2015-2016 National Monitoring Programs
Annual Report
(UATMP, NATTS, and CSATAM)
Final Report
EPA Contract No. EP-D-14-030
Prepared for:
Jeff Yane and David Shelow
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Prepared by:
Eastern Research Group, Inc.
601 Keystone Park Drive, Suite 700
Morrisville, NC 27560
July 2018
-------
DISCLAIMER
Through its Office of Air Quality Planning and Standards, the U.S. Environmental Protection
Agency funded and managed the research described in this report under EPA Contract No.
EP-D-14-030 to Eastern Research Group, Inc. This report has been subjected to the Agency's
peer and administrative review and has been approved for publication as an EPA document.
Mention of trade names or commercial products in this report does not constitute endorsement or
recommendation for their use.
11
-------
TABLE OF CONTENTS
Page
List of Appendices xii
List of Figures xiii
List of Tables xxvi
List of Acronyms xxxiii
Abstract xxxv
1.0 Introduction 1-1
1.1 Background 1-1
1.2 The Report 1-2
2.0 The 2015-2016 National Monitoring Programs Network 2-1
2.1 Monitoring Locations 2-1
2.2 Analytical Methods and Pollutants Targeted for Monitoring 2-12
2.2.1 Sampling and Analytical Methods for Canister Samples (VOC,
SNMOC, and Methane) 2-14
2.2.2 Carbonyl Compound Sampling and Analytical Method 2-19
2.2.3 PAH Sampling and Analytical Method 2-20
2.2.4 Metals Sampling and Analytical Method 2-21
2.2.5 Hexavalent Chromium Sampling and Analytical Method 2-23
2.3 Sample Collection Schedules 2-23
2.4 Completeness 2-33
3.0 Summary of the 2015-2016 National Monitoring Programs Data Treatment and
Methods for Data Analysis 3-1
3.1 Approach to Data Treatment 3-1
3.2 Human Health Risk and the Pollutants of Interest 3-3
3.3 Additional Program-Level Data Analyses of the 2015-2016 National
Monitoring Programs Dataset 3-7
3.4 Additional Site-Specific Data Analyses 3-8
3.4.1 Site Characterization 3-8
3.4.2 Preliminary Risk-Based Screening and Pollutants of Interest 3-8
3.4.2.1 Site-Specific Comparison to Program-level Average
Concentrations 3-9
3.4.2.2 Site Trends Analysis 3-9
3.4.2.3 Cancer Risk and Noncancer Hazard Approximations 3-11
3.4.2.4 Risk-Based Emissions Assessment 3-12
in
-------
TABLE OF CONTENTS (Continued)
Page
4.0 Summary of the 2015-2016 National Monitoring Programs Data 4-1
4.1 Statistical Results 4-1
4.1.1 Target Pollutant Detection Rates 4-1
4.1.2 Concentration Range and Data Distribution 4-15
4.1.3 Central Tendency 4-15
4.2 Preliminary Risk-Based Screening and Pollutants of Interest 4-17
4.2.1 Concentrations of the Pollutants of Interest 4-22
4.2.2 Variability Analysis for the Pollutants of Interest 4-30
4.2.2.1 Inter-site Variability 4-30
4.2.2.2 Quarterly Variability Analysis 4-59
5.0 Sites in Arizona 5-1
5.1 Site Characterization 5-1
5.2 Pollutants of Interest 5-6
5.3 Concentrations 5-9
5.3.1 2015 and 2016 Concentration Averages 5-9
5.3.2 Concentration Comparison 5-14
5.3.3 Concentration Trends 5-22
5.4 Additional Risk-Based Screening Evaluations 5-41
5.4.1 Cancer Risk and Noncancer Hazard Approximations 5-41
5.4.2 Risk-Based Emissions Assessment 5-45
5.5 Summary of the 2015-2016 Monitoring Data for PXSS and SPAZ 5-49
6.0 Sites in California 6-1
6.1 Site Characterization 6-1
6.2 Pollutants of Interest 6-11
6.3 Concentrations 6-12
6.3.1 2015 and 2016 Concentration Averages 6-13
6.3.2 Concentration Comparison 6-16
6.3.3 Concentration Trends 6-18
6.4 Additional Risk-Based Screening Evaluations 6-25
6.4.1 Cancer Risk and Noncancer Hazard Approximations 6-25
6.4.2 Risk-Based Emissions Assessment 6-27
6.5 Summary of the 2015-2016 Monitoring Data for the California Monitoring
Sites 6-33
7.0 Sites in Colorado 7-1
7.1 Site Characterization 7-1
iv
-------
TABLE OF CONTENTS (Continued)
Page
7.2 Pollutants of Interest 7-15
7.3 Concentrations 7-18
7.3.1 2015 and 2016 Concentration Averages 7-19
7.3.2 Concentration Comparison 7-27
7.3.3 Concentration Trends 7-41
7.4 Additional Risk-Based Screening Evaluations 7-72
7.4.1 Cancer Risk and Noncancer Hazard Approximations 7-72
7.4.2 Risk-Based Emissions Assessment 7-78
7.5 Summary of the 2015-2016 Monitoring Data for the Colorado Monitoring
Sites 7-88
8.0 Site in the District of Columbia 8-1
8.1 Site Characterization 8-1
8.2 Pollutants of Interest 8-5
8.3 Concentrations 8-6
8.3.1 2015 and 2016 Concentration Averages 8-7
8.3.2 Concentration Comparison 8-9
8.3.3 Concentration Trends 8-9
8.4 Additional Risk-Based Screening Evaluations 8-11
8.4.1 Cancer Risk and Noncancer Hazard Approximations 8-11
8.4.2 Risk-Based Emissions Assessment 8-13
8.5 Summary of the 2015-2016 Monitoring Data for WADC 8-16
9.0 Sites in Florida 9-1
9.1 Site Characterization 9-1
9.2 Pollutants of Interest 9-13
9.3 Concentrations 9-15
9.3.1 2015 and 2016 Concentration Averages 9-15
9.3.2 Concentration Comparison 9-20
9.3.3 Concentration Trends 9-25
9.4 Additional Risk-Based Screening Evaluations 9-39
9.4.1 Cancer Risk and Noncancer Hazard Approximations 9-39
9.4.2 Risk-Based Emissions Assessment 9-43
9.5 Summary of the 2015-2016 Monitoring Data for the Florida Monitoring
Sites 9-51
10.0 Sites in Illinois 10-1
10.1 Site Characterization 10-1
v
-------
TABLE OF CONTENTS (Continued)
Page
10.2 Pollutants of Interest 10-9
10.3 Concentrations 10-13
10.3.1 2015 and 2016 Concentration Averages 10-13
10.3.2 Concentration Comparison 10-21
10.3.3 Concentration Trends 10-32
10.4 Additional Risk-Based Screening Evaluations 10-58
10.4.1 Cancer Risk and Noncancer Hazard Approximations 10-58
10.4.2 Risk-Based Emissions Assessment 10-62
10.5 Summary of the 2015-2016 Monitoring Data for NBIL, SPIL, and ROIL 10-69
11.0 Sites in Indiana 11-1
11.1 Site Characterization 11-1
11.2 Pollutants of Interest 11-8
11.3 Concentrations 11-9
11.3.1 2015 and 2016 Concentration Averages 11-9
11.3.2 Concentration Comparison 11-12
11.3.3 Concentration Trends 11-14
11.4 Additional Risk-Based Screening Evaluations 11-20
11.4.1 Cancer Risk and Noncancer Hazard Approximations 11-20
11.4.2 Risk-Based Emissions Assessment 11-22
11.5 Summary of the 2015-2016 Monitoring Data for INDEM and WPIN 11-26
12.0 Sites in Kentucky 12-1
12.1 Site Characterization 12-1
12.2 Pollutants of Interest 12-18
12.3 Concentrations 12-24
12.3.1 2015 and 2016 Concentration Averages 12-25
12.3.2 Concentration Comparison 12-38
12.3.3 Concentration Trends 12-56
12.4 Additional Risk-Based Screening Evaluations 12-99
12.4.1 Cancer Risk and Noncancer Hazard Approximations 12-99
12.4.2 Risk-Based Emissions Assessment 12-108
12.5 Summary of the 2015-2016 Monitoring Data for the Kentucky Monitoring
Sites 12-121
13.0 Site in Massachusetts 13-1
13.1 Site Characterization 13-1
13.2 Pollutants of Interest 13-5
vi
-------
TABLE OF CONTENTS (Continued)
Page
13.3 Concentrations 13-7
13.3.1 2015 and 2016 Concentration Averages 13-7
13.3.2 Concentration Comparison 13-10
13.3.3 Concentration Trends 13-13
13.4 Additional Risk-Based Screening Evaluations 13-18
13.4.1 Cancer Risk and Noncancer Hazard Approximations 13-19
13.4.2 Risk-Based Emissions Assessment 13-21
13.5 Summary of the 2015-2016 Monitoring Data for BOMA 13-24
14.0 Site in Michigan 14-1
14.1 Site Characterization 14-1
14.2 Pollutants of Interest 14-6
14.3 Concentrations 14-7
14.3.1 2015 and 2016 Concentration Averages 14-8
14.3.2 Concentration Comparison 14-12
14.3.3 Concentration Trends 14-18
14.4 Additional Risk-Based Screening Evaluations 14-32
14.4.1 Cancer Risk and Noncancer Hazard Approximations 14-32
14.4.2 Risk-Based Emissions Assessment 14-34
14.5 Summary of the 2015-2016 Monitoring Data for DEMI 14-38
15.0 Site in Missouri 15-1
15.1 Site Characterization 15-1
15.2 Pollutants of Interest 15-5
15.3 Concentrations 15-7
15.3.1 2015 and 2016 Concentration Averages 15-8
15.3.2 Concentration Comparison 15-12
15.3.3 Concentration Trends 15-20
15.4 Additional Risk-Based Screening Evaluations 15-35
15.4.1 Cancer Risk and Noncancer Hazard Approximations 15-35
15.4.2 Risk-Based Emissions Assessment 15-37
15.5 Summary of the 2015-2016 Monitoring Data for S4MO 15-41
16.0 Sites in New Jersey 16-1
16.1 Site Characterization 16-1
16.2 Pollutants of Interest 16-13
16.3 Concentrations 16-17
16.3.1 2015 and 2016 Concentration Averages 16-17
vii
-------
TABLE OF CONTENTS (Continued)
Page
16.3.2 Concentration Comparison 16-25
16.3.3 Concentration Trends 16-35
16.4 Additional Risk-Based Screening Evaluations 16-61
16.4.1 Cancer Risk and Noncancer Hazard Approximations 16-61
16.4.2 Risk-Based Emissions Assessment 16-66
16.5 Summary of the 2015-2016 Monitoring Data for the New Jersey Monitoring
Sites 16-75
17.0 Sites in New York 17-1
17.1 Site Characterization 17-1
17.2 Pollutants of Interest 17-8
17.3 Concentrations 17-10
17.3.1 2015 and 2016 Concentration Averages 17-10
17.3.2 Concentration Comparison 17-14
17.3.3 Concentration Trends 17-19
17.4 Additional Risk-Based Screening Evaluations 17-29
17.4.1 Cancer Risk and Noncancer Hazard Approximations 17-29
17.4.2 Risk-Based Emissions Assessment 17-31
17.5 Summary of the 2015-2016 Monitoring Data for BXNY and ROCH 17-35
18.0 Sites in Oklahoma 18-1
18.1 Site Characterization 18-1
18.2 Pollutants of Interest 18-15
18.3 Concentrations 18-21
18.3.1 2015 and 2016 Concentration Averages 18-22
18.3.2 Concentration Comparison 18-34
18.3.3 Concentration Trends 18-52
18.4 Additional Risk-Based Screening Evaluations 18-83
18.4.1 Cancer Risk and Noncancer Hazard Approximations 18-83
18.4.2 Risk-Based Emissions Assessment 18-91
18.5 Summary of the 2015-2016 Monitoring Data for the Oklahoma Monitoring
Sites 18-102
19.0 Site in Rhode Island 19-1
19.1 Site Characterization 19-1
19.2 Pollutants of Interest 19-5
19.3 Concentrations 19-6
19.3.1 2015 and 2016 Concentration Averages 19-7
viii
-------
TABLE OF CONTENTS (Continued)
Page
19.3.2 Concentration Comparison 19-9
19.3.3 Concentration Trends 19-11
19.4 Additional Risk-Based Screening Evaluations 19-14
19.4.1 Cancer Risk and Noncancer Hazard Approximations 19-14
19.4.2 Risk-Based Emissions Assessment 19-16
19.5 Summary of the 2015-2016 Monitoring Data for PRRI 19-19
20.0 Site in Utah 20-1
20.1 Site Characterization 20-1
20.2 Pollutants of Interest 20-5
20.3 Concentrations 20-7
20.3.1 2015 and 2016 Concentration Averages 20-8
20.3.2 Concentration Comparison 20-12
20.3.3 Concentration Trends 20-19
20.4 Additional Risk-Based Screening Evaluations 20-34
20.4.1 Cancer Risk and Noncancer Hazard Approximations 20-34
20.4.2 Risk-Based Emissions Assessment 20-38
20.5 Summary of the 2015-2016 Monitoring Data for BTUT 20-42
21.0 Site in Vermont 21-1
21.1 Site Characterization 21-1
21.2 Pollutants of Interest 21-5
21.3 Concentrations 21-6
21.3.1 2015 and 2016 Concentration Averages 21-7
21.3.2 Concentration Comparison 21-10
21.3.3 Concentration Trends 21-11
21.4 Additional Risk-Based Screening Evaluations 21-14
21.4.1 Cancer Risk and Noncancer Hazard Approximations 21-14
21.4.2 Risk-Based Emissions Assessment 21-16
21.5 Summary of the 2015-2016 Monitoring Data for the Vermont Monitoring
Site 21-19
22.0 Site in Virginia 22-1
22.1 Site Characterization 22-1
22.2 Pollutants of Interest 22-5
22.3 Concentrations 22-6
22.3.1 2015 and 2016 Concentration Averages 22-7
22.3.2 Concentration Comparison 22-9
ix
-------
TABLE OF CONTENTS (Continued)
Page
22.3.3 Concentration Trends 22-10
22.4 Additional Risk-Based Screening Evaluations 22-11
22.4.1 Cancer Risk and Noncancer Hazard Approximations 22-11
22.4.2 Risk-Based Emissions Assessment 22-13
22.5 Summary of the 2015-2016 Monitoring Data for RIVA 22-16
23.0 Site in Washington 23-1
23.1 Site Characterization 23-1
23.2 Pollutants of Interest 23-5
23.3 Concentrations 23-7
23.3.1 2015 and 2016 Concentration Averages 23-7
23.3.2 Concentration Comparison 23-10
23.3.3 Concentration Trends 23-15
23.4 Additional Risk-Based Screening Evaluations 23-24
23.4.1 Cancer Risk and Noncancer Hazard Approximations 23-24
23.4.2 Risk-Based Emissions Assessment 23-26
23.5 Summary of the 2015-2016 Monitoring Data for SEWA 23-30
24.0 Data Quality 24-1
24.1 Completeness 24-1
24.2 Method Precision 24-2
24.2.1 VOC Method Precision 24-4
24.2.2 SNMOC Method Precision 24-14
24.2.3 Methane Method Precision 24-20
24.2.4 Carbonyl Compound Method Precision 24-21
24.2.5 PAH Method Precision 24-25
24.2.6 Metals Method Precision 24-27
24.2.7 Hexavalent Chromium Method Precision 24-28
24.3 Analytical Precision 24-29
24.3.1 VOC Analytical Precision 24-30
24.3.2 SNMOC Analytical Precision 24-40
24.3.3 Methane Analytical Precision 24-47
24.3.4 Carbonyl Compound Analytical Precision 24-47
24.3.5 PAH Analytical Precision 24-53
24.3.6 Metals Analytical Precision 24-57
24.3.7 Hexavalent Chromium Analytical Precision 24-60
24.4 Accuracy 24-61
x
-------
TABLE OF CONTENTS (Continued)
Page
25.0 Results, Conclusions, and Recommendations 25-1
25.1 Summary of Results 25-1
25.1.1 Program-level Results Summary 25-1
25.1.2 State-level Results Summary 25-2
25.1.3 Composite Site-level Results Summary 25-17
25.1.4 Data Quality Results Summary 25-23
25.2 Conclusions 25-24
25.3 Recommendations 25-26
26.0 References 26-1
XI
-------
List of Appendices
Appendix A
AQS Site Descriptions for the 2015-2016 NMP Monitoring Sites
Appendix B
2015-2016 VOC Raw Data
Appendix C
2015-2016 SNMOC Raw Data
Appendix D
2015-2016 Methane Raw Data
Appendix E
2015-2016 Carbonyl Compounds Raw Data
Appendix F
2015-2016 PAH Raw Data
Appendix G
2015-2016 Metals Raw Data
Appendix H
2015-2016 Hexavalent Chromium Raw Data
Appendix I
Summary of Invalidated 2015-2016 Samples
Appendix J
2015-2016 Summary Statistics for VOC Monitoring
Appendix K
2015-2016 Summary Statistics for SNMOC Monitoring
Appendix L
2015-2016 Summary Statistics for Methane Monitoring
Appendix M
2015-2016 Summary Statistics for Carbonyl Compounds Monitoring
Appendix N
2015-2016 Summary Statistics for PAH Monitoring
Appendix 0
2015-2016 Summary Statistics for Metals Monitoring
Appendix P
2015-2016 Summary Statistics for Hexavalent Chromium Monitoring
Appendix Q
Risk Factors Used Throughout the 2015-2016 NMP Report
Xll
-------
LIST OF FIGURES
Page
2-1 Locations of the 2015-2016 National Monitoring Programs Monitoring Sites 2-3
4-1 Inter-Site Variability for Acenaphthene 4-31
4-2 Inter-Site Variability for Acetaldehyde 4-32
4-3 Inter-Site Variability for Arsenic 4-34
4-4a Inter-Site Variability for Benzene - Method TO-15 4-35
4-4b Inter-Site Variability for Benzene - SNMOC 4-37
4-5a Inter-Site Variability for 1,3-Butadiene - Method TO-15 4-39
4-5b Inter-Site Variability for 1,3-Butadiene - SNMOC 4-41
4-6 Inter-Site Variability for Carbon Tetrachloride 4-43
4-7 Inter-Site Variability for/?-Dichlorobenzene 4-45
4-8 Inter-Site Variability for 1,2-Dichloroethane 4-47
4-9a Inter-Site Variability for Ethylbenzene - Method TO-15 4-49
4-9b Inter-Site Variability for Ethylbenzene - SNMOC 4-51
4-10 Inter-Site Variability for Fluorene 4-53
4-11 Inter-Site Variability for Formaldehyde 4-54
4-12 Inter-Site Variability for Hexachloro-l,3-butadiene 4-56
4-13 Inter-Site Variability for Naphthalene 4-58
4-14 Comparison of Quarterly Average Acenaphthene Concentrations 4-60
4-15 Comparison of Quarterly Average Acetaldehyde Concentrations 4-61
4-16a Comparison of Quarterly Average Arsenic (PMio) Concentrations 4-64
4-16b Comparison of Quarterly Average Arsenic (TSP) Concentrations 4-65
4-17 Comparison of Quarterly Average Benzene Concentrations 4-66
4-18 Comparison of Quarterly Average 1,3-Butadiene Concentrations 4-68
4-19 Comparison of Quarterly Average Carbon Tetrachloride Concentrations 4-70
4-20 Comparison of Quarterly Average/?-Dichlorobenzene Concentrations 4-72
4-21 Comparison of Quarterly Average 1,2-Dichloroethane Concentrations 4-74
4-22 Comparison of Quarterly Average Ethylbenzene Concentrations 4-76
4-23 Comparison of Quarterly Average Fluorene Concentrations 4-78
4-24 Comparison of Quarterly Average Formaldehyde Concentrations 4-80
4-25 Comparison of Quarterly Average Hexachloro-1,3-butadiene Concentrations 4-82
4-26 Comparison of Quarterly Average Naphthalene Concentrations 4-84
5-1 Phoenix, Arizona (PXSS) Monitoring Site 5-2
5-2 South Phoenix, Arizona (SPAZ) Monitoring Site 5-3
5-3 NEI Point Sources Located Within 10 Miles of PXSS and SPAZ 5-4
5-4 Program vs. Site-Specific Average Acetaldehyde Concentrations 5-15
5-5 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 5-15
5-6 Program vs. Site-Specific Average Benzene Concentrations 5-16
5-7 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 5-17
5-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 5-18
5-9 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 5-19
5-10 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 5-20
5-11 Program vs. Site-Specific Average Ethylbenzene Concentrations 5-21
5-12 Program vs. Site-Specific Average Formaldehyde Concentrations 5-21
xiii
-------
LIST OF FIGURES (Continued)
Page
5-13 Program vs. Site-Specific Average Naphthalene Concentrations 5-22
5-14 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS 5-23
5-15 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PXSS 5-24
5-16 Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS 5-25
5-17 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PXSS 5-26
5-18 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
PXSS 5-27
5-19 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
PXSS 5-29
5-20 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
PXSS 5-30
5-21 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS 5-31
5-22 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PXSS 5-32
5-23 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS 5-33
5-24 Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ 5-34
5-25 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPAZ 5-36
5-26 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPAZ 5-37
5-27 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
SPAZ 5-38
5-28 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPAZ 5-39
5-29 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at SPAZ 5-40
6-1 Los Angeles, California (CELA) Monitoring Site 6-2
6-2 NEI Point Sources Located Within 10 Miles of CELA 6-3
6-3 Rubidoux, California (RUCA) Monitoring Site 6-4
6-4 NEI Point Sources Located Within 10 Miles of RUCA 6-5
6-5 San Jose, California (SJJCA) Monitoring Site 6-6
6-6 NEI Point Sources Located Within 10 Miles of SJJCA 6-7
6-7 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 6-16
6-8 Program vs. Site-Specific Average Naphthalene Concentrations 6-17
6-9 Program vs. Site-Specific Average Nickel (PMio) Concentrations 6-18
6-10 Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA 6-19
6-11 Yearly Statistical Metrics for Naphthalene Concentrations Measured at RUCA 6-20
6-12 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at SJJCA 6-21
6-13 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SJJCA 6-23
6-14 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at SJJCA 6-24
7-1 Grand Junction, Colorado (GPCO) Monitoring Site 7-2
7-2 NEI Point Sources Located Within 10 Miles of GPCO 7-3
7-3 Battlement Mesa, Colorado (BMCO) Monitoring Site 7-4
7-4 Silt, Colorado (BRCO) Monitoring Site 7-5
7-5 Parachute, Colorado (PACO) Monitoring Site 7-6
7-6 Rifle, Colorado (RICO) Monitoring Site 7-7
7-7 NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and RICO 7-8
xiv
-------
LIST OF FIGURES (Continued)
Page
7-8 Glenwood Springs, Colorado (GSCO) Monitoring Site 7-9
7-9 Carbondale, Colorado (RFCO) Monitoring Site 7-10
7-10 NEI Point Sources Located Within 10 Miles of GSCO and RFCO 7-11
7-11 Program vs. Site-Specific Average Acenaphthene Concentrations 7-28
7-12 Program vs. Site-Specific Average Acetaldehyde Concentrations 7-29
7-13 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 7-30
7-14a Program vs. Site-Specific Average Benzene (Method TO-15) Concentrations 7-31
7-14b Program vs. Site-Specific Average Benzene (SNMOC) Concentrations 7-32
7-15a Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15) Concentrations ... 7-33
7-15b Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations 7-34
7-16 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 7-35
7-17 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 7-36
7-18 Program vs. Site-Specific Average Dichloromethane Concentrations 7-37
7-19a Program vs. Site-Specific Average Ethylbenzene (Method TO-15) Concentrations.... 7-37
7-19b Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentrations 7-38
7-20 Program vs. Site-Specific Average Fluoranthene Concentrations 7-39
7-21 Program vs. Site-Specific Average Formaldehyde Concentrations 7-40
7-22 Program vs. Site-Specific Average Naphthalene Concentrations 7-41
7-23 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at GPCO 7-42
7-24 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GPCO 7-44
7-25 Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO 7-45
7-26 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GPCO 7-46
7-27 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GPCO 7-47
7-28 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GPCO 7-48
7-29 Yearly Statistical Metrics for Dichloromethane Concentrations Measured at GPCO... 7-49
7-30 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at GPCO 7-50
7-31 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at GPCO 7-51
7-32 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GPCO 7-53
7-33 Yearly Statistical Metrics for Naphthalene Concentrations Measured at GPCO 7-54
7-34 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BRCO 7-55
7-35 Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO 7-56
7-36 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BRCO 7-57
7-37 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PACO 7-58
7-38 Yearly Statistical Metrics for Benzene Concentrations Measured at PACO 7-59
7-39 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at PACO 7-61
7-40 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at PACO 7-62
7-41 Yearly Statistical Metrics for Benzene Concentrations Measured at RFCO 7-63
7-42 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RFCO 7-64
7-43 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at RICO 7-65
7-44 Yearly Statistical Metrics for Benzene Concentrations Measured at RICO 7-66
7-45 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at RICO 7-68
7-46 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at RICO 7-69
7-47 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at RICO 7-71
xv
-------
LIST OF FIGURES (Continued)
Page
8-1 Washington, D.C. (WADC) Monitoring Site 8-2
8-2 NEI Point Sources Located Within 10 Miles of WADC 8-3
8-3 Program vs. Site-Specific Average Naphthalene Concentrations 8-9
8-4 Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC 8-10
9-1 St. Petersburg, Florida (AZFL) Monitoring Site 9-2
9-2 Pinellas Park, Florida (SKFL) Monitoring Site 9-3
9-3 NEI Point Sources Located Within 10 Miles of AZFL and SKFL 9-4
9-4 Valrico, Florida (SYFL) Monitoring Site 9-5
9-5 NEI Point Sources Located Within 10 Miles of SYFL 9-6
9-6 Winter Park, Florida (ORFL) Monitoring Site 9-7
9-7 Orlando, Florida (PAFL) Monitoring Site 9-8
9-8 NEI Point Sources Located Within 10 Miles of ORFL and PAFL 9-9
9-9 Program vs. Site-Specific Average Acetaldehyde Concentrations 9-21
9-10 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 9-22
9-11 Program vs. Site-Specific Average Formaldehyde Concentrations 9-23
9-12 Program vs. Site-Specific Average Naphthalene Concentrations 9-24
9-13 Program vs. Site-Specific Average Nickel (PMio) Concentrations 9-24
9-14 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at AZFL 9-26
9-15 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at AZFL 9-27
9-16 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SKFL 9-28
9-17 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SKFL 9-30
9-18 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL 9-31
9-19 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SYFL 9-32
9-20 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SYFL 9-33
9-21 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ORFL 9-35
9-22 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ORFL 9-36
9-23 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at PAFL 9-37
9-24 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at PAFL 9-38
9-25 Pollution Rose for Formaldehyde Concentrations Measured at AZFL 9-42
10-1 Northbrook, Illinois (NBIL) Monitoring Site 10-2
10-2 Schiller Park, Illinois (SPIL) Monitoring Site 10-3
10-3 NEI Point Sources Located Within 10 Miles of NBIL and SPIL 10-4
10-4 Roxana, Illinois (ROIL) Monitoring Site 10-5
10-5 NEI Point Sources Located Within 10 Miles of ROIL 10-6
10-6 Program vs. Site-Specific Average Acenaphthene Concentrations 10-21
10-7 Program vs. Site-Specific Average Acetaldehyde Concentrations 10-22
10-8 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 10-23
10-9 Program vs. Site-Specific Average Benzene Concentrations 10-24
10-10 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 10-25
10-11 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 10-26
10-12 Program vs. Site-Specific Average/>-Dichlorobenzene Concentrations 10-27
10-13 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 10-28
10-14 Program vs. Site-Specific Average Fluoranthene Concentrations 10-29
10-15 Program vs. Site-Specific Average Fluorene Concentrations 10-29
xvi
-------
LIST OF FIGURES (Continued)
Page
10-16 Program vs. Site-Specific Average Formaldehyde Concentrations 10-30
10-17 Program vs. Site-Specific Average Naphthalene Concentrations 10-31
10-18 Program vs. Site-Specific Average Trichloroethylene Concentrations 10-32
10-19 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at NBIL 10-33
10-20 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBIL 10-34
10-21 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at NBIL 10-36
10-22 Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL 10-37
10-23 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBIL 10-38
10-24 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBIL 10-40
10-25 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
NBIL 10-41
10-26 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBIL 10-42
10-27 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at NBIL 10-44
10-28 Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL 10-45
10-29 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBIL 10-46
10-30 Yearly Statistical Metrics for Naphthalene Concentrations Measured at NBIL 10-48
10-31 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SPIL 10-49
10-32 Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL 10-50
10-33 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SPIL 10-51
10-34 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SPIL 10-53
10-35 Yearly Statistical Metrics for/;-Dichlorobenzene Concentrations Measured at
SPIL 10-54
10-36 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SPIL 10-55
10-37 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SPIL 10-56
10-38 Yearly Statistical Metrics for Trichloroethylene Concentrations Measured at SPIL.. 10-57
11-1 Gary, Indiana (INDEM) Monitoring Site 11-2
11-2 NEI Point Sources Located Within 10 Miles of INDEM 11-3
11-3 Indianapolis, Indiana (WPIN) Monitoring Site 11-4
11-4 NEI Point Sources Located Within 10 Miles of WPIN 11-5
11-5 Program vs. Site-Specific Average Acetaldehyde Concentrations 11-12
11-6 Program vs. Site-Specific Average Formaldehyde Concentrations 11-13
11-7 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at INDEM.... 11-14
11-8 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at INDEM... 11-16
11-9 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at WPIN 11-18
11-10 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at WPIN 11-19
12-1 Ashland, Kentucky (ASKY) Monitoring Site 12-2
12-2 Ashland, Kentucky (ASKY-M) Monitoring Site 12-3
12-3 NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M 12-4
12-4 Grayson, Kentucky (GLKY) Monitoring Site 12-5
12-5 NEI Point Sources Located Within 10 Miles of GLKY 12-6
xvii
-------
LIST OF FIGURES (Continued)
Page
12-6 Baskett, Kentucky (BAKY) Monitoring Site 12-7
12-7 NEI Point Sources Located Within 10 Miles of BAKY 12-8
12-8 Calvert City, Kentucky (ATKY) Monitoring Site 12-9
12-9 Smithland, Kentucky (BLKY) Monitoring Site 12-10
12-10 Calvert City, Kentucky (TVKY) Monitoring Site 12-11
12-11 NEI Point Sources Located Within 10 Miles of ATKY, BLKY, and TVKY 12-12
12-12 Lexington, Kentucky (LEKY) Monitoring Site 12-13
12-13 NEI Point Sources Located Within 10 Miles of LEKY 12-14
12-14 Program vs. Site-Specific Average Acetaldehyde Concentrations 12-39
12-15 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 12-40
12-16 Program vs. Site-Specific Average Benzene Concentrations 12-41
12-17 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 12-43
12-18 Program vs. Site-Specific Average Cadmium (PMio) Concentrations 12-45
12-19 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 12-46
12-20 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations 12-47
12-21 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 12-48
12-22 Program vs. Site-Specific Average Formaldehyde Concentrations 12-49
12-23 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 12-50
12-24 Program vs. Site-Specific Average Lead (PMio) Concentrations 12-51
12-25 Program vs. Site-Specific Average Manganese (PMio) Concentrations 12-52
12-26 Program vs. Site-Specific Average Nickel (PMio) Concentrations 12-53
12-27 Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentrations 12-54
12-28 Program vs. Site-Specific Average Vinyl Chloride Concentrations 12-55
12-29 Yearly Statistical Metrics for Benzene Concentrations Measured at ASKY 12-56
12-30 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ASKY 12-57
12-31 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ASKY 12-58
12-32 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ASKY 12-59
12-33 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at ASKY 12-60
12-34 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
ASKY-M 12-61
12-35 Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
ASKY-M 12-62
12-36 Yearly Statistical Metrics for Lead (PMio) Concentrations Measured at ASKY-M... 12-63
12-37 Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
ASKY-M 12-64
12-38 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at ASKY-M. 12-65
12-39 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at GLKY 12-66
12-40 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at GLKY ... 12-67
12-41 Yearly Statistical Metrics for Benzene Concentrations Measured at GLKY 12-68
12-42 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at GLKY 12-69
12-43 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
GLKY 12-70
xviii
-------
LIST OF FIGURES (Continued)
Page
12-44 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GI.KY 12-71
12-45 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at GLKY 12-72
12-46 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BAKY... 12-73
12-47 Yearly Statistical Metrics for Benzene Concentrations Measured at ATKY 12-74
12-48 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ATKY 12-75
12-49 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ATKY 12-76
12-50 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ATKY 12-77
12-51 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at ATKY 12-78
12-52 Yearly Statistical Metrics for 1,1,2-Trichloroethane Concentrations Measured at
ATKY 12-79
12-53 Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at ATKY.... 12-80
12-54 Yearly Statistical Metrics for Benzene Concentrations Measured at BLKY 12-81
12-55 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BLKY 12-82
12-56 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BLKY 12-83
12-57 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BLKY 12-84
12-58 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at BLKY 12-85
12-59 Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at BLKY.... 12-86
12-60 Yearly Statistical Metrics for Benzene Concentrations Measured at TVKY 12-87
12-61 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TVKY 12-88
12-62 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TVKY 12-89
12-63 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TVKY 12-90
12-64 Yearly Statistical Metrics for 1,1,2-Trichloroethane Concentrations Measured at
TVKY 12-91
12-65 Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at TVKY.... 12-92
12-66 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at LEKY.... 12-93
12-67 Yearly Statistical Metrics for Benzene Concentrations Measured at LEKY 12-94
12-68 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at LEKY 12-95
12-69 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
LEKY 12-96
12-70 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
LEKY 12-97
12-71 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
LEKY 12-98
12-72 Pollution Rose for 1,2-Dichloroethane Concentrations Measured at TVKY 12-107
13-1 Boston, Massachusetts (BOMA) Monitoring Site 13-2
13-2 NEI Point Sources Located Within 10 Miles of BOMA 13-3
xix
-------
LIST OF FIGURES (Continued)
Page
13-3 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 13-10
13-4 Program vs. Site-Specific Average Cadmium (PMio) Concentrations 13-11
13-5 Program vs. Site-Specific Average Naphthalene Concentrations 13-11
13-6 Program vs. Site-Specific Average Nickel (PMio) Concentrations 13-12
13-7 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BOMA... 13-13
13-8 Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
BOMA 13-15
13-9 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BOMA 13-16
13-10 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BOMA.... 13-17
14-1 Dearborn, Michigan (DEMI) Monitoring Site 14-2
14-2 NEI Point Sources Located Within 10 Miles of DEMI 14-3
14-3 Program vs. Site-Specific Average Acenaphthene Concentrations 14-12
14-4 Program vs. Site-Specific Average Acetaldehyde Concentrations 14-13
14-5 Program vs. Site-Specific Average Benzene Concentrations 14-13
14-6 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 14-14
14-7 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 14-14
14-8 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 14-15
14-9 Program vs. Site-Specific Average Ethylbenzene Concentrations 14-16
14-10 Program vs. Site-Specific Average Fluorene Concentrations 14-16
14-11 Program vs. Site-Specific Average Formaldehyde Concentrations 14-17
14-12 Program vs. Site-Specific Average Naphthalene Concentrations 14-17
14-13 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at DEMI 14-18
14-14 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at DEMI 14-20
14-15 Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI 14-21
14-16 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at DEMI 14-22
14-17 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
DEMI 14-24
14-18 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
DEMI 14-25
14-19 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at DEMI 14-26
14-20 Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI 14-28
14-21 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at DEMI 14-29
14-22 Yearly Statistical Metrics for Naphthalene Concentrations Measured at DEMI 14-31
15-1 St. Louis, Missouri (S4MO) Monitoring Site 15-2
15-2 NEI Point Sources Located Within 10 Miles of S4MO 15-3
15-3 Program vs. Site-Specific Average Acenaphthene Concentrations 15-13
15-4 Program vs. Site-Specific Average Acetaldehyde Concentrations 15-13
15-5 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 15-14
15-6 Program vs. Site-Specific Average Benzene Concentrations 15-15
15-7 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 15-15
15-8 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 15-16
15-9 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations 15-16
15-10 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 15-17
15-11 Program vs. Site-Specific Average Ethylbenzene Concentrations 15-18
xx
-------
LIST OF FIGURES (Continued)
Page
15-12 Program vs. Site-Specific Average Fluorene Concentrations 15-18
15-13 Program vs. Site-Specific Average Formaldehyde Concentrations 15-19
15-14 Program vs. Site-Specific Average Naphthalene Concentrations 15-20
15-15 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at S4MO 15-21
15-16 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at S4MO 15-22
15-17 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at S4MO.... 15-23
15-18 Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO 15-24
15-19 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at S4MO 15-26
15-20 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
S4MO 15-27
15-21 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
S4MO 15-28
15-22 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
S4MO 15-30
15-23 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at S4MO 15-31
15-24 Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO 15-32
15-25 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at S4MO 15-33
15-26 Yearly Statistical Metrics for Naphthalene Concentrations Measured at S4MO 15-34
16-1 Camden, New Jersey (CSNJ) Monitoring Site 16-2
16-2 NEI Point Sources Located Within 10 Miles of CSNJ 16-3
16-3 Chester, New Jersey (CHNJ) Monitoring Site 16-4
16-4 NEI Point Sources Located Within 10 Miles of CHNJ 16-5
16-5 Elizabeth, New Jersey (ELNJ) Monitoring Site 16-6
16-6 North Brunswick, New Jersey (NBNJ) Monitoring Site 16-7
16-7 East Brunswick, New Jersey (NRNJ) Monitoring Site 16-8
16-8 NEI Point Sources Located Within 10 Miles of ELNJ, NBNJ, and NRNJ 16-9
16-9 Program vs. Site-Specific Average Acetaldehyde Concentrations 16-26
16-10 Program vs. Site-Specific Average Benzene Concentrations 16-27
16-11 Program vs. Site-Specific Average Bromomethane Concentrations 16-28
16-12 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 16-29
16-13 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 16-30
16-14 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations 16-31
16-15 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 16-32
16-16 Program vs. Site-Specific Average Ethylbenzene Concentrations 16-33
16-17 Program vs. Site-Specific Average Formaldehyde Concentrations 16-34
16-18 Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations 16-35
16-19 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at CHNJ 16-36
16-20 Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ 16-37
16-21 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at CHNJ 16-39
16-22 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
CHNJ 16-40
16-23 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
CHNJ 16-41
16-24 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at CHNJ 16-42
xxi
-------
LIST OF FIGURES (Continued)
Page
16-25 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at ELNJ 16-44
16-26 Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ 16-45
16-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at ELNJ 16-47
16-28 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
ELNJ 16-48
16-29 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
ELNJ 16-49
16-30 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at ELNJ 16-50
16-31 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at ELNJ 16-51
16-32 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at NBNJ 16-52
16-33 Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ 16-54
16-34 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at NBNJ 16-55
16-35 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
NBNJ 16-57
16-36 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
NBNJ 16-58
16-37 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at NBNJ 16-59
17-1 Bronx, New York (BXNY) Monitoring Site 17-2
17-2 NEI Point Sources Located Within 10 Miles of BXNY 17-3
17-3 Rochester, New York (ROCH) Monitoring Site 17-4
17-4 NEI Point Sources Located Within 10 Miles of ROCH 17-5
17-5 Program vs. Site-Specific Average Acenaphthene Concentrations 17-14
17-6 Program vs. Site-Specific Average Benzo(a)pyrene Concentrations 17-15
17-7 Program vs. Site-Specific Average Fluoranthene Concentrations 17-16
17-8 Program vs. Site-Specific Average Fluorene Concentrations 17-17
17-9 Program vs. Site-Specific Average Naphthalene Concentrations 17-18
17-10 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at BXNY 17-20
17-11 Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at BXNY.. 17-21
17-12 Yearly Statistical Metrics for Fluorene Concentrations Measured at BXNY 17-22
17-13 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BXNY 17-23
17-14 Yearly Statistical Metrics for Acenaphthene Concentrations Measured at ROCH 17-24
17-15 Yearly Statistical Metrics for Fluoranthene Concentrations Measured at ROCH 17-25
17-16 Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH 17-26
17-17 Yearly Statistical Metrics for Naphthalene Concentrations Measured at ROCH 17-27
18-1 Public Works, Tulsa, Oklahoma (TOOK) Monitoring Site 18-2
18-2 Fire Station, Tulsa, Oklahoma (TMOK) Monitoring Site 18-3
18-3 Riverside, Tulsa, Oklahoma (TROK) Monitoring Site 18-4
18-4 NEI Point Sources Located Within 10 Miles of TMOK, TOOK, and TROK 18-5
18-5 Oklahoma City, Oklahoma (OCOK) Monitoring Site 18-6
18-6 Near-road, Oklahoma City, Oklahoma (NROK) Monitoring Site 18-7
18-7 Yukon, Oklahoma (YUOK) Monitoring Site 18-8
18-8 NEI Point Sources Located Within 10 Miles of NROK, OCOK, and YUOK 18-9
18-9 Bradley, Oklahoma (BROK) Monitoring Site 18-10
18-10 NEI Point Sources Located Within 10 Miles of BROK 18-11
XXll
-------
LIST OF FIGURES (Continued)
Page
18-11 Program vs. Site-Specific Average Acetaldehyde Concentrations 18-35
18-12 Program vs. Site-Specific Average Arsenic (TSP) Concentrations 18-37
18-13 Program vs. Site-Specific Average Benzene Concentrations 18-38
18-14 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 18-40
18-15 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 18-42
18-16 Program vs. Site-Specific Average/?-Dichlorobenzene Concentrations 18-43
18-17 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 18-45
18-18 Program vs. Site-Specific Average Ethylbenzene Concentrations 18-46
18-19 Program vs. Site-Specific Average Formaldehyde Concentrations 18-48
18-20 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 18-49
18-21 Program vs. Site-Specific Average Manganese (TSP) Concentrations 18-50
18-22 Program vs. Site-Specific Average Nickel (TSP) Concentrations 18-51
18-23 Program vs. Site-Specific Average Propionaldehyde Concentrations 18-52
18-24 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TOOK 18-53
18-25 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TOOK 18-54
18-26 Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK 18-55
18-27 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TOOK 18-56
18-28 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TOOK 18-57
18-29 Yearly Statistical Metrics for /;-Dichlorobenzene Concentrations Measured at
TOOK 18-58
18-30 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TOOK 18-59
18-31 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TOOK 18-60
18-32 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TOOK 18-62
18-33 Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK 18-63
18-34 Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at TOOK 18-64
18-35 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at TMOK 18-65
18-36 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at TMOK .... 18-66
18-37 Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK 18-67
18-38 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at TMOK 18-68
18-39 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
TMOK 18-69
18-40 Yearly Statistical Metrics for />Dichlorobenzene Concentrations Measured at
TMOK 18-70
18-41 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
TMOK 18-71
18-42 Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at TMOK 18-72
18-43 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at TMOK.... 18-73
18-44 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at TMOK 18-74
18-45 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at OCOK 18-75
18-46 Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at OCOK 18-76
18-47 Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK 18-77
18-48 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at OCOK 18-78
xxiii
-------
LIST OF FIGURES (Continued)
Page
18-49 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
OCOK 18-79
18-50 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
OCOK 18-80
18-51 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at OCOK 18-81
18-52 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at OCOK 18-82
19-1 Providence, Rhode Island (PRRI) Monitoring Site 19-2
19-2 NEI Point Sources Located Within 10 Miles of PRRI 19-3
19-3 Program vs. Site-Specific Average Fluorene Concentrations 19-10
19-4 Program vs. Site-Specific Average Naphthalene Concentrations 19-11
19-5 Yearly Statistical Metrics for Fluorene Concentrations Measured at PRRI 19-12
19-6 Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI 19-13
20-1 Bountiful, Utah (BTUT) Monitoring Site 20-2
20-2 NEI Point Sources Located Within 10 Miles of BTUT 20-3
20-3 Program vs. Site-Specific Average Acetaldehyde Concentrations 20-12
20-4 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 20-13
20-5 Program vs. Site-Specific Average Benzene Concentrations 20-13
20-6 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 20-14
20-7 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 20-15
20-8 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 20-15
20-9 Program vs. Site-Specific Average Formaldehyde Concentrations 20-16
20-10 Program vs. Site-Specific Average Hexachloro-1,3-butadiene Concentrations 20-17
20-11 Program vs. Site-Specific Average Naphthalene Concentrations 20-18
20-12 Program vs. Site-Specific Average Nickel (PMio) Concentrations 20-18
20-13 Program vs. Site-Specific Average Propionaldehyde Concentrations 20-19
20-14 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at BTUT 20-20
20-15 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at BTUT.... 20-21
20-16 Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT 20-22
20-17 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at BTUT 20-24
20-18 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
BTUT 20-25
20-19 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
BTUT 20-26
20-20 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at BTUT 20-28
20-21 Yearly Statistical Metrics for Hexachloro-1,3-butadiene Concentrations Measured
at BTUT 20-29
20-22 Yearly Statistical Metrics for Naphthalene Concentrations Measured at BTUT 20-30
20-23 Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at BTUT 20-32
20-24 Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at BTUT.. 20-33
20-25 Pollution Rose for Formaldehyde Concentrations Measured at BTUT 20-37
21-1 Underhill, Vermont (UNVT) Monitoring Site 21-2
21-2 NEI Point Sources Located Within 10 Miles of UNVT 21-3
xxiv
-------
LIST OF FIGURES (Continued)
Page
21-3 Program vs. Site-Specific Average Benzo(a)pyrene Concentrations 21-10
21-4 Program vs. Site-Specific Average Naphthalene Concentrations 21-11
21-5 Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at UNVT .. 21-12
21-6 Yearly Statistical Metrics for Naphthalene Concentrations Measured at UNVT 21-13
22-1 East Highland Park, Virginia (RIVA) Monitoring Site 22-2
22-2 NEI Point Sources Located Within 10 Miles of RIVA 22-3
22-3 Program vs. Site-Specific Average Naphthalene Concentrations 22-9
22-4 Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA 22-10
23-1 Seattle, Washington (SEWA) Monitoring Site 23-2
23-2 NEI Point Sources Located Within 10 Miles of SEWA 23-3
23-3 Program vs. Site-Specific Average Acetaldehyde Concentrations 23-11
23-4 Program vs. Site-Specific Average Arsenic (PMio) Concentrations 23-11
23-5 Program vs. Site-Specific Average Benzene Concentrations 23-12
23-6 Program vs. Site-Specific Average 1,3-Butadiene Concentrations 23-12
23-7 Program vs. Site-Specific Average Carbon Tetrachloride Concentrations 23-13
23-8 Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations 23-14
23-9 Program vs. Site-Specific Average Formaldehyde Concentrations 23-14
23-10 Program vs. Site-Specific Average Naphthalene Concentrations 23-15
23-11 Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at SEWA 23-16
23-12 Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SEWA 23-17
23-13 Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA 23-18
23-14 Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at SEWA 23-19
23-15 Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured at
SEWA 23-20
23-16 Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
SEWA 23-21
23-17 Yearly Statistical Metrics for Formaldehyde Concentrations Measured at SEWA 23-22
23-18 Yearly Statistical Metrics for Naphthalene Concentrations Measured at SEWA 23-23
xxv
-------
LIST OF TABLES
Page
1-1 Organization of the 2015-2016 National Monitoring Programs Report 1-4
2-1 2015-2016 National Monitoring Programs Sites and Past Program Participation 2-4
2-2 Site Characterizing Information for the 2015-2016 National Monitoring Programs
Sites 2-8
2-3 2015-2016 VOC Method Detection Limits 2-16
2-4 2015-2016 SNMOC Method Detection Limits 2-17
2-5 2015-2016 Methane Method Detection Limit 2-18
2-6 2015-2016 Carbonyl Compound Method Detection Limits 2-20
2-7 2015-2016 PAH Method Detection Limits 2-21
2-8 2015-2016 Metals Method Detection Limits 2-22
2-9 2015-2016 Hexavalent Chromium Method Detection Limits 2-23
2-10 2015 Sampling Schedules and Completeness Rates 2-25
2-11 2016 Sampling Schedules and Completeness Rates 2-29
3-1 Overview and Organization of Data Presented 3-1
3-2 NATTS MQO Core Analytes 3-7
3 -3 POM Groups for PA I Is 3-14
4-1 Statistical Summaries of the VOC Concentrations 4-3
4-2 Statistical Summaries of the SNMOC Concentrations 4-6
4-3 Statistical Summaries of the Methane Concentrations 4-10
4-4 Statistical Summaries of the Carbonyl Compound Concentrations 4-11
4-5 Statistical Summaries of the PAH Concentrations 4-12
4-6 Statistical Summaries of the Metals Concentrations 4-13
4-7 Statistical Summary of the Hexavalent Chromium Concentrations 4-14
4-8 Results of the Program-Level Preliminary Risk-Based Screening Process 4-18
4-9 Site-Specific Risk-Based Screening Comparison 4-20
4-10 Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of
Interest 4-23
4-11 Annual Average Concentration Comparison of the Carbonyl Compound Pollutants
of Interest 4-24
4-12 Annual Average Concentration Comparison of the PAH Pollutants of Interest 4-25
4-13 Annual Average Concentration Comparison of the Metals Pollutants of Interest 4-26
5-1 Geographical Information for the Arizona Monitoring Sites 5-5
5-2 2015-2016 Risk-Based Screening Results for the Arizona Monitoring Sites 5-7
5-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Arizona Monitoring Sites 5-10
5-4 Risk Approximations for the Arizona Monitoring Sites 5-42
5-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Arizona Monitoring Sites 5-46
5-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Arizona Monitoring
Sites 5-47
xxvi
-------
LIST OF TABLES (Continued)
Page
6-1 Geographical Information for the California Monitoring Sites 6-8
6-2 2015-2016 Risk-Based Screening Results for the California Monitoring Sites 6-11
6-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
California Monitoring Sites 6-14
6-4 Risk Approximations for the California Monitoring Sites 6-26
6-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the California Monitoring Sites 6-29
6-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the California Monitoring
Sites 6-31
7-1 Geographical Information for the Colorado Monitoring Sites 7-12
7-2 2015-2016 Risk-Based Screening Results for the Colorado Monitoring Sites 7-16
7-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Colorado Monitoring Sites 7-20
7-4 Risk Approximations for the Colorado Monitoring Sites 7-73
7-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Colorado Monitoring Sites 7-79
7-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Colorado Monitoring
Sites 7-83
8-1 Geographical Information for the Washington, D.C. Monitoring Site 8-4
8-2 2015-2016 Risk-Based Screening Results for the Washington, D.C. Monitoring Site... 8-6
8-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington, D.C. Monitoring Site 8-8
8-4 Risk Approximations for the Washington, D.C. Monitoring Site 8-12
8-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site 8-14
8-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington, D.C.
