 laboratories during WS studies 34 to 41. The category of "other" contains those methods that
 were unknown or unidentified by the participating laboratories. As shown in Figure 43,
 laboratories participating in studies WS 34 to 41 used EPA 524.2 with increasing frequency over
 EPA 502.2. No laboratories used EPA 551.1 for PE determinations of trichlorpethylene.
    Figure 43. Distribution of Analytical Techniques by WS Study: Trichloroethylene

                                    Trichloroethylene
     &
     D)
     *35
     
-------
Table 75.  Evaluation of Trichloroethylene Data from. WS Studies Using the ± 20% or ±
          40% Acceptance Limits (in Order of Increasing Concentration)
ws#
40
35
26
41
24
I 31
37
34
30
25
32
38
27
33
29
39
36
Spiked "True"
Value (ng/L)
5,80
6,13
6,63
6.87
7,36
.7.46
8.70
8.89
9.45
10.4
11,2
12.4
, 14.0
14.9
15.9
16.4
17.4
# Results from EPA
and State Labs
58
34
59
4!
5?
36 ' .
48
60
38
37
63
55
3?
34
34
44 ,
61
% Labs Passing ± 20%
Acceptance Limits









83.8
95.2
. 94.5
91.9 - "
94.1
85.3
95.5
96.7
% Labs Passing ± 40%
Acceptance Limits
98.3
97, J
100
100
100
97,2
97,9
100
100








    Table 75 reveals that the percentage of laboratories passing the acceptance limit averaged
over 90 percent. Because the laboratories exceeded the standard 75 percent criterion used to
estimate the PQL, the typical regression method could not be successfully employed to estimate a
new PQL value. Another limitation of the data was that the true value concentrations in the
available studies were all greater than 0.5 p-g/L, preventing evaluation of laboratory performance
at concentrations below the current PQL, However, high laboratory passing rates at
concentrations slightly above the current PQL of 5 fig/L suggest that a lower PQL may be
possible.  .

Conclusion, for .Trichloroethylene

    The method comparison results show that, since the promulgation of analytical methods
under the original NPDWR for trichloroehtyiene, two of these methods are no longer approved
Methods'Support Document for Six-Year Review
139
Draft - March 2002

-------
 for determination of this contaminant. While EPA 551.1 is the most sensitive of the three
 currently approved methods, this method is not currently used by EPA or State laboratories
 according to the available WS data. Instead, EPA 524.2, the least sensitive of the three current
 methods, has been the primary method of choice. Evaluation of the quantitative PE data shows
 mat the majority of the laboratories conducting WS analyses had surpassed the 75 percent
 criterion.  Because  of the high percentage of laboratories passing, a re-evaluation of the PQL
 could not be performed using this approach. However, the high laboratory passing rates at
 concentrations slightly above 5 jxg/L are suggestive of a change in the PQL,


 Vinyl Chloride

 Results of the Method Comparison

    With the promulgation of NPDWRs for Phase I VOCs (proposed November 1985, 50 FR
 46905; finalized July 1987, 52 FR 25690), EPA Methods 502.1, 502.2, 524.1, and 524.2 were
 listed as approved methods for the determination of vinyl chloride in drinking water.  Since
 promulgation of this rule, EPA Methods 502.1 and 524.1 were removed.  As shown in Table 76,
 the MDLs of the current methods are comparable in sensitivity to previously approved methods.


 Table 76.  Results  of the Analytical Methods Comparison for Vinyl Chloride
MCL = 2|ag/L Current PQL = 2 fig/JL 01/ = 0,5 ug/L Acceptance Limit* ** ± 40%
Methods Approved At Promulgation
Method
EPA 502. 1!
EPA 502.2'
EPA 524. 11
EPA 524.2'
Technique
Purge and Trap GC
Purge and Trap GC
GC/MS
GC/MS
MDL°
0.01
0.01-
0.18*
0.3
0.04
Currently Approved Methods
Method
EPA 502.22
EPA 524.21