Monitoring Site 8-15
9-1 Geographical Information for the Florida Monitoring Sites 9-10
9-2 2015-2016 Risk-Based Screening Results for the Florida Monitoring Sites 9-14
9-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Florida Monitoring Sites 9-16
9-4 Risk Approximations for the Florida Monitoring Sites 9-41
9-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Florida Monitoring Sites 9-45
9-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Florida Monitoring
Sites 9-48
10-1 Geographical Information for the Illinois Monitoring Sites 10-7
10-2 2015-2016 Risk-Based Screening Results for the Illinois Monitoring Sites 10-10
xxvii
-------
LIST OF TABLES (Continued)
Page
10-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Illinois Monitoring Sites 10-14
10-4 Risk Approximations for the Illinois Monitoring Sites 10-59
10-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Illinois Monitoring Sites 10-63
10-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Illinois Monitoring
Sites 10-65
11-1 Geographical Information for the Indiana Monitoring Sites 11-6
11-2 2015-2016 Risk-Based Screening Results for the Indiana Monitoring Sites 11-8
11-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Indiana Monitoring Sites 11-11
11-4 Risk Approximations for the Indiana Monitoring Sites 11-21
11-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Indiana Monitoring Sites 11-23
11-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Indiana Monitoring
Sites 11-24
12-1 Geographical Information for the Kentucky Monitoring Sites 12-15
12-2 Overview of Pollutant Groups Sampled for at the Kentucky Monitoring Sites 12-19
12-3 2015-2016 Risk-Based Screening Results for the Kentucky Monitoring Sites 12-19
12-4 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Kentucky Monitoring Sites 12-26
12-5 Risk Approximations for the Kentucky Monitoring Sites 12-100
12-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Kentucky Monitoring Sites 12-109
12-7 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Kentucky Monitoring
Sites 12-113
13-1 Geographical Information for the Massachusetts Monitoring Site 13-4
13-2 2015-2016 Risk-Based Screening Results for the Massachusetts Monitoring Site 13-6
13-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Massachusetts Monitoring Site 13-8
13-4 Risk Approximations for the Massachusetts Monitoring Site 13-20
13-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Massachusetts Monitoring Site 13-22
13-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Massachusetts
Monitoring Site 13-23
14-1 Geographical Information for the Michigan Monitoring Site 14-4
14-2 2015-2016 Risk-Based Screening Results for the Michigan Monitoring Site 14-6
xxviii
-------
LIST OF TABLES (Continued)
Page
14-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Michigan Monitoring Site 14-9
14-4 Risk Approximations for the Michigan Monitoring Site 14-33
14-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Michigan Monitoring Site 14-35
14-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Michigan Monitoring
Site 14-36
15-1 Geographical Information for the Missouri Monitoring Site 15-4
15-2 2015-2016 Risk-Based Screening Results for the Missouri Monitoring Site 15-6
15-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Missouri Monitoring Site 15-9
15-4 Risk Approximations for the Missouri Monitoring Site 15-36
15-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Missouri Monitoring Site 15-38
15-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Missouri Monitoring
Site 15-39
16-1 Geographical Information for the New Jersey Monitoring Sites 16-10
16-2 2015-2016 Risk-Based Screening Results for the New Jersey Monitoring Sites 16-14
16-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New Jersey Monitoring Sites 16-18
16-4 Risk Approximations for the New Jersey Monitoring Sites 16-62
16-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New Jersey Monitoring Sites 16-67
16-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New Jersey Monitoring
Sites 16-70
17-1 Geographical Information for the New York Monitoring Sites 17-6
17-2 2015-2016 Risk-Based Screening Results for the New York Monitoring Sites 17-9
17-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
New York Monitoring Sites 17-11
17-4 Risk Approximations for the New York Monitoring Sites 17-30
17-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the New York Monitoring Sites 17-32
17-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the New York Monitoring
Sites 17-33
18-1 Geographical Information for the Oklahoma Monitoring Sites 18-12
18-2 2015-2016 Risk-Based Screening Results for the Oklahoma Monitoring Sites 18-16
18-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Oklahoma Monitoring Sites 18-23
XXIX
-------
LIST OF TABLES (Continued)
Page
18-4 Risk Approximations for the Oklahoma Monitoring Sites 18-84
18-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Oklahoma Monitoring Sites 18-92
18-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Oklahoma Monitoring
Sites 18-96
19-1 Geographical Information for the Rhode Island Monitoring Site 19-4
19-2 2015-2016 Risk-Based Screening Results for the Rhode Island Monitoring Site 19-6
19-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Rhode Island Monitoring Site 19-8
19-4 Risk Approximations for the Rhode Island Monitoring Site 19-15
19-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Rhode Island Monitoring Site 19-17
19-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Rhode Island
Monitoring Site 19-18
20-1 Geographical Information for the Utah Monitoring Site 20-4
20-2 2015-2016 Risk-Based Screening Results for the Utah Monitoring Site 20-6
20-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Utah Monitoring Site 20-9
20-4 Risk Approximations for the Utah Monitoring Site 20-35
20-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Utah Monitoring Site 20-39
20-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Utah Monitoring
Site 20-40
21-1 Geographical Information for the Vermont Monitoring Site 21-4
21-2 2015-2016 Risk-Based Screening Results for the Vermont Monitoring Site 21-6
21-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Vermont Monitoring Site 21-8
21-4 Risk Approximations for the Vermont Monitoring Site 21-15
21-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Vermont Monitoring Site 21-17
21-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Vermont
Monitoring Site 21-18
22-1 Geographical Information for the Virginia Monitoring Site 22-4
22-2 2015-2016 Risk-Based Screening Results for the Virginia Monitoring Site 22-6
22-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Virginia Monitoring Site 22-8
22-4 Risk Approximations for the Virginia Monitoring Site 22-12
XXX
-------
LIST OF TABLES (Continued)
Page
22-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Virginia Monitoring Site 22-14
22-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Virginia Monitoring
Site 22-15
23-1 Geographical Information for the Washington Monitoring Site 23-4
23-2 2015-2016 Risk-Based Screening Results for the Washington Monitoring Site 23-6
23-3 Quarterly and Annual Average Concentrations of the Pollutants of Interest for the
Washington Monitoring Site 23-8
23-4 Risk Approximations for the Washington Monitoring Site 23-25
23-5 Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations
for Pollutants with Cancer UREs for the Washington Monitoring Site 23-27
23-6 Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard
Approximations for Pollutants with Noncancer RfCs for the Washington
Monitoring Site 23-28
24-1 Method Precision by Analytical Method 24-4
24-2 VOC Method Precision: Coefficient of Variation Based on Duplicate and
Collocated Samples by Site and Pollutant 24-6
24-3 SNMOC Method Precision: Coefficient of Variation Based on Collocated and
Duplicate Samples by Site and Pollutant 24-15
24-4 Methane Method Precision: Coefficient of Variation Based on Duplicate Samples
by Site and Pollutant 24-21
24-5 Carbonyl Compound Method Precision: Coefficient of Variation Based on
Duplicate and Collocated Samples by Site and Pollutant 24-22
24-6 PAH Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 24-26
24-7 Metals Method Precision: Coefficient of Variation Based on Collocated Samples
by Site and Pollutant 24-27
24-8 Hexavalent Chromium Method Precision: Coefficient of Variation Based on
Collocated Samples by Site 24-28
24-9 Analytical Precision by Analytical Method 24-30
24-10 VOC Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-32
24-11 SNMOC Analytical Precision: Coefficient of Variation Based on Replicate
Analyses by Site and Pollutant 24-41
24-12 Methane Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-47
24-13 Carbonyl Compound Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site and Pollutant 24-48
24-14 PAH Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-54
24-15 Metals Analytical Precision: Coefficient of Variation Based on Replicate Analyses
by Site and Pollutant 24-58
XXXI
-------
LIST OF TABLES (Continued)
Page
24-16 Hexavalent Chromium Analytical Precision: Coefficient of Variation Based on
Replicate Analyses by Site 24-61
24-17 TO-15 NATTS PT Audit Samples 24-62
24-18 I O-l 1A NATTS PT Audit Samples 24-62
24-19 T0-13A NATTS PT Audit Samples 24-63
24-20 Metals NATTS PT Audit Samples 24-63
24-21 2015 Lead NAAQS Quarterly Audit Samples 24-64
24-22 2016 Lead NAAQS Quarterly Audit Samples 24-64
25-1 Summary of Site-Specific Pollutants of Interest 25-19
XXXll
-------
LIST OF ACRONYMS
AADT Annual Average Daily Traffic
AQS Air Quality System
ASE Accelerated Solvent Extractor
CBSA Core-Based Statistical Area(s)
CFR Code of Federal Regulations
CNG Compressed Natural Gas
COC Chain of Custody
CSATAM Community-Scale Air Toxics Ambient Monitoring
CV Coefficient of Variation
DNPH 2,4-Dinitrophenylhydrazine
DQI Data Quality Indicator(s)
DQO Data Quality Objective(s)
EPA U.S. Environmental Protection Agency
ERG Eastern Research Group, Inc.
FAC Federal Advisory Committee
FEM Federal Equivalent Method
GC/MS-FID Gas Chromatography/Mass Spectrometry and Flame Ionization Detection
GC-FID Gas Chromatography incorporating Flame Ionization Detection
HAP Hazardous Air Pollutant(s)
HPLC High-Performance Liquid Chromatography
HQ Hazard Quotient
IC Ion Chromatography
ICP-MS Inductively Coupled Plasma/Mass Spectrometry
MDL Method Detection Limit
mg/m3 Milligrams per cubic meter
mL Milliliter
MQO Measurement Quality Objective(s)
NAAQS National Ambient Air Quality Standard
NATA National-Scale Air Toxics Assessment
NATTS National Air Toxics Trends Stations
NCEI National Centers for Environmental Information
ND Non-detect
NEI National Emissions Inventory
ng/m3 Nanograms per cubic meter
NMOC Non-Methane Organic Compound(s)
NMP National Monitoring Programs
NOAA National Oceanic and Atmospheric Administration
NWS National Weather Service
PAH Polycyclic Aromatic Hydrocarbon(s)
PM Particulate Matter
PMio Particulate Matter less than 10 microns
POM Polycyclic Organic Matter
ppbC Parts per billion carbon
xxxiii
-------
LIST OF ACRONYMS (Continued)
ppbv
Parts per billion by volume
ppmC
Parts per million carbon
PT
Proficiency Test
PUF
Polyurethane Foam
QAPP
Quality Assurance Project Plan
RfC
Reference Concentration(s)
SATMP
School Air Toxics Monitoring Program
SIM
Selected Ion Monitoring
SNMOC
Speciated Nonmethane Organic Compound(s)
TAD
Technical Assistance Document
TNMOC
Total Nonmethane Organic Compound(s)
tpy
Tons per year
TSP
Total Suspended Particulate
UATMP
Urban Air Toxics Monitoring Program
Hg/m3
Micrograms per cubic meter
HL
Microliter
URE
Unit Risk Estimate(s)
UY
Ultraviolet
VOC
Volatile Organic Compound(s)
xxxiv
-------
Abstract
This report presents the results and conclusions from the ambient air monitoring conducted
as part of the 2015 and 2016 National Monitoring Programs (NATTS, UATMP, and CSATAM) -
three individual programs with different goals, but together result in a better understanding and
appreciation of the nature and extent of toxic air pollution. The 2015-2016 NMP includes data
from samples collected at 53 monitoring sites that collected 24-hour air samples, typically on a
l-in-6 or l-in-12 day sampling schedule, and analyzed by the national contract laboratory.
Twenty-seven sites sampled for 59 volatile organic compounds (VOCs); 31 sites sampled for 15
carbonyl compounds; nine sites sampled for 80 speciated nonmethane organic compounds
(SNMOCs); 19 sites sampled for 22 polycyclic aromatic hydrocarbons (PAHs); 19 sites sampled
for 11 metals; 2 sites samples for methane; and 1 site sampled for hexavalent chromium. More
than 445,000 ambient air concentrations were measured during the 2015-2016 NMP under the
national contract. This report uses various graphical, numerical, and statistical analyses to put the
vast amount of ambient air monitoring data collected into perspective. Not surprisingly, the
ambient air concentrations measured during the program varied from city-to-city, season-to-
season, and year-to-year.
The ambient air monitoring data collected during the 2015 and 2016 NMP under the
national contract serve a wide range of purposes. Not only do these data allow for the
characterization of the nature and extent of air pollution close to the 53 individual monitoring
sites participating in these programs, but they also exhibit trends and patterns that may be
common to urban and rural environments and across the country. Therefore, this report presents
results that are specific to particular monitoring locations and presents other results that are
common to all environments. The results presented provide additional insight into the complex
nature of air pollution. The raw data are included in the appendices of this report.
xxxv
-------
1.0 Introduction
Air pollution contains many components that originate from a wide range of stationary,
mobile, and natural emissions sources. Because some of these components include air toxics that
are known or suspected to have the potential for negative human health effects, the
U.S. Environmental Protection Agency (EPA) encourages state, local, and tribal agencies to
understand and appreciate the nature and extent of toxic air pollution in their respective locations
or areas of administration. To achieve this goal, EPA sponsors the National Monitoring
Programs (NMP), which includes the Urban Air Toxics Monitoring Program (UATMP),
National Air Toxics Trends Stations (NATTS) network, Community-Scale Air Toxics Ambient
Monitoring (CSATAM) Program, and monitoring for other pollutants such as speciated Non-
Methane Organic Compounds (NMOCs). The UATMP, the NATTS, and the CSATAM
programs include longer-term monitoring efforts (durations of one year or more) at specific
locations. These programs have the following program-specific objectives (EPA, 2014):
• The primary technical objective of the UATMP is to characterize the composition and
magnitude of air toxics pollution through ambient air monitoring.
http://www.epa.gov/ttnamtil/uatm.html
• The primary technical objective of the NATTS network is to obtain a statistically
significant quantity of high-quality representative air toxics measurements such that
long-term trends can be identified, http://www.epa.gov/ttnamtil/natts.html
• The primary technical objective of the CSATAM Program is to conduct local-scale
investigative ambient air toxics monitoring projects.
http://www.epa.gov/ttnamtil/local.html
1.1 Background
The UATMP was initiated by EPA to meet the increasing need for information on air
toxics. Over the years, the program has grown in both participation and targeted pollutants (EPA,
2014). The program has allowed for the identification of compounds that are prevalent in
ambient air and for participating agencies to screen air samples for concentrations of air toxics
that could potentially result in adverse human health effects.
The NATTS network was created to generate long-term ambient air toxics concentration
data at specific fixed sites across the country. The 10-City Pilot Program (LADCO, 2003) was
developed and implemented during 2001 and 2002, leading to the development and initial
implementation of the NATTS network during 2003 and 2004. The goal of the program is to
estimate the concentrations of air toxics on a national level from fixed sites that remain active
over an extended period of time such that concentration trends (i.e., any substantial increase or
1-1
-------
decrease over a period of time) may be identified (EPA, 2014). The data generated are also used
for validating modeling results and emissions inventories, assessing current regulatory
benchmarks, and assessing the potential for developing cancerous and noncancerous health
effects (EPA, 2017a). The initial site locations were based on existing infrastructure of
monitoring site locations (e.g., PM2.5 network) and results from preliminary air toxics programs
such as the 1996 National-scale Air Toxics Assessment (NATA), which used air toxics
emissions data to model ambient monitoring concentrations across the nation. Monitoring sites
were placed in both urban and rural locations. Urban areas were chosen to measure population
exposure, while rural areas were chosen to determine background levels of air pollution and to
assess impacts to non-urban areas (EPA, 2009a). Currently, 27 NATTS sites are strategically
placed across the country (EPA, 2017a).
The CSATAM Program began in 2004 and is intended to support state, local, and tribal
agencies in conducting ambient monitoring projects of approximately 2-year durations via
periodic grant competitions. The objectives of the CSATAM Program include identifying and
profiling sources of air toxics; developing and evaluating emerging measurement methods;
characterizing the degree and extent of local air toxics problems; and tracking progress
attributable to air toxics reduction activities (EPA, 2014).
1.2 The Report
Many environmental and health agencies have participated in these programs to assess
the sources, effects, and changes in air pollution within their jurisdictions. This report
summarizes and interprets measurements collected at monitoring sites participating in the
UATMP, NATTS, and CSATAM programs in 2015 and 2016. Included in this report are data
from sites whose operating agencies have opted to have their samples analyzed by EPA's
national contract laboratory, Eastern Research Group, Inc. (ERG). Agencies operating sites
under the NMP are not required to have their samples analyzed by ERG or may not have samples
for all methods analyzed by ERG, as they may have their own laboratories or use other
laboratories. In these cases, data are generated by sources other than ERG and are not included in
this report. In addition, a state, local, or tribal agency may opt to use the national contract for a
specialized, more targeted air toxics monitoring study in which their data are included in the
report as well. The purpose of this report is to summarize and characterize those data generated
by the contract laboratory over the 2015 and 2016 monitoring efforts.
1-2
-------
In past reports, measurements from UATMP, NATTS, and CSATAM monitoring sites
have been presented together and referred to as "UATMP sites." In more recent reports,
including the 2015-2016 report, a distinction has been made among the three programs due to the
increasing number of sites covered under each program. Thus, it is appropriate to describe each
program; to distinguish among their purposes and scopes; and to integrate the data, which allows
each program's objectives and goals to complement one another.
Included in this report are data collected at 53 monitoring
sites around the country. The 53 sites whose data are included in
this report are located in or near 30 urban or rural locations in 18
states and the District of Columbia, including 28 metropolitan or
micropolitan statistical areas (collectively referred to as core-based
statistical areas or CBSAs).
This report provides both a qualitative overview of air toxics pollution at participating
urban and rural locations and a quantitative data analysis of the factors that appear to most
significantly affect the behavior of air toxics in urban and rural areas. This report also focuses on
data summaries and characterizations for each of the 53 different air monitoring locations, a site-
specific approach that allows for a much more detailed evaluation of the factors (e.g., emissions
sources, natural sources, meteorological influences) that affect air quality differently from one
location to the next. Much of the data analysis and interpretation contained in this report focuses
on pollutant-specific risk potential.
This report offers participating agencies relevant information and insight into important
air quality issues. For example, participating agencies can use trends and patterns in the
monitoring data to determine whether levels of air pollution as reported present public health
concerns, to identify which emissions sources contribute most to air pollution, and/or to forecast
whether proposed pollution control initiatives could (or have) significantly improved air quality.
Monitoring data may also be compared to modeling results, such as from EPA's NATA. Policy-
relevant questions that the monitoring data may help answer include the following:
• Which anthropogenic sources substantially affect air quality?
• Have pollutant concentrations decreased as a result of regulations (or increased
despite regulation)?
• Which pollutants contribute the greatest health risk on a short-term, intermediate-
term, and long-term basis?
1-3
The pu rpuso nf ill is report is
In suniiiKin/.c ciikI
cIkiiiicIci i/.c lliosc dulu
jjciK'mk'il h> llic conimcl
kilnmlin'\ dui'iml! iIk- 2d 15
iind 2<>|(i monilonntj cl'lorls
-------
The data analyses contained in this report are applied to each participating UATMP,
NATTS, or C SAT AM monitoring site, depending upon pollutants sampled and duration of
monitoring. Although many types of data analyses are presented, state and local environmental
agencies are encouraged to perform additional evaluations of the monitoring data so that the
many factors that affect their specific ambient air quality can be understood fully.
To facilitate examination of the 2015-2016 UATMP, NATTS, and CSATAM monitoring
data, henceforth referred to as NMP data, the complete set of measurements is presented in the
appendices of this report. In addition, these data are publicly available in electronic format from
EPA's Air Quality System (AQS) (EPA, 2017b).
The 2015-2016 report is organized into 26 sections and 17 appendices. While each state
section is designed to be a stand-alone section to allow those interested in a particular site or
state to understand the associated data analyses without having to read the entire report, it is
recommended that Sections 1 through 4 (Introduction, Monitoring Programs Network, Data
Treatments and Methods, and Summary of NMP Data) and Sections 24 and 25 (Data Quality and
Results, Conclusions, and Recommendations) be read as complements to the individual state
sections. Table 1-1 highlights the contents of each section.
Table 1-1. Organization of the 2015-2016 National Monitoring Programs Report
Report
Section
Section Title
Overview of Contents
1
Introduction
This section serves as an introduction to the background, objectives,
and scope of specific element of EPA's NMP (specifically, the
UATMP, NATTS, and CSATAM Programs).
2
The 2015-2016 National
Monitoring Programs Network
This section provides an overview on the 2015-2016 NMP
monitoring effort, including:
• Monitoring locations
• Pollutants selected for monitoring
• Sampling and analytical methods
• Sampling schedules
• Completeness of the air monitoring programs.
3
Summary of the 2015-2016
National Monitoring Programs Data
Treatments and Methods
This section presents and discusses the data treatments applied to the
2015-2016 NMP data to determine significant trends and
relationships in the data, characterize data based on how ambient air
concentrations varied with monitoring location and with time,
interpret the significance of the observed spatial and temporal
variations, and evaluate human health risk.
4
Summary of the 2015-2016
National Monitoring Programs Data
This section presents and discusses the results of the data analyses
from the 2015-2016 NMP data.
5
Sites in Arizona
Monitoring results for the sites in the Phoenix-Mesa-Scottsdale, AZ
CBS A (PXSS and SPAZ)
6
Sites in California
Monitoring results for the sites in the Los Angeles-Long Beach-
Anaheim, CA CBSA (CELA), the Riverside-San Bernardino-
Ontario, CA CBSA (RUCA), and the San Jose-Sunnyvale-Santa
Clara, CA CBSA (SJJCA)
1-4
-------
Table 1-1. Organization of the 2015-2016 National Monitoring Programs Report (Continued)
Report
Section
Section Title
Overview of Contents
7
Sites in Colorado
Monitoring results for the sites in the Grand Junction CO CBS A
(GPCO) and the Glenwood Springs, CO CBS A (BMCO, BRCO,
GSCO, PACO, RFCO, and RICO)
8
Site in the District of Columbia
Monitoring results for the site in the Washington-Arlington-
Alexandria, DC-VA-MD-WV CBS A (WADC)
9
Sites in Florida
Monitoring results for the sites in the Orlando-Kissimmee-Sanford,
FL CBSA (ORFL and PAFL) and the Tampa-St. Petersburg-
Clearwater, FL CBSA (AZFL, SKFL, and SYFL)
10
Sites in Illinois
Monitoring results for the sites in the Chicago-Naperville-Elgin IL-
IN-WI CBSA (NBIL and SPIL) and the St. Louis, MO-IL CBSA
(ROIL)
11
Sites in Indiana
Monitoring results for the sites in the Chicago- Naperville-Elgin, IL-
IN-WI CBSA (INDEM) and the Indianapolis-Carmel-Anderson, IN
CBSA (WPIN)
12
Sites in Kentucky
Monitoring results for the sites in the Huntington-Ashland, WV-KY-
OH CBSA (ASKY and ASKY-M), the Lexington-Fayette, KY
CBSA (LEKY), the Evansville, IN-KY CBSA (BAKY), the
Paducah KY-IL CBSA (BLKY), and the sites in Marshall County
(ATKY and TVKY) and Carter County (GLKY)
13
Site in Massachusetts
Monitoring results for the site in the Boston-Cambridge-Newton
MA-NH CBSA (BOMA)
14
Site in Michigan
Monitoring results for the site in the Detroit-Warren-Dearborn, MI
CBSA (DEMI)
15
Site in Missouri
Monitoring results for the site in the St. Louis, MO-IL CBSA
(S4MO)
16
Sites in New Jersey
Monitoring results for the sites in the New York-Newark-Jersey
City, NY-NJ-PA CBSA (CHNJ, ELNJ, NBNJ, and NRNJ) and the
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD CBSA (CSNJ)
17
Sites in New York
Monitoring results for the sites in the New York-Newark-Jersey
City, NY-NJ-PA CBSA (BXNY) and the Rochester, NY CBSA
(ROCH)
18
Sites in Oklahoma
Monitoring results for the sites in the Tulsa, OK CBSA (TOOK,
TMOK, and TROK), and the Oklahoma City, OK CBSA (BROK,
NROK, OCOK, and YUOK)
19
Site in Rhode Island
Monitoring results for the site in the Providence-Warwick, RI-MA
CBSA (PRRI)
20
Site in Utah
Monitoring results for the site in the Ogden-Clearfield, UT CBSA
(BTUT)
21
Site in Vermont
Monitoring results for the site in the Burlington-South Burlington
VT CBSA (UNVT)
22
Site in Virginia
Monitoring results for the site in the Richmond, VA CBSA (RIVA)
23
Site in Washington
Monitoring results for the site in the Seattle-Tacoma-Bellevue, WA
CBSA (SEWA)
24
Data Quality
This section defines and discusses the general concepts of precision
and accuracy. Based on quantitative and qualitative analyses, this
section comments on the specific precision and accuracy of the
2015-2016 NMP ambient air monitoring data.
25
Results, Conclusions, and
Recommendations
This section summarizes the most significant findings of the report
and makes several recommendations for future projects that involve
ambient air monitoring.
26
References
This section lists the references cited throughout the report.
1-5
-------
2.0 The 2015-2016 National Monitoring Programs Network
Agencies operating UATMP, NATTS, or C SAT AM sites may choose to have their
samples analyzed by EPA's contract laboratory, ERG, in Morrisville, North Carolina. This report
report summarizes and characterizes data generated for agencies that chose to have ERG provide
analytical services and monitoring support. Data included in this report are from 53 monitoring
sites that collected 24-hour integrated ambient air samples for up to 24 months, at l-in-6 or 1-in-
12 day sampling intervals, and sent them to ERG for analysis. Samples were analyzed for
concentrations of the following suites of pollutants:
• selected hydrocarbons, halogenated
hydrocarbons, and polar compounds from
canister samples for Speciated Non-Methane
Organic Compounds (SNMOCs) and/or
Volatile Organic Compounds (VOCs) using
EPA Compendium Method TO-15,
• carbonyl compounds from sorbent cartridge
samples using EPA Compendium Method
TO-11 A,
• polycyclic aromatic hydrocarbons (PAHs)
from polyurethane foam (PUF) and XAD-2® resin samples using EPA Compendium
Method TO-13 A,
• trace metals from filters using EPA Compendium Method IO-3.5/Federal Equivalency
Methods (FEM) EQL-0512-201 or EQL-0512-202, and
• hexavalent chromium from sodium bicarbonate-coated filters using ASTM D7614.
Two sites participating in the NMP during 2015 and 2016 also submitted their canister
samples for methane analysis as part of a special study. While not an official part of the NMP,
additional information regarding this sampling methodology, along with the other methods listed
above, is provided in Section 2.2.
The following sections review the monitoring locations, pollutants selected for
monitoring, sampling and analytical methods, collection schedules, and completeness of the
2015-2016 NMP dataset.
2.1 Monitoring Locations
For the NATTS network, monitor siting is based on the need to assess population
exposure and/or background-level concentrations. For the UATMP and C SAT AM programs,
representatives from the state, local, and tribal agencies that voluntarily participate in the
Agencies operating sites under the
NMP are not required to have their
samples analyzed by ERG. They may
have samples for only select methods
analyzed by ERG, as they may have
their own laboratory capabilities for
other methods. In these cases, data are
generated by sources other than ERG
and are therefore not included in this
report.
V
2-1
-------
programs select the monitoring locations based on specific siting criteria and study needs.
Among these programs, monitors were placed in urban areas near the centers of heavily
populated cities (e.g., Chicago, Illinois and Phoenix, Arizona), while others were placed in
moderately or sparsely populated rural areas (e.g., Grayson, Kentucky and Underhill, Vermont).
Figure 2-1 shows the locations of the 53 monitoring sites participating in the 2015 and
2016 monitoring programs under the national contract, which encompass 30 different urban and
rural areas. Outlined in Figure 2-1 are the associated CBSAs, as designated by the U.S. Census
Bureau, where each site is located (Census Bureau, 2015). A CBSA refers to either a
metropolitan (an urban area with 50,000 or more people) or micropolitan (an urban area with at
least 10,000 people but less than 50,000 people) statistical area (Census Bureau, 2017). Table 2-1
lists the respective monitoring program and the years of program participation under the national
contract for the 53 monitoring sites. Most of the monitoring sites have been included in at least
one previous NMP annual report; only BROK, GSCO, NRNJ, and NROK did not participate in
the NMP prior to 2015.
As Figure 2-1 and Table 2-1 show, the 2015-2016 NMP sites are widely distributed
across the country. Detailed information about the monitoring sites is provided in Table 2-2,
Appendix A, and the individual state sections (Sections 5 through 23). Monitoring sites that are
designated as part of the NATTS network are indicated by bold italic type in Table 2-1 and
subsequent tables throughout this report. Table 2-2 shows that the locations of the monitoring
sites vary significantly, depending on the individual program's technical objectives. These sites
are located in areas of differing elevation, population, land use, climatology, and topography. A
more detailed look at each monitoring site's surroundings is provided in the individual state
sections.
For record-keeping and reporting purposes, each site was assigned the following:
• A unique four or five-letter site code used to track samples from the monitoring site to
the ERG laboratory.
• A unique nine-digit AQS site code used to index monitoring results in the AQS
database.
This report cites the four or five-letter site code when presenting monitoring results. For
reference, each site's AQS site code is provided in Table 2-2.
2-2
-------
Figure 2-1. Locations of the 2015-2016 National Monitoring Programs Monitoring Sites1
J }J3r3C r!'.ltO L-.
•i'vt-'i: CO:
,70C C; i:l: ::i;;:DwL.U
Ca'l)0:i;iai-3. CO
io:l7CO
'• J • OI • Cs iJ
>\V
1 Includes monitoring sites participating under the NMP with the national contract laboratory.
-------
Table 2-1. 2015-2016 National Monitoring Programs Sites and Past Program Participation1
Monitoring Location
(and Site Name)
Program
2006 and Earlier
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Ashland. KY (ASKY)
UATMP
~
~
~
~
~
Ashland, KY (ASKY-M)
UATMP
~
~
~
~
~
Baskett KY (BAKY)
UATMP
~
~
~
~
~
Battlement Mesa, CO (BMCO)
UATMP
~
~
~
~
~
~
~
Boston MA (BOMA)
NATTS
2003-2006
~
~
~
~
~
~
~
~
~
~
Bountiful, UT (BTUT)
NATTS
2003-2006
~
~
~
~
~
~
~
~
~
~
Bradley, OK (BROK)
UATMP
~
~
Calvert City, KY (ATKY)
UATMP
~
~
~
~
~
Calvert City, KY (TVKY)
UATMP
~
~
~
~
~
Camden, NJ (CSNJ)
UATMP
~
~
~
~
Carbondale, CO (RFCO)
UATMP
~
~
~
~
~
Chester, NJ (CHNJ)
UATMP
2001-2006
~
~
~
~
~
~
~
~
~
~
Dearborn, MI (DEMI)
NATTS
2001-2006
~
~
~
~
~
~
~
~
~
~
East Brunswick, NJ (NRNJ)
UATMP
~
East Highland Park, VA (RIVA)
NATTS
~
~
~
~
~
~
~
~
~
Elizabeth, NJ (ELNJ)
UATMP
1999-2006
~
~
~
~
~
~
~
~
~
~
Gary, IN (INDEM)
UATMP
2004-2006
~
~
~
~
~
~
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2015-2016 National Monitoring Programs Sites and Past Program Participation1 (Continued)
Monitoring Location
(and Site Name)
Program
2006 and Earlier
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Glenwood Springs, CO (GSCO)
UATMP
~
~
Grand Junction CO (GPCO)
NATTS
2004-2006
~
~
~
~
~
~
~
~
~
~
Grayson, KY (GLKY)
NATTS
~
~
~
~
~
~
~
~
~
Indianapolis, IN (WPIN)
UATMP
2006
~
~
~
~
~
~
~
~
~
~
Lexington KY (LEKY)
UATMP
~
~
~
~
~
Los Angeles, CA (CELA)
NATTS
~
~
~
~
~
~
~
~
~
~
Bronx, NY (BXNY)
NATTS
2006
~
~
~
~
~
~
~
~
~
North Brunswick, NJ (NBNJ)
UATMP
2001-2006
~
~
~
~
~
~
~
~
~
Northbrook, IL (NBIL)
NATTS
2003-2006
~
~
~
~
~
~
~
~
~
~
Oklahoma City, OK (OCOK)
UATMP
~
~
~
~
~
~
~
~
Oklahoma City, OK (NROK)
UATMP
~
Orlando, FL (PAFL)
UATMP
~
~
~
~
~
~
~
~
~
Parachute, CO (PACO)
UATMP
~
~
~
~
~
~
~
~
~
Phoenix, AZ (PXSS)
NATTS
2001-2004, 2006
~
~
~
~
~
~
~
~
~
~
Phoenix, AZ (SPAZ)
UATMP
2001
~
~
~
~
~
~
~
~
~
~
Pinellas Park, FL (SKFL)
NATTS
2004-2006
~
~
~
~
~
~
~
~
~
~
Providence, RI (PRRI)
NATTS
2005-2006
~
~
~
~
~
~
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2015-2016 National Monitoring Programs Sites and Past Program Participation1 (Continued)
Monitoring Location
(and Site Name)
Program
2006 and Earlier
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Rifle, CO (RICO)
UATMP
~
~
~
~
~
~
~
~
~
Rochester, NY (ROCH)
NATTS
2006
~
~
~
~
~
~
~
~
~
~
Roxana, IL (ROIL)
UATMP
~
~
~
~
Rubidoux, CA (RUCA)
NATTS
~
~
~
~
~
~
~
~
~
~
San Jose, CA (SJJCA)
NATTS
~
~
~
~
~
~
~
~
~
Schiller Park, IL (SPIL)
UATMP
2003-2006
~
~
~
~
~
~
~
~
~
~
Seattle, WA (SEWA)
NATTS
2005-2006
~
~
~
~
~
~
~
~
~
~
Silt, CO (BRCO)
UATMP
~
~
~
~
~
~
~
~
~
Smithland, KY (BLKY)
UATMP
~
~
~
~
~
St. Louis, MO (S4MO)
NATTS
2002, 2003-2006
~
~
~
~
~
~
~
~
~
~
St. Petersburg, FL (AZFL)
UATMP
1991-1992, 2001-
2006
~
~
~
~
~
~
~
~
~
~
Tulsa, OK (TMOK)
UATMP
~
~
~
~
~
~
~
~
Tulsa, OK (TOOK)
UATMP
2006
~
~
~
~
~
~
~
~
~
~
Tulsa, OK (TROK)
UATMP
~
~
~
~
Underhill, VT (UNVT)
NATTS
2002, 2005-2006
~
~
~
~
~
~
~
~
~
~
Valrico, FL (SYFL)
NATTS
2004-2006
~
~
~
~
~
~
~
~
~
~
Washington D.C. (WADC)
NATTS
2005-2006
~
~
~
~
~
~
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
-------
Table 2-1. 2015-2016 National Monitoring Programs Sites and Past Program Participation1 (Continued)
Monitoring Location
(and Site Name)
Program
2006 and Earlier
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Winter Park, FL(ORFL)
UATMP
1990-1991.2003-
2006
~
~
~
~
~
~
~
~
~
~
Yukon OK (YUOK)
UATMP
~
~
~
~
BOLD ITALICS = EPA-designated NATTS site
1 Includes monitoring sites participating under the NMP with the national contract laboratory
to
-------
Table 2-2. Site Characterizing Information for the 2015-2016 National Monitoring Programs Sites
Site
Code
AQS
Code
Location
Land Use
Location Setting
Estimated
Daily Traffic,
AADTa
(Year)
County-level
Stationary Source
HAP Emissionsb
(tpy)
County-level
Mobile Source
HAP Emissionsb
(tpy)
ASKY
21-019-0017
Ashland, KY
Residential
Suburban
5,934
(2014)
170.82
140.50
ASKY-M
21-019-0002
Ashland, KY
Industrial
Urban/City Center
13,241
(2015)
170.82
140.50
ATKY
21-157-0016
Calvert City, KY
Industrial
Suburban
3,672
(2015)
1,100.01
494.44
AZFL
12-103-0018
St. Petersburg, FL
Residential
Suburban
39,000
(2016)
1,513.94
2,866.11
BAKY
21-101-0014
Baskett, KY
Commercial
Rural
929
(2015)
541.67
232.81
BLKY
21-139-0004
Smithland, KY
Agricultural
Rural
2,011
(2016)
21.88
124.88
BMCO
08-045-0019
Battlement Mesa, CO
Commercial
Suburban
1,880
(2014)
4,509.05
246.90
BOMA
25-025-0042
Boston MA
Commercial
Urban/City Center
27,654
(2010)
771.94
978.39
BRCO
08-045-0009
Silt, CO
Agricultural
Rural
1,182
(2014)
4,509.05
246.90
BROK
40-051-0065
Bradley, OK
Residential
Rural
3,100
(2015)
736.53
208.62
BTUT
49-011-0004
Bountiful, UT
Residential
Suburban
133,965
(2014)
461.96
792.60
BXNY
36-005-0110
Bronx, NY
Residential
Urban/City Center
100,898
(2015)
1,203.95
974.59
CELA
06-037-1103
Los Angeles, CA
Residential
Urban/City Center
231,000
(2015)
12,908.63
11,950.01
CHNJ
34-027-3001
Chester, NJ
Agricultural
Rural
11,215
(2012)
717.64
1,270.98
CSNJ
34-007-0002
Camden, NJ
Industrial
Urban/City Center
3,231
(2012)
672.99
849.53
BOLD ITALICS = EPA-designated NATTS site
a Individual references provided in each state section.
bReference: 2014 NEI, version 1 (EPA, 2016)
0 GPCO's metals collection system is at a separate, but adjacent, location; thus, this site has two AQS codes.
d S4MO's emissions are city-level + county-level data.
-------
Table 2-2. Site Characterizing Information for the 2015-2016 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location Setting
Estimated
Daily Traffic,
AADTa
(Year)
County-level
Stationary Source
HAP Emissionsb
(tpy)
County-level
Mobile Source
HAP Emissionsb
(tpy)
DEMI
26-163-0033
Dearborn MI
Industrial
Suburban
86,600
(2015)
5,424.72
4,590.35
ELNJ
34-039-0004
Elizabeth, NJ
Industrial
Suburban
250,000
(2006)
801.13
969.16
GLKY
21-043-0500
Grayson, KY
Residential
Rural
303
(2012)
43.80
103.35
GPCO
08-077-0017
08-077-0018
Grand Junction, CO
Commercial
Urban/City Center
12,000
(2015)
525.20
509.82
GSCO
08-045-0020
Glenwood Springs, CO
Commercial
Suburban
27,000
(2015)
4,509.05
246.90
INDEM
18-089-0022
Gary, IN
Industrial
Urban/City Center
41,860
(2016)
1,128.12
1,605.34
LEKY
21-067-0012
Lexington KY
Residential
Suburban
18,993
(2014)
398.52
1,124.81
NBIL
17-031-4201
Northbrook, IL
Residential
Suburban
115,100
(2014)
13,088.01
10,072.89
NBNJ
34-023-0006
North Brunswick, NJ
Agricultural
Rural
114,322
(2010)
1,139.75
1,562.28
NRNJ
34-023-0011
East Brunswick, NJ
Agricultural
Rural
22,297
(2014)
1,139.75
1,562.28
NROK
40-109-0097
Oklahoma City, OK
Commercial
Urban/City Center
167,600
(2015)
1,508.04
2,790.29
OCOK
40-109-1037
Oklahoma City, OK
Residential
Suburban
52,500
(2015)
1,508.04
2,790.29
ORFL
12-095-2002
Winter Park, FL
Commercial
Urban/City Center
33,000
(2016)
2,204.15
3,938.70
PACO
08-045-0005
Parachute, CO
Residential
Urban/City Center
17,000
(2015)
4,509.05
246.90
PAFL
12-095-1004
Orlando, FL
Commercial
Suburban
50,000
(2016)
2,204.15
3,938.70
BOLD ITALICS = EPA-designated NATTS site
a Individual references provided in each state section.
bReference: 2014 NEI, version 1 (EPA, 2016)
0 GPCO's metals collection system is at a separate, but adjacent, location; thus, this site has two AQS codes.
d S4MO's emissions are city-level + county-level data.
-------
Table 2-2. Site Characterizing Information for the 2015-2016 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location Setting
Estimated
Daily Traffic,
AADTa
(Year)
County-level
Stationary Source
HAP Emissionsb
(tpy)
County-level
Mobile Source
HAP Emissionsb
(tpy)
PRRI
44-007-0022
Providence, RI
Residential
Urban/City Center
148,000
(2015)
874.24
1,353.77
PXSS
04-013-9997
Phoenix, AZ
Residential
Urban/City Center
35,103
(2010)
4,268.89
9,747.67
RFCO
08-045-0018
Carbondale, CO
Residential
Rural
18,000
(2015)
4,509.05
246.90
RICO
08-045-0007
Rifle, CO
Commercial
Urban/City Center
16,000
(2015)
4,509.05
246.90
RIVA
51-087-0014
East Highland Park, VA
Residential
Suburban
80,000
(2016)
721.62
838.70
ROCH
36-055-1007
Rochester, NY
Residential
Urban/City Center
85,833
(2015)
3,485.92
1,703.05
ROIL
17-119-9010
Roxana, IL
Industrial
Suburban
6,850
(2015)
1,119.13
896.15
RUCA
06-065-8001
Rubidoux, CA
Residential
Suburban
166,000
(2015)
2,253.98
2,699.06
S-IMOl
29-510-0085
St. Louis, MO
Residential
Urban/City Center
57,558
(2015)
2,109.40
3,912.01
SEWA
53-033-0080
Seattle, WA
Residential
Urban/City Center
186,000
(2015)
3,294.34
6,232.04
SJJCA
06-085-0005
San Jose, CA
Commercial
Urban/City Center
126,000
(2015)
1,561.19
1,852.36
SKFL
12-103-0026
Pinellas Park, FL
Residential
Suburban
4,000
(2016)
1,513.94
2,866.11
SPAZ
04-013-4003
Phoenix, AZ
Residential
Urban/City Center
21,601
(2015)
4,268.89
9,747.67
SPIL
17-031-3103
Schiller Park, IL
Mobile
Suburban
193,800
(2013)
13,088.01
10,072.89
SYFL
12-057-3002
Valrico, FL
Residential
Rural
3,900
(2016)
5,295.52
3,909.38
BOLD ITALICS = EPA-designated NATTS site
a Individual references provided in each state section.
bReference: 2014 NEI, version 1 (EPA, 2016)
0 GPCO's metals collection system is at a separate, but adjacent, location; thus, this site has two AQS codes.
d S4MO's emissions are city-level + county-level data.
-------
Table 2-2. Site Characterizing Information for the 2015-2016 National Monitoring Programs Sites (Continued)
Site
Code
AQS
Code
Location
Land Use
Location Setting
Estimated
Daily Traffic,
AADTa
(Year)
County-level
Stationary Source
HAP Emissionsb
(tpy)
County-level
Mobile Source
HAP Emissionsb
(tpy)
TMOK
40-143-1127
Tulsa, OK
Residential
Urban/City Center
4,400
(2015)
1,297.42
2,416.72
TOOK
40-143-0235
Tulsa, OK
Industrial
Urban/City Center
66,800
(2015)
1,297.42
2,416.72
TROK
40-143-0179
Tulsa, OK
Industrial
Urban/City Center
55,400
(2015)
1,297.42
2,416.72
TVKY
21-157-0014
Calvert City, KY
Industrial
Suburban
1,458
(2014)
1,100.01
494.44
UNVT
50-007-0007
Underhill, VT
Forest
Rural
970
(2014)
420.26
439.52
WADC
11-001-0043
Washington. D.C.
Commercial
Urban/City Center
3,600
(2014)
557.33
874.26
WPIN
18-097-0078
Indianapolis, IN
Residential
Suburban
24,917
(2016)
2,042.54
3,429.15
YUOK
40-017-0101
Yukon OK
Commercial
Suburban
42,900
(2015)
901.24
358.36
BOLD ITALICS = EPA-designated NATTS site
a Individual references provided in each state section.
bReference: 2014 NEI, version 1 (EPA, 2016)
0 GPCO's metals collection system is at a separate, but adjacent, location; thus, this site has two AQS codes.
d S4MO's emissions are city-level + county-level data.
-------
The proximity of the monitoring sites to different emissions sources, especially industrial
facilities and heavily traveled roadways, often explains the observed spatial variations in ambient
air quality. To provide a first approximation of the potential contributions of stationary and
mobile source emissions on ambient air quality at each site, Table 2-2 also lists the following:
• The number of vehicles passing the nearest available representative roadway to the
monitoring site, generally expressed as annual average daily traffic (AADT).
• Stationary and mobile source hazardous air pollutant (HAP) emissions for the
monitoring site's residing county, according to version 1 of the 2014 National
Emissions Inventory (NEI). (Version 2 of the 2014 NEI was published in the middle of
the production of the 2015-2016 NMP report and will be utilized in the 2017 NMP
report.)
This information is discussed in further detail in the individual state sections (Sections 5
through 23).
2.2 Analytical Methods and Pollutants Targeted for Monitoring
Air pollution typically contains hundreds of components, including, but not limited to,
VOCs, metals, and particulate matter (PM). Because the sampling and analysis required to
monitor for every component of air pollution has been prohibitively expensive, the NMP focuses
on specific pollutants that are analyzed at the laboratory using methods based on the EPA-
approved methods, as listed below:
• Compendium Method TO-15 was used to measure ambient air concentrations of
59 VOCs.
• EPA-approved SNMOC Method was used to measure 80 ozone precursors plus total
NMOC. This method can be performed concurrently with Method TO-15.
• Compendium Method TO-11A was used to measure ambient air concentrations of
15 carbonyl compounds.
• A combination of Compendium Method TO-13A and ASTMD6209 was used to
measure ambient air concentrations of 22 PAHs.
• A combination of Compendium Method 10-3.5 and EPA FEMEQL-0512-201
or EQL-0512-202 was used to measure ambient air concentrations of 11 metals.
• ASTM Method D7614 was used to measure ambient air concentrations of hexavalent
chromium.
• EPA-approved Methane Method was used to measure ambient air concentrations of
methane.
2-12
-------
The target pollutants and methods utilized varied from monitoring site to monitoring site.
The sample collection equipment at each site was installed either as a stand-alone collection
system or in a temperature-controlled enclosure (usually a trailer or a shed) with the sampling
probe inlet exposed to the ambient air. With these common setups, most monitoring sites
sampled ambient air at heights approximately 5 feet to 20 feet above local ground level.
The detection limits of the analytical methods must be considered carefully when
interpreting the corresponding ambient air monitoring data. By definition, method detection
limits (MDLs) represent the lowest concentrations at which laboratory equipment have been
experimentally determined to quantify concentrations of selected pollutants to a specific
confidence level. If a pollutant's concentration in ambient air is less than the method sensitivity
(as gauged by the MDL), the analytical method might not differentiate the pollutant from other
pollutants in the sample or from the random "noise" inherent in the analyses. While
quantification less than the MDL is possible, the measurement reliability is lower. Therefore,
when pollutants are present at concentrations less than their respective detection limits, multiple
analyses of the same sample may lead to a wide range of measurement results, including highly
variable concentrations or "non-detect" observations (i.e., the pollutant was not detected by the
instrument). Data analysts should exercise caution when interpreting monitoring data with a high
percentage of reported concentrations at levels near or less than the corresponding detection
limits.
MDLs are determined annually at the ERG laboratory using procedures outlined in the
U.S. Code of Federal Regulations (CFR), specifically 40 CFR, Part 136 Appendix B (EPA,
1986), in accordance with the specifications presented in the NATTS Technical Assistance
Document (TAD) (EPA, 2009a). This procedure involves analyzing at least seven replicate
standards spiked onto the appropriate sampling media and extracted (per analytical method).
Instrument-specific detection limits (replicate analysis of standards in solution) are not
determined because sampling media background and preparation variability would not be
considered. (Note that the 2016 NATTS TAD revisions went into effect at the end of 2017, and
thus, the updates to MDL determination are not applicable to the 2015-2016 NMP dataset).
MDLs for metals samples were calculated using the procedure described by
"Appendix D: DQ FAC Single Laboratory Procedure v2.4" (FAC, 2007), with the exception of
the arsenic MDL for Teflon® filters. The Federal Advisory Committee (FAC) MDL procedure
involves using historical blank filter data to calculate MDLs for each pollutant. For arsenic, the
2-13
-------
procedure described in 40 CFR was used to calculate the MDL rather than the FAC procedure
because this metal is not present at a high enough level in the background on the filters. In the
fall of 2015, the ERG laboratory obtained and employed a new Inductively Coupled
Plasma/Mass Spectrometry (ICP-MS) instrument. During this time, the FAC procedure for
determining MDLs was implemented for all 11 target analytes, such that all MDLs were
determined using the FAC approach.
Tables 2-3 through 2-9 identify the specific target pollutants for each analytical method
and their experimentally determined MDLs, as determined at the ERG laboratory for 2015 and
2016. For individual samples, the MDLs for VOC and SNMOC analyses do not change unless
the sample was diluted.
The following discussion presents an overview of the sampling and analytical methods.
For detailed descriptions of the methods, refer to EPA's original documentation of the
Compendium Methods (EPA, 1998; EPA, 1999a; EPA, 1999b; EPA, 1999c; EPA, 1999d; EPA,
2012; ASTM, 2012; ASTM, 2013; and SAE, 2011).
2.2.1 Sampling and Analytical Methods for Canister Samples (VOC, SNMOC, and
Methane)
VOC and SNMOC sampling and analysis are performed using methodology based on
EPA Compendium Method TO-15 (EPA, 1999a) and the procedure presented in EPA's
"Technical Assistance Document for Sampling and Analysis of Ozone Precursors" (EPA, 1998),
respectively. Because the TO-15 and SNMOC methods can be employed at the same time to
analyze the same canister sample, this report may refer to the SNMOC method as the
"concurrent" SNMOC method. Ambient air samples for VOC and/or SNMOC analysis were
collected in passivated stainless steel canisters. The ERG laboratory distributed the prepared
canisters (i.e., cleaned and evacuated) to the monitoring sites before each scheduled sample
collection event, and site operators connected the canisters to air sampling equipment prior to
each sample day. Prior to field sampling, the passivated canisters had internal pressures much
lower than atmospheric pressure. Using this pressure differential, ambient air flowed into the
canisters automatically once an associated system solenoid valve was opened. A mass flow
controller on the sampling device inlet ensured that ambient air entered the canister at an
integrated constant rate across the collection period. At the end of the 24-hour sampling period,
the solenoid valve automatically closed and stopped ambient air from flowing into the canister.
2-14
-------
Site operators recovered and returned the canisters, along with the Chain of Custody (COC)
forms and all associated documentation, to the ERG laboratory for analysis.
By analyzing each sample with gas chromatography incorporating mass spectrometry
(operating in the Selected Ion Monitoring (SIM) mode) and flame ionization detection
(GC/MS-FID), laboratory staff determined ambient air concentrations of 59 VOCs and/or
80 SNMOCs, and calculated the total non-methane organic compounds (TNMOC) concentration.
TNMOC is the sum of all hydrocarbon concentrations within the sample. Because w-xylene and
/^-xylene elute from the GC column at the same time, both the VOC and SNMOC analytical
methods report only the sum concentration for these two isomers, and not the separate
concentration for each isomer. Raw data for both methods are presented in Appendices B and C.
Table 2-3 presents the experimentally-determined detection limits for the VOC target
pollutants for 2015 and 2016 using Method TO-15. VOC detection limits are expressed in parts
per billion volume (ppbv). Table 2-4 presents the experimentally-determined detection limits for
the SNMOC target pollutants for 2015 and 2016. SNMOC detection limits are expressed in parts
per billion Carbon (ppbC).
2-15
-------
Table 2-3. 2015-2016 VOC Method Detection Limits
Pollutant
2015
MDL
(ppbv
2016
MDL
(ppbv)
Pollutant
2015
MDL
(ppbv)
2016
MDL
(ppbv)
Acetonitrile
0.031
0.051
Dichloromethane
0.019
0.021
Acetylene
0.020
0.029
1,2-Dichloropropane
0.017
0.019
Acrolein
0.030
0.120
cis-1,3 -Dichloropropene
0.017
0.020
Acrylonitrile
0.017
0.030
trans-1,3 -Dichloropropene
0.021
0.027
tert-Amyl Methyl Ether
0.009
0.017
Dichlorotetrafluoroethane
0.019
0.031
Benzene
0.039
0.021
Ethyl Acrylate
0.011
0.027
Bromochloromethane
0.015
0.013
Ethyl tert-Butyl Ether
0.008
0.012
Bromodichloromethane
0.019
0.019
Ethylbenzene
0.019
0.019
Bromoform
0.017
0.024
Hexachloro-1,3 -Butadiene
0.034
0.042
Bromomethane
0.009
0.025
Methyl Isobutyl Ketone
0.014
0.022
1.3 -Butadiene
0.014
0.026
Methyl Methacrylate
0.028
0.027
Carbon Disulfide
0.012
0.020
Methyl tert-Butyl Ether
0.014
0.009
Carbon Tetrachloride
0.010
0.017
n-Octane
0.017
0.018
Chlorobenzene
0.018
0.020
Propylene
0.032
0.054
Chloroethane
0.009
0.029
Styrene
0.016
0.021
Chloroform
0.016
0.012
1,1,2,2-Tetrachloroethane
0.018
0.030
Chloromethane
0.011
0.034
T etrachloroethylene
0.014
0.016
Cliloroprene
0.012
0.010
Toluene
0.018
0.017
Dibromoclilorometliane
0.013
0.021
1,2,4 -Trie hlorobenzene
0.050
0.035
1,2 -Dibromoethane
0.017
0.021
1,1,1 -T richloroethane
0.013
0.015
w;-Diclilorobenzene
0.025
0.024
1,1,2 -T richloroethane
0.017
0.020
o-Diclilorobenzene
0.025
0.027
T richloroethylene
0.017
0.017
p-Dichlorobcnzcnc
0.026
0.023
T richlorofluoro methane
0.008
0.020
Dichlorodifluoromethane
0.008
0.0202
T richlorotrifluoroethane
0.009
0.017
1,1 -Dicliloroethane
0.015
0.013
1,2,4 -T rimethylbenzene
0.021
0.024
1,2 -Dicliloroethane
0.013
0.013
1,3,5 -T rimethylbenzene
0.021
0.023
1,1 -Dichloroethene
0.007
0.023
Vinyl Chloride
0.008
0.032
67.Y- 1.2-Dichlorocthvlcnc
0.014
0.014
m,p-Xylene1
0.028
0.040
trans-1,2 -Dicliloroethy lene
0.012
0.013
o-Xylene
0.016
0.020
1 The VOC analytical method reports the sum concentration for w;-xylene and /^-xylene because these
isomers elute from the GC column at the same time.
2-16
-------
Table 2-4. 2015-2016 SNMOC Method Detection Limits1
Pollutant
2015
MDL
(ppbC)
2016
MDL
(ppbC)
Pollutant
2015
MDL
(ppbC)
2016
MDL
(ppbC)
Pollutant
2015
MDL
(ppbC)
2016
MDL
(ppbC)
Acetylene
0.086
0.027
//-Heptane
0.108
0.080
1-Octene
0.188
0.096
Benzene
0.083
0.096
1-Heptene
0.094
0.098
n-Pentane
0.109
0.053
1.3 -Butadiene
0.109
0.123
n-Hexane
0.124
0.074
1-Pentene
0.084
0.048
//-Butane
0.123
0.061
1-Hexene
0.105
0.052
6/.v-2-Pentcne
0.085
0.024
1-Butene
0.101
0.059
67.Y-2-Hc\cnc
0.067
0.071
trans-2 -Pentene
0.080
0.045
67.Y-2-Butcnc
0.060
0.025
tr <7/75-2 -Hexene
0.074
0.069
«-Pincnc
0.208
0.154
;ra«.v-2-B lite nc
0.071
0.027
Isobutane
0.125
0.039
/>-Pi nc nc
1.033
0.527
Cyclohexane
0.108
0.116
Isobutylene
0.124
0.052
Propane
0.256
0.122
Cyclopentane
0.095
0.074
Isopentane
0.116
0.033
/7-Propylbenzene
0.144
0.090
Cyclopentene
0.548
0.014
Isoprene
0.110
0.033
Propylene
0.163
0.090
w-Decane
0.971
0.175
Isopropylbenzene
0.147
0.088
Propyne
0.039
0.021
1-Decene
0.381
0.382
2-Methyl-1 -Butene
0.119
0.108
Styrene
0.803
0.756
///-Diethylbenzene
0.253
0.336
3 -Methyl-1 -Butene
0.170
0.337
Toluene
0.130
0.093
p-Diethylbenzene
0.194
0.257
2-Methyl-1 -Pentene
0.126
0.070
/7-Tridecane
0.329
0.542
2,2-Dimethylbutane
0.128
0.030
4-Methyl-1 -Pentene
0.094
0.060
1-Tridecene
0.449
0.376
2,3 -Dimethylbutane
0.098
0.019
2 -Methy 1-2 -B utene
0.108
0.061
1,2,3 -Trimethylbenzene
0.222
0.198
2,3 -Dimethylpentane
0.141
0.046
Methylcyclohexane
0.112
0.049
1,2,4-Trimethylbenzene
0.381
0.382
2,4-Dimethylpentane
0.137
0.054
Methylcyclopentane
0.097
0.066
1,3,5 -T rimethy lbenzene
0.228
0.141
n-Dodecane
0.228
0.789
2-Methylheptane
0.199
0.126
2,2,3 -T rimethy lpentane
0.131
0.053
1-Dodecene
0.804
1.005
3-Methylheptane
0.120
0.076
2,2,4-Trimethylpentane
0.122
0.078
Ethane
0.352
0.189
2-Methylhexane
0.291
0.262
2,3,4-Trimethylpentane
0.130
0.090
2-Ethyl-l-butene
0.094
0.080
3-Methylhexane
0.553
0.224
n-Undecane
0.241
0.446
Ethylbenzene
0.157
0.093
2-Methylpentane
0.220
0.105
1-Undecene
0.269
0.620
Ethylene
0.170
0.153
3-Methylpentane
0.093
0.043
/w-Xylene/p-Xylene2
0.196
0.152
///-Ethyltoluene
0.212
0.124
n-Nonane
0.304
0.080
o-Xylene
0.127
0.084
o-Ethyltoluene
0.213
0.164
1-Nonene
0.143
0.087
SumofKnowns, Sun of Unknowns, and
TNMOC have no applicable MDLs
/?-Ethvltolucnc
0.148
0.141
n-Octane
0.264
0.082
1 Concentration in ppbC = concentration in ppbv * number of carbon atoms in the compound.
2 The SNMOC analytical method reports the sum concentration for ///-xylene and /?-xylene because these isomers elute from the GC column at the same time.
-------
Methane sampling and analysis was performed with a vacuum well interface attached to a
gas chromatograph using methodology based on SAE J1151: Methane Measurement Using Gas
Chromatography (SAE, 2011). Ambient air samples for methane analysis were collected in
passivated stainless steel canisters. The ERG laboratory distributed the prepared canisters
(i.e., cleaned and evacuated) to the monitoring sites before each scheduled sample collection
event, and site operators connected the canisters to air sampling equipment prior to each sample
day. Prior to field sampling, the passivated canisters had internal pressures much lower than
atmospheric pressure. Using this pressure differential, ambient air flowed into the canisters
automatically once an associated system solenoid valve was opened. A mass flow controller on
the sampling device inlet ensured that ambient air entered the canister at an integrated constant
rate across the collection period. At the end of the 24-hour sampling period, the solenoid valve
automatically closed and stopped ambient air from flowing into the canister. Site operators
recovered and returned the canisters, along with the COC forms and all associated
documentation, to the ERG laboratory for analysis.
By analyzing each sample with gas chromatography incorporating flame ionization
detection (GC-FID), laboratory staff determined ambient air concentrations of methane. Because
of the use of a vacuum well, the pressure of each sample must be taken immediately prior to
analysis and used as a correction factor for the final result. Methane samples were collected at
two sites in 2015 and 2016, BROK and NROK, as part of a special study. Raw data for methane
are presented in Appendix D.
Table 2-5 presents the 2015 and 2016 MDLs for the laboratory analysis of methane
samples. Methane detection limits are expressed in parts per million Carbon (ppmC).
Table 2-5. 2015-2016 Methane Method Detection Limit
Pollutant
2015
MDL
(ppmC)
2016
MDL
(ppmC)
Methane
0.104
0.101
2-18
-------
2.2.2 Carbonyl Compound Sampling and Analytical Method
Sampling and analysis for carbonyl compounds was performed using methodology based
on EPA Compendium Method TO-11A (EPA, 1999b). Ambient air samples for carbonyl
compound analysis were collected by passing ambient air through an ozone scrubber and then
through cartridges containing silica gel coated with 2,4-dinitrophenylhydrazine (DNPH), a
compound known to react selectively and reversibly with many aldehydes and ketones. Carbonyl
compounds in ambient air are retained in the sampling cartridge, while other compounds pass
through without reacting with the DNPH-coated matrix. The ERG laboratory distributed the
DNPH cartridges to the monitoring sites prior to each scheduled sample collection event and site
operators connected the cartridges to the air sampling equipment. After each 24-hour sampling
period, site operators recovered the cartridges and returned them, along with the COC forms and
all associated documentation, to the ERG laboratory for analysis.
To quantify concentrations of carbonyl compounds in the sampled ambient air, laboratory
analysts extracted the exposed DNPH cartridges with acetonitrile. High-performance liquid
chromatography (HPLC) analysis and ultraviolet (UV) detection of these solutions determined
the relative amounts of individual carbonyl compounds present in the original air sample.
Because the three tolualdehyde isomers co-elute from the HPLC column, only the sum
concentration for these isomers, and not the separate concentrations for each isomer, are
reported. Raw data for Method TO-11A are presented in Appendix E.
Table 2-6 presents the experimentally-determined detection limits for the carbonyl
compound target pollutants for 2015 and 2016. Detection limits for carbonyl compounds are
expressed in ppbv.
2-19
-------
Table 2-6. 2015-2016 Carbonyl Compound Method Detection Limits1
2015
2016
MDL
MDL
Pollutant
(ppbv)
(ppbv)
Acetaldehyde
0.006
0.006
Acetone
0.013
0.053
Benzaldehyde
0.003
0.004
2-Butanone
0.003
0.005
Butyraldehyde
0.003
0.006
Crotonaldehyde
0.006
0.004
2,5 -Dimethylbenzaldehyde
0.002
0.003
Formaldehyde
0.012
0.010
Hexaldehyde
0.002
0.005
Isovaleraldehyde
0.004
0.003
Propionaldehyde
0.003
0.004
Tolualdehydes2
0.004
0.008
Valeraldehyde
0.002
0.004
1 Assumes a volume of 1,000 m3.
2 The three tolualdehyde isomers elute from the HPLC column at
the same time; thus, the analytical method reports only the sum
concentration for these three isomers and not the individual
concentrations.
2.2.3 PAH Sampling and Analytical Method
PAH sampling and analysis was performed using methodology based on EPA
Compendium Method TO-13A (EPA, 1999c) and ASTM D6209 (ASTM, 2013). The ERG
laboratory prepared sampling media and supplied them to the sites before each scheduled sample
collection event. The clean sampling PUF/XAD-2® cartridge and glass fiber filter were installed
in a high volume collection system for a 24-hour sampling period. Sample collection modules,
COC forms, and all associated documentation were returned to the ERG laboratory after sample
collection. Within 14 days of sampling, the filter and cartridge are extracted together using a
toluene in hexane solution using the Dionex Accelerated Solvent Extractor (ASE) 350 or
ASE 300. The sample extract is concentrated to a final volume of 1.0 milliliter (mL). A volume
of 0.6 microliter ([j,L) is injected into the GC/MS operating in the SIM mode to analyze for
22 PAHs. Raw data for Method TO-13A are presented in Appendix F.
Table 2-7 presents the experimentally-determined detection limits for the 22 PAH target
pollutants for 2015 and 2016. PAH detection limits are expressed in nanograms per cubic
meter (ng/m3).
2-20
-------
Table 2-7. 2015-2016 PAH Method Detection Limits1
Pollutant
2015
MDL
(ng/m3)
2016
MDL
(ng/m3)
Acenaphthene
0.082
0.047
Acenaphthylene
0.067
0.015
Anthracene
0.072
0.076
Benzo(a)anthracene
0.082
0.060
Benzo(a)pyrene
0.132
0.063
Benzo(b)fluoranthene
0.092
0.074
Benzo(e)pyrene
0.096
0.046
Benzo(g,h,i)perylene
0.068
0.042
Benzo(k)fluoranthene
0.098
0.059
Chrysene
0.069
0.074
Coronene
0.095
0.008
Cvclopcnta|cd|pvrcnc
0.125
0.006
Dibenz(a,h)anthracene
0.089
0.017
Fluoranthene
0.098
0.114
Fluorene
0.189
0.113
9-Fluorenone
0.132
0.025
Indeno( 1,2,3 -cd)pyrene
0.083
0.045
Naphthalene
0.166
0.791
Perylene
0.065
0.013
Phenanthrene
0.156
0.066
Pyrene
0.090
0.094
Retene
0.093
0.083
1 Assumes a volume of 300 m3.
2.2.4 Metals Sampling and Analytical Method
Ambient air samples for metals analysis were collected by passing ambient air through
either 47mm Teflon® filters or 8" x 10" quartz filters, depending on the separate and distinct
sampling apparatus used to collect the sample; the 47mm Teflon® filter is used for low-volume
collection systems, whereas the 8" x 10" quartz filter is used for high-volume collection systems.
EPA provided the filters to the monitoring sites. Sites sampled for either particulate matter less
than 10 microns (PMio) or total suspended particulate (TSP). Particulates in ambient air were
collected on the filters and, after a 24-hour sampling period, site operators recovered and sent the
filters, along with the COC forms and all associated documentation, to the ERG laboratory for
analysis.
2-21
-------
Extraction and analysis for the determination of speciated metals in or on particulate
matter was performed using a combination of EPA Compendium Method 10-3.5 and EPA FEM
Methods EQL-0512-201 and EQL-0512-202 (EPA, 1999d; EPA, 2012). Upon receipt at the
laboratory, whole filters (47mm Teflon®) or filter strips (8" x 10" quartz) were digested using a
dilute nitric acid, hydrochloric acid, hydrogen peroxide and/or hydrofluoric acid (Teflon® only)
solution. The digestate was then quantified using ICP-MS to determine the concentration of
individual metals present in the original air sample. Raw data for speciated metals are presented
in Appendix G.
Table 2-8 presents the experimentally-determined detection limits for metals samples for
2015 and 2016, as reported by the ERG laboratory. Due to the difference in sample volume/filter
collection media, there are two sets of MDLs listed in Table 2-8, one for each filter type.
Table 2-8. 2015-2016 Metals Method Detection Limits
Pollutant
2015
MDL
(ng/m3)
2016
MDL
(ng/m3)
Pollutant
2015
MDL
(ng/m3)
2016
MDL
(ng/m3)
47mm Teflon®1
8x10" Quartz2
Antimony
0.041
0.017
Antimony
0.012
0.114
Arsenic
0.153
0.039
Arsenic
0.057
0.013
Beryllium
0.017
0.001
Beryllium
0.002
0.0005
Cadmium
0.011
0.002
Cadmium
0.006
0.006
Chromium
12.0
4.18
Chromium
2.56
1.74
Cobalt
0.012
0.097
Cobalt
0.043
0.058
Lead
0.039
0.034
Lead
0.113
0.150
Manganese
0.127
0.143
Manganese
0.218
0.446
Mercury
0.036
0.017
Mercury
0.006
0.006
Nickel
0.286
0.204
Nickel
0.572
0.479
Selenium
0.292
0.054
Selenium
0.031
0.012
Assumes a volume of 24.04 m3.
2 Assumes a volume of 2,000 m3.
2-22
-------
2.2.5 Hexavalent Chromium Sampling and Analytical Method
Hexavalent chromium was measured using the method described in ASTM D7614
(ASTM, 2012). Ambient air samples of hexavalent chromium from TSP were collected by
passing ambient air through sodium bicarbonate impregnated acid-washed cellulose filters. ERG
prepared and distributed the filters secured in Teflon® cartridges or in petri dishes, per the
specific collection system used at each site, to the monitoring sites prior to each scheduled
sample collection event. Site operators connected the cartridges (or installed the filters) to the air
sampling equipment. After a 24-hour sampling period, site operators recovered the cartridges (or
filters) and returned them, along with the COC forms and all associated documentation, to the
ERG laboratory for analysis. Upon receipt at the laboratory, the filters were extracted using a
sodium bicarbonate solution. Ion chromatography (IC) analysis using Ultraviolet-Visible
detection of the extracts determined the amount of hexavalent chromium present in each sample.
Raw data for the hexavalent chromium method are presented in Appendix H.
Table 2-9 presents the experimentally-determined detection limit for hexavalent
chromium samples for 2015 and 2016, as reported by the ERG laboratory, which are expressed
in ng/m3.
Table 2-9. 2015-2016 Hexavalent Chromium Method Detection Limits1
Pollutant
2015
MDL
(ng/m3)
2016
MDL
(ng/m3)
Hexavalent Chromium
0.0038
0.0037
1 Assumes a volume of 21.6 m3.
2.3 Sample Collection Schedules
Table 2-10 presents the first and last date upon which sample collection occurred for each
monitoring site sampling under the NMP in 2015. Table 2-11 presents similar information for the
2016 sampling year. The first sample date for each site is generally at the beginning of January
and sampling continued through the end of December, although there are exceptions, including:
• Concurrent VOC and SNMOC sampling was initiated at RFCO at the beginning of
2015. VOC sampling was discontinued in September 2015.
• Sampling at BMCO was discontinued at the beginning of February 2015. The
instrumentation was redeployed at GSCO within a week's time, where sampling
reconvened and continued for just over one year. The instrumentation was returned to
BMCO in mid-March 2016, where sampling resumed.
2-23
-------
• The state of Oklahoma initiated two new monitoring sites, one in Bradley, Oklahoma
(BROK) in April 2015 and a near-road site in Oklahoma City, Oklahoma (NROK) in
May 2016. Methane samples were collected at these sites on the same schedule as the
other pollutant groups sampled for at these locations.
• Sampling at the site in Roxana, Illinois (ROIL) was discontinued at the end of July
2015, after completing a 3-year monitoring study.
• The instrumentation at one of the UATMP's longest running sites, NBNJ, was moved
to a new location, NRNJ, at the beginning of 2016. The new location is less than a
mile from the old location.
• Sampling at two long-term Orange County, Florida sites (PAFL and ORFL) was
discontinued at the end of September 2016.
• Hexavalent chromium sampling was discontinued at RIVA at the end of June 2016.
• VOC sampling was discontinued at LEKY at the end of July 2016.
According to the NMP schedule, 24-hour integrated samples were collected at each
monitoring site on a l-in-6 day schedule and sample collection began and ended at midnight,
local standard time. However, there were some exceptions, as some sites collected samples on a
l-in-12 day schedule, dependent upon location and monitoring objectives. The sites sampling on
a l-in-12 day schedule are denoted in Tables 2-10 and 2-11 and include:
• SNMOC samples were collected on a l-in-6 day schedule while carbonyl compounds
were collected on a l-in-12 day schedule at BMCO, BRCO, GSCO, PACO, and
RICO. Sampling at RFCO was conducted on a l-in-12 day schedule for both VOCs
and SNMOCs.
• The South Phoenix, Arizona site (SPAZ) collected VOC samples on a l-in-12 day
schedule.
• The Orlando, Florida site (PAFL) collected metals samples on a l-in-12 day schedule.