Technique
Purge and Trap GC
GC/MS


MDL*
0.0 1 -0,04 •
0.04 -OJ 7


1 "Methods for the Determination of Oiganic Compounds in Drinking Wrter," EPA/600/4-88/039, December
1988.
• "Methods for the Determination of Oiganic Compounds in Drinking \\ta--Supplement HI." EPA/600/JR-95-
131, August 1995,
'' The MDLs of the original methods for this contaminant ranged from 0,2 - 0.5 ^g/L accordin* to the July 1987
Federal Register notice promulgating NPDWRs for the VOCs (52 FR 25690). However the 1 988 methods
manual cited in footnote 1 lists the MDLs shown above.
* Multiple method detection limit (MDL) values result from variability of reagents, instrumentation and/or
laboratory/analyst performance,
* Regulatory DLs for Oiganic compounds are listed at 40 CFR § 14 1 ,24(f)(17)(i).
t Acceptance limits for vinyl chloride are listed at 40 CFR § 14 1 ,24(00 7)0).
Methods Support Document for Six-Year Review
140
                          Draft. - March 2002

-------
Results of the Analysis of the WS Data

a.  Method Usage Over Time

   The distribution of methods used by EPA and State laboratories during WS studies 34 to 41
is plotted in. Figure 44, The category of "other" contains those methods that were unknown or
unidentified by the participating laboratories. As shown in Figure 44, over the course of the past
eight WS Studies,' the use of EPA 524.2 has grown while the use of EPA 502.2 has decreased.
EPA 524.2 has consistently remained the most commonly used method for vinyl chloride
determination.
     Figure 44.  Distribution of Analytical Techniques by WS Study: Vinyl Chloride

                                      Vinyl Chloride
        100% -

         90% •

         80%

      •g  70% -
      £
      1  60% -

         50%-
      jg  40% -
      co
      g?  30% -
         20% •

         10% -I

          0%
                                       JS 502,21

                                       (0524.2 j

                                       (•other i
               WS 34   WS 35   WS 36
WS 37   WS 38   WS 39

  Water Study
WS 40   WS 41
b.  Results of the PQL Analysis

    The original PQL of 2 jxg/L for vinyl chloride was determined by using multi-laboratory
performance data, rather than the multiplier procedure used for other VOCs (52 FR 25700). A
re-evaluation of the PQL was attempted using more recent PE data from. WS 24 to 41. Table' 77
summarizes the results of these studies, providing the study number, the spiked value for the WS
sample, the number of laboratory results, and the percentage of laboratories passing the
performance evaluation within the + 40 percent acceptance limits (40 CFR § 141.24(f)(17)(ii)),
Methods Support Document for Six-Year Review
    141
      Draft -March 2002

-------
 Table 77. Evaluation of Vinyl Chloride Data from WS Studies Using ± 40% Acceptance
           Limits (in Order of Increasing Concentration)
ws#
32
27
24
35
30
39
33
26
36
31
25
34
29 '
37
3S
41
40
Spiked "True" Value (ng/L)
2.57
3,57
4.35
4.91
5.48
6,19
7.35
8.70
9.47
11.9
12.4
14.1
14.6
14.8
17.9
22.3
27.2
# Results from EPA and
State Labs
43
39'
57
36
58
45
34
59
60
39
38
59
38
50
• 54
41
58
% Labs Passing ± 40%
Acceptance Limits
79.1
79.5
80,7
1.00
82.8
91.1
79.4
• 86.4
88.3
79.5
. 86.8
93,2
78,9
88.0
96.3
92,7
94.8
    EPA prefers to estimate the PQL by choosing the spiked value at which 75 percent of
laboratories can determine the concentration within the appropriate 40 percent acceptance
window. In the case of vinyl chloride, the results from participating laboratories listed in Table
77 could not be used to re-evaluate the PQL because the laboratory success rate exceeded this 75
percent criterion.  The other limitation of these data was the high range of spiked concentrations
exhibited during the WS studies, which exceeded the current PQL of 2 u-g/L.  Therefore, Hie
available data do not provide sufficient data to recalculate the PQL. However, passing rates of
79 to 80 percent for the three lowest concentrations-(2.57, 3.57 and 4.35 p-g/L) above the current
PQL of 2 p,g/L  suggest that the current PQL is in the appropriate range and unlikely to change.
Methods Support Document for Six-Year Review
142
Draft - March 2002