2-24
-------
Table 2-10. 2015 Sampling Schedules and Completeness Rates
to
to
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
ASKY
1/6/15
12/26/15
_
_
„
60
60
100
_
„
„
_
_
„
„
_
„
„
„
_
ASKY-M
1/6/15
12/26/15
_
_
„
_
„
_
_
„
„
55
60
92
„
_
„
„
„
_
ATKY
1/6/15
12/26/15
_
_
„
59
60
98
_
„
„
_
_
„
„
_
„
„
„
_
AZFL
1/6/15
12/26/15
58
60
97
_
„
_
_
„
„
_
_
„
„
_
„
„
„
_
BAKY
1/6/15
12/26/15
_
_
„
_
„
_
_
„
„
56
60
93
„
_
„
„
„
_
BLKY
1/6/15
12/26/15
_
_
„
59
60
98
_
„
„
51
60
85
„
_
„
„
„
_
BMCO
1/6/15
2/5/15
3
3
1002
_
„
_
_
„
„
_
_
„
6
6
100
„
„
_
BOMA
1/6/15
12/26/15
_
_
„
_
„
_
_
„
„
60
60
100
„
_
„
60
60
100
BRCO
1/6/15
12/26/15
28
30
932
_
„
_
_
„
„
_
_
„
52
60
87
„
„
_
BROK3
4/12/15
12/26/15
40
44
91
39
44
89
_
„
„
_
_
„
39
44
89
„
„
_
BTUT
1/6/15
12/26/15
54
60
90
50
60
83
_
„
„
60
60
100
49
60
82
56
60
93
BXNY
1/6/15
12/26/15
_
_
„
_
„
_
_
„
„
_
_
„
„
_
„
59
60
98
CELA
1/6/15
12/26/15
_
_
„
_
„
_
_
„
„
_
_
„
„
_
„
57
60
95
CHNJ
1/6/15
12/26/15
34
60
57
55
60
92
_
„
„
_
_
„
„
_
„
„
„
_
CSNJ
1/6/15
12/26/15
59
60
98
61
60
>100
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2015 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-10. 2015 Sampling Schedules and Completeness Rates (Continued)
to
to
On
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
DEMI
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
_
_
„
„
_
„
62
60
>100
ELNJ
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
_
_
„
„
_
„
„
„
_
GLKY
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
58
60
97
„
_
„
59
60
98
GPCO
1/6/15
12/28/15
55
60
92
50
60
83
_
„
„
57
60
95
„
_
„
55
60
92
GSCO
2/11/15
12/26/15
26
27
96
„
„
_
_
„
„
_
_
„
52
54
96
„
„
_
INDEM
1/6/15
12/26/15
60
60
100
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
LEKY
1/6/15
12/26/15
„
„
_
53
60
88
_
„
„
56
60
93
„
_
„
„
„
_
NBIL
1/6/15
12/26/15
57
60
95
54
60
90
_
„
„
56
60
93
54
60
90
60
60
100
NBNJ
1/6/15
12/26/15
59
60
98
59
60
98
_
„
„
_
_
„
„
_
„
„
„
_
OCOK
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
60
60
100
„
_
„
„
„
_
ORFL
1/6/15
12/26/15
58
60
97
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
PACO
1/6/15
12/26/15
18
30
602
„
„
_
_
„
„
_
_
„
54
60
90
„
„
_
PAFL2
1/12/15
12/26/15
„
„
_
„
„
_
_
„
„
30
30
100
„
_
„
„
„
_
PRRI
1/6/15
12/26/15
„
„
_
„
„
_
_
„
„
_
_
„
„
_
„
60
60
100
PXSS
1/6/15
12/26/15
33
60
55
58
60
97
59
60
98
55
60
92
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2015 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-10. 2015 Sampling Schedules and Completeness Rates (Continued)
to
to
-J
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
RFCO2
1/6/15
12/26/15
„
„
„
20
22
91
_
„
„
_
_
„
26
30
87
„
„
_
RICO
1/6/15
12/26/15
26
30
872
„
„
_
_
„
„
_
_
„
46
60
77
„
„
_
RIVA
1/6/15
12/26/15
„
„
„
„
„
_
59
60
98
_
_
„
„
_
„
56
60
93
ROCH
1/6/15
12/26/15
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
56
60
93
ROIL
1/6/15
7/29/15
33
35
94
32
35
91
_
„
„
_
_
„
„
_
„
„
„
_
RUCA
1/6/15
12/26/15
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
59
60
98
S4MO
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
60
60
100
„
_
„
58
60
97
SEWA
1/6/15
12/26/15
59
60
98
57
60
95
_
„
„
58
60
97
„
_
„
57
60
95
SJJCA
1/6/15
12/26/15
„
„
„
„
„
_
_
„
„
59
60
98
„
_
„
58
60
97
SKFL
1/6/15
12/26/15
58
60
97
„
„
_
_
„
„
_
_
„
„
_
„
59
60
98
SPAZ2
1/6/15
12/26/15
„
„
„
32
30
>100
_
„
„
_
_
„
„
_
„
„
„
_
SPIL
1/6/15
12/26/15
61
60
>100
60
60
100
_
„
„
_
_
„
„
_
„
„
„
_
SYFL
1/6/15
12/26/15
57
60
95
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
TMOK
1/6/15
12/26/15
60
60
100
60
60
100
_
„
„
59
60
98
„
_
„
„
„
_
TOOK
1/6/15
12/26/15
60
60
100
59
60
98
60
60
100
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2015 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-10. 2015 Sampling Schedules and Completeness Rates (Continued)
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
TROK
1/6/15
12/26/15
59
60
98
60
60
100
_
„
„
60
60
100
„
_
„
„
„
_
TVKY
1/6/15
12/26/15
„
„
„
62
60
>100
_
„
„
_
_
„
„
_
„
„
„
_
UNVT
1/6/15
12/26/15
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
60
60
100
WADC
1/6/15
12/26/15
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
60
60
100
WPIN
1/6/15
12/26/15
59
60
98
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
YUOK
1/6/15
12/26/15
59
60
98
59
60
98
__
__
59
60
98
__
__
__
__
Total
1,533
1,639
94
1,458
1,511
96
59
60
98
1,073
1,110
97
378
434
87
1,106
1,140
97
A = Number of valid samples collected.
Is) B = Number of valid samples that should be collected in 2015 based on sample schedule and start/end date of sampling.
Is) C = Completeness (%).
00 1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-11. 2016 Sampling Schedules and Completeness Rates
to
to
VO
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
ASKY
1/1/16
12/26/16
_
_
„
60
61
98
_
„
„
_
_
„
„
_
„
„
„
_
ASKY-M
1/1/16
12/26/16
_
_
„
_
„
_
_
„
„
55
61
90
„
_
„
„
„
_
ATKY
1/1/16
12/26/16
_
_
„
61
61
100
_
„
„
_
_
„
„
_
„
„
„
_
AZFL
1/1/16
12/26/16
61
61
100
_
„
_
_
„
„
„
_
„
„
_
„
„
„
_
BAKY
1/1/16
12/26/16
_
_
„
_
„
_
_
„
„
58
61
95
„
_
„
„
„
_
BLKY
1/1/16
12/26/16
_
_
„
60
61
98
_
„
„
59
61
97
„
_
„
„
„
_
BMCO
3/13/16
12/26/16
24
24
1002
_
„
_
_
„
„
_
_
„
48
49
98
„
„
_
BOMA
1/1/16
12/26/16
_
_
„
_
„
_
_
„
„
58
61
95
„
_
„
61
61
100
BRCO
1/1/16
12/26/16
26
30
872
_
„
_
_
„
„
_
_
„
58
61
95
„
„
_
BROK3
1/1/16
12/26/16
60
61
98
59
61
97
_
„
„
_
_
„
59
61
97
„
„
_
BTUT
1/1/16
12/26/16
59
61
97
59
61
97
_
„
„
55
61
90
59
61
97
61
61
100
BXNY
1/1/16
12/26/16
_
_
„
_
„
_
_
„
„
_
_
„
„
_
„
61
61
100
CELA
1/1/16
12/26/16
_
_
„
_
„
_
_
„
„
_
_
„
„
_
„
61
61
100
CHNJ
1/1/16
12/26/16
60
61
98
58
61
95
_
„
„
_
_
„
„
_
„
„
„
_
CSNJ
1/1/16
12/26/16
60
61
98
55
61
90
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2016 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-11. 2016 Sampling Schedules and Completeness Rates (Continued)
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
DEMI
1/1/16
12/26/16
61
61
100
60
61
98
_
„
„
_
_
„
„
_
„
61
61
100
ELNJ
1/1/16
12/26/16
61
61
100
60
61
98
_
„
„
_
_
„
„
_
„
„
„
_
GLKY
1/1/16
12/26/16
61
61
100
61
61
100
_
„
„
59
61
97
„
_
„
61
61
100
GPCO
1/1/16
12/26/16
59
61
97
62
61
>100
_
„
„
60
61
98
„
_
„
61
61
100
GSCO
1/1/16
3/7/16
5
6
83
„
„
_
_
„
„
_
_
„
12
12
100
„
„
_
INDEM
1/1/16
12/26/16
59
61
97
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
LEKY
1/1/16
12/26/16
„
„
_
29
36
81
_
„
„
47
61
77
„
_
„
„
„
_
NBIL
1/1/16
12/26/16
59
61
97
59
61
97
_
„
„
57
61
93
59
61
97
56
61
92
NRNJ
1/1/16
12/26/16
60
61
98
60
61
98
_
„
„
_
_
„
„
_
„
„
„
_
NROK3
5/18/16
12/26/16
38
38
100
38
38
100
_
„
„
_
_
„
38
38
100
„
„
_
OCOK
1/1/16
12/26/16
60
61
98
61
61
100
_
„
„
60
61
98
„
_
„
„
„
_
ORFL
1/1/16
9/27/16
41
46
89
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
PACO
1/7/16
12/26/16
29
30
972
„
„
_
_
„
„
_
_
„
54
61
89
„
„
_
PAFL2
1/1/16
9/27/16
„
„
_
„
„
_
_
„
„
24
23
>100
„
_
„
„
„
_
PRRI
1/1/16
12/26/16
60
61
98
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2016 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-11. 2016 Sampling Schedules and Completeness Rates (Continued)
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
PXSS
1/1/16
12/29/16
58
61
95
60
61
98
_
„
„
61
61
100
„
_
„
51
61
84
RFCO2
1/7/16
12/20/16
„
„
„
„
„
_
_
„
„
_
_
„
27
30
90
„
„
_
RICO
1/7/16
12/26/16
30
30
1002
„
„
_
_
„
„
_
_
„
62
61
>100
„
„
_
RIVA
1/1/16
12/26/16
„
„
„
„
„
_
27
31
87
_
_
„
„
_
„
60
61
98
ROCH
1/1/16
12/26/16
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
60
61
98
RUCA
1/1/16
12/26/16
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
59
61
97
S4MO
1/1/16
12/26/16
60
61
98
60
61
98
_
„
„
61
61
100
„
_
„
60
61
98
SEWA
1/1/16
12/26/16
61
61
100
61
61
100
_
„
„
58
61
95
„
_
„
61
61
100
SJJCA
1/1/16
12/26/16
„
„
„
„
„
_
_
„
„
57
61
93
„
_
„
61
61
100
SKFL
1/1/16
12/28/16
59
61
97
„
„
_
_
„
„
_
_
„
„
_
„
59
61
97
SPAZ2
1/1/16
12/26/16
„
„
„
31
31
100
_
„
„
_
_
„
„
_
„
„
„
_
SPIL
1/1/16
12/26/16
57
61
93
58
61
95
_
„
„
_
_
„
„
_
„
„
„
_
SYFL
1/1/16
12/26/16
56
61
92
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
TMOK
1/1/16
12/26/16
60
61
98
60
61
98
_
„
„
61
61
100
„
_
„
„
„
_
TOOK
1/1/16
12/26/16
61
61
100
61
61
100
61
61
100
A = Number of valid samples collected.
B = Number of valid samples that should be collected in 2016 based on sample schedule and start/end date of sampling.
C = Completeness (%).
1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
Table 2-11. 2016 Sampling Schedules and Completeness Rates (Continued)
Site
Monitoring Period1
Carbonyl
Compounds
VOCs
Hexavalent
Chromium
Metals
SNMOCs
PAHs
First
Sample
Last
Sample
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
A
B
C
TROK
1/1/16
12/26/16
61
61
100
61
61
100
_
„
„
61
61
100
„
_
„
„
„
_
TVKY
1/1/16
12/26/16
„
„
„
61
61
100
_
„
„
_
_
„
„
_
„
„
„
_
UNVT
1/1/16
12/26/16
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
61
61
100
WADC
1/1/16
12/26/16
„
„
„
„
„
_
_
„
„
_
_
„
„
_
„
58
61
95
WPIN
1/1/16
12/26/16
57
61
93
„
„
_
_
„
„
_
_
„
„
_
„
„
„
_
YUOK
1/1/16
12/26/16
61
61
100
61
61
100
__
__
61
61
100
__
__
__
__
Total
1,624
1,668
97
1,476
1,508
98
27
31
87
1,073
1,121
96
476
495
96
1,133
1,159
98
A = Number of valid samples collected.
Is) B = Number of valid samples that should be collected in 2016 based on sample schedule and start/end date of sampling,
w C = Completeness (%).
1x0 1 Begins with first sample collected and ends with last sample collected; date range presented may not be representative of each method-specific date range.
2 Sampling schedule was a l-in-12 day schedule rather than a l-in-6 schedule.
3 Methane samples were also collected at this site and will be discussed in later sections.
BOLD ITALICS = EPA-designated NATTS site.
Orange shading indicates that completeness is less than the MQO of 85 percent.
-------
The number of sites at which samples were collected for each method varied only slightly
between 2015 and 2016, as shown in Tables 2-10 and 2-11:
• 27 sites collected VOC samples in 2015 compared to 26 sites in 2016.
• 9 sites collected SNMOC samples in 2015 compared to 10 sites in 2016.
• 31 sites collected carbonyl compound samples in both 2015 and 2016.
• 19 sites collected PAH samples in both 2015 and 2016.
• 19 sites collected metals samples in both 2015 and 2016.
• 1 site collected hexavalent chromium samples in both 2015 and 2016.
• 1 site collected methane samples in 2015 compared to 2 sites in 2016.
As part of the sampling schedule, site operators were instructed to collect duplicate (or
collocated) samples on roughly 10 percent of the sample days for select methods when duplicate
(or collocated) collection systems were available. Field blanks were collected once per month for
carbonyl compounds, hexavalent chromium, metals, and PAHs. Sampling calendars were
distributed to help site operators schedule the collection of samples, duplicates, and field blanks.
In cases where a valid sample was not collected on a given scheduled sample day, site operators
were instructed to reschedule or "make up" samples on other days. This practice explains why
some monitoring sites periodically strayed from the l-in-6 or l-in-12 day sampling schedule.
The l-in-6 or l-in-12 day sampling schedule provides cost-effective approaches to data
collection for trends characterization of toxic pollutants in ambient air and ensures that sample
days are evenly distributed among the seven days of the week to allow weekday/weekend
comparison of air quality. Because the l-in-6 day schedule yields twice the number of
measurements than the l-in-12 day schedule, data characterization based on this schedule tends
to be more representative.
2.4 Completeness
Completeness refers to the number of valid samples collected and analyzed compared to
the number of total samples expected based on a l-in-6 or l-in-12 day sample schedule.
Monitoring programs that consistently generate valid samples have higher completeness than
programs that consistently have invalid samples. The completeness of an air monitoring
program, therefore, can be a qualitative measure of the reliability of air sampling and laboratory
analytical equipment as well as a measure of the competency of the field and laboratory staff
2-33
-------
involved and the efficiency with which the program is managed. The completeness for each
monitoring site and method sampled is presented in Tables 2-10 (for 2015) and 2-11 (for 2016).
The measurement quality objective (MQO) for completeness based on the EPA-approved
Quality Assurance Project Plan (QAPP) specifies that at least 85 percent of samples from a given
monitoring site must be collected and analyzed successfully to be considered sufficient for data
trends analysis (ERG, 2015 and ERG, 2016). The data in Tables 2-10 and 2-11 show that 204 of
the 215 datasets (107 from 2015 and 108 from 2016) met the 85 percent completeness MQO;
11 datasets (seven from 2015 and four from 2016) from the 2015-2016 NMP monitoring effort
did not meet this MQO (orange shaded cells in Tables 2-10 and 2-11):
• BTUT VOCs and SNMOCs in 2015 - A number of invalid canister samples scattered
throughout the year a completeness resulted in a completeness less than 85 percent for
BTUT in 2015.
• CHNJ carbonyl compounds in 2015 - Issues with the collection system were
discovered at CHNJ and resulted in the invalidation of carbonyl compound samples
collected between March 31, 2015 and September 3, 2015, after which a new
collection system was installed.
• GPCO VOCs in 2015 - A number of invalid VOC samples between February and
April 2015, many of which were related to a shortened sampling duration, resulted in a
completeness less than 85 percent for GPCO in 2015.
• PACO carbonyl compounds in 2015 - A series of carbonyl compound samples
resembling field blanks were collected at PACO during the first half of 2015, resulting
in the invalidation of one-third of the samples collected in 2015.
• PXSS carbonyl compounds in 2015 - An instrument contamination issue with the
primary carbonyl compound collection system at PXSS resulted in the invalidation of
many samples collected on the primary collection system between January and May
2015 and also between August and November 2015.
• RICO SNMOCs in 2015 - A number of SNMOC samples did not run properly at
RICO in June and first part of July 2015; this combined with other invalid samples
throughout the year resulted in a completeness less than 85 percent for RICO in 2015.
• GSCO carbonyl compounds in 2016 - Because the instrumentation at GSCO was
moved back to BMCO in March 2016, the one invalid sample was enough to reduce
the completeness to less than 85 percent for GSCO's carbonyl compounds.
• LEKY metals and VOCs in 2016 - A number of metals samples collected at LEKY in
March and April 2016 had QA-related issues according to the state of Kentucky.
Operator errors combined with other invalidated samples also resulted in a VOC
completeness less than 85 percent for LEKY.
2-34
-------
• PXSS PAHs in 2016 - Issues with the collection system during late summer and early
fall 2016 led to a PAH completeness less than 85 percent for PXSS.
The percent completeness for each of these datasets varies from just less than the MQO of
85 percent (between 80 percent and 85 percent for each) to 55 percent (PXSS carbonyl
compounds in 2015). Appendix I identifies samples that were invalidated and lists the reason for
invalidation, based on the applied AQS null code.
Also of note, a contaminated internal standard used at the laboratory for Method TO-15
resulted in unusual analytical results for hexachloro-1,3-butadiene and 1,2,4-trichlorobenzene at
VOC sites sampling in 2015. It was determined that the internal standard in use was
contaminated and these results were invalidated. Affected samples were collected at the end of
February 2015 or early March 2015 through mid-December 2015. As this affected only two of
the VOCs for which measurements were collected, this invalidation is not reflected in
Table 2-10.
A second, separate contaminated internal standard was in use during the fall of 2016, and
further investigation led to the correction of data for eight VOCs (bromochloromethane,
chloroethane, chloromethane, dichlorodifluoromethane, dichlorotetrafluoroethane, propylene,
trichlorofluoromethane, and vinyl chloride) for samples collected near the end of August 2016 or
early September 2016 through the end of the year.
Method-specific completeness is presented at the bottom of Tables 2-10 and 2-11.
Method specific completeness was greater than 85 percent for all methods performed under the
2015 and 2016 NMP monitoring efforts and ranged from 87.10 percent (2015 SNMOC and 2016
hexavalent chromium) to 98.33 percent (2015 hexavalent chromium).
Because methane is not an official part of the NMP, completeness is not provided for
methane in Tables 2-10 and 2-11. Methane samples were collected at two sites in Oklahoma, one
in 2015 (BROK) and two in 2016 (BROK and NROK). Methane completeness for BROK was
89 percent in 2015 and 93 percent in 2016; completeness for NROK was 100 percent in 2016.
This yields a method completeness for methane of 89 percent in 2015 and 96 percent in 2016.
2-35
-------
3.0 Summary of the 2015-2016 National Monitoring Programs Data Treatment and
Methods for Data Analysis
This section summarizes the data treatments
employed and approaches used to analyze the data
generated from samples collected during the 2015-2016
NMP sampling years. These data were analyzed on a
program-wide basis as well as a site-specific basis.
A total of 445,119 valid air toxics concentrations
(including non-detects and analyses for duplicate,
replicate, and collocated samples) were produced from
15,321 valid samples collected at 53 monitoring sites
during the 2015-2016 reporting years. A tabular
presentation of the raw data are found in Appendices B through H and statistical summaries are
presented in Appendices J through P, as shown in Table 3-1.
The 2015-2016 NMP report
includes data from samples
collected at monitoring sites
participating under the NMP and
supported by the national contract
laboratory.
Results from the program-wide
data analyses are presented in
Section 4 while results from the
site-specific data analyses are
presented in the individual state
sections, Sections 5 through 23.
7
Table 3-1. Overview and Organization of Data Presented
Pollutant Group
Number of Sites
Appendix
2015
2016
Raw Data
Statistical Summary
VOCs
27
26
B
J
SNMOCs
9
10
C
K
Methane
1
2
D
L
Carbonyl Compounds
31
31
E
M
PAHs
19
19
F
N
Metals
19
19
G
O
Hexavalent Chromium
1
1
H
P
3.1 Approach to Data Treatment
This section examines the various statistical tools employed to analyze and characterize
the data collected during the 2015-2016 sampling year. Certain data analyses were performed at
the program-level, other data analyses were performed at both the program-level and on a site-
specific basis, and still other approaches were reserved for site-specific data analyses only.
Regardless of the data analysis employed, it is important to understand how the monitoring data
were treated. The following paragraphs describe techniques used to prepare this large quantity of
concentration data for data analysis.
3-1
-------
Considerable care is taken to ensure that there is a single concentration for each target
pollutant per site, sample date, and analytical method. In cases where a site has primary,
duplicate (or collocated), and/or replicate measurements, the primary sample result is used for
data analysis. For instances in which the primary sample was invalid, the duplicate or collocate
results were used. This is referred to as thepreprocesseddaily measurement. This approach
represents a change from past NMP reports, in which the primary, duplicate (or collocated), and
replicate measurements were averaged together to obtain a preprocessed daily measurement.
Concentrations of m,p-xylene and o-x ylene were summed together and are referred to as
"total xylenes," or simply "xylenes" throughout the remainder of this report, with a few
exceptions. One exception is Section 4.1, which examines the results of basic statistical
calculations performed on the dataset. Table 4-1 and Table 4-2, which are the method-specific
statistical summaries for VOCs and SNMOCs, respectively, present the xylenes results retained
as m,p-xylene and o-xylene species. Data for these isomers are also presented individually in the
Data Quality section (Section 24).
For the 2015-2016 NMP, where statistical parameters are calculated based on the
preprocessed daily measurements, zeros have been substituted for non-detect results. This
approach agrees with how data are loaded into AQS, as directedy by the NATTS TAD (EPA,
2009a), and is consistent with other EPA air toxics monitoring programs, such as the School Air
Toxics Monitoring Program (SATMP) (EPA, 2011), and other associated reports, including the
NATTS Network Assessment (EPA, 2017c). The substitution of zeros for non-detects results in
lower average concentrations of pollutants that are rarely measured at or above the associated
MDL and/or have a relatively high MDL.
In order to compare concentrations across multiple sampling methods, all concentrations
have been converted to a common unit of measure: microgram per cubic meter ([j,g/m3).
However, whenever a particular sampling method is isolated from others, such as in Tables 4-1
through 4-7, the statistical parameters are presented in the unit of measure associated with the
particular sampling method. Thus, it is important to pay close attention to the unit of measure
associated with each data analysis discussed in this and subsequent sections of the report.
This report presents various time-based averages to summarize the measurements for a
specific site. Where applicable, quarterly and annual concentration averages were calculated for
each site. The quarterly average concentration of a particular pollutant is simply the average
3-2
-------
concentration of the preprocessed daily measurements over a given calendar quarter. Quarterly
averages include the substitution of zeros for all non-detects. Quarterly averages for the first
quarter of the calendar year include measurements collected in January, February, and March;
the second quarter includes April, May, and June samples; the third quarter includes July,
August, and September samples; and the fourth quarter includes October, November, and
December samples. A minimum of 75 percent of the total number of samples possible within a
given calendar quarter must be valid to have a quarterly average presented. For sites sampling on
a l-in-6 day sampling schedule, 12 samples meet the 75 percent criteria; for sites sampling on a
l-in-12 day schedule, six samples meet the 75 percent criteria. Sites that do not meet this
minimum requirement do not have a quarterly average concentration presented. Sites may not
meet this minimum requirement due to invalidated or missed samples or because of a shortened
sampling duration.
An annual average concentration includes all measured detections and substituted zeros
for non-detects for a given calendar year. Annual average concentrations were calculated for
monitoring sites where three quarterly averages could be calculated and where method
completeness, as presented in Section 2.4, is greater than or equal to 85 percent. Sites that do not
meet these requirements do not have an annual average concentration presented, although site-
specific statistical summaries are provided in the Appendices of this report.
The concentration averages presented in this report are often provided with their
associated 95 percent confidence intervals. Confidence intervals represent the interval within
which the true average concentration falls 95 percent of the time. The confidence interval
includes an equal amount of quantities less than and greater than the concentration average
(EPA, 2011). For example, an average concentration may be written as 1.25 ± 0.25 |ig/m3; thus,
the interval over which the true average would be expected to fall would be between 1.00 |ig/m3
and 1.50 |ig/m3.
3.2 Human Health Risk and the Pollutants of Interest
A practical approach to making an assessment on a large number of air monitoring
measurements is to focus on a subset of pollutants based on the end-use of the dataset. Thus, a
subset of pollutants is selected for further data analyses for each annual NMP report. Health risk-
based calculations have been used to identify "pollutants of interest" for several years, including
the 2015-2016 NMP report. The following paragraphs provide an overview of health risk terms
3-3
-------
and concepts and outline how the pollutants of interest are determined and then used throughout
the remainder of the report.
EPA defines risk as "the probability that damage to life, health, or the environment will
occur as a result of a given hazard (such as exposure to a toxic chemical)" (EPA, 2015a). Human
health risk can be further defined in terms of time. Chronic effects develop from repeated
exposure over long periods of time; acute effects develop from a single exposure or from
exposures over short periods of time (EPA, 2010a). Health risk is also route-specific; that is, risk
varies depending upon route of exposure (i.e., oral vs. inhalation). Because this report covers air
toxics in ambient air, only the inhalation route is considered.
Health risks are typically divided into cancer and noncancer effects when referring to
human health risk. Cancer risk is defined as the likelihood of developing cancer as a result of
exposure to a given concentration over a 70-year period, and is presented as the number of
people at risk for developing cancer per million people. Noncancer health effects are those other
than cancer, including conditions such as asthma. Noncancer health risks are presented as a
hazard quotient (HQ), the ratio of a given concentration level and the value at which adverse
health effects are not expcted. An HQ less than or equal to 1.0 indicates that adverse health
effects are not expected (EPA, 2015a). Cancer risk is presented as a probability while the hazard
quotient is a ratio and thus, a unitless value.
Hazardous air pollutants (HAPs) are those pollutants "known or suspected to cause
cancer or other serious health effects, such as reproductive effects or birth defects, or adverse
environmental effects" (EPA, 2017d). In order to assess health risk, EPA and other agencies
develop toxicity factors, such as cancer unit risk estimates (UREs) and noncancer reference
concentrations (RfCs), to estimate cancer and noncancer risks and to identify (or screen) where
HAP concentrations may present a human health risk. EPA has published a guidance document
outlining a risk-based screening approach for performing an initial screen of ambient air toxics
monitoring datasets (EPA, 2010a). The preliminary risk-based screening process provided in
this report is an adaptation of that approach and is a risk-based methodology for analysts and
interested parties to identify which pollutants may pose a health risk in their area. For this
process, cancer UREs and noncancer RfCs are converted into screening values. The cancer
screening value is the cancer URE converted to |ig/m3 and divided by one million. The
noncancer screening value is one-tenth of the noncancer RfC and converted from milligram per
cubic meter (mg/m3) to |ig/m3. The final screening value used in this report is the lower of the
3-4
-------
two screening values. Not all pollutants analyzed under the NMP have screening values; of the
pollutants sampled under the NMP, 71 pollutants have screening values. The screening values
used in this data analysis are presented in Appendix Q1.
The preprocessed daily measurements of the target pollutants were compared to these
chronic risk screening values in order to identify pollutants of interest across the program. The
following risk-based screening process was used to identify pollutants of interest:
1. The TO-15 and SNMOC methods have 12 pollutants in common. If a pollutant was
measured by both the TO-15 and SNMOC methods at the same site, the TO-15
results were used. (This is because the TO-15 pollutants are calibrated individually
whereas the SNMOC pollutants are based on propane-based response factors.) The
purpose of this data treatment is to have one concentration per pollutant for each
sample day. The exception to this is for RFCO. Concurrent analysis was performed
for RFCO from January 2015 through September 2015, after which, only the
SNMOC analysis continued. Thus, for RFCO, the SNMOC results were used over the
TO-15 results for the 12 pollutants in common between the methods.
2. Each preprocessed daily measurement was compared to its associated risk screening
value. Concentrations that are greater than the risk screening value are described as
"failing the screen."
3. The number of failed screens was summed for each applicable pollutant.
4. The percent contribution of the number of failed screens to the total number of failed
screens program-wide was calculated for each applicable pollutant.
5. The pollutants contributing to the top 95 percent of the total failed screens were
identified as pollutants of interest.
In regards to Step 5 above, the actual cumulative contribution may exceed 95 percent in
order to include all pollutants contributing to the minimum 95 percent criteria (refer to
acenaphthene in Table 4-8 for an example). In addition, if the 95 percent cumulative criterion is
reached, but the next pollutant contributed equally to the number of failed screens, that pollutant
was also designated as a pollutant of interest. Results of the program-wide risk-based screening
process are provided in Section 4.2.
A note regarding measurements of acetonitrile, acrylonitrile, carbon disulfide, and
acrolein: acetonitrile concentrations may be artificially high (or non-existent) due to site
conditions and potential cross-contamination with concurrent sampling of carbonyl compounds
1 The risk-based screening process used in this report is an adaption of guidance from EPA Region 4's report "A
Preliminary Risk-Based Screening Approach for Air Toxics Monitoring Datasets" but the screening values
referenced in that report have since been updated (EPA, 2017e; EPA, 2015b; EPA, 2015c).
3-5
-------
using Method TO-11 A. Similarly, acrylonitrile and carbon disulfide concentrations may also be
artificially high due to potential contamination of the collection systems using Method TO-15.
Additionally, questions about the consistency and reliability of acrolein measurements have been
raised during other monitoring projects, such as the SATMP (EPA, 2010b). The inclusion of
acetonitrile, acrylonitrile, carbon disulfide, and acrolein in data analyses must be determined on a
site-specific basis by the agency responsible for the site. Thus, results for these pollutants are
excluded from program-wide and site-specific data analyses related to risk.
The NATTS TAD (EPA, 2009a) identifies 19 "MQO Core Analytes" that participating
sites are required to sample and analyze for under the NATTS program. Table 3-2 presents these
MQO Core Analytes. Monitoring for these pollutants is required because they are major health
risk drivers according to EPA (EPA, 2009a). Many of the pollutants listed in Table 3-2 are
identified as pollutants of interest via the risk-based screening process. Note that hexavalent
chromium was removed from the list of required pollutants for which to sample under the
NATTS program beginning in July 2013. As a result, many NATTS sites discontinued sampling
hexavalent chromium. During the 2015 and 2016 sampling years, RIVA was the only NATTS
site at which sampling for this pollutant continued; however, sampling for hexavalent chromium
at RIVA was discontinued in June 2016.
The "pollutants of interest" designation is reserved for pollutants targeted for sampling
through the NMP that meet the identified criteria. As discussed in Section 2.0, agencies
operating monitoring sites that participate under the UATMP, NATTS, or CSATAM
programs are not required to have their samples analyzed by ERG or may measure pollutants
other than those targeted under the NMP. In these cases, data are generated by sources other than
ERG and are not included in the preliminary risk-based screening process or any other data
analysis contained in this report.
3-6
-------
Table 3-2. NATTS MQO Core Analytes
Pollutant
Class/Method
Acrolein
Benzene
1.3 -Butadiene
Carbon Tetrachloride
VOCs/TO-15
Chloroform
T etrachloroethylene
T richloroethylene
Vinyl Chloride
Acetaldehyde
Carbonyl Compounds/
Formaldehyde
TO-11A
Naphthalene
PAHs/
Benzo(a)pyrene
TO-13A
Arsenic
Beryllium
Cadmium
Metals/
10-3.5 and EQL-0512-
201/202
Manganese
Lead
Nickel
Hexavalent chromium1
Metals/ASTM D7614
1 Hexavalent chromium was removed from the Core Analytes list in July 2013.
3.3 Additional Program-Level Data Analyses of the 2015-2016 National Monitoring
Programs Dataset
This section summarizes additional data analyses performed on the 2015-2016 NMP
dataset at the program level. Additional program-level analyses include a review of how
concentrations vary among the sites and from quarter-to-quarter. The results of these data
analyses are presented in Section 4.2.
Variability refers to the degree of difference among values in a dataset. Two types of
variability are analyzed for this report. The first type of variability assessed in this report is inter-
site variability. For this data analysis, the annual average concentrations for each site are plotted
in the form of a bar graph for each program-wide pollutant of interest. The criteria for calculating
an annual average concentration are discussed in Section 3.1 and sites that do not meet these
requirements do not have an annual average concentration presented. This assessment allows the
reader to visualize how concentrations varied across the sites for a particular pollutant of interest.
In order to further this data analysis, the program-level average concentration for each polluant,
as presented in Tables 4-1 through 4-7 in Section 4.1, is plotted against the site-specific annual
3-7
-------
averages. This allows the reader to see how the site-specific annual averages compare to the
program-level average for each pollutant. Note that the average concentrations shown for VOCs,
SNMOCs, and carbonyl compounds in Tables 4-1, 4-2, and 4-4 are presented in method-specific
units, but have been converted to a common unit of measurement (|ig/m3) for the purposes of this
data analysis.
Quarterly variability is the second type of variability assessed in this report. The
concentration data for each site were divided into the four quarters of each year, as described in
Section 3.1. The completeness criteria, also described in Section 3.1, are maintained here as well.
The site-specific quarterly average concentrations are illustrated by bar graphs for each program-
level pollutant of interest. This data analysis allows for the potential determination of a quarterly
(or seasonal) correlation with the magnitude of concentrations for a specific pollutant.
3.4 Additional Site-Specific Data Analyses
In addition to the analyses described in the preceding sections, the state-specific sections
contain additional analyses that are applicable only at the local level. This section provides an
overview of these analyses but does not discuss their results. Results of these site-specific data
analyses are presented in the individual state-specific sections (Sections 5 through 23).
3.4.1 Site Characterization
For each site participating in the 2015-2016 NMP, a site characterization was performed.
This characterization includes a review of the nearby area surrounding the monitoring site; the
plotting of emissions sources surrounding the monitoring site; and providing traffic data and
other characterizing information. For the 2015-2016 NMP report, the locations of point sources
located near the monitoring sites were obtained from Version 1 of the 2014 NEI (EPA, 2016).
Sources for other site-characterizing data are provided in the individual state sections.
3.4.2 Preliminary Risk-Based Screening and Pollutants of Interest
The preliminary risk-based screening process described in Section 3.2 and applied at the
program-level was also completed for each individual monitoring site to determine site-specific
pollutants of interest. Once these were determined, the time-period averages (quarterly and
annual) described in Section 3.1 were calculated for each site and were used for various data
analyses at the site-specific level, as described below:
• Comparison to the program-level average concentrations
• Trends analysis
3-8
-------
• The calculation of cancer risk and noncancer hazard approximations, including the
emission tracer analysis
• Risk-based emissions assessment.
3.4.2.1 Site-Specific Comparison to Program-level Average Concentrations
To better understand how an individual site's measurements compare to the program-
level results, as presented in Section 4.1 in Tables 4-1 through 4-7, the site-specific and program-
level concentrations are presented together graphically for each site-specific pollutant of interest
identified via the risk-based screening process. This data analysis is an extension of the data
analysis discussed in Section 3.3 and utilizes box and whisker plots, or simply box plots, to
visually show this comparison. These box plots were created in Microsoft Excel, using the
Peltier Tech Charts for Excel 3.0 utility (Peltier, 2016). Note that for sites sampling VOCs (or
SNMOCs), pollutants are shown only in comparison to other sites sampling VOCs (or SNMOCs)
to match the program-level averages presented in Section 4.1 in Tables 4-1 and 4-2.
The box plots used in this data analysis overlay the site-specific minimum, annual
average, and maximum concentrations over several program-level statistical metrics. For the
program-level statistics, the first, second (median), third, and fourth (maximum) quartiles are
shown as colored segments on a "bar" where the color changes correspond to the exact
numerical value of the quartile. The thin vertical line represents the program-level average
concentration. The site-specific annual average is shown as a black (2015) or white (2016) circle
plotted on top of the bar and the horizontal lines extending outward from the circles represent the
minimum and maximum concentration measured at the site. An example of this figure is shown
in Figure 5-6. Note that the program-level average concentrations shown for VOCs, SNMOCs,
and carbonyl compounds in Tables 4-1, 4-2, and 4-4 are presented in method-specific units, but
have been converted to a common unit of measurement (|ig/m3) for the purposes of this data
analysis. These graphs are presented in Sections 5 through 23, and are grouped by pollutant
within each state section. This allows for both a "site vs. program" comparison as well as an
inter-site comparison for sites within a given state.
3.4.2.2 Site Trends Analysis
Table 2-1 presents current monitoring sites that have participated in the NMP in previous
years. A site-specific trends analysis was conducted for sites with at least 5 consecutive years of
method-specific data analyzed under the NMP. The trends analysis was conducted for each of
the site-specific pollutants of interest identified via the risk-based screening process. Forty-five
3-9
-------
of the 53 monitoring sites have sampled at least one pollutant group long enough for the trends
analysis to be conducted. The approach to this trends analysis is described below and the results
are presented in the individual state sections (Sections 5 through 23).
Five individual 1-year statistical metrics were calculated for this data analysis and are
presented as box and whisker plots, an example of which can be seen in Figure 5-17. The
statistical metrics shown include the minimum and maximum concentration measured during
each year of sampling (as shown by the upper and lower value of the lines extending from the
box); the 5th percentile, 50th percentile (or median), and 95th percentile (as shown by the
y-values corresponding with the bottom of the box, the thick blue line, and top of the box,
respectively); and the average (or mean) concentration (as denoted by the orange diamond). Each
of the statistical metrics incorporates all measurements collected during that 1-year period. For
each 1-year period, there must be a minimum of 85 percent completeness, which corresponds to
roughly 51 valid samples or approximately 10 months of sampling (for a site sampling on a
l-in-6 day sampling schedule) for an average concentration to be presented. For cases where
sampling began mid-year or ended early, a minimum of 6 months of sampling is required. In
these cases, a 1-year average is not provided but the concentration range and quartiles are still
presented.
Historical data used in this analysis were downloaded from EPA's AQS database (EPA,
2017b) in order to ensure the use of the most up-to-date data available. Similar to other analyses
presented in this report, zeros representing non-detects were incorporated into the statistical
calculations.
In NMP reports prior to 2014, results from sample days with precision data (duplicates,
collocates, and/or replicates) were averaged together to allow for the determination of a single
concentration per pollutant and date for each site. For the 2014 NMP report, duplicate and
replicate data were not downloaded from AQS due to a change in the availability of this data in
AQS. However, for collocated results, the averaging schema was retained. This is also true for
the 2015-2016 NMP report.
3-10
-------
3.4.2.3 Cancer Risk and Noncancer Hazard Approximations
Risk was further examined by calculating cancer risk and noncancer hazard
approximations for each of the site-specific pollutants of interest. The cancer risk approximations
presented in this report estimate the cancer risk due to exposure to a given pollutant at the annual
average concentration over a 70-year period (not the risk resulting from exposure over the time
period covered in this report). A cancer risk approximation less than 1 in-a-million is considered
negligible; a cancer risk greater than 1 in-a-million but less than 100 in-a-million is generally
considered acceptable; and a cancer risk greater than 100 in-a-million is considered significant
(EPA, 2009b). The noncancer hazard approximation is presented as the Noncancer Hazard
Quotient (HQ), which is a unitless value. According to EPA, "A hazard quotient less than or
equal to one indicates that adverse noncancer effects are not likely to occur, and thus can be
considered to have negligible hazard." (EPA, 2015a).
The toxicity factors applied to calculate the cancer risk and noncancer hazard
approximations are typically UREs (for cancer) or RfCs (for noncancer), which are developed by
EPA. However, UREs and RfCs are not available for all pollutants. In the absence of EPA
values, toxicity factors developed by agencies with credible methods and that are similar in
scope and definition were used (EPA, 2015b). Cancer URE and noncancer RfC toxicity factors
can be applied to the annual average concentrations to approximate risk based on ambient
monitoring data. While the cancer risk and noncancer hazard approximations do not incorporate
human activity patterns and therefore do not reflect true human inhalation exposure, they may
allow analysts to further refine their focus by identifying concentrations of specific pollutants
that may present health risks. Cancer UREs and/or noncancer RfCs, site-specific annual
averages, and corresponding annual average-based cancer risk and noncancer hazard
approximations are presented in each state section (Sections 5 through 23).
To further this data analysis, pollution roses were created for each of the site-specific
pollutants of interest that have cancer risk approximations greater than 75 in-a-million and/or a
noncancer hazard approximation greater than 1.0, where applicable. This data analysis is
performed to help identify the geographical area where the emissions sources of these pollutants
may have originated. A pollution rose is a plot of the ambient concentration versus the wind
speed and direction; high concentrations may be shown in relation to the direction of potential
emissions sources.
3-11
-------
There are, however, limitations to this data analysis. Wind data are typically obtained
from the National Centers for Environmental Information (NCEI), part of the National Oceanic
and Atmospheric Administration (NOAA), for the nearest observation station (NOAA, 2017).
These are hourly observations while concentrations from this report are 24-hour measurements.
Thus, the wind data must be averaged for comparison to the concentrations data. Wind speed and
direction can fluctuate throughout a given day or change dramatically if a frontal system moves
through. Thus, the average calculated wind data may not be completely representative of a given
day. This can be investigated more thoroughly if the need arises.