-------
Conclusion for Vinyl Chloride

   As shown by the method comparison table, EPA Methods 502,2 and 524,2 were approved
with the Phase I Rule promulgation for VOCs and continue to be approved today.  The sensitivity
of newer versions of the methods are comparable older versions. According to the plot of
methods usage over time, the laboratories who participated in WS 34 to 41 employed EPA 524.2
more frequently than EPA 502.2. The available PE data were reviewedbut did not provide
sufficient data to recalculate the PQL, Evaluation of the available WS data suggest that the
current PQL of 2 u.g/L vinyl chloride is unlikely to change.


VI.    Conclusion     .'                .      .

   As part of the 1996-2002 Six-Year Review of National Primary Drinking Water Regulations,
EPA's Office of Ground Water and Drinking Water re-evaluated the analytical feasibility for 38
selected NPDWRs. Table 78 summarizes the results of the analytical feasibility analysis.  Upon
review, EPA found that the majority of the available WS data were insufficient for the
recalculation of the PQL for many of the 38 contaminants of interest. The data were considered.
insufficient because either the true value of the spike-concentrations used in the WS studies were
above the concentration of interest and/or the percentages of labs passing exceeded the 75
percent criterion used to calculate a PQL, However, for many of the 38 contaminants, the
available data were sufficient to indicate whether the PQL might change or if the current PQL is
still appropriate. Of the 38 NPDWRs evaluated, the available information indicates that the PQL
for 23 may possibly be lower. The PQL for the remaining 15 appears to still be appropriate.

   For the 23 analytes where the WS data indicate that a lower PQL may exist, EPA used the
information about method usage over time, the MDLs for these methods, and the 10 x MDL
multiplier to estimate what the potentially lower PQL might be. These estimates are shown, in
Appendix A. Pending the outcome of the health effects review, the majority of these  estimated
values will be used as thresholds in the occurrence and exposure analyses to determine whether
an improvement in public health protection might be possible if BPA were to consider gathering
more definitive data to recalculate the PQL and possibly lower the MCL.
Methods Support Document for Six-Year Review
143
Draft - March 2002

-------
Table 78. Summary of Results from the Methods Comparison
and WS Analysis
SDWA Chemical
Contaminant
t
2
3
4
5
j
7
8
9
10
11
2
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Alachlor
Benzene
Bcnzo(a)pyrcne
Beryllium
Bis (2-cthyIhexyl)phthalate
(also known as Di(2-ethylhcxyl)
phthabtc or DEHP)
Cadmium
Carbofuran
Carbon tetrachloride
Chlordane
Chromium (total - Cr HI and VI)
1 ,2-Dibromo-3-chloropropanc
(DBCP)
1,4-Dichlorobenzcne (para)
1,2-DichIoroethanc
1 , 1-Dichlorocthylene
Dichloromethanc (methylcnc
chloride)
1 ,2-Dichloropropanc
Dioxin - 2,3,7,8-TCDD
Diquat
Ethylene dibromide
Fluoride
Glyphosate
Hepladilor
leptachlor upoxide
Icxachlorobcnzcne
Icxachlorocyclopcntadicnc
Mercury
Methoxyehlor
Oxamyl
'olychlorinated biphenyls (PCBs)
as decachlorobiphenyl)
Pentaehlorophenol
Current PQL (mg/L)
(PQL at the time, of the original
promulgation)
0.002
0,005
0.0002
0.001
0.006
0.002
0.007
0.005
0.002
0.01
0.0002
0.005
0.005
0.005
0.005
0.005
3 x 10-8
0.004
0.00005
0.5
0,0(5
0.0004
0,0002
0.001
0.001
0,002
0.01
0,02
0.0005
0.001
Result of the Six- Year Analytical
Feasibility Reassessment
Current PQL still appropriate
WS Data indicative of change
Current PQL still appropriate
Current PQL still appropriate
Current PQL still appropriate
WS Data indicative of a change
WS Data indicative of a change
WS Data indicative of change
WS Data indicative of change
Current PQL still appropriate
WS. Data indicative of change
WS Data indicative of a change
WS Data indicative of change
WS Data indicative of change
WS Data indicative of change
WS Data indicative of change
PQL most likely still appropriate - no data but
unlikely to change since no new method
approved
Current PQL still appropriate
Current PQL still appropriate
Current PQL still appropriate
Current PQL still appropriate
WS Data indicative of change
WS Data indicative of change
WS Data indicative of change
WS Data indicative of change
Current PQL still appropriate .
WS Data indicative of change
PQL could range from 0,02 to 0.04 mg/L
Current PQL still appropriate
Current PQL still appropriate
Methods Support Document for Six-Year Review
144
                            Draft ~ March 2002

-------
Table 78. Summary of Results from the Methods Comparison
and WS Analysis
SDWA Chemical
Contaminant
31
32
33
34
35
36
37
38
Picloram
Tetrachloroethylena
Thallium
(Non-zero MCLG = 0.0005 rog/L)
Toxaphene
1,1,1 -Triehloroethane
1 , 1,2-Tricliloroetliajie
(Non-zero MCLG = 0,003 mg/L)
Trich loroethy letie
Vinvl chloride
Current PQL (mg/L)
(PQf, at the time of the original
promulgation)
0,001
0.005
0.002
0,003
0.005
0.005
0.005
0.002
Result of the Six- Year Analytical
Feasibility Reassessment
Current PQL still appropriate
WS Data indicative of change
WS Data Indicative of change
WS Data indicative of change •
WS Data indicative of change
WS Data indicative of change
WS Data indicative of change
Current PQL still appropriate
Methods Support Document for Six-Year Review
145
Dfaft-March 2002

-------
References

USEPA.  1985.  National Primary Drinking Water Regulations; Volatile Synthetic Organic
    Chemicals; Final Rule and Proposed Rule.  Federal Register. Vol. 50, No. 219, p. 46880,
    November 13, 1985.

USEPA.  1987.  National Primary Drinking Water Regulations - Synthetic Organic Chemicals;
    Monitoring for Unregulated Contaminants; Final Rule. Federal Register. Vol. 52, No. 130,
    p. 25690, July 8,1987.

USEPA.  1989.  National Primary and Secondary Drinking Water Regulations; Proposed Rule,
    Federal Register. Vol. 54, No. 97, p. 22062, May 22,1989.

USEPA.  1991.  National Primary Drinking Water Regulations - Synthetic Organic Chemicals
    and Inorganic Chemicals; Monitoring for Unregulated Contaminants; National Primary
    Drinking Water Regulations Implementation; National Secondary Drinking Water
    Regulations; Final Rule. Federal Register. Vol. 56, No. 30, p. 3526, January 30,1991.

USEPA.  1996.  Performance Evaluation Studies  Supporting Administration of the Clean Water
    Act and the Safe Drinking Water Act. Federal Register. Vol. 61, No. 139,  p. 37464, July
    18,1996.

USEPA.  1997.  National Primary Drinking Water Regulations: Analytical Methods for
    Radionuclides; Final Rule and Proposed Rule. Federal Register. Vol. 62, No. 43, p. 10167,
    March 5,1997.

USEPA.  1997.  Performance Evaluation Studies  Supporting Administration of the Clean Water
    Act and the Safe Drinking Water Act. Federal Register. Vol. 62, No. 113, p. 32112, June
    12,1997.

USEPA.  1999.  National Primary and Secondary Drinking Water Regulations: Analytical
    Methods for Chemical and Microbiological Contaminants and Revisions to Laboratory
    Certification Requirements; Final Rule. Federal Register, Vol. 64, No. 230, p. 67449,
    December 1,1999.

USEPA.  200 la.  Guidelines Establishing Test Procedures for the Analysis of Pollutants Under
    the Clean Water Act; National Primary Drinking Water Regulations; and National Secondary
    Drinking Water Regulations; Methods Update; Proposed Rule. Federal Register. Vol. 66,, p,
    3 526, January 16,2001.