3.4.2.4 Risk-Based Emissions Assessment
A pollutant emitted in high quantities does not necessarily present a higher risk to human
health than a pollutant emitted in very low quantities. The more toxic the pollutant, the more risk
associated with its emissions in ambient air. The development of various health-based toxicity
factors, as discussed in previous sections, has allowed analysts to apply weight to the emissions
of pollutants based on toxicity rather than mass emissions. This approach considers both a
pollutant's toxicity potential and the quantity emitted.
This assessment compares county-level emissions to toxicity-weighted emissions based
on the EPA-approved approach described below (EPA, 2007). The 10 pollutants with the highest
total mass emissions and the 10 pollutants with the highest associated toxicity-weighted
emissions for pollutants with cancer and noncancer toxicity factors are presented in each state
section. While the absolute magnitude of the pollutant-specific toxicity-weighted emissions is
not meaningful, the relative magnitude of toxicity-weighted emissions is useful in identifying the
order of potential priority for air quality managers. Higher values suggest greater priority;
however, even the highest values may not reflect potential cancer effects greater than the level of
concern (100 in-a-million) or potential noncancer effects above the level of concern
(e.g., HQ greater than or equal to 1.0). The pollutants exhibiting the 10 highest annual average-
based risk approximations for cancer and noncancer effects are also presented in each state
section. The results of this data analysis may help state, local, and tribal agencies better
understand which pollutants emitted, from a toxicity basis, are of the greatest concern and
whether or not these pollutants are already being monitoring or perhaps should be monitored in
the future.
3-12
-------
The toxicity-weighted emissions approach consists of the following steps:
1. Obtain HAP emissions data for all anthropogenic sectors (nonpoint, point, onroad,
and nonroad) from the NEI. For point sources, sum the process-level emissions to the
county-level. Biogenic emissions are not included in this data analysis.
2. Apply the mass extraction speciation profiles to extract metal and cyanide mass.
3. Apply weight to the emissions derived from the steps above based on their toxicity.
The results of the toxicity-weighting process are unitless.
a. To apply weight based on cancer toxicity, multiply the emissions of each
pollutant by its cancer URE.
b. To apply weight based on noncancer toxicity, divide the emissions of each
pollutant by its noncancer RfC.
The PAHs measured using Method TO-13A are a sub-group of Polycyclic Organic
Matter (POM). Because these compounds are often not speciated into individual compounds in
the NEI, the PAHs are grouped into POM Groups in order to assess risk attributable to these
pollutants. Thus, emissions data and toxicity-weighted emissions for many of the PAHs are
presented by POM Groups for this data analysis. Table 3-3 presents the 22 PAHs measured with
Method TO-13A and their associated POM Groups, if applicable.
The POM groups are sub-grouped in Table 3-3 because toxicity research has led to the
refining of UREs for certain PAHs. With the release of the 2011 NAT A, the POM Groups have
been renamed, although the grouping is still based on the same risk levels. For simplicity's sake,
the original names are provided in the data analysis, but both names are provided in Table 3-3.
Note the following in regard to Table 3-3:
• naphthalene emissions are reported to the NEI individually; therefore, it is not
included in one of the POM Groups;
• four pollutants analyzed using Method TO-13 A and listed in Table 3-3 do not have
assigned POM Groups;
• anthracene, phenanthrene, and pyrene used to be part of POM Group 2 (2d) but have
been removed.
3-13
-------
Table 3-3. POM Groups for PAHs1
Pollutant
POM
Group
POM
Subgroup
New POM
Grouping
Acenaphthene
Group 2
Group 2b
PAH 880E5
Acenaphthylene
Group 2
Group 2b
PAH 880E5
Anthracene
NA
PAH 000E0
Benzo(a)anthracene
Group 6
PAH 176E4
Benzo(a)pyrene
Group 5
Group5a
PAH 176E3
Benzo(b)fluoranthene
Group 6
PAH 176E4
Benzo(e)pyrene
Group 2
Group 2b
PAH 880E5
Bcnzo(g.h.i)pcrvlcnc
Group 2
Group 2b
PAH 880E5
Benzo(k)fluoranthene
Group 6
PAH 176E4
Clirysene
Group 7
PAH 176E5
Coronene
NA
Cyclopenta | cdlpyrene
NA
Dibenz(a,h)anthracene
Group 5
Group5b
PAH 192E3
Fluoranthene
Group 2
Group 2b
PAH 880E5
Fluorene
Group 2
Group 2b
PAH 880E5
9-Fluorenone
NA
Indeno( 1,2,3 -cd)pyrene
Group 6
PAH 176E4
Naphthalene*
NA
Perylene
Group 2
Group 2b
PAH 880E5
Phenantlirene
NA
PAH 000E0
Pyrene
NA
PAH 000E0
Retene
NA
1 Reference: EPA, 2015c
* Emissions for naphthalene are reported to the NEI individually; therefore, naphthalene is not
included in one of the POM Groups.
NA = POM Group not assigned.
3-14
-------
4.0 Summary of the 2015-2016 National Monitoring Programs Data
This section summarizes the results of the data analyses performed on the NMP dataset,
as described in Section 3.
4.1 Statistical Results
This section examines the following statistical parameters for the target pollutants of each
analytical method: 1) detection rates, 2) concentration ranges and data distribution, and 3) central
tendency statistics. Tables 4-1 through 4-6 present statistical summaries for the target pollutants
and Sections 4.1.1 through 4.1.3 review the basic findings of these statistical calculations.
The 2015-2016 NMP report includes data from samples collected at monitoring sites
participating under the NMP with the support of the national contract laboratory.
p
4.1.1 Target Pollutant Detection Rates
There is an experimentally-determined MDL for every target pollutant, as described in
Section 2.2. Quantification less than the MDL is possible, although the measurement's reliability
is lower. If a concentration does not exceed the MDL, it does not mean that the pollutant is not
present in the air. If the instrument does not generate a numerical concentration, the
measurement is marked as "ND," or "non-detect." As explained in Section 2.2, data analysts
should exercise caution when interpreting monitoring data with a high percentage of reported
concentrations at levels near or less than the corresponding MDL. A thorough review of the
number of measured detections, the number of non-detects, and the total number of samples is
beneficial to understanding the representativeness of the interpretations made.
Tables 4-1 through 4-7 summarize the number of times each target pollutant was detected
out of the number of valid samples collected and analyzed. Approximately 54 percent of the
reported measurements (based on the preprocessed daily measurements) were equal to or greater
than their respective MDLs across the program. The following list provides the percentage of
measurements that were greater than the MDLs for each of the target pollutant groups:
• 40 percent for VOCs
• 52 percent for SNMOCs
• 100 percent for methane
• 83 percent for carbonyl compounds
4-1
-------
• 61 percent for PAHs
• 85 percent for metals
• 69 percent for hexavalent chromium.
Some pollutants were detected in every valid sample collected while others were
infrequently detected or not detected at all. Ten VOCs (benzene, carbon disulfide, carbon
tetrachloride, chloromethane, dichlorodifluoromethane, dichloromethane, propylene, toluene,
trichlorofluoromethane, and trichlorotrifluoroethane) were detected in every valid VOC sample
collected (2,934), based on the preprocessed daily measurements across both years of sampling.
Eight pollutants (acetylene, //-butane, ethane, ethylene, //-heptane, propane, propylene, and
toluene) were detected in every valid SNMOC sample collected (854). Methane was detected in
all 134 samples collected. Formaldehyde and acetone were detected in every valid carbonyl
compound sample collected (3,157). Fluoranthene, naphthalene, phenanthrene, and pyrene were
detected in every valid PAH sample collected (2,239). Antimony, cadmium, cobalt, lead, and
manganese were detected in every valid speciated metals sample collected (2,146). Hexavalent
chromium was detected in 59 of the 84 valid samples collected.
BTUT and NBIL have the greatest number of measured detections by a considerable
margin (more than 13,000 for each). But they are the only two NMP sites that collected samples
for at least five analytical methods/pollutant groups. They are also among the five sites to sample
both VOCs and SNMOCs but are the only two to sample throughout both 2015 and 2016.
However, the detection rates for BTUT and NBIL (67 percent and 65 percent, respectively) were
not as high as other sites. The detection rate at a given site may result from multiple factors that
must be considered when reviewing such a metric. Concentration levels are the primary factor,
certainly, although which pollutant groups are sampled for also plays a role. For example, metals
were rarely reported as non-detects. As a result, sites that sampled only metals (such as
ASKY-M and PAFL) would be expected to have higher detection rates. The detection rate for
each of these sites is nearly 100 percent. Conversely, VOCs had one of the lowest percentages of
concentrations greater than the MDLs (40 percent). A site measuring only VOCs would be
expected to have relatively low detection rates, such as ASKY or SPAZ (both approximately
53 percent).
4-2
-------
Table 4-1. Statistical Summaries of the VOC Concentrations
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
¦k
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
Table 4-1. Statistical Summaries of the VOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations
On
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-Pi nc nc3
844
5
0
1.21
8.30
0.019
0
0
0
0.323
Propane
0
854
0
1.05
476
32.6
17.1
8.52
31.7
49.2
w-Propylbenzene
557
297
26
0.049
2.02
0.090
0
0
0.157
0.165
Propylene
0
854
0
0.169
5.56
0.963
0.784
0.523
1.20
0.635
Propyne
832
22
0
0.062
0.258
0.003
0
0
0
0.022
Styrene3
590
112
35
0.108
21.8
0.626
0
0
0
2.24
Toluene
0
854
0
0.573
255
5.93
3.53
2.28
5.66
14.9
n-Tridecane
658
196
154
0.086
46.4
0.248
0
0
0
2.14
1-Tridecene
854
0
0
Not Detected
1,2,3 -T rimethy lbenzene
399
455
240
0.078
6.65
0.163
0.100
0
0.213
0.326
1,2,4-Trimethy lbenzene
14
840
186
0.088
19.6
0.805
0.603
0.395
0.979
0.898
1.3.5 -T rimethy lbenzene
315
539
105
0.074
3.37
0.209
0.172
0
0.333
0.246
2,2,3-Trimethylpentane
779
75
1
0.046
0.812
0.025
0
0
0
0.095
2,2,4-Trimethy lpentane3
302
488
1
0.072
79.3
0.685
0.306
0
0.768
2.93
2,3,4-Trimethy lpentane3
167
685
61
0.057
3.62
0.293
0.212
0.096
0.364
0.342
n-Undecane
203
651
474
0.066
16.8
0.271
0.171
0.087
0.292
0.691
1-Undecene
779
75
66
0.070
2.65
0.024
0
0
0
0.139
w-Xylene/p-Xylene
3
851
4
0.120
9.03
1.45
1.18
0.677
1.90
1.07
1 Out of 854 valid samples.
2 Excludes zeros for non-detects.
3 The total number of concentrations may not add up to 854 for some compounds where no value could be reported due to co-elution.
NA = Not applicable for these parameters.
-------
Table 4-2. Statistical Summaries of the SNMOC Concentrations (Continued)
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-3. Statistical Summaries of the Methane Concentrations1
#
# of
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects
Detections2
-------
Table 4-4. Statistical Summaries of the Carbonyl Compound Concentrations
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-5. Statistical Summaries of the PAH Concentrations
# of
#
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum2
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects1
Detections1
-------
Table 4-6. Statistical Summaries of the Metals Concentrations
Pollutant
#
of Non-
Detects1
# of
Measured
Detections1
# of
Measured
Detections
-------
Table 4-7. Statistical Summary of the Hexavalent Chromium Concentrations1
#
# of
# of
Measured
Arithmetic
First
Third
Standard
of Non-
Measured
Detections
Minimum3
Maximum
Mean
Median
Quartile
Quartile
Deviation
Pollutant
Detects2
Detections2
-------
4.1.2 Concentration Range and Data Distribution
The concentrations measured during the 2015-2016 NMP exhibit a wide range of
variability. The minimum and maximum concentrations measured (excluding zeros substituted
for non-detects) for each target pollutant are presented in Tables 4-1 through 4-7 (in respective
pollutant group units). Some pollutants, such as dichloromethane, were measured across a wide
range of concentrations, while other pollutants, such as dichlorotetrafluoroethane, were not, even
though they were both detected frequently. For each method-specific pollutant group, the
pollutant with the largest range in concentrations measured is as follows:
• For VOCs, acetonitrile (0.026 ppbv to 1,940 ppbv)
• For SNMOCs, 2-methylpentane (0.317 ppbC to 483 ppbC)
• For methane, concentrations ranged from 1.94 ppmC to 4.01 ppmC
• For carbonyl compounds, formaldehyde (0.017 ppbv to 20.7 ppbv)
• For PAHs, naphthalene (0.446 ng/m3 to 403 ng/m3)
• For metals in PMio, manganese (0.005 ng/m3 to 202 ng/m3)
• For metals in TSP, manganese (1.91 ng/m3 to 92.5 ng/m3)
• For hexavalent chromium, concentrations ranged from 0.0025 ng/m3 to 0.116 ng/m3.
4.1.3 Central Tendency
In addition to the number of measured detections and the concentration ranges,
Tables 4-1 through 4-7 also present several central tendency and data distribution statistics
(arithmetic mean or average, median, first and third quartiles, and standard deviation) for each of
the pollutants measured during the 2015-2016 NMP, in respective pollutant group units. A
multitude of observations can be made from these tables. The pollutants with the three highest
average concentrations for each method-specific pollutant group are provided below, with
respective confidence intervals (the 95 percent confidence intervals are not provided in the
tables).
The three VOCs with the highest average concentrations, as presented in Table 4-1, are:
• Acetonitrile (5.63 ± 1.50 ppbv)
• Acetylene (0.992 ± 0.069 ppbv)
• Dichloromethane (0.975 ± 0.436 ppbv).
4-15
-------
The three SNMOCs with the highest average concentrations, as presented in Table 4-2,
are:
• Ethane (44.8 ± 3.73 ppbC)
• Propane (32.6 ±3.31 ppbC)
• w-Butane (16.6 ± 1.83 ppbC).
The average concentration of methane, as presented in Table 4-3, is 2.52 ± 0.07 ppmC.
The three carbonyl compounds with the highest average concentrations, as presented
in Table 4-4, are:
• Formaldehyde (2.47 ± 0.07 ppbv)
• Acetone (1.19 ± 0.04 ppbv).
• Acetaldehyde (0.923 ± 0.023 ppbv).
The three PAHs with the highest average concentrations, as presented in Tables 4-5, are:
• Naphthalene (61.2 ± 1.96 ng/m3)
• Phenanthrene (10.7 ± 0.84 ng/m3)
• Acenaphthene (4.36 ± 0.73 ng/m3).
The three metals with the highest average concentrations for both PMio and TSP
fractions, as presented in Table 4-6, are;
• Manganese (PMio = 8.51 ± 0.51 ng/m3, TSP = 17.8 ± 1.78 ng/m3)
• Total chromium (PMio = 3.51 ± 0.21 ng/m3, TSP = 4.01 ± 0.47 ng/m3)
• Lead (PMio = 3.07 ± 0.17 ng/m3, TSP = 3.28 ± 0.31 ng/m3).
The average concentration of hexavalent chromium, as presented in Table 4-7, is
0.011 ± 0.004 ng/m3.
Appendices J through P present statistical calculations on a site-specific basis, like those
presented in Tables 4-1 through 4-7.
4-16
-------
4.2 Preliminary Risk-Based Screening and Pollutants of Interest
Based on the preliminary risk-based screening process described in Section 3.2, Table 4-8
identifies the pollutants that failed at least one screen; summarizes each pollutant's total failed
screens, total number of measured detections, percentage of screens failed, and cumulative
percentage of failed screens; and highlights those pollutants contributing to the top 95 percent of
failed screens (shaded in gray) and thereby designated as program-wide pollutants of interest.
The results in this table are provided over both years of sampling. The number of failed screens,
the number of measured detections, and the failure rate must all be considered when reviewing
the results of the preliminary risk-based screening process.
The results in Table 4-8 are listed in descending order by number of screens failed.
Table 4-8 shows that benzene failed the greatest number of screens (3,403). Carbon tetrachloride,
formaldehyde, acetaldehyde, and 1,2-dichloroethane each failed greater than 2,700 screens. Each
of these pollutants were among those with the greatest number of measured detections, among
pollutants shown in Table 4-8, each with a detection rate greater than 95 percent. Seven
pollutants listed in Table 4-8 failed only one screen each (1,2-dichloropropane, antimony,
benzo(a)anthracene, benzo(b)fluoranthene, beryllium, dibenz(a,h)anthracene, and hexavalent
chromium). The number of measured detections for these pollutants varied significantly. Several
of these pollutants were detected in greater than 2,000 samples each, while 1,2-dichloropropane
was detected in fewer than 6 percent of sample collected. Two pollutants exhibited a failure rate
of 100 percent (1,2-dibromoethane and chloroprene). These pollutants were infrequently detected
(1,2-dibromoethane and chloroprene were detected in 3 percent and less than 1 percent of
samples collected, respectively).
4-17
-------
Table 4-8. Results of the Program-Level Preliminary Risk-Based Screening Process
Screening
# of
# of
%of
%of
Cumulative
Value
Failed
Measured
Failed
Total
%
Pollutant
(Hg/m3)
Screens
Detections
Screens
Failures
Contribution
Benzene
0.13
3,403
3,406
99.91
14.29
14.29
Formaldehyde
0.077
3,150
3,157
99.78
13.23
27.52
Acetaldehyde
0.45
3,005
3,155
95.25
12.62
40.13
Carbon Tetrachloride
0.17
2,927
2,934
99.76
12.29
52.42
1,2-Dichloroethane
0.038
2,720
2,741
99.23
11.42
63.84
1.3 -Butadiene
0.03
2,385
2,812
84.82
10.01
73.86
Arsenic
0.00023
1,910
2,141
89.21
8.02
81.88
Naphthalene
0.029
1,653
2,239
73.83
6.94
88.82
Ethylbenzene
0.4
525
3,338
15.73
2.20
91.02
p-Dichlorobenzene
0.091
388
1,266
30.65
1.63
92.65
Hexachloro-1,3 -butadiene
0.045
298
343
86.88
1.25
93.90
Fluorene
0.011
199
1,592
12.50
0.84
94.74
Acenaphthene
0.011
197
1,829
10.77
0.83
95.57
Nickel
0.0021
183
2,144
8.54
0.77
96.33
Vinyl chloride
0.11
161
914
17.61
0.68
97.01
Manganese
0.03
127
2,146
5.92
0.53
97.54
1,2-Dibromoethane
0.0017
97
97
100.00
0.41
97.95
Fluoranthene
0.011
87
2,239
3.89
0.37
98.32
Propionaldehyde
0.8
87
3,098
2.81
0.37
98.68
Tricliloroethylene
0.2
62
627
9.89
0.26
98.94
1.1,2-Trichloroethane
0.0625
51
88
57.95
0.21
99.16
Cadmium
0.00056
46
2,146
2.14
0.19
99.35
Benzo(a)pyrene
0.00057
44
1,788
2.46
0.18
99.53
Lead
0.015
29
2,146
1.35
0.12
99.66
Bro mo methane
0.5
28
2,898
0.97
0.12
99.77
Dichloromethane
60
23
2,934
0.78
0.10
99.87
Cliloroform
9.8
8
2,881
0.28
0.03
99.90
Acenaphthylene
0.011
4
1,252
0.32
0.02
99.92
Cliloroprene
0.0021
4
4
100.00
0.02
99.94
1,1 -Dichloroethane
0.625
3
193
1.55
0.01
99.95
Tetracliloroethylene
3.8
3
2,298
0.13
0.01
99.96
Xylenes
10
2
3,406
0.06
0.01
99.97
1,2-Dichloropropane
0.4
1
165
0.61
0.00
99.97
Antimony
0.02
1
2,146
0.05
0.00
99.98
Benzo(a)antliracene
0.0057
1
2,054
0.05
0.00
99.98
Benzo(b)fluoranthene
0.0057
1
2,116
0.05
0.00
99.99
Beryllium
0.00042
1
2,046
0.05
0.00
99.99
Dibenz(a,h)anthracene
0.00052
1
1,367
0.07
0.00
100.00
Hexavalent Chromium
0.000083
1
59
1.69
0.00
100.00
Total
23,816
74,205
32.09
4-18
-------
The program-level pollutants of interest, as indicated by the shading in Table 4-8, are
identified as follows:
• Acenaphthene • 1,2-Dichloroethane
• Acetaldehyde • Ethylbenzene
• Arsenic • Fluorene
• Benzene • Formaldehyde
• 1,3-Butadiene • Hexachloro-1,3-butadiene
• Carbon Tetrachloride • Naphthalene.
• /;-Dichlorobenzene
The pollutants of interest identified via the preliminary risk-based screening process for
2015 and 2016 are similar to the pollutants identified in previous years. Nickel is the only
pollutant that was a program-wide pollutant of interest for in the 2014 NMP report but is not on
the list for the 2015-2016 report. Nickel is the first pollutant just outside the 95 percent criteria,
as shown in Table 4-8, and therefore is not a pollutant of interest for 2015-2016. Acenaphthene
and fluorene were not on the list for 2014 but are for 2015-2016. Both of these have been
identified as pollutants of interest in previous reports.
Of the pollutants that have corresponding screening values, concentrations of 39
pollutants failed at least one screen. Of these, a total of 23,816 concentrations out of 74,205
concentrations (or 32 percent) failed screens. If all pollutants with screening values are
considered (including those that did not fail any screens), the percentage of concentrations failing
screens is less (23,816 of 111,639, or 21 percent). Note that these percentages exclude acrolein,
acetonitrile, acrylonitrile, and carbon disulfide measurements per the explanations provided in
Section 3.2; these pollutants are excluded from all risk-related analyses contained in the report
from this point forward.
Table 4-9 presents the total number of failed screens per site, in descending order, as a
means of comparing the results of the preliminary risk-based screening process across the sites.
In addition to the number of failed screens, Table 4-9 also provides the total number of screens
conducted (one screen per valid preprocessed daily measurement for each site for all pollutants
with screening values). The failure rate, as a percentage, was determined from the number of
4-19
-------
failed screens and the total number of screens conducted (based on applicable measured
detections) and is also provided in Table 4-9.
Table 4-9. Site-Specific Risk-Based Screening Comparison
Site
# of
Failed
Screens
Total # of
Measured
Detections1
%of
Failed
Screens
# of Pollutant
Groups
Analyzed
S4MO
1,101
5,417
20.32
4
PXSS
1,047
4,862
21.53
4
NBIL
1,023
5,171
19.78
5
TOOK
990
3,546
27.92
3
DEMI
959
4,119
23.28
3
GPCO
955
5,068
18.84
4
TMOK
939
3,566
26.33
3
TROK
938
3,582
26.19
3
BTUT
913
4,672
19.54
5
SEWA
841
4,705
17.87
4
OCOK
825
3,519
23.44
3
CSNJ
819
2,499
32.77
2
YUOK
818
3,470
23.57
3
SPIL
783
2,419
32.37
2
ELNJ
776
2,406
32.25
2
GLKY
770
4,578
16.82
4
BLKY
610
3,158
19.32
2
CHNJ
556
2,000
27.80
2
TVKY
551
2,150
25.63
1
BROK
547
1,816
30.12
4
ATKY
534
2,160
24.72
1
ASKY
476
1,960
24.29
1
LEKY
429
2,307
18.60
2
NRNJ
374
1,319
28.35
2
NBNJ
344
1,156
29.76
2
SPAZ
341
1,017
33.53
1
SKFL
316
1,947
16.23
2
RICO
308
847
36.36
2
NROK
276
821
33.62
4
INDEM
238
357
66.67
1
WPIN
232
348
66.67
1
AZFL
231
357
64.71
1
SYFL
226
335
67.46
1
ROCH
207
1,700
12.18
1
ASKY-M
197
1,097
17.96
1
ORFL
196
294
66.67
1
PACO
196
737
26.59
2
BOMA
194
2,848
6.81
2
1 Total number of measured detections for all pollutants with screening
values, not just those failing screens. Also excludes acrolein,
acetonitrile, acrylonitrile, and carbon disulfide results.
BOLD ITALICS = EPA-designated NATTS Site
4-20
-------
Table 4-9. Site-Specific Risk-Based Screening Comparison (Continued)
# of
Total # of
%of
# of Pollutant
Failed
Measured
Failed
Groups
Site
Screens
Detections1
Screens
Analyzed
ROIL
194
620
31.29
2
SJJCA
186
2,501
7.44
2
BRCO
179
681
26.28
2
BXNY
163
1,759
9.27
1
GSCO
137
438
31.28
2
CELA
117
1,557
7.51
1
BAKY
113
1,134
9.96
1
WADC
109
1,534
7.11
1
RIVA
106
1,484
7.14
2
PRRI
105
1,684
6.24
1
RFCO
96
464
20.69
2
BMCO
93
347
26.80
2
RUCA
85
1,470
5.78
1
PAFL
55
529
10.40
1
UNVT
2
1,107
0.18
1
Total
23,816
111,639
21.33
1 Total number of measured detections for all pollutants with screening
values, not just those failing screens. Also excludes acrolein,
acetonitrile, acrylonitrile, and carbon disulfide results.
BOLD ITALICS = EPA-designated NATTS Site
As shown, S4MO has the largest number of failed screens (1,101), followed by PXSS
(1,047); these two sites also had the largest number of failed screens in 2014. Conversely,
concentrations measured at UNVT failed relatively few screens (2). Every NMP site had at least
one concentration fail a screen. The total number of screens and the number of pollutant groups
measured by each site must be considered when interpreting the results in Table 4-9. Sites
sampling four or five pollutant groups tended to have a higher number of failed screens due, at
least in part, to the higher number of pollutants (with screening values) sampled. For sites
sampling only one or two pollutant groups, it depends on the pollutant group sampled as the
number of pollutants analyzed varies from one (hexavalent chromium) to 80 (SNMOCs).
Although SYFL, ORFL, INDEM and WPIN have the highest failure rates (67 percent each),
these sites sampled only one pollutant group (carbonyl compounds). Three pollutants measured
with Method TO-11A (carbonyl compounds) have screening values (acetaldehyde,
formaldehyde, and propionaldehyde) and two of these pollutants typically fail all or most of the
screens conducted, as shown in Table 4-8. Thus, sites sampling only carbonyl compounds have
higher failure rates. Conversely, sites that sampled several pollutant groups tended to have lower
failure rates due to the larger number of HAPs screened, as is the case with GLKY and SEWA.
4-21
-------
These sites both sampled four pollutant groups and have a failure rate less than 20 percent. Of
course, the magnitude of concentrations measured greatly factors into this as well.
The following sections from this point forward focus primarily on those pollutants
designated as program-level pollutants of interest.
4.2.1 Concentrations of the Pollutants of Interest
Concentrations of the program-level pollutants of interest vary significantly, among the
pollutants and across the sites. Tables 4-10 through 4-13 present the top 10 annual average
concentrations and 95 percent confidence intervals by site for each of the program-level
pollutants of interest (for VOC/SNMOCs, carbonyl compounds, PAHs, and metals,
respectively). As described in Section 3.1, an annual average concentration is the average
concentration of all measured detections and zeros substituted for non-detects for a given year.
An annual average is only calculated where at least three quarterly averages could be calculated
for a given year and where the site-specific method completeness is at least 85 percent. The
annual average concentrations in Tables 4-10 and 4-11, for VOC/SNMOCs and carbonyl
compounds, respectively, are reported in |ig/m3 while the annual average concentrations for
PAHs and metals, in Tables 4-12 and 4-13, respectively, are reported in ng/m3 for ease of
viewing. Note that not all sites sampled each pollutant group; thus, the list of possible sites
presented in Tables 4-10 through 4-13 is limited to those sites sampling each pollutant. For
instance, only five sites sampled TSP metals; thus, these would be the only sites to appear in
Table 4-13 for each metal (TSP) pollutant of interest shown. Annual average concentrations for
2015 are shaded in gray in Tables 4-10 through 4-13, while annual average concentrations for
2016 are shown in white. Thus, sites sampling during 2015 and 2016 could appear in each table
more than once.
4-22
-------
Table 4-10. Annual Average Concentration Comparison of the VOC/SNMOC Pollutants of Interest1
Carbon
P-
1,2-
Hexachloro-1,3-
Benzene
1,3-Butadiene
Tetrachloride
Dichlorobenzene
Dichloroethane
Ethylbenzene
Butadiene
Rank
Oig/m3)
frig/m3)
Oig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)
SPAZ
TVKY
TVKY
SPAZ
TVKY
SPAZ
BTUT
1
1.33 ±0.32
0.35 ±0.19
0.85 ±0.13
0.30 ±0.06
3.75 ± 1.56
0.74 ±0.14
0.04 ± 0.04
SPAZ
SPAZ
TVKY
SPAZ
TVKY
SPAZ
TMOK
2
1.28 ±0.31
0.27 ±0.10
0.80 ±0.10
0.25 ± 0.06
3.49 ± 1.36
0.66 ±0.15
0.03 ±0.01
PACO
SPAZ
BLKY
S4MO
BLKY
PXSS
TROK
3
1.21 ±0.16
0.23 ± 0.08
0.73 ±0.04
0.22 ±0.10
1.89 ± 1.55
0.65 ±0.11
0.03 ±0.01
PACO
PXSS
SEWA
PXSS
ATKY
PXSS
NRNJ
4
1.20 ±0.14
0.22 ±0.05
0.70 ±0.02
0.20 ±0.03
0.90 ±0.55
0.54 ±0.12
0.03 ±0.01
PXSS
PXSS
ATKY
PXSS
BLKY
TOOK
BLKY
5
1.13 ± 0.19
0.20 ± 0.05
0.69 ±0.03
0.15 ±0.03
0.72 ±0.29
0.47 ±0.08
0.03 ±0.01
RICO
TVKY
ATKY
S4MO
ATKY
CSNJ
PXSS
6
1.10 ± 0.14
0.16 ±0.12
0.67 ±0.03
0.14 ±0.04
0.41 ±0.19
0.47 ±0.09
0.03 ±0.01
TOOK
ELNJ
SEWA
NBIL
TMOK
TMOK
TOOK
7
1.09 ±0.12
0.12 ±0.02
0.67 ±0.02
0.13 ± 0.11
0.10 ±0.01
0.44 ± 0.08
0.02 ±0.01
TOOK
SPIL
BLKY
TMOK
TOOK
CSNJ
ATKY
8
1.08 ±0.13
0.12 ±0.02
0.67 ±0.03
0.08 ±0.02
0.10 ±0.01
0.42 ± 0.06
0.02 ±0.01
PXSS
ELNJ
DEMI
CSNJ
TROK
TROK
CSNJ
9
1.04 ±0.21
0.12 ±0.02
0.67 ±0.02
0.06 ± 0.02
0.10 ±0.01
0.41 ±0.07
0.02 ±0.01
TVKY
SPIL
DEMI
TMOK
BROK
TOOK
NBIL
10
1.04 ±0.30
0.11 ±0.01
0.66 ±0.03
0.06 ±0.01
0.09 ±0.01
0.40 ± 0.06
0.02 ±0.01
BOLD ITALICS = EPA-designated NATTS Site
1 Annual average concentrations for 2015 are shaded in gray, while those for 2016 are in white.
-------
Table 4-11. Annual Average Concentration Comparison of the
Carbonyl Compound Pollutants of Interest1
Acet aldehyde
Formaldehyde
Rank
(ug/m3)
(ug/m3)
BROK
BTUT
1
4.06 ± 1.42
8.42 ± 1.37
BTUT
AZFL
2
3.64 ±0.48
7.31 ± 1.85
PXSS
BTUT
3
2.75 ±0.28
5.68 ±0.72
CSNJ
BROK
4
2.65 ±0.37
4.95 ± 1.59
BTUT
SKFL
5
2.62 ±0.31
4.72 ± 1.29
ELNJ
ELNJ
6
2.50 ±0.26
4.43 ±0.41
ELNJ
ELNJ
7
2.49 ±0.31
4.38 ±0.68
SPIL
CSNJ
8
2.45 ±0.61
4.06 ±0.43
SPIL
SPIL
9
2.43 ±0.52
3.85 ±0.45
NBNJ
PXSS
10
2.03 ± 0.20
3.80 ±0.27
BOLD ITALICS = EPA-designated NATTS Site
1 Annual average concentrations for 2015 are
shaded in gray, while those for 2016 are in white.
4-24
-------
Table 4-12. Annual Average Concentration Comparison of the
PAH Pollutants of Interest1
Acenaphthene
Fluorene
Naphthalene
Rank
(ng/m3)
(ng/m3)
(ng/m3)
NBIL
NBIL
DEMI
1
18.93 ±6.15
19.19 ±6.42
116.18 ± 15.46
NBIL
NBIL
BXNY
2
17.48 ±5.88
16.27 ±5.41
113.05 ± 12.40
ROCH
ROCH
DEMI
3
17.37 ±4.25
13.07 ±3.08
107.01 ± 14.51
ROCH
ROCH
BXNY
4
13.81 ±3.80
11.71 ± 3.14
93.29 ± 10.96
DEMI
DEMI
GPCO
5
10.30 ±3.05
9.57 ±2.69
91.01 ± 12.89
DEMI
DEMI
NBIL
6
8.86 ±2.14
7.93 ± 1.69
89.32 ±22.07
GPCO
GPCO
NBIL
7
8.29 ±3.97
7.33 ±3.80
79.55 ± 17.47
S4MO
S4MO
CELA
8
6.51 ± 1.47
6.67 ± 1.81
78.40 ± 9.06
S4MO
BXNY
S4MO
9
6.13 ± 1.71
6.61 ± 1.30
78.32 ± 11.77
GPCO
S4MO
CELA
10
5.67 ± 1.20
6.40 ± 1.31
76.85 ± 9.42
BOLD ITALICS = EPA-designated NATTS Site
1 Annual average concentrations for 2015 are shaded in gray, while
those for 2016 are in white.
4-25
-------
Table 4-13. Annual Average Concentration Comparison of the
Metals Pollutants of Interest1
Arsenic
Arsenic
(PMio)
(TSP)
Rank
(ng/m3)
(ng/m3)
ASKY-M
TROK
1
1.38 ±0.32
0.95 ±0.20
ASKY-M
TOOK
2
1.13 ±0.23
0.89 ± 0.11
BAKY
TROK
3
0.97 ±0.19
0.85 ±0.14
NBIL
TOOK
4
0.94 ±0.27
0.78 ±0.08
BAKY
TMOK
5
0.92 ±0.15
0.67 ±0.08
S4MO
TMOK
6
0.90 ±0.13
0.64 ±0.07
S4MO
OCOK
7
0.88 ±0.12
0.56 ±0.07
NBIL
YUOK
8
0.87 ±0.13
0.55 ±0.09
LEKY
YUOK
9
0.81 ±0.17
0.54 ±0.08
SEWA
OCOK
10
0.77 ±0.16
0.51 ±0.06
BOLD ITALICS = EPA-designated NATTS Site
1 Annual average concentrations for 2015 are
shaded in gray, while those for 2016 are in
white.
Observations from Tables 4-10 through 4-13 include the following:
• The highest annual average concentration among the program-wide pollutants of
interest was calculated for formaldehyde for BTUT for 2015 (8.42 ± 1.37 |ig/m3).
BTUT's 2016 annual average is less (5.68 ± 0.72 |ig/m3), but still ranks third highest
among annual average formaldehyde concentrations. Annual average concentrations
of formaldehyde for BTUT have topped this list for the last several NMP reports.
Formaldehyde accounts for 25 of the 29 annual average concentrations greater than
3.0 |ig/m3 shown in Tables 4-10 through 4-13 (with 1,2-dichloroethane and
acetaldehyde accounting for two each).
• Among the VOCs shown in Table 4-10, the highest annual average concentrations
were calculated for 1,2-dichloroethane for TVKY (3.75 ± 1.56 |ig/m3 for 2015 and
3.49 ± 1.36 |ig/m3 for 2016). Only one other NMP site sampling this pollutant has an
annual average concentration greater than 1 |ig/m3 (BLKY, 1.89 ± 1.55 |ig/m3 for
2016) and no NMP site outside of Calvert City, Kentucky has an annual average
concentration of this pollutant greater than 0.10 |ig/m3. While the Calvert City,
Kentucky sites (ATKY, BLKY, and TVKY) account for the six highest annual
average concentrations of this pollutant in Table 4-10, their averages are also quite
variable.
4-26
-------
Benzene is the only other VOC shown in Table 4-10 with annual average
concentrations greater than 1 |ig/m3. In fact, all 10 annual average concentrations of
benzene shown in Table 4-10 are greater than 1 |ig/m3. The annual average
concentrations for both years appear in Table 4-10 for most of the sites shown. For
example, SPAZ has the two highest annual average benzene concentrations,
1.33 ± 0.32 |ig/m3 for 2016 and 1.28 ± 0.31 |ig/m3 for 2015. PACO, PXSS, and
TOOK also appear for both years. RICO (shown for 2015) and TVKY (shown for
2015) are the exceptions. Note that the annual average benzene concentrations for
SPAZ have the largest confidence intervals associated with them. It is worth noting
that VOC samples were collected on a l-in-12 day sampling schedule at SPAZ,
compared to a l-in-6 day schedule for the other sites.
The highest annual average concentration of 1,3-butadiene (0.35 ± 0.19 |ig/m3) was
calculated for TVKY for 2015 and is very similar to the annual average calculated for
this site for 2014. TVKY's 2016 annual average concentration of 1,3-butadiene
(0.16 ± 0.12 |ig/m3) is less than half the 2015 annual average for this site but still
ranks sixth highest among sites sampling this pollutant. Both the 2015 and 2016
annual average concentrations of 1,3-butadiene rank in the top 10 for the five sites
shown in Table 4-10. Note the relatively large confidence intervals associated with
the annual average concentrations for TVKY. This site has the highest measurements
of 1,3-butadiene across the program; of the 10 1,3-butadiene concentrations greater
than 1 |ig/m3 measured across the program, nine were measured at TVKY (seven in
2015 and two in 2016). The annual average concentrations of 1,3-butadiene for the
two Phoenix, Arizona sites rank second, third, fourth, and fifth (each of which lies
between 0.20 |ig/m3 and 0.30 |ig/m3) among sites sampling this pollutant.
The highest annual average concentrations of carbon tetrachloride were also
calculated for TVKY (0.85 ± 0.13 |ig/m3 for 2015 and 0.80 ± 0.10 |ig/m3 for 2016).
Calvert City, Kentucky sites account for six of the 10 highest annual average
concentrations of carbon tetrachloride. Most of the annual average concentrations of
carbon tetrachloride do not vary significantly across NMP sites; less than 0.10 |ig/m3
separates most of the annual average carbon tetrachloride concentrations across the
program. Only TVKY has an annual average concentration greater than 0.75 |ig/m3,
with most lying between 0.6 |ig/m3 and 0.7 |ig/m3. Measurements of carbon
tetrachloride collected at Calvert City sites account for the 28 highest carbon
tetrachloride concentrations measured across the program, including 22
measurements greater than 1 |ig/m3. Annual average concentrations for SEWA and
DEMI account for the remaining annual averages of carbon tetrachloride shown in
Table 4-10.
Similar to 2014, and previous years, SPAZ has the highest annual average
concentrations of />dichlorobenzene among NMP sites sampling this pollutant. The
two Phoenix, Arizona sites account for four of the five highest annual average
concentrations of p-dichlorobenzene shown in Table 4-10.
The three Calvert City, Kentucky sites account for the six highest annual average
concentrations of 1,2-dichloroethane, although the averages vary significantly among
them, ranging from 3.75 ± 1.56 |ig/m3 for TVKY for 2015 to 0.41 ± 0.19 |ig/m3 for
ATKY for 2015. All other NMP sites have annual average 1,2-dichloroethane
4-27
-------
concentrations of 0.10 |ig/m3 or less, including the four Oklahoma sites rounding out
the top 10 annual averages shown in Table 4-10. The three sites Calvert City account
for the all but one of the 176 measurements of 1,2-dichloroethane greater than
0.25 |ig/m3 measured across the program, with these measurements ranging from
0.251 |ig/m3 to 45.8 |ig/m3.
The Phoenix, Arizona sites also have the four highest annual average concentrations
of ethylbenzene across the program, with the remaining annual average
concentrations shown in Table 4-10 less than 0.5 |ig/m3. These sites also ranked
highest for ethylbenzene in the 2014 NMP report. The only other sites with annual
average concentrations of ethylbenzene greater than or equal to 0.4 |ig/m3 are located
in Tulsa, Oklahoma (TOOK TMOK, or TROK) or Camden, New Jersey (CSNJ).
The annual average concentrations of hexachloro-1,3-butadiene shown in Table 4-10
were calculated based on 2016 data. This is due to a standard contamination issue that
was found in 2015, as discussed in Section 2.4, resulting in the invalidation of a large
portion of the 2015 data for this pollutant. Hexachloro-l,3-butadiene is the only VOC
in Table 4-10 that does not have at least one annual average concentration greater
than 0.1 |ig/m3, BTUT has the highest annual average concentration of this pollutant
(0.04 ± 0.04 |ig/m3), although the range of annual average concentrations of
hexachloro-1,3-butadiene is relatively small, varying by 0.02 |ig/m3 across the sites
shown and by 0.04 |ig/m3 across all NMP sites.