USEPA.  200Ib. National Primary Drinking Water Regulations; Arsenic and Clarifications to
    Compliance and New Source  Contaminants Monitoring; Final Rule,  Federal Register. Vol.
    66, No. 14, p. 6976, January 22,2001,               .
Methods Support Document for Six-Year Review
146
Draft - March 3002

-------
Appendix A                                                        '             ,

   After re-evaluating more recent Water Supply data for the Six-Year Review, EPA found that  *
insufficient data were available around the 75 percent criterion to actually recalculate the PQL.
However, in many cases, the passing rates for the EPA Regional and State laboratories exceeded
the 75 percent at values close to the current PQL. If the passing rates were greater than 80 to 85
percent at spike concentrations close to the current PQL, then this information was considered to
be indicative of a possible change in the PQL. If data indicated a possible change in .the PQL,
EPA then evaluated the distribution of the analytical, methods used to analyze the spike samples
in the WS studies. Evaluation of the method-usage over time allowed EPA to determine the
analytical methods that appear to be the most widely used for the analysis of a particular
contaminants. Knowledge of which analytical methods are the most widely used, along with the
MDL for. these methods, and. a ten times MDL multiplier allowed EPA .to estimate where the
potential lower limit of quantitation may lie today. These values are shown in Table 79,  Most of
these estimated PQLs have or will be used as a threshold value in the occurrence analysis to help
the Agency determine if there may be a significant gain in public health protection if EPA were
to consider gathering the information needed to recalculate the PQL.
Table 79. Estimated PQLs Based on Method Usage and 10 x MDL Multiplier
Estimated values to use in the Occurrence and Exposure (O/E) analyses
Chemical analyte
1
2
3
4
Benzene . '
Cadmium
Carboftiran
Carbon tetrachloride
Current PQL
(mg/L)
(PQL at the time
of the original
promulgation)
0.005
0,002
0.00?
0.005
Most commonly used methods with
published MDL (mg/L)
EPA 524,2
EPA 502,2
BPA 200.7
EPA 200.8,
EPA 200.9
all used equally
EPA 531.1
EPA 524.2
EPA 502.2
0.00004 (upper MDL)
0.00001
0.001
0.0005
0.00005
0.00052
0.00021
0.00002
Estimated value for O/E
(mg/L)
0.00.04
0.0001
Use upper value of 0.0004 mg/L
0.01
0-005
0.0005
If use upper or intermediate value -
these are higher than current MCL.
Could use 0.0005 mg/L as a value
for O/E but this would probably be
the lower edge of the quantitation
limit.
0.0052
Round to 0.005 mg/L
0.0021
0.0002
Use upper value of 0.0021 and
since this -value is close to one-half
MCL, use 0.0025 mg/L.
Methods Support Document, for Six-Year Review
14?
Draft - March 2002

-------
II
Table 79. Estimated PQLs Based on Method Usage and 10 xMDL Multiplier
Estimated values to use in the Occurrence and Exposure fO/E) analvses

\
JJ6

8
9
10
11
12
Chemical analytc
Chlordanc
l,2-Dibromo-3-
chloropropanc (DBCP)
1,4-Dichlorobenzcnc (para)
1 ,2-DichIorocthanc
1, 1-Dichlorocthylcnc
Dichloromethanc
methylene chloride)
,2-Dichloropropane
leptachlor
Current JPQL
(mg/L)
(PQL at the timt,
of the original
promulgation)
0.002
0,0002
0.005
0.005
0.005
0.005
0.005
.0004
Most commonly used methods with
published MDt (mg/L)
EPA 505
EPA 508
EPA 504.1
EPA 502.2
EPA 524.2
EPA 502.2
EPA 524.2
EPA 524.2
EPA 502.2
EPA 524.2
EPA 502.2
EPA 524.2
EPA 502.2
EPA 525,2
«:PA sos
EPA 508
0.00014
0.0000041
0.00001
0.00029
0.00004 •
0.00007
0.00006
0.00012
0,00007
0.00003
0.00002
0.00004
0.00003
0,00015 (upper MDL)
0.000003
0.0000015
Estimated value for O/E
(mg/L)
0.0014
0.000041
Because of an order of magnitude
difference between these two
values, use the average and round
"P.
Average <= 0,00072 mg/L
Round up to 0.001 mg/L
0,0001 mg/L
0,0029
0.0004 ,
Use the upper value of 0.0029
tog/L. Since close t'o one-half
MCL, could also use 0.0025 mg/L.
0.0007
0.0006
Could use either since so close and
round to 0.001 mg/L
0,0012
0.0007
Could use either both round to
0.001 mg/L
0.0003
0.0002 :
Used average for these two since
close to one-half MCL :
Average = 0.00025 tng/L <
0.0004
0.0003 :
Use upper value 0,0004 mg/L
0.0015
0.00003
0.000015
Because of widespread between
hese - use the intermediate value
nd round to 0.0001 mg/L to be.
onservative
Mtithods Support Document for Six-Year Review
148
                                                                            Draft - March 2002