Many of the sites shown in Table 4-11 for the highest annual average concentrations
of acetaldehyde are the same as the sites shown for formaldehyde. For example,
BTUT's 2015 annual average concentration of formaldehyde ranks highest among
sites sampling this pollutant; BTUT's 2015 annual average concentration of
acetaldehyde ranks second highest among sites sampling this pollutant. BTUT's 2016
annual average concentrations of both pollutants also rank the in top five in
Table 4-11. There are exceptions, however. BROK's 2015 annual average is the
highest annual average concentration of acetaldehyde shown in Table 4-11, and
BROK's 2015 annual average concentration formaldehyde ranks fourth highest, yet
neither 2016 annual average appears in Table 4-11. For both pollutants, the annual
average concentration for BROK for 2015 is more than twice the annual average for
2016 (4.06 ± 1.42 |ig/m3 for 2015 vs. 1.46 ± 0.14 |ig/m3 for 2016 for acetaldehyde
and 4.95 ± 1.59 |ig/m3 for 2015 vs. 2.07 ± 0.30 |ig/m3 for 2016 for formaldehyde).
The differences in the confidence intervals for the annual averages for each year
indicates that the concentrations measured in 2015 are highly variable and are likely
influences by outliers. Similarly, AZFL's 2016 annual average concentration of
formaldehyde (7.31 ±1.85 |ig/m3) is more than four times greater than its 2015
annual average concentration of formaldehyde (1.79 ±0.18 |ig/m3). The significant
difference between the two annual average concentrations for AZFL (and BROK) is
discussed in detail in the individual state sections.
Annual average concentrations of acetaldehyde shown in Table 4-11 vary from
4.06 ± 1.42 |ig/m3 for BROK (2015) to 2.03 ± 0.20 |ig/m3 for NBNJ (2015). Four
individual acetaldehyde concentrations greater than 15 |ig/m3 were measured in 2015,
three at BROK and one at SPIL; none were measured in 2016 (although an
acetaldehyde concentration of 14.8 |ig/m3 was measured at SPIL in 2016).
4-28
-------
Annual average formaldehyde concentrations shown in Table 4-11 vary from
8.42 ± 1.37 |ig/m3 for BTUT (2015) to 3.80 ± 0.27 |ig/m3 for PXSS (2016). As
shown, there are eight annual average concentrations of formaldehyde greater than
4 |ig/m3, yet this is only true for two NMP sites for both years (BTUT and ELNJ).
The 2015 and 2016 annual average concentrations of formaldehyde for ELNJ are
similar to each other while the annual averages for BTUT are less so.
There are three PAH program-wide pollutant of interest, as shown in Table 4-12. All
sites sampling PAHs under the NMP in 2015 and 2016 were NATTS sites.
There is considerable agreement in the ranking of the highest annual average
concentrations of acenaphthene and fluorene among the sites shown. For example,
NBIL's 2016 annual average concentrations of both pollutants rank highest, followed
by NBIL's 2015 annual averages; ROCH's 2015 annual average concentrations of
both pollutants rank third highest, followed by ROCH's 2016 annual averages; and
DEMI's 2016 annual average concentrations of both pollutants rank fifth highest,
followed by DEMI's 2015 annual averages.
Table 4-12 shows that the range of the 10 highest annual average concentrations of
naphthalene varies considerably, from 116.18 ± 15.46 ng/m3 for DEMI (2015) to
76.85 ± 9.42 ng/m3 for CELA (2015), with three annual average concentrations of
naphthalene greater than 100 ng/m3. DEMI also had the highest annual average
concentration of naphthalene in the 2014 NMP report. The three highest individual
naphthalene concentrations, including one greater than 400 ng/m3, were measured at
NBIL. In total, six measurements of naphthalene greater than 300 ng/m3 were
measured across the program (five in 2015 and two in 2016).
ASKY-M has the highest annual average concentration of arsenic, similar to the 2014
and 2013 NMP reports. This site has the only annual average concentrations of
arsenic greater than 1 ng/m3 (1.38 ± 0.32 ng/m3 for 2015 and 1.13 ± 0.23 ng/m3 for
2016). Three of the five Kentucky sites sampling PMio metals (and where annual
average concentrations could be calculated) appear in Table 4-13 for arsenic (BLKY
and GLKY are the exceptions). Both years' annual average concentrations of arsenic
for S4MO and NBIL also appear in Table 4-13. Annual averages of arsenic for S4MO
consistently rank among the highest in past annual reports.
Among the Oklahoma sites sampling TSP metals, the annual average concentrations
of arsenic for the three Tulsa sites ranked higher than the annual averages for the
Oklahoma City sites. The 2016 annual average concentration of arsenic for TROK
(0.95 ± 0.20 ng/m3) is the highest among the sites sampling TSP metals.
Annual average concentrations for PXSS appear in Tables 4-10 through 4-13 the
most, a total of 11 times, followed by DEMI, NBIL, and SPAZ (at 8 appearances
each), TVKY and S4MO (at 7 each), and TOOK, CSNJ, and ELNJ (at 6 each). The
two Phoenix, Arizona sites appear in Table 4-10 a combined 19 times; the three
Calvert City, Kentucky sites appear in Table 4-10 a combined 17 times; and the
Tulsa, Oklahoma sites appear a combined 14 times.
4-29
-------
4.2.2 Variability Analysis for the Pollutants of Interest
This section presents the results of the two variability analyses described in Section 3.3.
4.2.2.1 Inter-site Variability
Figures 4-1 through 4-12 are bar graphs depicting the site-specific annual averages for
each year (in gray for 2015 and in white for 2016) overlain on the program-level averages, for
both years combined (indicated by the solid shading), as presented in Section 4.1. For each
program-level pollutant of interest, the inter-site variability graphs allow the reader to see how
the individual site-specific annual average concentrations feed into the program-level averages
(e.g., if a specific site(s) is driving the program average). In addition, the confidence intervals
provided on the inter-site variability graphs are an indication of the amount of variability
contained within the site-specific dataset and thus, annual average concentrations. The published
MDL for each year from the ERG laboratory is also plotted on the graph as an indication of how
the data fall in relation to the MDL. The preliminary risk-based screening values are also plotted
on the graphs.
Several items to note about these figures: Some sites do not have annual average
concentrations presented on the inter-site variability graphs because they did not meet the criteria
for the calculation of annual averages specified in Section 3.1. For the sites sampling metals, the
program-level average for sites collecting PMio samples is presented in green while the program-
level average for sites collecting TSP samples is presented in pink. For benzene, 1,3-butadiene,
and ethylbenzene, the three pollutants sampled and analyzed with two methods (VOC and
SNMOC) and identified as program-level pollutants of interest, two graphs are presented, one for
each method. Note that BTUT and NBIL have their canister samples analyzed using both TO-15
and SNMOC methods. While both results are shown in this section, only the VOC results are
discussed throughout the remainder of this report, as described in Section 3.2. The exception is
for RFCO; canister samples collected at RFCO were analyzed with both methods between
January and September 2015, after which only the SNMOC analysis was performed. This too is
discussed in Section 3.2.
4-30
-------
Figure 4-1. Inter-Site Variability for Acenaphthene
,
y
ill
II
w
_
t
i, r
II.
t
ii i _.ji
BOM A BTUT BXNY CELA DEMI GLKY GPCO NBIL PRRI PXSS RIVA ROCH RUCA S4MO SEWA SJJCA SKFL UNVT WADC
Program Average 2015 Average I 12016 Ave rage 2015 MDL 2016 MDL Risk
(0.082 ng/m3) (0.047 ng/m3) (11 ng/m3)
Observations from Figure 4-1 include the following:
• Figure 4-1 presents the program-level and site-specific annual average concentrations
of acenaphthene.
• The program-level average concentration of acenaphthene is4.36±0.37 ng/m3, as
shown in orange in Figure 4-1. Site-specific annual average concentrations range
from 0.12 ± 0.06 ng/m3 (UNVT, 2015) to 18.93 ± 6.15 ng/m3 (NBIL, 2016).
• Both annual average concentrations of acenaphthene for NBIL are more than four
times the program-level average concentration for acenaphthene. Both of ROCH's
annual average concentrations are also considerably greater than the program-level
average concentration of acenaphthene. The confidence intervals associated with
these (and several other) annual average concentrations indicate that there is
considerable variability within the measurements.
• Other sites with annual average concentrations greater than the program-level average
include BXNY, DEMI, GPCO, and S4MO.
• Sites with relatively low annual average concentrations (less than 1 ng/m3) other than
UNVT include GLKY and SJJCA.
• An annual average concentration could not be calculated for PXSS for 2016 due to
issues with the collection system resulting in relatively low completeness.
4-31
-------
Figure 4-2. Inter-Site Variability for Acetaldehyde
k
1
t
AZFL BMCO BRCO BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO GSCO IN DEM NBIL NBNJ NRNJ NROK
i i Program Average 2015 Average i 12016 Average
2015 MDL
(0.010 jjg/m3)
2016 MDL
(0.012 |Jg/m3) (0.45
¦Risk
|ag/m3)
T
T
i
L It
1
L
ft
J
|
r 1
f
l
OCOK ORFL PACO PXSS RICO ROIL S4MO SEWA SKFL SPIL SYFL TMOK TOOK TROK WPIN YUOK
i i Pro gr am Ave ra g e 2015 Ave rage i i 2016 Ave r age ................. 2015 M D L ......... 2016 M D L R is k
(0.010 Mg/m3) (0.012 pg/m3) (0.45 |jg/m3)
4-32
-------
Observations from Figure 4-2 include the following:
• Figure 4-2 presents the program-level and site-specific annual average concentrations
of acetaldehyde.
• The program-level average concentration of acetaldehyde is 1.67 ± 0.04 |ig/m3, as
shown in purple in Figure 4-2.
• Site-specific annual average concentrations range from 0.22 ± 0.08 |ig/m3 (BRCO,
2016) to 4.06 ± 1.42 |ig/m3 (BROK, 2015).
• The 2015 annual average concentration of acetaldehyde for BROK is nearly two and
half times the program-level average concentration for acetaldehyde. BROK's annual
average for 2015 is nearly three times greater than the annual average for 2016 for
this site. The confidence intervals associated with the 2015 annual average
concentration of acetaldehyde for BROK indicate that there are likely outliers
affecting this dataset.
• Other sites with annual average concentrations greater than the program-level average
include BTUT, CSNJ, DEMI, ELNJ, GPCO (2016 only), NBNJ (2015 only), OCOK
(2015 only), PXSS (2016 only), SPIL, TMOK, TOOK, TROK, and WPIN (2015
only).
• Besides BROK (in 2015), SPIL, BTUT, CSNJ, and ELNJ have the most variability
associated with their measurements, as indicated by the confidence intervals shown.
• Sites with relatively low annual average concentrations (less than 1 |ig/m3) include
BRCO, BMCO, GSCO, PACO, SEW A, AZFL (2015), RICO, and GLKY.
• Annual averages could not be calculated for BMCO, CHNJ, NRNJ, PACO, PXSS,
and RICO for 2015; GSCO and NBNJ for 2016; and NROK and ROIL for either
year. Note, however, the relocation of the NBNJ collection system to NRNJ in 2016,
the relocation of the BMCO collection system to GSCO in 2015 and back again in
2016, the initiation of sampling at NROK in 2016, and the discontinuation of
sampling at ROIL in mid-2015.
4-33
-------
Figure 4-3. Inter-Site Variability for Arsenic
ASKY-M BAKY BLKY BOM A BTUT GLKY GPCO LEKY NBIL PAFL PXSS S4MO SEWA SJJCA OCOK TMOK TOOK TROK YUOK
ZZ1 Program PM10 Average Program TSP Average 2015 Average i i 2016 Average 2015 Teflon MDL
(0.153 ng/m3)
2015 Quartz MDL 2016Teflon MDL 2016 Quartz MDL Risk
(0.057 ng/m3) (0.039 ng/m3) (0.013 ng/m3) (0.230 ng/m3)
Observations from Figure 4-3 include the following:
• Figure 4-3 presents the inter-site variability graph for arsenic, which also includes a
comparison of PMio results (green) and TSP results (pink). Note that only sites from
Oklahoma are using TSP collection systems.
• The program-level average concentration of arsenic in PMio is similar to the program-
level average concentration of arsenic in TSP, with a PMio average of
0.703 ± 0.032 ng/m3 and a TSP average of 0.695 ± 0.036 ng/m3.
• Site-specific annual average arsenic concentrations for PMio range from
0.28 ± 0.03 ng/m3 (GPCO, 2015) to 1.38 ± 0.32 ng/m3 (ASKY-M, 2015) and from
0.51 ± 0.06 ng/m3 (OCOK, 2016) to 0.95 ± 0.20 ng/m3 (TROK, 2016) for TSP.
• ASKY-M, NBIL, and BTUT have the most variability in the PMio measurements,
while TROK has the most variability in the TSP measurements.
• Most of the annual average concentrations of arsenic are within a 0.5 ng/m3 window
(between 0.4 ng/m3 and 0.9 ng/m3). Those sites with annul averages greater than
0.9 ng/m3 include ASKY-M, BAKY, NBIL (2015 only), and TROK (2016 only).
GPCO is the only site with annual average concentrations less than 0.4 ng/m3.
• An annual average could not be calculated for LEKY in 2016 because the
completeness criteria was not met.
4-34
-------
Figure 4-4a. Inter-Site Variability for Benzene - Method TO-15
E
~3B
u 1.0
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average I I 2016 Average 2015 MDL 2016 MDL Risk
(0.125 jag/m3) (0.069 pg/m3) (0.13 |jg/m3)
E
"52
u 1.0
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ SPIL TMOK TOOK TROK TVKY YUOK
Program Average
12015 Average
] 2016 Average
2015 MDL
(0.125 |Jg/m3)
2016 MDL Risk
(0.069 |Jg/m3) (0.13 |ag/m3)
4-35
-------
Observations from Figure 4-4a include the following:
• Figure 4-4a is the inter-site variability graph for benzene, as measured with
Method TO-15. (Figure 4-4b presents the inter-site variability graph for benzene, as
measured with the SNMOC method.)
• The program-level average concentration of benzene (TO-15 only) is
0.72 ± 0.02 |ig/m3.
• Site-specific annual average benzene concentrations range from 0.34 ± 0.04 |ig/m3
(CHNJ, 2016) to 1.33 ± 0.32 |ig/m3 (SPAZ, 2016).
• Other sites measuring benzene with Method TO-15 with annual average
concentrations greater than 1 |ig/m3 include RFCO (2015), PXSS, TOOK, and TVKY
(2015). Sites with relatively low annual average concentrations of benzene (less than
0.5 |ig/m3) include CHNJ, GLKY, NBIL (2016), SEWA (2016), and BLKY (2016).
• RFCO (2015), ATKY (2015), SPAZ, TVKY (2015), and PXSS have the most
variability associated with the benzene measurements collected, as indicated by the
relatively large confidence intervals shown in Figure 4-4a.
• Annual averages could not be calculated for BROK, BTUT, GPCO, and NRNJ for
2015; LEKY and NBNJ for 2016; and NROK and ROIL for either year. Note,
however, the relocation of the NBNJ collection system to NRNJ in 2016, the
initiation of sampling at BROK in 2015 and NROK in 2016, and the discontinuation
of sampling at ROIL in mid-2015 and at LEKY in mid-2016.
• Sampling and analysis of benzene was performed with both Methods TO-15 and
SNMOC for canister samples collected at RFCO between January and September
2015, after which only the SNMOC method was used. As a result, the annual average
benzene concentration for RFCO with Method TO-15 is presented here, although the
annual average concentration of benzene presented for this site for the remainder of
the report is from the SNMOC method.
4-36
-------
Figure 4-4b. Inter-Site Variability for Benzene - SNMOC
1.4
BMCO BRCO BROK BTUT GSCO NBIL NROK PACO RFCO RICO
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
(0.044Mg/m3) (0.0511_ig/m3) (0.13 (.ig/m3)
Observations from Figure 4-4b include the following:
• Figure 4-4b is the inter-site variability graph for benzene, as measured with the
SNMOC method. Canister samples collected at 10 sites are analyzed with this
method.
• The program4evel average concentration of benzene (SNMOC only) is
0.84 ± 0.04 |ig/m3. Site-specific annual average concentrations of benzene (SNMOC
only) range from 0.45 ± 0.08 |ig/m3 (RFCO, 2016) to 1.21 ± 0.16 |ig/m3 (PACO,
2015).
• PACO's annual average concentrations of benzene for both 2015 and 2016 are similar
to each other, both of which are greater than 1 |ig/m3. RICO's annual average for
2016 is also greater than 1 |ig/m3, but no annual average could be calculated for 2015.
Annual average concentrations for GSCO, NBIL, and RFCO (2016) are considerably
less.
• Note the differences in the annual average concentrations of benzene for RFCO. The
annual average for 2016 (0.45 ± 0.08 |ig/m3) is half the magnitude of the annual
average for 2015 (0.92 ± 0.68 |ig/m3). The confidence intervals associated with the
annual average concentration for 2015 suggest that outliers may be affecting this
annual average concentration. RFCO's average for 2015 shown in Figure 4-4a for
Method TO-15 also reflects a significant level of variability.
4-37
-------
• Less than half of the sites shown in Figure 4-4b have annual average concentrations
of benzene shown for both years. Note the initiation of sampling at BROK in 2015
and NROK in 2016, and the relocation of the BMCO collection system to GSCO in
2015 and back again in 2016. RICO experienced issues with the collection system
resulting in relatively low completeness in 2015. In addition, co-elution affected some
of the samples during analysis, such that some benzene concentrations could not be
determined.
• Note that canisters from BTUT and NBIL were analyzed using both analytical
methods and their annual average benzene concentrations are similar, although
slightly higher, using the SNMOC method. The annual average concentrations of
benzene presented for these two sites for the remainder of the report is from Method
TO-15.
• Canisters were also analyzed with both methods for RFCO between January and
September 2015. However, the because the time frame of collection is different
(SNMOC - all year, TO-15 - January through September 2015 only), the averages
shown should not be compared directly. The annual average concentrations of
benzene presented for this site for the remainder of the report are from the SNMOC
method.
4-38
-------
Figure 4-5a. Inter-Site Variability for 1,3-Butadiene - Method TO-15
E
~3B
Hh
f
1
1
1
1 n
¦ 1
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
(0.032 jag/m3) (0.058 jag/m3) (0.030 |Jg/m3)
E
~53
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ
M Program Average 2015 Average i 12016 Average
SPIL TMOK TOOK TROK TVKY YUOK
2015 MDL 2016 MDL Risk
(0.032 |ig/m3) (0.058 |Jg/m3) (0.030 |Jg/m3)
4-39
-------
Observations from Figure 4-5a include the following:
• Figure 4-5a is the inter-site variability graph for 1,3-butadiene, as measured with
Method TO-15. (Figure 4-5b presents the inter-site variability graph for
1,3-butadiene, as measured with the SNMOC method.)
• The program-level average concentration of 1,3-butadiene (TO-15 only) is
0.086 ± 0.006 |ig/m3.
• Site-specific annual average 1,3-butadiene concentrations range from
0.02 ± <0.01 |ig/m3 (CHNJ, 2016) to 0.35 ± 0.19 |ig/m3 (TVKY, 2015).
• Figure 4-5a shows that the annual average concentrations for a few sites are
considerably higher than most other sites. The annual average concentrations for
PXSS, SPAZ, and TVKY stand out the most in this figure. These sites also have the
most variability associated with their 1,3-butadiene measurements, as indicated by the
confidence intervals shown in Figure 4-5a.
• Many sites' annual average concentrations are less than the program-level average
concentration, including some whose annual average is also less than the MDLs
shown, including BROK (2016) and CHNJ. Note the difference between the 2015
(0.032 |ig/m3) and 2016 (0.058 |ig/m3) MDLs for this pollutant.
• Annual averages could not be calculated for BROK, BTUT, GPCO, and NRNJ for
2015; LEKY and NBNJ for 2016; and NROK and ROIL for either year. Note,
however, the relocation of the NBNJ collection system to NRNJ in 2016, the
initiation of sampling at BROK in 2015 and NROK in 2016, and the discontinuation
of sampling at ROIL in mid-2015 and at LEKY in mid-2016.
• Sampling and analysis of 1,3-butadiene was performed with both Methods TO-15 and
SNMOC for canisters collected at RFCO between January and September 2015, after
which only the SNMOC method was used. As a result, the annual average
1,3-butadiene concentration for RFCO with Method TO-15 is presented here,
although the annual average concentration of 1,3-butadiene presented for this site for
the remainder of the report is from the SNMOC method.
4-40
-------
Figure 4-5b. Inter-Site Variability for 1,3-Butadiene - SNMOC
fr~s ^ It
BMCO BRCO BROK BTUT GSCO NBIL NROK PACO RFCO RICO
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
(0.061 |jg/m3) (0.068 |jg/m3) (0.030 |Jg/m3)
Observations from Figure 4-5b include the following:
• Figure 4-5b is the inter-site variability graph for 1,3-butadiene, as measured with the
SNMOC method. Canister samples collected at 10 sites are analyzed with this
method. Note that the scale in this figure is aligned with Figure 4-5a.
• The program-level average concentration of 1,3-butadiene (SNMOC only) is
0.026 ± 0.004 |ig/m3. Site-specific annual average concentrations of 1,3-butadiene
(SNMOC only) range from 0 |ig/m3 (i.e., no detects at BRCO in 2015 and BMCO in
2016) to 0.074 ± 0.020 |ig/m3 (RICO, 2016). RICO's annual average concentration of
1,3-butadiene for 2016 is the only annual average greater than the MDLs for this
pollutant.
• Few of the sites in Figure 4-5b have annual average concentrations of 1,3-butadiene
shown for either or both years. In some cases, as noted above, this is due to a lack of
measured detections of 1,3-butadiene (i.e., the average is at or close to zero). In other
cases, this is due to a lack of or change in sampling. Note the initiation of sampling at
BROK in 2015 and NROK in 2016, and the relocation of the BMCO collection
system to GSCO in 2015 and back again in 2016. In still other cases, this is due to not
meeting the completeness criteria established in Section 3.2. BTUT and RICO
experienced issues with the collection system resulting in relatively low completeness
in 2015.
4-41
-------
• Canisters were analyzed with both methods for RFCO between January and
September 2015. However, because the time frame of collection is different (SNMOC
- all year, TO-15 - January through September 2015 only), the averages shown
should not be compared directly. The annual average concentrations of 1,3-butadiene
presented for this site for the remainder of the report are from the SNMOC method.
4-42
-------
Figure 4-6. Inter-Site Variability for Carbon Tetrachloride
„ 0.9
E
O 0.6
ft
ft
dh
ih
ffl
(h
ft
ft
ft
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average I 12016 Average 2015 MDL 2016 MDL Risk
(0.066 |jg/m3) (0.104 |jg/m3) (0.17 |Jg/m3)
E
~3B
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ SPIL TMOK TOOK TROK TVKY YUOK
Pro gr a m Ave ra g e 2015 Ave r ag e i 12016 Ave r ag e ......... 2015 M D L R is k ................. 2016 M D L
(0.066|jg/m3) (0.104|ig/m3) (0.17 |ag/m3)
4-43
-------
Observations from Figure 4-6 include the following:
• Figure 4-6 is the inter-site variability graph for carbon tetrachloride, as measured with
Method TO-15.
• The program-level average concentration of carbon tetrachloride is
0.64 ± 0.01 |ig/m3, as shown in blue in Figure 4-6.
• For most sites, the annual average concentrations are either slightly less or slightly
more than the program-level average concentration and the associated confidence
intervals are relatively small. This indicates that there is little variability in the carbon
tetrachloride concentrations measured across the program. This uniformity is
expected. Carbon tetrachloride is a pollutant that was used worldwide as a refrigerant.
However, it was identified as an ozone-depleting substance in the stratosphere and its
use was banned by the Montreal Protocol (EPA, 2017f). This pollutant has a long
lifetime in the atmosphere, but slowly degrades over time. Today, its concentration in
ambient air is fairly ubiquitous regardless of where it is measured.
• The annual average carbon tetrachloride concentrations for several of the Calvert
City, Kentucky sites are greater than annual averages for the remaining sites,
particularly for TVKY. Most of the annual average concentrations of carbon
tetrachloride range from 0.60 |ig/m3 to 0.70 |ig/m3, with annual averages for only
TVKY and BLKY (2016) falling outside this range. In addition, the confidence
intervals shown for these sites are relatively large, particularly for TVKY, indicating
a higher level of variability in the measurements compared to most other NMP sites.
• Annual averages could not be calculated for a number of sites (and years), including
BROK, BTUT, GPCO, and NRNJ for 2015; LEKY and NBNJ for 2016; and NROK
and ROIL for either year. Note, however, the relocation of the NBNJ collection
system to NRNJ in 2016, the initiation of sampling at BROK in 2015 and NROK in
2016, and the discontinuation of sampling at ROIL in mid-2015 and at LEKY in mid-
2016.
4-44
-------
Figure 4-7. Inter-Site Variability for /7-Dichlorobenzene
i a
1*1 fe m A PI fa 1
ibifl ^ ft i 1
hi t\
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
(0.154 |Jg/m3) (0.139 |ag/m3) (0.091 |ag/m3)
t i\ fa i
n*i 1
1
i 1 1 ft ft « »
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ SPIL TMOK TOOK TROK TVKY YUOK
Program Average 2015 Average I i 2016 Average 2015 MDL 2016 MDL Risk
(0.154 |jg/m3) (0.139 |jg/m3) (0.091 |ag/m3)
4-45
-------
Observations from Figure 4-7 include the following:
• Figure 4-7 is the inter-site variability graph for p-dichlorobenzene, as measured with
Method TO-15.
• The program-level average concentration (0.047 ± 0.004 |ig/m3) and most of the site-
specific annual average concentrations are less than the MDLs shown for this
pollutant (0.154 |ig/m3 for 2015, 0.139 |ig/m3 for 2016), as indicated by the dashed
blue lines. This indicates that many of the measurements are either non-detects or less
than the detection limit. Table 4-1 shows that more than half of the 2015-2016
measurements of p-dichlorobenzene are non-detects and of the measured detections,
83 percent were less than the MDL.
• SPAZ is the only site for which both annual average concentrations of
/;-dichlorobenzene are greater than both MDLs for this pollutant. The annual average
concentrations for SPAZ, PXSS, and S4MO (both years) and NBIL (2015 only) are
considerably higher than the other annual averages shown in Figure 4-7. Each of
these annual average concentrations has a considerable level of variability associated
with each average, as indicated by the large confidence intervals.
• PXSS is the only site with more than 100 measured detections of />dichlorobenzene
(112), although S4MO is close (97). Other sites with a relatively higher number of
measured detections include the three Tulsa, Oklahoma sites, CSNJ, ELNJ, and
SPAZ.
• The maximum /;-dichlorobenzene concentration measured across the program was
measured at NBIL (2.78 |ig/m3); four additional />dichlorobenzene concentrations
greater than 1 |ig/m3 were measured at S4MO (ranging from 1.02 |ig/m3 to
1.80 |ig/m3). Concentrations of />dichlorobenzene greater than 0.5 |ig/m3 were
measured at only five sites, S4MO (9), NBIL (4), SPAZ (3), PXSS (2), and BTUT
(1).
• Annual averages could not be calculated for a number of sites (and years), including
BROK, BTUT, GPCO, and NRNJ for 2015; LEKY and NBNJ for 2016; and NROK
and ROIL for either year. Note, however, the relocation of the NBNJ collection
system to NRNJ in 2016, the initiation of sampling at BROK in 2015 and NROK in
2016, and the discontinuation of sampling at ROIL in mid-2015 and at LEKY in mid-
2016.
4-46
-------
Figure 4-8. Inter-Site Variability for 1,2-Dichloroethane
3.5
3.0
2.5
o
1.0
0.5
0.0
3.5
3.0
2.5
o
1.0
0.5
0.0
Program Average 2015 Average i i 7016 Average 2015 MDL Risk 2016MDL
T
II
1
1
n n ¦-! B-I ¦ ¦ n ¦ an ¦ n
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
TVKY's 2015 Annual Avg = 3.75+ 1.56 |jg/m3
TVKY's 2016 Annual Avg = 3.49+ 1.36 Mg/m3
¦1 ¦-! ¦-! - *1 ¦-! M-] ¦-!
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ SPIL TMOK TOOK TROK TVKY YUOK
4-47
-------
Observations from Figure 4-8 include the following:
• Figure 4-8 is the inter-site variability graph for 1,2-dichloroethane, as measured with
Method TO-15.
• The annual average concentrations of 1,2-dichloroethane calculated for the Calvert
City, Kentucky sites are significantly higher than the annual averages for other NMP
sites. Excluding the Calvert City sites, annual average concentrations of
1,2-dichloroethane range from 0.057 ± 0.004 |ig/m3 (GLKY, 2016) to
0.101 ± 0.001 |ig/m3 (TMOK, 2016). The annual average concentrations of
1,2-dichloroethane for the three Calvert City sites range from 0.41 ± 0.02 |ig/m3
(ATKY, 2015) to 3.75 ± 1.56 |ig/m3 (TVKY, 2015). The confidence intervals for
these annual average concentrations are large, indicating there is considerable
variability in the measurements collected at these sites. These measurements are
discussed further in the Kentucky section (Section 12).
• 1,2-Dichloroethane concentrations measured at the Calvert City sites are driving the
program-level average concentration (0.30 ± 0.06 |ig/m3), which was a similar
finding in the 2012, 2013, 2014 NMP reports. The three Calvert City sites account for
the 174 highest concentrations of 1,2-dichloroethane measured under the NMP in
2015 and 2016. Without the Calvert City sites, the program-level average
concentration would be 0.08 ± <0.01 |ig/m3.
• Annual averages could not be calculated for a number of sites (and years), including
BROK, BTUT, GPCO, and NRNJ for 2015; LEKY and NBNJ for 2016; and NROK
and ROIL for either year. Note, however, the relocation of the NBNJ collection
system to NRNJ in 2016, the initiation of sampling at BROK in 2015 and NROK in
2016, and the discontinuation of sampling at ROIL in mid-2015 and at LEKY in mid-
2016.
4-48
-------
Figure 4-9a. Inter-Site Variability for Ethylbenzene - Method TO-15
O 0.4
13
¦ llf, IN '
ASKY ATKY BLKY BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO LEKY NBIL NBNJ NRNJ
Program Average 2015 Average I 12016 Average 2015 MDL 2016 MDL Risk
(0.081 |ig/m3) (0.083 |Jg/m3) (0.4 |Jg/m3)
NROK OCOK PXSS RFCO ROIL S4MO SEWA SPAZ SPIL TMOK TOOK TROK TVKY YUOK
Program Average 2015 Average I 12016 Average 2015 MDL 2016 MDL Risk
(0.081 Mg/m3) (0.083 |jg/m3) (0.4 |jg/m3)
4-49
-------
Observations from Figure 4-9a include the following:
• Figure 4-9a is the inter-site variability graph for ethylbenzene, as measured with
Method TO-15. (Figure 4-9b presents the inter-site variability graph for ethylbenzene,
as measured with the SNMOC method.)
• The program-level average concentration of ethylbenzene (TO-15 only) is
0.26 ± 0.01 |ig/m3, which is similar to the program-level average for 2014.
• Site-specific annual average ethylbenzene concentrations range from
0.07 ± 0.01 |ig/m3 (GLKY, 2015) to 0.74 ± 0.14 |ig/m3 (SPAZ, 2016).
• The Phoenix, Arizona sites (PXSS, SPAZ) have annual average concentrations of
ethylbenzene more than twice the program-level average concentration. Other sites
measuring ethylbenzene with higher annual average concentrations (using the risk
level as the cut-off) include CSNJ and the three Tulsa, Oklahoma sites. Sites with
relatively low annual average concentrations of ethylbenzene (using the MDLs as a
cut-off) include CHNJ and several of the Kentucky sites (GLKY, BLKY, and
TVKY).
• SPAZ, PXSS, DEMI, and CSNJ have the most variability associated with the
ethylbenzene measurements collected, as indicated by the relatively large confidence
intervals shown in Figure 4-9a.
• Annual averages could not be calculated for a number of sites (and years), including
BTUT, GPCO, and NRNJ for 2015; LEKY and NBNJ for 2016; and NROK and
ROIL for either year. Note, however, the relocation of the NBNJ collection system to
NRNJ in 2016, the initiation of sampling at BROK in 2015 and NROK in 2016, and
the discontinuation of sampling at ROIL in mid-2015 and at LEKY in mid-2016.
• Sampling and analysis of ethylbenzene was performed with both Methods TO-15 and
SNMOC for canister samples collected at RFCO between January and September
2015, after which only the SNMOC method was used. As a result, the annual average
ethylbenzene concentration for RFCO with Method TO-15 is presented here,
although the annual average concentration of ethylbenzene presented for this site for
the remainder of the report is from the SNMOC method.
4-50
-------
Figure 4-9b. Inter-Site Variability for Ethylbenzene - SNMOC
1.0
£
1
c
ro
c
-------
• Note that canisters from BTUT and NBIL were analyzed using both methods and
their annual average ethylbenzene concentrations are similar although slightly higher
using the SNMOC method. The annual average concentrations of ethylbenzene
presented for these two sites for the remainder of the report is from Method TO-15.
• Canisters were also analyzed with both methods for RFCO between January and
September 2015. However, because the time frame of collection is different (SNMOC
- all year, TO-15 - January through September 2015 only), the averages shown
should not be compared directly. The annual average concentration of ethylbenzene
presented for this site for the remainder of the report is from the SNMOC method.
4-52
-------
Figure 4-10. Inter-Site Variability for Fluorene
I
fejJLii
BOM A BTUT BXNY CELA DEMI GLKY GPCO NBIL PRRI PXSS RIVA ROCH RUCA S4MO SEWA SJJCA SKFL UNVT WADC
Program Average 2015 Average i i 2016 Average 2015 MDL 2016 MDL Risk
(0.189 ng/m3) (0.113 ng/m3) (11 ng/m3)
Observations from Figure 4-10 include the following:
• Figure 4-10 presents the program-level and site-specific annual average
concentrations of fluorene.
• The program-level average concentration of fluorene is4.36±0.35 ng/m3, as shown
in orange in Figure 4-10.
• Site-specific annual average concentrations range from 0.30 ± 0.10 ng/m3 (UNVT,
2016) to 19.19 ± 6.42 ng/m3 (NBIL, 2016).
• Both annual average concentrations of fluorene for NBIL are more than four times
greater than the program-level average concentration for fluorene. ROCH's annual
average concentrations are more than two (2016) and three (2015) times greater than
the program-level average. Other sites with annual average concentrations greater
than the program-level average include BXNY, DEMI, GPCO, and S4MO.
• Sites with relatively low annual average concentrations of fluorene (less than
1 ng/m3) include UNVT, GLKY, SJJCA, and PXSS.
• An annual average concentration could not be calculated for PXSS for 2016 due to
issues with the collection system resulting in relatively low completeness.
4-53
-------
12
10
8
6
4
2
0
12
10
8
6
4
2
0
Figure 4-11. Inter-Site Variability for Formaldehyde
i Hh
1
f\
AZFL BMCO BRCO BROK BTUT CHNJ CSNJ DEMI ELNJ GLKY GPCO GSCO IN DEM NBIL NBNJ NRNJ NROK
I I Program Average 2015 Average I 12016 Average 2015 MDL 2016 MDL Risk
(0.014 |jg/m3) (0.012 |jg/m3) (0.077 |Jg/m3)
OCOK ORFL PACO PXSS RICO ROIL S4MO SEWA SKFL SPIL SYFL TMOK TOOK TROK WPIN YUOK
i i Program Average 2015 Average i 12016 Average 2015 MDL 2016 MDL Risk
(0.014 |Jg/m3) (0.012 |Jg/m3) (0.077 |Jg/m3)
4-54
-------
Observations from Figure 4-11 include the following:
• Figure 4-11 presents the program-level and site-specific annual average
concentrations of formaldehyde.
• The program-level average concentration of formaldehyde is 3.05 ± 0.09 |ig/m3, as
shown in purple in Figure 4-11.
• Site-specific annual average concentrations range from 0.37 ± 0.14 |ig/m3 (BRCO,
2016) to 8.42 ± 1.37 |ig/m3 (BTUT, 2015). 2015 is the fifth year in a row (2011-
2015) that BTUT has had the highest annual average concentration of formaldehyde
among NMP sites.
• Although nearly 2 |ig/m3 separates BTUT's 2015 (8.42 ± 1.37 |ig/m3) and 2016
(5.68 ± 0.72 |ig/m3) annual average concentrations of formaldehyde, this site has the
highest and third highest annual averages among NMP sites sampling this pollutant.
Only AZFL also has an annual average concentration greater than 5 |ig/m3 (2016,
7.31 ± 1.85 |ig/m3). The 2016 annual average concentration of formaldehyde for
AZFL is more than four times this site's 2015 annual average concentration
(1.79 ±0.18 |ig/m3). Other sites besides AZFL and BTUT exhibiting this disparity
between their two annual averages include BROK and, to a lesser extent, SKFL.
• Sites with relatively low annual average concentrations of formaldehyde (less than
1 |ig/m3) include BMCO (2016), BRCO, GSCO (2015), and SEW A.
• Annual averages could not be calculated for a number of sites (and years), including
BMCO, CHNJ, NRNJ, PACO, PXSS, and RICO for 2015; GSCO and NBNJ for
2016; and NROK and ROIL for either year. Note, however, the relocation of the
NBNJ collection system at NRNJ in 2016, the relocation of the BMCO collection
system at GSCO in 2015 and back again in 2016, the initiation of sampling at NROK
in 2016, and the discontinuation of sampling at ROIL in mid-2015.
4-55
-------
Figure 4-12. Inter-Site Variability for Hexachloro-l,3-butadiene
E
~3B
rti
* ft
ft
PI
ASKY ATKY BLKY
Program Average
BROK BTUT CHNJ
2015 Average
CSNJ DEMI ELNJ
I 2016 Average
2015 MDL
(0.362 |Jg/m3)
2016 MDL
(0.447 |Jg/m3)
NBNJ NRNJ
Risk
(0.045 |Jg/m3)
„ 0.3
E
OCOK PXSS
Program Average
RFCO ROIL
2015 Average
SEW A SPAZ
D 2016 Average
2015 MDL
(0.362 pg/m3)
2016 MDL
(0.447 |ig/m3)
Risk
(0.045 |ig/m3)
4-56
-------
Observations from Figure 4-12 include the following:
• Figure 4-12 presents the program-level and annual average concentrations of
hexachloro-1,3-butadiene. Annual average concentrations of hexachloro-1,3-
butadiene for 2015 could not be calculated due to a standard contamination issue that
resulted in the invalidation of a large portion of the 2015 data for this pollutant.
• The program-level average concentration (0.017 ± 0.002 |ig/m3) and all the site-
specific annual average concentrations shown are considerably less than the MDLs
for this pollutant, despite the difference between them (0.362 |ig/m3 for 2015 and
0.447 |ig/m3 for 2016), as indicated by the dashed blue lines. This indicates that many
of the measurements are either non-detects or less than the detection limits. Table 4-1
shows that 80 percent of the measurements of hexachloro-l,3-butadiene were non-
detects and that only one of the measured detections was greater than the MDL. This
concentration was measured at BTUT on January 7, 2016 (1.02 |ig/m3) and is more
than six times greater than the next highest hexachloro-1,3-butadiene concentration
measured across the two years of sampling. The effects of this outlier can be seen in
the relatively large confidence interval associated with BTUT's 2016 annual average
hexachl oro-1,3 -butadi ene concentrati on.
• Site-specific annual average concentrations range from 0 |ig/m3 for SEWA and SPAZ
for 2016 (indicating that this pollutant was not detected at these sites in 2016) to
0.041 ± 0.035 |ig/m3 (BTUT, 2016). Over the two years of sampling, the number of
measured detections varied from 23 (BLKY) to none (RFCO, SEWA, and SPAZ).
• Annual averages could not be calculated for a number of sites (and years), including
BROK, BTUT, GPCO, and NRNJ for 2015; LEKY and NBNJ for 2016; and NROK
and ROIL for either year. Note, however, the relocation of the NBNJ collection
system to NRNJ in 2016, the initiation of sampling at BROK in 2015 and NROK in
2016, and the discontinuation of sampling at ROIL in mid-2015 and at LEKY in mid-
2016.
4-57
-------
Figure 4-13. Inter-Site Variability for Naphthalene
BOM A BTUT BXNY CELA DEMI GLKY GPCO NBIL PRRI PXSS RIVA ROCH RUCA S4MO SEWA SJJCA SKFL UNVT WADC
Program Average
12015 Average
D2016 Average
2015 MDL
(0.166 ng/m3)
2016 MDL
(0.791 ng/m3)
Risk
(29 ng/m3)
Observations from Figure 4-13 include the following:
• Figure 4-13 presents the program-level and site-specific annual average
concentrations of naphthalene.
• The program-level average concentration of naphthalene is 61.23 ± 1.96 ng/m3, as
shown in orange in Figure 4-13.
• Site-specific annual average concentrations range from 8.39 ±1.17 ng/m3 (UNVT,
2016) to 116.18 ± 15.46 ng/m3 (DEMI, 2015). Aside from DEMI, the only other site
with an annual average concentration greater than 100 ng/m3 is BXNY
(113.05 ± 12.40 ng/m3, 2015). Sites with annual average concentrations less than
29 ng/m3 (the risk level shown in Figure 4-13) are UNVT and GLKY.
• The site with the most variability in the measurements, as indicated by the magnitude
of the confidence intervals, is NBIL. Concentrations measured at NBIL span three
orders of magnitude, ranging from 0.446 ng/m3 (the minimum naphthalene
concentration measured across the program) to 403 ng/m3.
• An annual average concentration could not be calculated for PXSS for 2016 due to
issues with the collection system resulting in relatively low completeness.
4-58
-------
4.2.2.2 Quarterly Variability Analysis
Figures 4-14 through 4-26 provide a graphical display of the site-specific quarterly
average concentrations for each of the program-level pollutants of interest. Quarterly average
concentrations are calculated based on the criteria specified in Section 3.1. For each metal
pollutant of interest, there are two graphs, one for PMio and one for TSP, the scales for which are
the same.
The design of these figures changed for the 2015-2016 NMP report, from a two-
dimensional version to a three-dimensional version. The benefits of this change are two-fold: 1)
the latest version allows for the plotting of multiple years of data on a fairly easy-to-view graph,
and 2) quarterly average concentrations of zero (resulting from the substitution of zeros for non-
detects) can be easily identified in the graphs (in the 2-D version, they appeared to be "missing").
In addition, quarterly average benzene concentrations for sites whose canisters are analyzed
using different methods (i.e., Method TO-15 and SNMOC) are provided on the same graph, such
that they match the quarterly average concentrations provided in the individual state sections.
This is also true for 1,3-butadiene and ethylbenzene. The MDLs and risk factors presented in the
previous section were not added to the quarterly variability graphs for this report, so not the
convolute the graphs.
"Missing" quarterly average concentrations in the figures can be attributed to several
reasons. One reason for missing quarterly averages is due to the sampling duration of each site.
Some sites started late or ended early in the year, which may result in a lack of quarterly
averages. Additionally, the criteria specified in Section 3.1 require a site to have 75 percent of
the possible samples within a given calendar quarter (12 for a site sampling on a l-in-6 day
schedule) for a quarterly average concentration to be calculated. A quarterly average
concentration is not presented for sites that did not meet this criterion. Co-elution can also affect
whether a site has a quarterly average concentration for different pollutants measured and
analyzed by different methods.
Comparing the quarterly average concentrations may provide insight on the detection rate
of the pollutants of interest. Comparing quarterly average concentrations for sites with four valid
quarterly averages in a given year may reveal a temporal trend for some pollutants, such as
formaldehyde, the quarterly averages for which tend to be highest for the summer months, based
on this and previous reports. Trends in quarterly average concentrations are discussed below and
in more detail in the state sections (Sections 5 through 23). The quarterly average concentration
4-59
-------
comparison also allows for the identification of sites with unusually high concentrations of the
pollutants of interest compared to other sites and when those high concentrations were measured;
if concentrations measured at a specific site are significantly lower than other sites; when there is
little variability in the quarterly averages across other sites; and whether inter-state trends exist.
Figure 4-14. Comparison of Quarterly Average Acenaphthene Concentrations
S4MO
ROCH
RUCA
SEW A
SUC A
WADC
UNVT
2015 | 2016 2015 | 2016
BOMA BTUT
2015 | 2016
BXNY
2015 | 2016 2015 | 2016
CELA DEMI
Monitoring Site & Sampling Year
~ Q1 OQ2 ~ Q3 QQ4
Monitoring Site&Sampling Year
~ Q1 OQ2 QQ3 C3Q4
4-60
-------
Observations from Figure 4-14 include the following:
• Figure 4-14 presents the site-specific quarterly average concentrations of
acenaphthene.
• Quarterly average concentrations of acenaphthene range from 0 ng/m3 (GLKY and
UNVT, first quarter 2015) to 46.29 ± 11.43 ng/m:' (NBIL, third quarter 2016).
• The highest quarterly average concentrations for acenaphthene for many of the sites
were calculated for the second and third quarters of each year (during the warmer
months of the years), as indicated by the yellow and blue bars in Figure 4-14.
• NBIL and ROCFI have the highest quarterly average concentrations of acenaphthene;
the third quarter averages for both sites for both years are greater than 30 ng/m3.
These sites' second quarter averages are also among the highest calculated. Other
sites with quarterly average concentrations of acenaphthene greater than 10 ng/m3
include DEMI, GPCO, and S4MO, all of which were calculated for either the second
or third quarter of a given year.
Figure 4-15. Comparison of Quarterly Average Acetaldehyde Concentrations
BMCO
BRCO
BROK
CHNJ
GPCO
BTUT
Monitoring Site & Sampling Year
~ Q.1 n02 OQ3 DQ4
4-61
-------
Figure 4-15. Comparison of Quarterly Average Acetaldehyde Concentrations (Continued)
2015 | 2016
NRNJ
2015 | 2016
6 SCO
2015 | 2016
NDEM
2015 | 2016
NBIL
2015 2016
2015 | 2016
OCOK
2015 | 2016
ORFL
2015 | 2016
PACO
2015 | 2016
RICO
nBNJ
NROK
Monitoring Site & Sampling Year
~ Q1 DQ2 QQ3 OQ4
E 5.0
~II ||
ml
nio
2015 | 2016
ROIL
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
S4MO
SEW A
TMOK
TOOK
TROK
YUOK
WPIN
Monitoring Site & Sampling Year
E01 DQ2 DQ3 DQ4
4-62
-------
Observations from Figure 4-15 include the following:
• Figure 4-15 presents the site-specific quarterly average concentrations of
acetaldehyde.
• Quarterly average concentrations of acetaldehyde range from 0.11 ± 0.06 |ig/m3
(BRCO, third quarter 2016) to 6.23 ± 3.78 |ig/m3 (BROK, third quarter 2015), which
are plotted side-by-side in Figure 4-15.
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., BTUT), consistently lower quarterly average concentration
(e.g., GLKY), or where concentrations exhibited considerable variability among the
calendar quarters (e.g., BROK). BROK's quarterly average concentrations exhibit the
most variability among the sites, particularly for 2015, varying by more than 5 |ig/m3.
Other sites besides BROK who's quarterly average concentrations of acetaldehyde
vary by more than 2 |ig/m3 include BTUT, CSNJ, and SPIL. Conversely, sites who's
quarterly average concentrations of acetaldehyde vary by less than 0.5 |ig/m3 include
GLKY, GSCO, BMCO, NROK, PACO, RICO, SEW A, and SKFL.
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (45 site-year combinations), the third quarter
average concentrations were most often higher than the other quarterly averages
(24 in total, 13 for 2016 and 11 for 2015). These can be seen by the blue bars
extending higher in Figure 4-15 than the others; examples include CSNJ, ELNJ,
OCOK, TOOK, TROK, and YUOK.
4-63
-------
Figure 4-16a. Comparison of Quarterly Average PMio Arsenic Concentrations
„ 2.0
in
2015 | 2016
ASKY-M
2015 | 2016
BAKY
2015 | 2016
BLKY
2015 2016
BOM A
Monitoring Site & Sampling Year
EQ1 DQ2 003 DQ4
2015 | 2016
BTUT
2015 | 2016
GLKY
2015 | 2016
GPCO
sr 2,0
"k
2015 | 2016
LEKY
2015 | 2016
NBIL
2015 | 2016
PAFL
2015 2016
PXSS
Monitoring Site & Sampling Year
2015 | 2016
S4MO
2015 | 2016
SEWA
2015 | 2016
SJJC A
~ Q1 D02 DQ3 D04
4-64
-------
Figure 4-16b. Comparison of Quarterly Average TSP Arsenic Concentrations
OCOK
TMOK
TOOK
TROK
YUOK
2.0
Monitoring Site & Sampling Year
~ 01 nQ2 ~ Q3 n04
Observations from Figures 4-16a and 4-16b include the following:
• Figures 4-16a and 4-16b present the quarterly average concentrations of arsenic for
sites sampling speciated metals, first for PMio then for TSP.
• These figures show that the quarterly average concentrations of arsenic vary more for
those sampling the PMio fraction compared to TSP fraction. This is not altogether
unexpected, given that the sampling locations are more varied among the PMio sites
(the five TSP sites are located in either Tulsa or Oklahoma City, Oklahoma).
• Quarterly average concentrations of arsenic range from 0.22 ± 0.04 ng/m3 (GPCO,
third quarter 2015) to 1.94 ± 0.96 ng/m3 (ASKY-M, second quarter 2015). ASKY-M
and NBIL are the only sites for which a quarterly average concentration of arsenic
greater than 1.5 ng/m3 was calculated (1.94 ± 0.96 ng/m3, second quarter 2015 for
ASKY-M and 1.55 ± 1.08 ng/m3, third quarter 2015 for NBIL).
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., ASKY-M), consistently lower quarterly average concentrations
(e.g., GPCO), or where concentrations exhibited considerable variability among the
calendar quarters (e.g., BTUT). ASKY-M not only has some of the highest quarterly
averages, but this site's quarterly average concentrations exhibit the most variability
among the sites, particularly for 2015. Other sites besides ASKY-M who's quarterly
average concentrations of arsenic vary by more than 1 ng/m3 include BTUT, BAKY,
and NBIL. Conversely, sites who's quarterly average concentrations of arsenic vary
by less than 0.30 ng/m3 include BLKY, GLKY, GPCO, OCOK, and YUOK.
4-65
-------
Figure 4-17. Comparison of Quarterly Average Benzene Concentrations
I)
2015 | 2016
ASKY
2015 | 2016
ATKY
2015 | 2016
BLKY
2015 2016
201b 2016
2015 | 2016
BROK
2015 | 2016
BTUT
2015 | 2016
CHNJ
2015 | 2016
CSNJ
2015 | 2016
DEMI
2015 | 2016
ELNJ
BMCO
BRCO
Monitoring Site & Sampling Year
~ Q1 DQ2 DQ3 DQ4
2015 | 2016
GLKY
2015 | 2016
GPCO
2015 | 2016
GSCO
2015 | 2016
LEKY
2015 201b
2015 2016
2015 | 2016
NRNJ
2015 | 2016
NROK
2015 | 2016
OCOK
2015 | 2016
PACO
2015 | 2016
PXSS
\J BNJ
Monitoring Site & Sampling Year
~ Q1 DQ2 EJQ.3 QQ4
4-66
-------
Figure 4-17. Comparison of Quarterly Average Benzene Concentrations (Continued)
2015 2016
2015|2016
SPIL
2015|2016
TVKY
2015 2016
2015|2016
RiCO
2015 12016
S4MO
2015|2016
SEW A
2015|2016
SPAZ
2015 | 2016
TMOK
2015|2016
TOOK
2015|2016
TROK
2015 | 2016
YUOK
RFCO
Monitoring Site & Sampling Year
13 01 DQ2 G3Q3 DQ4
Observations from Figure 4-17 include the following:
• Figure 4-17 presents the site-specific quarterly average concentrations of benzene.
• Quarterly average concentrations of benzene range from 0.21 ± 0.03 |ig/ni3 (CFtNJ,
third quarter 2016) to 1.93 ± 0.74 jig/m3 (SPAZ, first quarter 2016, with similar
averages for the first and fourth quarters of 2015).
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., SPAZ), consistently lower quarterly average concentration
(e.g., RFCO), or where concentrations exhibited considerable variability among the
quarters (e.g., PXSS, RICO). The quarterly average concentrations forPXSS and
SPAZ exhibit the most variability among the sites sampling benzene. For both sites,
the first and fourth quarter average concentrations were considerably higher than the
second and third quarter averages. This is true for both years. RICO is the only other
site besides PXSS and SPAZ who's quarterly average concentrations of benzene vary
by more than 1 jig/m3. Conversely, sites who's quarterly average concentrations of
benzene vary by less than 0.25 ug/'m3 include BLKY, LEKY, RFCO, SPIL, and
YUOK
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (46 site-year combinations), the first and
fourth quarter average concentrations tended to be higher than the other quarterly
averages (40 in total, 22 for the first quarter and 18 for the fourth quarter). These can
be seen by the gray and orange bars extending higher in Figure 4-17 than the others;
examples in the figure where this can readily be seen include ELNJ, NBNJ, NRNJ,
PXSS, and SPAZ.
4-67
-------
Figure 4-18. Comparison of Quarterly Average 1,3-Butadiene Concentrations
q 0.2
2015 2016
2015 | 2016
ASKY
2015 | 2016
ATKY
2015 | 2016
BLKY
2015 2016
2015 2016
2015 | 2016
BROK
2015 2016
2015 | 2016
CSNJ
2015 | 2016
DEMI
2015 | 2016
ELNJ
BMCO
BRCO
3TU
CHNJ
Monitoring Site & Sampling Year
~ Q1 DQ2 QQ3 DQ4
0.4
1=
ao
3.
c 0.3
W.
m
2015 | 2016
GPCO
2015 | 2016
6 SCO
2015 | 2016
LEKY
2015 | 2016
NBIL
2015 | 2016
NBNJ
2015 2016
2015 2016
2015 2016
2015 2016
2015 2016
NROK
QCOK
PACO
2015 | 2016
GLKY
Monitoring Site & Sampling Year
~ Q1 D02 DQ3 D04
4-68
-------
Figure 4-18. Comparison of Quarterly Average 1,3-Butadiene Concentrations (Continued)
m
2015 | 2016
SEWA
2015|2016
TOOK
2015 2016
2015 | 2016
SPAZ
2015 | 2016
SPIL
2015|2016
TMOK
2015|2016
TROK
2015 2016
2015|2016
YUOK
54MO
TVKY
2015|2016 2015|2016
RFCO RICO
2015 12016
ROIL
Monitoring Site & Sampling Year
001 DQ2 DQ3 QQ4
Observations from Figure 4-18 include the following:
• Figure 4-18 presents the quarterly average concentrations for sites sampling
1,3-butadiene.
• Quarterly average concentrations of 1,3-butadiene range from 0 ug/nv5 (several sites
and several quarters) to 0.52 ± 0.29 jig/m3 (SPAZ, first quarter 2016). Other sites
with relatively high quarterly average concentrations of 1,3-butadiene include PXSS
and TVKY.
• For sites sampling this pollutant with only the SNMOC method, the quarterly average
concentrations shown are very low, often appearing at or close to zero; sites sampling
1,3-butadiene exclusively with the SNMOC method include most of the Garfield
County, Colorado sites (BMCO, BRCO, GSCO, PACO, and RICO). The detection
rate of 1,3-butadiene with the SNMOC method is generally lower than the detection
rate for Method TO-15. Note the 1,3-butadiene was not detected at BMCO in 2015
(January and February only) or 2016 (February through December); 1,3-butadiene
was not detected at BRCO in 2015 and was detected only twice at this site in 2016.
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., TVKY), consistently lower quarterly average concentration
(e.g., Garfield County, Colorado, GLKY, YUOK), or where concentrations exhibited
considerable variability among the quarters (e.g., PXSS, SPAZ). The quarterly
average concentrations for PXSS, SPAZ, and TVKY exhibit the most variability
among the sites sampling 1,3-butadiene. Conversely, sites who's quarterly average
concentrations of benzene vary by less than 0.025 |ig/m3 include BRCO, BROK,
GLKY, and YUOK.
4-69
-------
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (46 site-year combinations), the first and
fourth quarter average concentrations were often higher than the other quarterly
averages (38 in total, 17 for the first quarter and 21 for the fourth quarter). These can
be seen by the gray and orange bars extending higher in Figure 4-18 than the others;
examples in the figure where this can readily be seen include BTUT, ELNJ, PXSS,
SPAZ, and TMOK.
Figure 4-19. Comparison of Quarterly Average Carbon Tetrachloride Concentrations
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
ATKY
BLKY
BROK
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
2015 | 2016
ASKY
Monitoring Site & Sampling Year
~ Q1 DQ2 QQ3 QQ4
Q4
Q3
Q2
01
2015 | 2016
GPCO
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
LEKY
NBIL
NBNJ
NRNJ
NROK
OCOK
PXSS
RFCO
ROIL
Monitoring Site & Sampling Year
~ Q1 QQ2 DQ3 QQ4
4-70
-------
Figure 4-19. Comparison of Quarterly Average Carbon Tetrachloride Concentrations (Continued)
2015 | 2016
SEW A
20lb 2016
2015 | 2016
SPIL
2015 | 2016
TMOK
2015 | 2016
TOOK
2015 | 2016
TROK
2015 | 2016
TVKY
SPAZ
2015 | 2016
S4MO
2015 | 2016
YUOK
Monitoring Site & Sampling Year
~ 01 DQ2 DQ3 D04
Observations from Figure 4-19 include the following:
• Figure 4-19 presents the site-specific quarterly average concentrations of carbon
tetrachloride.
• Quarterly average concentrations of carbon tetrachloride range from
0.40 ± 0.11 jig/m3 (BTUT, third quarter 2016) to 0.94 ± 0.42 jag/m5 (TVKY, third
quarter 2015, with similar averages for the first quarter of 2015 and second quarter
2016).
• Figure 4-19 shows that the quarterly average concentrations of carbon tetrachloride
for most monitoring sites vaiy by less than 0.25 |ig/m3, falling between 0.5 jag/m3 and
0.75 ug/m3. Only three sites have quarterly average concentrations of this pollutant
outside this range (BLKY, BTUT, and TVKY). The site with the largest difference in
its quarterly average concentrations of carbon tetrachloride is BTUT.
• For 2015, among the 20 sites with four quarterly average concentrations of carbon
tetrachloride, the third quarter average concentrations were the highest for 17 of them.
For 2016, among the 24 sites with four quarterly average concentrations of carbon
tetrachloride, the second quarter average concentrations were the highest for 23 of
them. However, the differences among the quarterly averages are so small, it makes
little difference.
4-71
-------
Figure 4-20. Comparison of Quarterly Average jP-Dichlorobenzene Concentrations
ATKY
BLKY
BROK
CHNJ
NROK
OCOK
Monitoring Site & Sampling Year
~ Q1 DQ2 DQ3 QQ4
Monitoring Site & Sampling Year
EQ1 DQ2 E1Q3 DQ4
4-72
-------
Figure 4-20. Comparison of Quarterly Average />-Dichlorobenzene Concentrations (Continued)
TMOK
Monitoring Site & Sampling Year
~ Q1 QQ2 QQ3 DQ4
2015 | 2016
S4MO
2015 | 2016
SEWA
2015 | 2016
SPAZ
2015
SPIL
2016
TOOK
2015 | 2016
TROK
2015 | 2016 2015 | 2016
TVKY YUOK
Observations from Figure 4-20 include the following:
• Figure 4-20 presents the site-specific quarterly average concentrations of
/;-dichlorobenzene.
• Quarterly average concentrations of p-dichlorobenzene range from 0 |ig/m3 (several
sites and several quarters) to 0.40 ±0.13 |ig/m3 (SPAZ, fourth quarter 2016).
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., SPAZ, PXSS), consistently lower quarterly average
concentrations (e.g., GLKY, SEWA), or where concentrations exhibited considerable
variability among the quarters (e.g., NBIL, S4MO). NBIL and S4MO are the only
sites who's quarterly average concentrations ofp-dichlorobenzene vary by more than
0.25 |ig/m3. Conversely, the quarterly average concentrations ofp-dichlorobenzene
vary by less than 0.01 |ig/m3 for NRNJ.
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (44 site-year combinations), the fourth
quarter average concentrations were most often the highest compared to other
quarterly averages (21 in total, 12 for 2015 and nine for 2016). These can be seen by
the orange bars extending higher in Figure 4-20 than the others; examples in the
figure where this can readily be seen include the two Phoenix, Arizona sites and the
three Tulsa, Oklahoma sites.
4-73
-------
Figure 4-21. Comparison of Quarterly Average 1,2-Dichloroethane Concentrations
ATKY
BLKY
BROK
BTUT
CHNJ
<3
-------
Figure 4-21. Comparison of Quarterly Average 1,2-Dichloroethane Concentrations (Continued)
4.0
E 2.0
2015 | 2016
S4MO
2015 | 2016
2015 [ 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
SEWA
SPAZ
SPIL
TMOK
TOOK
TROK
Q4
2015 | 2016
TVKY
2015 | 2016
YUOK
Q1
Monitoring Site & Sampling Year
~ Q1 IHQ2 DQ3 DQ4
Observations from Figure 4-21 include the following:
• Figure 4-21 presents the site-specific quarterly average concentrations of
1,2-dichloroethane.
• Quarterly average concentrations of 1,2-dichloroethane range from
0.013 ± 0.019 |ig/m3 (SPAZ, third quarter 2016) to 5.02 ± 3.76 |ig/m3 (TVKY,
second quarter 2015).
• This figure shows that besides the Calvert City, Kentucky sites (ATKY, BLKY, and
TVKY), no other NMP site has a quarterly average concentration of
1,2-dichloroethane greater than 0.15 |ig/m3. Higher concentrations of this pollutant
have consistently been measured at the Calvert City sites over the last several years.
Few NMP sites (seven) have a quarterly average concentration greater than
0.1 |ig/m3. Excluding the Calvert City sites, TOOK has the highest quarterly average
concentration of 1,2-dichloroethane (0.12 ± 0.02 |ig/m3, fourth quarter 2015, with
similar averages for the first and second quarters of 2016).
4-75
-------
Figure 4-22. Comparison of Quarterly Average Ethylbenzene Concentrations
BMCO
BRCO
GPCO
NBNJ
Monitoring Site & Sampling Year
~ Q1 DQ2 DQ3 DQ4
„ 0.9
2015
GLKY
Monitoring Site & Sampling Year
~ Q1 DQ2 OQ3 DQ4
_ 0.9
"E
2015 | 2016
ASKY
2015 | 2016
ATKY
2015
BLKY
2015 | 2016
BROK
2015 | 2016
BTUT
2015 | 2016
CHNJ
2015 | 2016
CSNJ
2015 | 2016
DEMI
2015 | 2016
ELNJ
PACO
4-76
-------
Figure 4-22. Comparison of Quarterly Average Ethylbenzene Concentrations (Continued)
g 0.3
<
JrI! lil itb
201512016 201512016 2015|2016 201512016 2015| 2016 2015
RFCO RICO ROIL S4MO SEWA SI
2015 2016
2015 | 2016
TMOK
2015 | 2016
TOOK
2015 | 2016
TROK
2015 | 2016
TVKY
2015 | 2016
YUOK
Monitoring Site & Sampling Year
~ Q1 DQ2 OQ3 D04
Observations from Figure 4-22 includes the following:
• Figure 4-22 presents the site-specific quarterly average concentrations of
ethylbenzene.
• Quarterly average concentrations of ethylbenzene range from 0.03 ± 0.03 |ig/m3
(BRCO, second quarter 2015, with a similar average for the second quarter of 2016)
to 1.04 ± 0.74 jig/m3 (SPAZ, fourth quarter 2016).
• Figure 4-22 shows which sites have consistently higher quarterly average
concentrations (e.g., PXSS, SPAZ), consistently lower quarterly average
concentration (e.g., BRCO, TVKY), and/or where concentrations exhibited
considerable variability among the quarters (e.g., the Phoenix sites, DEMI, TMOK).
The quarterly average concentrations for PXSS and SPAZ exhibit the most variability
among the sites sampling ethylbenzene. For both sites, the first and fourth quarter
average concentrations were considerably higher than the second and third quarter
averages. This is true for both years. These are the only sites who's quarterly average
concentrations of ethylbenzene vary by more than 0.5 jag/m3 (though the averages for
DEMI and TMOK just miss this cut-off). Conversely, sites who's quarterly average
concentrations of ethylbenzene vary by less than 0.05 Lig/m5 include BLKY, BMCO,
BRCO, GLKY, and TVKY.
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (46 site-year combinations), the third and
fourth quarter average concentrations were often higher than the other quarterly
averages (38 in total, 17 for the third quarter and 21 for the fourth quarter). These can
be seen by the blue and orange bars extending higher in Figure 4-22 than the others;
examples in the figure where this can readily be seen include ASKY, ELNJ, OCOK,
TROK, YUOK.
4-77
-------
50
40
30
20
10
0
50
40
30
20
10
0
Figure 4-23. Comparison of Quarterly Average Fluorene Concentrations
BOMA
GPCO
BTUT
BXNY
DEMI
ROCH
RUCA
S4MO
SEW A
SJJC A
WADC
UNVT
Monitoring Site & Sampling Year
4-78
-------
Observations from Figure 4-23 include the following:
• Figure 4-23 presents the site-specific quarterly average concentrations of fluorene.
This figure resembles Figure 4-14 for acenaphthene.
• Quarterly average concentrations of fluorene range from 0 ng/m3 (several sites for
several quarters) to 46.67 ± 12.88 ng/m3 (NBIL, third quarter 2016).
• The highest quarterly average concentrations for fluorene for many of the sites were
calculated for the second and third quarters of each year (during the warmer months
of the years), as indicated by the yellow and blue bars in Figure 4-23.
• NBIL has the highest quarterly average concentrations of fluorene; the third quarter
averages for both years are greater than 30 ng/m3. NBIL and ROCH are the only
NMP sites for which quarterly average concentrations of fluorene greater than
20 ng/m3 were calculated. Other sites with quarterly average concentrations greater
than 10 ng/m3 include BXNY, DEMI, GPCO, and S4MO, all of which were
calculated for either the second or third quarter of a given year.
4-79
-------
14
12
10
8
6
4
2
0
14
12
10
8
6
4
2
0
Figure 4-24. Comparison of Quarterly Average Formaldehyde Concentrations
AZFL
BMCO
BRCO
BROK
CHNJ
DEMI
GPCO
BTUT
Monitoring Site & Sampling Year
~ Q1 m02 HQ3 DQ4
NROK
OCOK
ORFL
PACO
INDEM
4-80
-------
Figure 4-24. Comparison of Quarterly Average Formaldehyde Concentrations (Continued)
S4MO
SEW A
TMOK
TOOK
TROK
YUOK
WPIN
Monitoring Site & Sampling Year
~ 01 m02 HQ3 DQ4
Observations from Figure 4-24 include the following:
• Figure 4-24 presents the site-specific quarterly average concentrations of
formaldehyde.
• Quarterly average concentrations of formaldehyde range from 0.16 ± 0.09 jag/m5
(BRCO, third quarter 2016) to 13.30 ± 2.55 |ig/m3 (AZFL, third quarter 2016).
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., BTUT), consistently lower quarterly average concentration
(e.g., BRCO, SEW A), or where concentrations exhibited considerable variability
among the quarters (e.g., AZFL, BROK). AZFL's quarterly average concentrations
exhibit the most variability among the sites, particularly for 2016. Other sites besides
AZFL who's quarterly average concentrations of formaldehyde vary by more than
5 |ig/nr include BROK, BTUT, and SKFL. Conversely, sites who's quarterly average
concentrations of formaldehyde vary by less than 1 |ag/m3 include the Garfield
County, Colorado sites (BMCO, BRCO, GSCO, PACO, and RICO), GPCO, and
SEWA.
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (45 site-year combinations), the third quarter
average concentrations were often higher than the other quarterly averages (35 in
total, 17 for 2015 and 18 for 2016). These can be seen by the blue bars extending
higher in Figure 4-24 than the others; examples include AZFL, BROK, ELNJ,
INDEM, OCOK, TOOK, TMOK, TROK, and YUOK.
4-81
-------
Figure 4-25. Comparison of Quarterly Average Hexachloro-1,3-butad iene Concentrations
GPCO
NROK
OCOK
RFCO
Monitoring Site & Sampling Year
~ Q1 DQ2 DQ3 QQ4
Monitoring Site & Sampling Year
~ Q1 DQ2 OQ3 QQ4
$ 0.04
4-82
-------
Figure 4-25. Comparison of Quarterly Average Hexachloro-l,3-butadiene Concentrations
(Continued)
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
2015 | 2016
S4MO
SEWA
SPAZ
SPIL
TMOK
TOOK
TROK
TVKY
YUOK
Monitoring Site & Sampling Year
~ Q1 D02 DQ3 QQ4
Observations from Figure 4-25 include the following:
• Figure 4-25 presents the site-specific quarterly average concentrations of hexachloro-
1,3-butadiene. Concentrations of hexachloro-1,3-butadiene for 2015 were affected by
a standard contamination issue that resulted in the invalidation of a large portion of
the 2015 data for this pollutant. Thus, no quarterly average concentrations of
hexachloro-1,3-butadiene are presented for 2015 in Figure 4-25.
• Quarterly average concentrations of hexachloro-1,3-butadiene range from 0 |ig/m3
(several sites for several quarter) to 0.08 ± 0.14 |ig/m3 (BTUT, first quarter 2016).
This quarterly average concentration for BTUT is the only quarterly average
concentration greater than 0.05 |ig/m3. The maximum concentration of hexachloro-
1,3-butadiene across the program was measured at BTUT and is more than six times
higher than the next highest hexachloro-1,3-butadiene concentration measured across
the two years of sampling.
• Sites who's quarterly average concentrations of hexachloro-1,3-butadiene vary by
more than 0.035 |ig/m3 include BTUT, CHNJ, and NBIL. Conversely, sites who's
quarterly average concentrations of hexachloro-l,3-butadiene vary by less than
0.005 |ig/m3 include DEMI and SPIL.
• Many of the measurements of hexachloro-1,3-butadiene are either non-detects or less
than the detection limit (80 percent of the measurements of hexachloro-1,3-butadiene
were non-detects). This indicates that a large number of substituted zeroes are
included in the quarterly average calculations, including several sites where this
pollutant was not detected at all (e.g., SPAZ and SEW A).
4-83
-------
Figure 4-26. Comparison of Quarterly Average Naphthalene Concentrations
£ 100
2015 | 2016
BTUT
2015 | 2016
CELA
2015 | 2016
GLKY
2015 | 2016
NBIL
2015 | 2016
BOMA
2015 | 2016
BXNY
2015 | 2016
DEMI
2015 | 2016
GPCO
2015 | 2016
PRRI
2015 | 2016
PXSS
Q4
Q3
Monitoring Site & Sampling Year
~ Q1 DQ2 DQ3 DQ4
= 100
Q1
2015 | 2016
RIVA
2015 | 2016
ROCH
2015 | 2016
RUCA
2015 | 2016
S4MO
2015 | 2016
SEW A
2015 | 2016
SUC A
2015 | 2016
SKFL
2015 | 2016
UNVT
2015 | 2016
WADC
Monitoring Site & Sampling Year
~ Q1 EJQ2 DQ3 QQ4
4-84
-------
Observations from Figure 4-26 include the following:
• Figure 4-26 presents the site-specific quarterly average concentrations of naphthalene.
• Quarterly average concentrations of naphthalene range from 4.87 ± 0.81 ng/m3
(UNVT, third quarter 2016) to 179.35 ± 54.50 ng/m3 (NBIL, third quarter 2015).
Eight NMP sites have at least one quarterly average concentration of naphthalene
greater than 100 ng/m3. Conversely, UNVT is the only NMP sites with quarterly
average concentrations of naphthalene less than 10 ng/m3.
• This figure shows which sites have consistently higher quarterly average
concentrations (e.g., BXNY, DEMI), consistently lower quarterly average
concentration (e.g., GLKY, UNVT), or where concentrations exhibited considerable
variability among the quarters (e.g., NBIL). Other sites besides NBIL who's quarterly
average concentrations of naphthalene vary by more than 75 ng/m3 include DEMI and
PXSS. Conversely, sites who's quarterly average concentrations of naphthalene vary
by less than 15 ng/m3 include GLKY, SEW A, and UNVT.
• Among the sites sampling this pollutant that have four quarterly average
concentrations available for a given year (35 site-year combinations), the fourth
quarter average concentrations were often higher than the other quarterly averages
(17 in total, 7 for 2015 and 10 for 2016). These can be seen by the orange bars
extending higher in Figure 4-26 than the others; examples include RIVA, S4MO,
SJJCA, and WADC. Note that for the sites with the highest quarterly average
naphthalene concentrations (NBIL and DEMI), the highest quarterly averages were
calculated for the third quarter of both years.
4-85
-------
5.0 Sites in Arizona
This section summarizes those data from samples collected at the NATTS and UATMP
sites in Arizona and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines
Sections 1 through 4 for detailed discussions and 1 ~ "
definitions regarding the various data analyses presented below.
5.1 Site Characterization
This section characterizes the Arizona monitoring sites by providing a description of the
nearby area surrounding the monitoring sites; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for the sites.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
The Arizona monitoring sites are located in Phoenix, Arizona. Figures 5-1 and 5-2
present composite satellite images retrieved from ArcGIS Explorer showing the monitoring sites
and their immediate surroundings. Figure 5-3 identifies nearby point source emissions locations
by source category, as reported in the 2014 NEI for point sources, version 1. Note that only
sources within 10 miles of the sites are included in the facility counts provided in Figure 5-3.
A 10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring sites. Further, this boundary provides both the proximity of emissions sources to the
monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile boundaries are still visible on the map for reference but have been
grayed out to emphasize emissions sources within the boundary. Table 5-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
the spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
IXiln Liciicralcd h> sources oilier
I hail I !R( i. LPA's conliucl
kilior;iior\ I'oiihc \\ll\ cuv noi
included in I lie lUiki aiuil> ses
coiiLiinci.1 in llns iV|toi'l
5-1
-------
Figure 5-1. Phoenix, Arizona (PXSS) Monitoring Site
i «jrg m
=— W.Mariposa,St . 1
¦ hi W Glenn
-------
Figure 5-2. South Phoenix, Arizona (SPAZ) Monitoring Site
U>
; ;-'v niJI
Southgate Ave*WScxjthgateAve
W 5 «| - »jflK3Hjpu
^3.5 ft
-------
Figure 5-3. NEI Point Sources Located Within 10 Miles of PXSS and SPAZ
Maricopa
County
Legend
~ PXSS NATTS site ~
112°5'0"W 112WW 111 °55'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
SPAZ UATMP site Q 10 mile radius
County boundary
Source Category Group (No. of Facilities)
*
Aerospace/Aircraft Manufacturing Facility (2)
E
Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (1)
T
Airport/Airline/Airport Support Operations (34)
I
Foundry, Iron and Steel (1)
£
Asphalt Production/Hot Mix Asphalt Plant (1)
Metal Can, Box, and Other Metal Container Manufacturing Facility (1)
<8>
Battery Manufacturing Facility (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
nn
Brick, Structural Clay, or Clay Ceramics Plant (2)
<•>
Metals Processing/Fabrication Facility (5)
B
Bulk Terminal/Bulk Plant (4)
?
Miscellaneous Commercial/Industrial Facility (6)
C
Chemical Manufacturing Facility (4)
R
Plastic, Resin, or Rubber Products Plant (5)
e
Electrical Equipment Manufacturing Facility (3)
X
Rail Yard/Rail Line Operations (3)
f
Electricity Generation Facility (4)
w
Woodwork, Furniture, Millwork and Wood Preserving Facility (2)
5-4
-------
Table 5-1. Geographical Information for the Arizona Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
PXSS
04-013-9997
Phoenix
Maricopa
Phoenix-Mesa-
Scottsdale, AZ
33.503833,
-112.095767
Residential
Urban/City
Center
35,103
W Camelback Rd, on either side of
N 19th Ave
SPAZ
04-013-4003
Phoenix
Maricopa
Phoenix-Mesa-
Scottsdale, AZ
33.403160
-112.075330
Residential
Urban/City
Center
21,601
Central Ave, south of
W Tamarisk St
1AADT reflects 2010 data for PXSS and 2015 data for SPAZ (AZ DOT, 2017)
BOLD ITALICS = EPA-designated NATTS Site
-------
PXSS is located in central Phoenix. Figure 5-1 shows that PXSS is located in a residential
area on North 17th Avenue. The Grand Canal is shown along the bottom of Figure 5-1. The
monitoring site is approximately three-quarters of a mile east of 1-17 and 2 miles north of 1-10.
Figure 5-2 shows that SPAZ is located in South Phoenix near the intersection of West Tamarisk
Street and South Central Avenue. SPAZ is surrounded by residential properties to the west and
south and commercial properties to the east. SPAZ is located approximately 1 mile south of
I-17/1-10.
PXSS is located approximately 7 miles north of SPAZ. The majority of emissions sources
are located between the sites, to the south of PXSS and north of SPAZ, as shown in Figure 5-3.
The source category with the greatest number of emissions sources near these monitoring sites is
the airport source category, which includes airports and related operations as well as small
runways and heliports, such as those associated with hospitals or television stations. The
emissions source nearest PXSS is a hospital heliport while the source nearest SPAZ is a heliport
at a police station.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 5-1 also contains traffic volume information for each site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. PXSS
experiences a higher traffic volume compared to SPAZ, although the traffic volumes near both
sites rank in the middle of the range compared to traffic volumes near other NMP sites. These
traffic volumes were obtained for roadways fairly close to PXSS and SPAZ (West Camelback
Road and Central Avenue, respectively).
5.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
Arizona site to identify site-specific "pollutants of interest," which allows analysts and readers to
focus on a subset of pollutants through the context of risk. For each site, each pollutant's
preprocessed daily measurement was compared to its associated risk screening value. If the
concentration was greater than the risk screening value, then the concentration "failed the
screen." The site-specific results of this risk-based screening process are presented in Table 5-2
and incorporate measurements from both 2015 and 2016. Pollutants of interest are those for
which the individual pollutant's total failed screens contribute to the top 95 percent of the site's
total failed screens and are shaded in gray in Table 5-2. It is important to note which pollutants
5-6
-------
were sampled for at each site when reviewing the results of this analysis. VOCs, carbonyl
compounds, PAHs, and metals (PMio) were sampled for at PXSS; VOCs were the only
pollutants sampled for at SPAZ.
Table 5-2. 2015-2016 Risk-Based Screening Results for the Arizona Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Phoenix, Arizona - PXSS
Benzene
0.13
118
118
100.00
11.27
11.27
Carbon Tetrachloride
0.17
118
118
100.00
11.27
22.54
Arsenic (PMio)
0.00023
114
120
95.00
10.89
33.43
1.3 -Butadiene
0.03
114
118
96.61
10.89
44.32
1,2-Dichloroethane
0.038
108
110
98.18
10.32
54.63
Acetaldehyde
0.45
91
91
100.00
8.69
63.32
Formaldehyde
0.077
91
91
100.00
8.69
72.02
p-Dichlorobenzene
0.091
88
112
78.57
8.40
80.42
Naphthalene
0.029
88
106
83.02
8.40
88.83
Ethylbenzene
0.4
65
118
55.08
6.21
95.03
Hexachloro-1,3 -butadiene
0.045
16
18
88.89
1.53
96.56
Nickel (PMio)
0.0021
15
120
12.50
1.43
97.99
Manganese (PMio)
0.03
7
120
5.83
0.67
98.66
Propionaldehyde
0.8
5
91
5.49
0.48
99.14
Benzo(a)pyrene
0.00057
3
76
3.95
0.29
99.43
Cadmium (PMio)
0.00056
2
120
1.67
0.19
99.62
Beryllium (PMio)
0.00042
1
120
0.83
0.10
99.71
Chloroprene
0.0021
1
1
100.00
0.10
99.81
1,2-Dibromoethane
0.0017
1
1
100.00
0.10
99.90
Lead (PMio)
0.015
1
120
0.83
0.10
100.00
Total
1,047
1,889
55.43
South Phoenix, Arizona - SPAZ
Benzene
0.13
63
63
100.00
18.48
18.48
Carbon Tetrachloride
0.17
63
63
100.00
18.48
36.95
1.3 -Butadiene
0.03
61
62
98.39
17.89
54.84
p-Dichlorobenzene
0.091
54
62
87.10
15.84
70.67
1,2-Dichloroethane
0.038
52
56
92.86
15.25
85.92
Ethylbenzene
0.4
46
63
73.02
13.49
99.41
1,2-Dichloropropane
0.4
1
1
100.00
0.29
99.71
Trichloroethylene
0.2
1
9
11.11
0.29
100.00
Total
341
379
89.97
5-7
-------
Observations from Table 5-2 include the following:
• The number of pollutants failing screens varied significantly between the two
monitoring sites; this is expected given the difference in pollutants measured at each
site.
• Concentrations of 20 pollutants failed at least one screen for PXSS; 55 percent of
concentrations for these 20 pollutants were greater than their associated risk screening
value (or failed screens).
• Concentrations of 10 pollutants contributed to 95 percent of failed screens for PXSS
and therefore were identified as pollutants of interest for PXSS. These 10 include two
carbonyl compounds, six VOCs, one PMio metal, and one PAH.
• PXSS failed the second highest number of screens (1,047) among NMP sites (refer to
Table 4-9 of Section 4.2), and is one of only three sites with more than 1,000 failed
screens. However, the failure rate for PXSS, when incorporating all pollutants with
screening values, is relatively low, at just less than 22 percent. This is due primarily
to the relatively high number of pollutants sampled for at this site, as discussed in
Section 4.2 and the previous section.
• Concentrations of eight pollutants failed screens for SPAZ; approximately 90 percent
of concentrations for these six pollutants were greater than their associated risk
screening value (or failed screens). This percentage is greater than the percentage for
PXSS. However, nearly all of the measured detections for the pollutants listed for
SPAZ failed screens, ranging from a 73 percent failure rate for ethylbenzene to a
100 percent failure rate for benzene and carbon tetrachloride; for PXSS, the
percentage of screens failed for each individual pollutant is more varied, ranging from
less than 1 percent for lead to 100 percent for six pollutants.