-------
Table 79. Estimated PQLs Based on Method Usage and 10 x MDL Multiplier
Estimated values to me in the Occurrence and Exposure (O/E) analyses
Chemical analyie
13
14
15
16
1?
18
19
20
21
Heptachlor epoxide
Hexachlorobenzene •
Hexachlorocyclopentadiene
Methoxycblor
Oxarnyl
Tetrachloroethylene
Thallium
(Non-zero MCLG = 0,0005
mg/L)
Toxaphene
1,1,1 -Trichloroethane
Current PQL
(.mg/L)
(PQL at the time
of the original
promulgation)
0,0002
0.001
0.001
0,01
0.02
0,005.
0,002
0.003
0.005
Most commonly used methods with
published MDL (mg/L)
EPA 525,2
EPA 508
EPA 505
EPA 508
EPA 505
EPA 525.2
EPA 508
EPA 505
EPA 525.2
EPA 505
EPA 525,2
EPA 508


EPA 525,2
EPA 502,2
EPA 200.9
EPA 200,8
EPA 508
EPA 505
EPA 524,2
EPA 502,2..
0.00013
0.0000059
0,000004
0,0000077
0,000002
0.000001
N/A
0,00013
0,0001
0.00096
0.0001
0.000022


0,00014
0.00005
0,0007
0.0003
N/A
0.00]
0.00008
0.00003 ,
Estimated value for 0/E
<«ig/L)
0,0013
0.000059
0.00004
Because of widespread between
these - use the intermediate value
and round to 0.000 1 mg/L to be
conservative
0,000077
0.00002
0.00001
Because these are close together -
use upper value of 0,000077 and
round up to 0.0001 mg/L
0.0013 (rounds to 0.001)
0,001
Use 0.00 1 mg/L
0,0096
0;001
0,00022
Use intermediate value of 0.001
mg/L
PQL couid range from 0.02 to 0.04
mg/L
0.0014
0,0005
In this case, used the lower value
of 0.0005 mg/L because of the 95-
100 % passing rates around the
current PQL.
0.007
0.003
These values higher than current
MCL, could not estimate using 10
multiplier.
0.0 i (higher than current MCL}
0.0008
0.0003
Average » 0,0005 mg/L
Methods Support Document for Six-Year Review
149
Draft - March 2002

-------
Table 79. Estimated PQLs Based on Method Usage and 10 x MDL Multiplier
Estimated values to use In the Occurrence and Exposure (O/E) analyses

22
23
Chemical analytc
1 , 1 ,2-1 richloroethane
(Non-zero MCLG = 0.003
mg/L)
Trichlorocthylene
Current PQL
{mg/L)
(PQL at the time
of the original
promulgation)
0.005
0.005
Most commonly used methods with
published MiDL (mg/L)
EPA 524.2
EPA' 502.2
EPA 502.2
EPA 524,2
0.00003
0.00003
0.0004
0.00019
Estimated value for O/E
(mg/L)
0.0003
0,0003
Estimated value is higher than
MCLG of 0,003 mg/L, So should
use MCLG value as threshold in
O/E analysis. j
0.004 ,
0.0019
Average => 0.0029 mg/L ' I
This is cjose to one-half MCL - so
use 0.0025 mg/L.
Methods Support Document for Six-Year Review
150
                            Draft -March 2002

-------