• Concentrations of six pollutants that failed screens for SPAZ contributed to
95 percent of failed screens for SPAZ and therefore were identified as pollutants of
interest for this site.
• Of the pollutants of interest in common for these sites (VOCs only), benzene and
carbon tetrachloride were detected in all valid samples collected and failed
100 percent of screens for each site. Other VOCs, such as 1,2-dichloroethane,
1,3-butadiene, and />dichlorobenzene were detected frequently and also failed the
majority of screens. Formaldehyde and acetaldehyde were detected in all of the valid
samples collected at PXSS and also failed 100 percent of screens for this site.
For each of the data analyses described in the remaining section, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
5-8
-------
5.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Arizona monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria is met (as specified in the
appropriate sections below) and is limited to the site-specific pollutants of interest. However,
site-specific statistical summaries for all pollutants sampled for at PXSS and SPAZ are provided
in Appendices J, M, N, and O.
5.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Arizona monitoring site, as described in Section 3.1. The quarterly
average concentration of a particular pollutant is simply the average concentration of the
preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated for a given year and where method completeness was greater than or equal to
85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
pollutants of interest for the Arizona monitoring sites are presented in Table 5-3, where
applicable. Note that concentrations of the PAHs and metals for PXSS are presented in ng/m3 for
ease of viewing. Also note that if a pollutant was not detected in a given calendar quarter, the
quarterly average simply reflects "0" because only zeros substituted for non-detects were
factored into the quarterly average concentration.
5-9
-------
Table 5-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Arizona Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
Oig/m3)
Q4
Avg
Oig/m3)
Annual
Avg
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Avg
frig/m3)
Phoenix, Arizona - PXSS
Acetaldehyde
33/33/33
NA
NA
NA
NA
NA
58/58/58
2.74
±0.64
2.66
±0.60
2.06
±0.28
3.47
±0.45
2.75
±0.28
Benzene
58/58/58
1.49
±0.50
0.56
±0.12
0.50
±0.14
1.62
±0.43
1.04
±0.21
60/60/60
1.40
±0.40
0.68
±0.18
0.62
±0.19
1.78
±0.39
1.13
±0.19
1.3 -Butadiene
58/56/58
0.28
±0.12
0.09
±0.03
0.06
±0.02
0.35
±0.13
0.20
±0.05
60/52/60
0.32
±0.12
0.11
±0.03
0.08
±0.03
0.36
±0.10
0.22
±0.05
Carbon Tetrachloride
58/58/58
0.63
±0.03
0.61
±0.06
0.67
±0.03
0.59
±0.05
0.63
±0.02
60/60/60
0.60
±0.04
0.67
±0.04
0.56
±0.09
0.61
±0.04
0.61
±0.03
p-Dichlorobenzene
53/26/58
0.18
±0.06
0.11
±0.03
0.07
±0.03
0.24
±0.06
0.15
±0.03
59/40/60
0.19
±0.05
0.15
±0.04
0.15
±0.04
0.30
±0.08
0.20
±0.03
1,2 -Dichloroethane
55/43/58
0.10
±0.03
0.08
±0.02
0.05
±0.01
0.10
±0.01
0.08
±0.01
55/55/60
0.11
±0.02
0.10
±0.01
0.05
±0.02
0.10
±0.01
0.09
±0.01
Ethylbenzene
58/58/58
0.69
±0.28
0.29
±0.08
0.31
±0.10
0.88
±0.26
0.54
±0.12
60/60/60
0.73
±0.23
0.39
±0.13
0.45
±0.13
1.02
±0.23
0.65
±0.11
Formaldehyde
33/33/33
NA
NA
NA
NA
NA
58/58/58
3.21
±0.57
4.34
±0.73
3.57
±0.32
4.17
±0.40
3.80
±0.27
Arsenic (PMi0)a
59/58/59
0.55
±0.16
0.49
±0.16
0.46
±0.14
0.78
±0.21
0.57
±0.09
61/61/61
0.83
±0.50
0.44
±0.09
0.57
±0.25
1.01
±0.28
0.71
±0.16
Naphthalene1
55/55/55
69.06
±40.51
42.38
± 10.11
NA
119.29
± 25.43
74.36
± 15.68
51/51/51
78.55
± 19.41
NA
NA
109.73
± 14.21
NA
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
-------
Table 5-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Arizona Monitoring Sites (Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
Oig/m3)
Q4
Avg
Oig/m3)
Annual
Avg
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Avg
frig/m3)
South Phoenix, Arizona - SPAZ
Benzene
32/32/32
1.92
±0.78
0.83
±0.35
0.64
±0.24
1.92
±0.64
1.28
±0.31
31/31/31
1.93
±0.74
0.81
±0.32
0.67
±0.31
1.82
±0.61
1.33
±0.32
1.3 -Butadiene
32/31/32
0.35
±0.22
0.12
±0.05
0.10
±0.05
0.40
±0.19
0.23
±0.08
30/26/31
0.52
±0.29
0.13
±0.06
0.07
±0.04
0.35
±0.16
0.27
±0.10
Carbon Tetrachloride
32/32/32
0.59
±0.07
0.57
±0.03
0.63
±0.04
0.61
±0.05
0.60
±0.02
31/31/31
0.61
±0.04
0.67
±0.05
0.60
±0.07
0.61
±0.03
0.62
±0.02
p-Dichlorobenzene
32/21/32
0.27
±0.12
0.22
±0.11
0.17
±0.08
0.38
±0.12
0.25
±0.06
30/26/31
0.34
±0.07
0.25
±0.11
0.22
±0.14
0.40
±0.13
0.30
±0.06
1,2 -Dichloroethane
30/24/32
0.08
±0.03
0.07
±0.03
0.05
±0.01
0.10
±0.02
0.07
±0.01
26/23/31
0.09
±0.01
0.09
±0.02
0.01
±0.02
0.09
±0.02
0.07
±0.01
Ethylbenzene
32/32/32
0.92
±0.43
0.42
±0.12
0.39
±0.18
0.96
±0.31
0.66
±0.15
31/31/31
0.88
±0.16
0.47
±0.22
0.54
±0.29
1.04
±0.33
0.74
±0.14
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
-------
Observations for PXSS from Table 5-3 include the following:
• The pollutants of interest with the highest annual average concentrations for PXSS for
2016 are formaldehyde (3.80 ± 0.27 |ig/m3), acetaldehyde (2.75 ± 0.28 |ig/m3), and
benzene (1.13 ± 0.19 |ig/m3). For 2015, annual average concentrations could not be
calculated for the carbonyl compounds due to a contamination issue that resulted in
the invalidation of samples collected on the primary collection system between
January and May and August and November. However, statistical summaries for
2015 are provided in Appendix M for the samples that were valid. For 2015, the
pollutant of interest with the highest annual average concentrations for PXSS is
benzene (1.04 ± 0.21 |ig/m3).
• A review of the quarterly average concentrations of benzene for PXSS shows that the
first and fourth quarter average concentrations of benzene are significantly higher
than the second and third quarter average concentrations, for both years. This
indicates that there is a seasonal tendency in these measurements, with higher
concentrations measured during the cooler months of the year. A similar observation
was made in the 2013 and 2014 NMP reports. A review of the benzene data shows
that all 18 benzene concentrations greater than 2 |ig/m3 were measured at PXSS
between January and March or October and December of either year. Further, of the
52 benzene concentrations greater than 1 |ig/m3, 44 were measured during the first or
fourth quarters of either year (19 during the first quarter and 25 during the fourth
quarter).
• This trend can also be seen in the quarterly average concentrations of 1,3-butadiene
and, to a lesser extent, /?-dichlorobenzene and ethylbenzene. For 1,3-butadiene, 30 of
the 31 highest concentrations, those greater than 0.25 |ig/m3, were measured during
the first or fourth quarters of either year, with 14 measured at PXSS during the first
quarters and 16 measured during the fourth quarters. All 15 concentrations greater
than 0.5 |ig/m3 were measured in November, December, January, and February.
• Although the fourth quarter average concentration of acetaldehyde for 2016 is
considerably higher than the next highest quarterly average shown, the confidence
intervals for most of the quarterly averages indicate that there is considerable
variability in the acetaldehyde measurements. Acetaldehyde concentrations measured
at PXSS in 2016 vary by an order of magnitude, from 0.589 |ig/m3 to 5.83 |ig/m3,
with a similar range if the 2015 data are included. Formaldehyde concentrations
measured in 2016 exhibit considerable variability as well, ranging from 0.96 |ig/m3 to
7.24 |ig/m3. The minimum formaldehyde concentration measured in 2015 is similar to
the minimum measured in 2016, but the maximum concentration measured in 2015
(18.3 |ig/m3) is among the highest formaldehyde concentrations measured across the
program.
• Among the available quarterly average concentrations of naphthalene, the fourth
quarter averages for both 2015 and 2016 are considerably higher than the other
quarterly average shown, both greater than 100 ng/m3. These quarterly averages and
their associated confidence intervals indicate that there is considerable variability in
the naphthalene concentrations measured at PXSS. Concentrations of naphthalene
measured at PXSS range from 3.33 ng/m3 to 213 ng/m3. Several of the highest
5-12
-------
naphthalene concentrations measured at PXSS were measured in November or
December of either year, including 21 greater than 100 ng/m3 (10 in 2015 and 11 in
2016), compared to 11 measured during the other 10 months of the years. An annual
average concentration for 2016 could not be calculated for naphthalene because
issues with the collection system resulting in the invalidation of samples collected
during late summer and early fall led to low completeness. However, statistical
summaries for 2016 are provided in Appendix N.
• Arsenic is the only metal pollutant of interest for PXSS. Arsenic concentrations
measured at PXSS range from 0.08 ng/m3 to 4.22 ng/m3, with all five concentrations
greater than 1.5 ng/m3 measured in 2016. Three of these were measured during the
fourth quarter of 2016, explaining (at least in part) the considerable differences in the
quarterly average concentrations for 2016. The quarterly average concentrations for
2015 are less variable than the ones calculated for 2016.
Observations for SPAZ from Table 5-3 include the following:
• The pollutant of interest with the highest annual average concentrations for SPAZ is
benzene (1.28 ± 0.31 |ig/m3, 2015, and 1.33 ± 0.32 |ig/m3, 2016). Benzene is the only
pollutant of interest with an annual average concentration greater than 1 |ig/m3.
Benzene concentrations measured at SPAZ range from 0.195 |ig/m3 to 4.19 |ig/m3.
• Similar to PXSS, benzene concentrations were highest during the first and fourth
quarters of 2015 and 2016 at SPAZ, as indicated by the quarterly averages shown in
Table 5-3. Nineteen of the 20 benzene concentrations greater than 1.5 |ig/m3 were
measured during the first or fourth quarters of either year. Conversely, all but one of
the 20 benzene concentrations less than 0.75 |ig/m3 were measured between April and
September. The first and fourth quarter average concentrations of benzene for SPAZ
are among the highest quarterly averages calculated across the program for this
pollutant. However, the confidence intervals calculated for these averages also
indicate that the concentrations measured are highly variable.
• The trend in the quarterly averages for benzene is also shown for 1,3-butadiene and
ethylbenzene. All 22 1,3-butadiene concentrations greater than or equal to 0.25 |ig/m3
were measured at SPAZ during the first or fourth quarters of either year. Similarly, all
but one of the 16 ethylbenzene concentrations greater than 1 |ig/m3 was measured at
SPAZ during the first or fourth quarters of either year.
• Concentrations of carbon tetrachloride exhibit the least variability, with the quarterly
average concentrations varying by approximately 0.1 |ig/m3.
• The fourth quarter 2016 average concentration of />dichlorobenzene for SPAZ is the
highest quarterly average concentration of this pollutant among sites sampling this
pollutant. SPAZ's fourth quarter 2015 average concentration of />dichlorobenzene
ranks second highest. In fact, quarterly average concentrations of />dichlorobenzene
for SPAZ account for seven of the 12 quarterly averages greater than 0.2 |ig/m3
(i.e., only one quarterly average for SPAZ is less than 0.2 |ig/m3). Together, the
Phoenix sites account for nine of the top 12 quarterly averages of this pollutant across
the program.
5-13
-------
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for PXSS and
SPAZ from those tables include the following:
• PXSS and SPAZ appear in Tables 4-10 through 4-13 a total of 19 times.
• SPAZ and PXSS have the highest annual average concentrations (both years) of
ethylbenzene among all NMP sites sampling VOCs. Although the highest individual
ethylbenzene concentrations across the program were not measured at these sites,
PXSS and SPAZ have the highest number of ethylbenzene concentrations greater
than 1 |ig/m3 (20 and 16, respectively) among sites sampling this pollutant (the next
highest site has seven).
• SPAZ also has the highest annual average concentrations (both years) of benzene and
p-dichlorobenzene, and the second and third highest annual average concentrations
(both years) of 1,3-butadiene (behind TVKY, 2015). The annual average
concentrations for PXSS for these pollutants also appear in Table 4-10, but of varying
ranks.
• PXSS's 2016 annual average concentration of acetaldehyde ranks third highest annual
among NMP sites sampling carbonyl compounds. An annual average could not be
calculated for 2015. PXSS's 2016 annual average concentration of formaldehyde
ranks tenth highest.
• PXSS's annual average concentrations of the PAH pollutants of interest do not appear
among the 10 highest annual average concentrations in Table 4-12. Annual averages
could not be calculated for 2016. PXSS also does not appear in Table 4-13 for its
annual average concentrations of arsenic.
5.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest. Thus, box plots were created for each of the pollutants listed in Table 5-3
for PXSS and SPAZ. Figures 5-4 through 5-13 overlay the sites' minimum, annual average, and
maximum concentrations for each year onto the program-level minimum, first quartile, median,
average, third quartile, and maximum concentrations, as described in Section 3.4.2.1, and are
discussed below. If an annual average concentration could not be calculated, the range of
concentrations are still provided in the figures that follow.
5-14
-------
Figure 5-4. Program vs. Site-Specific Average Acetaldehyde Concentrations
1
¦
0 2 4 6 8 10 12 14 16 18
Concentration (ug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 5-4 presents the box plot for acetaldehyde for PXSS and shows the following:
• While an annual average concentration could not be calculated for 2015, this figure
shows that the range of concentrations measured was fairly consistent across the two
years of sampling.
• PXSS's 2016 annual average concentration is greater than the program4evel average
concentration (1.67 |ig/m3) as well as the program4evel third quartile (2.11 |ig/m3).
PXSS has the third highest annual average acetaldehyde concentration among NMP
sites sampling this pollutant and where annual average concentrations could be
calculated. Acetaldehyde concentrations measured at PXSS span an order magnitude,
ranging from 0.668 |ig/m3 to 5.89 |ig/m3.
Figure 5-5. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 5-5 presents the box plot for arsenic for PXSS and shows the following:
• The maximum arsenic concentration measured at PXSS in 2016 is nearly three times
higher than the maximum concentration measured in 2015, both of which are greater
than the program-level third quartile.
• The annual average concentration of arsenic for 2016 is slightly higher than the
annual average for 2015, both of which fall between the program-level median
5-15
-------
(second quartile) and third quartile. The annual average for 2016 is similar to the
program-level average concentration 0.70 ng/m3.
Figure 5-6. Program vs. Site-Specific Average Benzene Concentrations
¦
¦
E
1%
¦
IP
vj
¦5
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 5-6 presents the box plots for benzene for both sites and shows the following:
• The range of benzene concentrations measured at PXSS in 2015 is similar to the
range measured at SPAZ. For 2016, a few higher benzene concentrations were
measured at SPAZ compared to PXSS.
• For both years, the annual average concentration for SPAZ is approximately
0.2 |ig/m3 greater than the annual averages for PXSS.
• Although the maximum benzene concentration measured at each Arizona site is
considerably less than the maximum benzene concentration measured across the
program, both sites' annual averages are greater than the program-level average
concentration and third quartile. SPAZ has the highest annual average concentrations
of benzene across the program, and PXSS's annual averages of benzene rank fifth
(2016) and ninth (2015) among the NMP sites sampling this pollutant.
5-16
-------
Figure 5-7. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
PXSS
! Program Max Concentration = 3.90 (xg/m3 j
¦
1 1
E i
Program Max Concentration = 3.90 (.ig/m3
I
A
1
1 *
Q
¦
¦ i
1
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (]ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 5-7 presents the box plots for 1,3-butadiene for both sites and shows the
following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plots in Figure 5-7 because the scale of the box plots would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plots has been reduced to 1.75 |ig/m3.
• The range of 1,3-butadiene concentrations measured at SPAZ in 2015 is fairly similar
to the range measured at PXSS. For 2016, the concentration range for SPAZ appears
considerably larger. This is due to a single non-detect on the low end of the
concentration range and a single concentration greater than 1 |ig/m3 on the upper end
of the range. If these two data points for SPAZ were excluded, the range of
measurements for 2016 would be more similar between the two sites.
• The annual average concentrations for SPAZ are slightly higher than the annual
average concentrations for PXSS, though not significantly so. The annual average
concentrations for these two sites are each two to the three times greater than the
program-level average concentrations (0.086 |ig/m3).
5-17
-------
Figure 5-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Concentration (|ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 5-8 presents the box plots for carbon tetrachloride for both sites and shows the
following:
• The program-level median (0.637 |ig/m3) and average (0.636 |ig/m3) concentrations
for carbon tetrachloride are similar and plotted nearly on top of each other.
• The maximum carbon tetrachloride concentrations measured at these sites are similar
to each other, while the minimum concentrations are more variable. Nine
concentrations measured at PXSS are less than the minimum concentration measured
at SPAZ.
• The annual average concentrations of carbon tetrachloride for the Arizona sites are
similar to each other, all four of which are just less than the program-level average
concentration.
5-18
-------
Figure 5-9. Program vs. Site-Specific Average />-Dichlorobenzene Concentrations
PXSS
1
r
Program Max Concentration =
2.78 |ig/m3 i
'
1
1 A
1
Program MaxConcentration =
2.78 |ig/m3
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 5-9 presents the box plots for p-dichlorobenzene for both sites and shows the
following:
• The program-level maximum /;-dichlorobenzene concentration (2.78 |ig/m3) is not
shown directly on the box plots in Figure 5-9 because the scale of the box plots would
be too large to readily observe data points at the lower end of the concentration range.
Thus, the scale of the box plots has been reduced. The program-level first and second
quartiles are both zero for p-dichlorobenzene and therefore not visible on the box
plots.
• The range of p-dichlorobenzene concentrations measured at SPAZ is larger than the
range measured at PXSS, though the maximum concentrations for both sites are
considerably less than the maximum concentration measured across the program.
However, the maximum concentrations measured at SPAZ and PXSS are still among
some of the highest measured across the program.
• SPAZ has the two highest annual average concentrations of p-dichlorobenzene among
the NMP sites sampling VOCs; PXSS has the fourth and fifth highest annual average
concentrations of this pollutant. The annual average concentrations for SPAZ and
PXSS are several times greater than the program-level average concentration
(0.047 |ig/m3).
• Six non-detects of p-dichlorobenzene were measured at PXSS over the two years of
sampling while a single non-detect was measured at SPAZ.
5-19
-------
Figure 5-10. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
PXSS
Program Max Concentration =45.8 p.g/m3 i
br
"
| Program Max Concentration = 45.8 (.ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (]ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 5-10 presents the box plots for 1,2-dichloroethane for both sites and shows the
following:
• The scale of the box plots in Figure 5-10 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• All of the concentrations of 1,2-dichloroethane measured at PXSS and SPAZ are less
than the program-level average concentration of 0.30 |ig/m3, which is being driven by
the measurements at the upper end of the concentration range.
• Each of the annual average concentrations for PXSS and SPAZ is less than 0.1 |ig/m3,
and three of the four annual averages are less than the program-level median
concentration (0.081 |ig/m3).
5-20
-------
Figure 5-11. Program vs. Site-Specific Average Ethylbenzene Concentrations
¦
I
w 1
Q
1 1
,
a
1
V
Q
¦
0 0.5 1 1.5 2 2.5 3
Concentration (|ig/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 5-11 presents the box plots for ethylbenzene for both sites and shows the
following:
• The range of ethylbenzene concentrations measured at SPAZ in 2015 is similar to the
range of ethylbenzene concentrations measured at PXSS for 2015. For 2016, a few
higher ethylbenzene concentrations were measured at PXSS compared to SPAZ.
Together, PXSS (20) and SPAZ (16) have the highest number of ethylbenzene
concentrations greater than 1 |ig/m3 among NMP sites sampling this pollutant (the
next highest is seven).
• The annual average concentrations of ethylbenzene for these two sites are two to
three times greater than the program-level average. These sites have the highest
annual average concentrations of ethylbenzene among NMP sites sampling this
pollutant.
Figure 5-12. Program vs. Site-Specific Average Formaldehyde Concentrations
1
,
¦
a
¦i
mr
0 3 6 9 12 15 18 21 24 27
Concentration (|ug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
5-21
-------
Figure 5-12 presents the box plot for formaldehyde for PXSS and shows the following:
• The minimum formaldehyde concentration measured at PXSS in 2015 is similar to
the minimum measured in 2016. This is not true for the maximum concentration.
Despite the limited number of valid samples for 2015, the maximum concentration
measured in 2015 is two and half times greater than the maximum concentration
measured in 2016. PXSS is one of only six sites where a formaldehyde concentration
greater than 15 |ig/m3 was measured (though there is only one).
• The annual average concentration of formaldehyde for 2016 is greater than the
program-level average concentration and similar to the program-level third quartile.
PXSS has the tenth highest annual average concentration of formaldehyde among
NMP sites sampling carbonyl compounds.
Figure 5-13. Program vs. Site-Specific Average Naphthalene Concentrations
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
• O
Figure 5-13 presents the box plot for naphthalene for PXSS and shows the following:
• The range of naphthalene concentrations measured at PXSS in 2015 is somewhat
larger than the range of concentrations measured in 2016. There is a considerable
difference in the minimum concentrations measured each year (3.33 ng/m3 for 2015
and 23.5 ng/m3 for 2016). PXSS is one of only five NMP sites with an individual
naphthalene concentration less than 5 ng/m3. A number of samples with consistently
low measurements was collected at PXSS during March 2015.
• The annual average naphthalene concentration for PXSS for 2015 falls between the
program-level average concentration (61.23 ng/m3) and the program-level third
quartile (82.15 ng/m3). An annual average concentration could not be calculated for
2016 because the completeness criteria was not met.
5.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
PXSS has sampled PMio metals under the NMP since 2006; in addition, SPAZ began sampling
VOCs and PXSS began sampling VOCs, carbonyl compounds, and PAHs under the NMP in
5-22
-------
2007. Thus, Figures 5-14 through 5-29 present the 1-year statistical metrics for each of the
pollutants of interest first for PXSS, then for SPAZ. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
Figure 5-14. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at PXSS
7.0
r
T
40
1
c
O
c
m
o
u
o-
o-
~
o
T
T
L?-
T
2007
2008
2009
2010 2
2011
Year
2012
2013
2014
2015
3
20
16
1
O
th Percen
ile
-
M'
nlmum — Mec
ian
-
Ma
ximum
O 95t
h Per
centile
•~••• Aver
age
1
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
2 Some statistical metrics are not presented because data from Feb 2010 to Mar 2011 was invalidated.
3 A 1-year average is not presented due to a contamination issue affecting numerous samples.
Observations from Figure 5-14 for acetaldehyde concentrations measured at PXSS
include the following:
• PXSS began sampling acetaldehyde under the NMP in July 2007. Because a full
year's worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided. In addition, much of the
data between February 2010 and March 2011 was invalidated due to maintenance
issues on the primary collection system. No statistical metrics are provided for 2010
due to the low number of valid measurements. The range of measurements is
provided for 2011, although a 1-year average is not provided. Similarly, no 1-year
average is provided for 2015.
5-23
-------
• The maximum acetaldehyde concentration (6.21 |ig/m3) was measured on
January 1, 2009, although this measurement is not significantly higher than maximum
concentrations measured in other years. Acetaldehyde concentrations greater than
5 |ig/m3 have been measured every year except 2008 (and 2010, for which no data is
provided).
• A distinct trend is hard to identify because several 1-year average concentrations
could not be calculated. However, the 1-year averages shown vary by less than
0.4 |ig/m3, ranging from 2.52 |ig/m3 (2014) to 2.90 |ig/m3 (2012). The median
concentrations exhibit more variability, ranging from 2.17 |ig/m3 (2015) to
3.24 |ig/m3 (2007).
Figure 5-15. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
PXSS
2
t
2006 2007
2011
Year
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
• Linear (Median)
Observations from Figure 5-15 for arsenic concentrations measured at PXSS include the
following:
• The maximum arsenic concentration (6.73 ng/m3) was measured on
December 26, 2007. The second highest concentration was measured on
January 1, 2016 (4.22 ng/m3). Only one additional concentration greater than 3 ng/m3
has been measured at PXSS (3.05 ng/m3, 2011). In total, 18 arsenic measurements
greater than or equal to 2 ng/m3 have been measured at PXSS, with at least one
measured each year of sampling except 2013 and 2015.
• After several years of decreasing slightly, the 1-year average concentration increased
significantly from 2010 to 2011, after which additional decreasing is shown through
5-24
-------
2013. The 1-year average concentration is at a minimum for 2013 (0.49 ng/m3). The
1-year average increases slightly each year following 2013, even for 2015, when the
smallest range of concentrations was measured.
Figure 5-16. Yearly Statistical Metrics for Benzene Concentrations Measured at PXSS
o-
¦O"
o
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-16 for benzene concentrations measured at PXSS include the
following:
• PXSS began sampling VOCs under the NMP in July 2007. Because a full year's
worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided.
• The maximum benzene concentration shown was measured on January 1, 2009
(5.22 |ig/m3), the same day the maximum acetaldehyde concentration was measured.
Four additional measurements greater than 4 |ig/m3 have been measured at this site
(one measured each year during 2007, 2008, 2009, and 2011).
• All but one of the 36 highest benzene concentrations (those greater than 3.0 |ig/m3)
were measured during the first or fourth quarter of any given year (and the exception
was measured in late September). Further, of the 128 benzene concentrations greater
than or equal to 2 |ig/m3, all but 10 were measured during the first or fourth quarters
of a given year; those other 10 were all measured in either April or September, or just
outside the first or fourth calendar quarters.
5-25
-------
• The median concentration increased significantly from 2008 to 2009 and is greater
than the 1-year average concentration for 2009. A review of the data shows that the
number of concentrations greater than 2 |ig/m3 increased from 15 in 2008 to 24 in
2009. After the increase from 2008 to 2009, the median benzene concentration has a
decreasing trend through 2015, and is at a minimum for 2014 and 2015 (both at
0.74 |ig/m3). The number of benzene concentrations greater than 2 |ig/m3 decreased
to 12 in 2010, with the number ranging from eight (2015) to 14 (2011) for each of the
remaining years.
• The 1-year average concentration exhibits a similar pattern as the median
concentration, with the 1-year average concentration (1.05 |ig/m3) at a minimum for
2014, although there is relatively little change between 2013 and 2015.
• Despite having one of the smallest range of concentrations measured, the 1-year
average and median concentrations for 2016 both exhibit increases. For the 1-year
average, the increase is slight (less than 0.1 |ig/m3) while the median concentration is
approaching 1 |ig/m3 for the first time since 2012.
Figure 5-17. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
PXSS
-o-
2008 2009
O 5th Percentile
— Minimum
2011 2012
Year
t Lr
2013 2014
"I"
2016
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
5-26
-------
Observations from Figure 5-17 for 1,3-butadiene concentrations measured at PXSS
include the following:
• The maximum 1,3-butadiene concentration (1.09 |ig/m3) was measured on
December 11, 2011. The only other concentration greater than 1.0 |ig/m3 was
measured at PXSS on January 1, 2009, the same day that the maximum benzene and
acetaldehyde concentrations were measured.
• All but one of the 138 1,3-butadiene concentrations greater than 0.30 |ig/m3 were
measured during the first or fourth quarters. The one concentration not measured
during the first or fourth quarters was measured in September.
• The 1-year average 1,3-butadiene concentration exhibits relatively little change over
the period shown, ranging from 0.20 |ig/m3 (both 2014 and 2015) to 0.23 |ig/m3 (both
2009 and 2011). The median concentration ranges from 0.10 |ig/m3 (2015) to
0.21 |ig/m3 (2007).
• The minimum concentrations, and in one case the 5th percentile, for several years are
zero, indicating the substitution of zeros for non-detects. Ten non-detects of
1,3-butadiene have been measured at PXSS since the onset of VOC sampling at
PXSS under the NMP. Five of these were measured in 2011, with one each measured
in 2007, 2013, and 2014, and two measured in 2010. Non-detects were not measured
in 2008, 2009, 2012, 2015, or 2016.
Figure 5-18. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at PXSS
1.4
Concentration (ng/m3)
3 O O h
\—<> |0
X
o
r-9-
pL
——
0-
r1
rj-
r«—|
T
1
¦¦
i
20071 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
5-27
-------
Observations from Figure 5-18 for carbon tetrachloride concentrations measured at PXSS
include the following:
• Seven concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured at PXSS since the onset of sampling in 2007, with five measured in 2008
and two measured in 2009.
• The box and whisker plots for 2007, 2010, 2011, 2014, and 2015 appear "inverted,"
with the minimum concentration extending farther away from the majority of the
measurements rather than the maximum concentration, which is more common (see
benzene or 1,3-butadiene as examples).
• All of the carbon tetrachloride concentrations measured in 2007 are less than the
1-year average and median concentrations calculated for 2008. However, the
concentrations measured in 2007 represent only one-half of the year.
• The 1-year average concentration exhibits a decreasing trend between 2008 and 2011.
Although the range of concentrations measured decreased for 2012, an increase is
shown for the 1-year average and median concentrations for 2012. This is mostly a
result of a change at the lower end of the concentration range. The number of
concentrations less than 0.6 |ig/m3 measured in 2011 is 23; the number of
concentrations less than 0.6 |ig/m3 measured in 2012 is five.
• All of the statistical parameters for carbon tetrachloride exhibit a decrease for 2013.
The majority of concentrations fall into a similar range between 2013 and 2015, as
little change is shown in most of the statistical parameters during this period.
• The 5th percentile for 2016 exhibits a considerable decrease compared to the 5th
percentiles for several of the previous years, and is at a minimum for the period of
sampling. This year has more carbon tetrachloride measurements less than 0.4 |ig/m3
(four) than any other year of sampling (prior to 2015, no other year has more than one
carbon tetrachloride concentration less than 0.4 |ig/m3; 2015 has two and 2016 has
four).
5-28
-------
Figure 5-19. Yearly Statistical Metrics for /7-Dichlorobenzene Concentrations
Measured at PXSS
§ 0.5
o..
O.
o
•O
2007 2008 2009
2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile Averc
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-19 for p-dichlorobenzene concentrations measured at PXSS
include the following:
• The three highest concentrations of p-dichlorobenzene were all measured in
November 2007 and are the only ones greater than 0.75 |ig/m3 measured at PXSS.
• The maximum, 95th percentile, 1-year average, and median concentrations all exhibit
a significant decreasing trend through 2010. The minimum concentration and 5th
percentile also decreased each year between 2008 and 2010.
• Each of the statistical parameters increased from 2010 to 2011, with the exceptions of
the minimum and 5th percentile, as several non-detects were measured in both years.
Although the range within which the majority of the concentrations fall tightened up
for 2012 and 2013, little change is shown for the 1-year average or median
concentrations between 2011 and 2013.
• Each of the statistical parameters decreased at least slightly for 2014, except the
minimum concentration, which has remained constant since 2010. Additional
decreases in most of the statistical parameters are shown for 2015, with the minimum
and 5th percentile remaining unchanged.
• All of the statistical parameters exhibit increases for 2016, including the minimum
concentration, which is greater than zero for the first time since 2009.
5-29
-------
• The number of non-detects has varied between none (several years) and nine (2010).
Figure 5-20. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at PXSS
O"
O"
200?
2011 2012
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-20 for 1,2-dichloroethane concentrations measured at PXSS
include the following:
• There were no measured detections of 1,2-dichloroethane in 2007. The number
increased gradually each year through 2011 (12), then significantly for 2012 (47),
with more than 50 measured detections in each of the last three years.
• The median concentration is zero for each year through 2011, indicating that at least
50 percent of the measurements were non-detects for the first 5 years of sampling.
• The number of measured detections increased markedly for 2012, and the median and
1-year average concentrations increased correspondingly. The median concentration
is greater than the 1-year average concentration for each year between 2012 and 2014.
This is because there were still many non-detects (or zeros) factoring into the 1-year
average concentration for 2012 (14), 2013 (23), and 2014 (8), which pull the 1-year
average concentrations down in the same manner that a maximum or outlier
concentration can drive the average up. The number of measured detections exceeds
90 percent for 2015 and 2016, and the 1-year average concentration for these two
years is greater than the median concentration.
5-30
-------
Figure 5-21. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at PXSS
J-J-
T
¦ O"
I
o
o
~
.o
T
—Q-
T
T
T
L2"
20071 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-21 for ethylbenzene concentrations measured at PXSS
include the following:
• The maximum concentration of ethylbenzene measured at PXSS (2.16 |ig/m3) was
measured on January 1, 2009, the same day that the maximum concentrations of
several pollutants were measured at this site. The next four highest concentrations
were all measured in November 2011, including the only other concentration greater
than 2 |ig/m3 measured at PXSS (2.01 |ig/m3).
• Similar to benzene and 1,3-butadiene, the highest ethylbenzene concentrations were
measured most often during the first and fourth quarters of the years. Of the 119
highest concentrations of ethylbenzene (those greater than 1.0 |ig/m3), 108 were
measured between January and March or October and December of any given year.
• The median ethylbenzene concentration decreases each year through 2009, then
returns to 2008 levels for 2010, and returns to 2007 levels for 2011. All of the
statistical parameters shown increased from 2010 to 2011. Nearly twice the number
of concentrations greater than 1 |ig/m3 were measured in 2011 (20) than in each the
previous years, which vary between nine (2008) and 11 (both 2007 and 2010).
• A significant decreasing trend in the 1-year average concentration is shown between
2011 and 2015, with the 1-year average concentration at a minimum for 2015
(0.55 |ig/m3). The median concentration is at a minimum for 2014 and 2015
(0.38 |ig/m3).
5-31
-------
• All of the statistical parameters except the 95th percentile exhibit an increase for
2016, with the concentration profile for 2016 resembling the concentration profile for
2013. The number of ethylbenzene concentrations between 0.5 |ig/m3 and 1.5 |ig/m3
measured nearly doubled between 2015 (17) and 2016 (31).
Figure 5-22. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at PXSS
Maximum Concentration
for 2015 is 18.34 ^g/m3
2011 2012
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
2 Some statistical metrics are not presented because data from Feb 2010 to Mar 2011 was invalidated.
3 A 1-year average is not presented due to a contamination issue affecting numerous samples.
Observations from Figure 5-22 for formaldehyde concentrations measured at PXSS
include the following:
• PXSS began sampling formaldehyde under the NMP in July 2007. Because a full
year's worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided. In addition, much of the
data between February 2010 and March 2011 was invalidated due to maintenance
issues on the primary collection system. No statistical metrics are provided for 2010
due to the low number of valid measurements. The range of measurements is
provided for 2011, although a 1-year average is not provided. Similarly, no 1-year
average is provided for 2015.
• The maximum formaldehyde concentration was measured at PXSS on November 26,
2015 (18.34 |ig/m3) and is more than twice the next highest formaldehyde
concentration (7.56 |ig/m3). The only formaldehyde concentrations greater than
7 |ig/m3 were measured in either 2007 (3), 2015 (1), and 2016 (1).
5-32
-------
• The median concentration for 2007 is nearly 5 |ig/m3 (the median concentrations for
all of the years that follow are all less than 4 |ig/m3) and is greater than the 95th
percentile for two of the years of sampling (2008 and 2014).
• Even though the maximum concentration measured in 2015 is considerably higher
than any other formaldehyde concentrations measured since the onset of sampling,
the median concentration (3.30 |ig/m3) is at a minimum for 2015. It is important to
note that a large number of carbonyl compound samples collected in 2015 were
invalidated due to contamination issues with the primary collection system. Only
about half of the samples for 2015 were retained after the contamination issue was
resolved.
Figure 5-23. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PXSS
•o
o
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile ••~~~~• Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
2 A 1-year average is not presented due to collection system issues affecting many samples collected in
2016."
Observations from Figure 5-23 for naphthalene concentrations measured at PXSS include
the following:
• PXSS began sampling PAHs under the NMP in July 2007. Because a full year's
worth of data is not available, a 1-year average for 2007 is not presented, although the
range of measurements is provided. Issues with the collection system resulted in
completeness less than 85 percent for 2016, and thus, a 1-year average concentration
for 2016 is not presented.
5-33
-------
• The maximum naphthalene concentration was measured at PXSS on
December 20, 2008 (400 ng/m3), with a concentration of similar magnitude measured
12 days later on January 1, 2009 (386 ng/m3). This is the same day that the maximum
concentrations of several of PXSS's pollutants of interest were measured. Two
additional naphthalene measurements greater than 300 ng/m3 have been measured at
PXSS, one in December 2012 and one in January 2014.
• Many of the statistical parameters are at a maximum for 2009. The median
concentration, or midpoint, for 2009 is 107 ng/m3; 2009 is the only year in which
naphthalene concentrations greater than 100 ng/m3 account for more than half of the
measurements. The median concentrations for the other years are less than 100 ng/m3,
ranging from 52.30 ng/m3 (2015) to 84.10 ng/m3 (2010), and have a steady decreasing
trend between 2009 and 2015. The 1-year average concentration is also at a maximum
for 2009 (120.17 ng/m3) and at a minimum for 2015 (74.36 ng/m3).
• Naphthalene concentrations measured in 2016 have the smallest range of
measurements shown, with 2016 the only full year of sampling for which less than
200 ng/m3 separates the minimum and maximum concentrations measured. This data
set missed the completeness criteria of 85 percent, as established in Section 2.4, by
only 1 percent. Although a 1-year average concentration was not calculated for 2016,
the median concentration shown in Figure 5-23 is at its highest since 2010.
Figure 5-24. Yearly Statistical Metrics for Benzene Concentrations Measured at SPAZ
T
2007 2008
2010 2011 2012 2013
Year
2014 2015
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
5-34
-------
Observations from Figure 5-24 for benzene concentrations measured at SPAZ include the
following:
• SPAZ also began sampling VOCs under the NMP in July 2007. Because a full year's
worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of concentrations measured is provided.
• The maximum benzene concentration shown was measured on January 27, 2011
(5.41 |ig/m3) and is the only benzene concentration greater than 5 |ig/m3 measured at
SPAZ. Six additional measurements greater than 4 |ig/m3 have been measured at this
site (one for each year of sampling prior to 2012 and one in 2016).
• Benzene concentrations measured at SPAZ exhibit a seasonal tendency; 65 of the 69
benzene concentrations greater than 2 |ig/m3 were measured at SPAZ during the first
or fourth quarters of any given year.
• The 1-year average and median concentrations are fairly similar to each other for all
years except 2011, when more than 0.5 |ig/m3 separates them. The largest range of
benzene concentrations was measured in 2011, spanning more than 5 |ig/m3, and the
maximum concentration for the period shown was measured in 2011. This year has
the highest number of benzene concentrations greater than 3 |ig/m3 (5) but also the
highest number of benzene concentrations less than 1.5 |ig/m3 (18) for the years prior
to 2013.
• After several years of increasing, the maximum and 95th percentile decreased
considerably for 2012 and again for 2013, with little change shown for 2014. The
range of benzene concentrations measured is at a minimum for 2013, spanning less
than 2 |ig/m3. The range of measurements increases for 2015 and again for 2016.
• The 1-year average concentrations changed little between 2009 and 2011, then
decreased significantly from 2011 to 2013, with little change for 2014; afterwards, the
1-year average concentration begins to increase, though not significantly. The 1-year
average concentrations range from 1.07 |ig/m3 (2013) to 1.69 |ig/m3 (2010) over the
period shown; the median concentration exhibits more variability during this time
frame.
5-35
-------
Figure 5-25. Yearly Statistical Metrics for 1,3-Butadiene Concentrations
Measured at SPAZ
o
2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median
Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-25 for 1,3-butadiene concentrations measured at SPAZ
include the following:
• The maximum 1,3-butadiene concentration was measured at SPAZ on January 1,
2016 (1.42 |ig/m3). The only other 1,3-butadiene concentration greater than 1 |ig/m3
was measured at SPAZ on January 27, 2011 (1.29 |ig/m3). Six additional
1,3-butadiene concentrations greater than 0.75 |ig/m3 have been measured at SPAZ
(one in 2007, two in 2010, one in 2011, and two in 2015).
• Ninety-five of the 99 concentrations greater than 0.25 |ig/m3 were measured at SPAZ
during the first or fourth quarters of any given year, similar to the seasonal tendency
in the benzene measurements.
• The maximum concentration and 95th percentile increased each year after 2008
through 2011, while the 5th percentile remained fairly static. This indicates that more
of the concentrations measured were on the higher end of the concentration range for
each of these years. For 2012, the maximum concentration and 95th percentiles are
lower, with the maximum concentration for 2012 less than the 95th percentile for
2011. This is also true for 2013, where the maximum concentration is less than the
95th percentile for the preceding year. The 95th percentile continued its decrease for
2014, although the maximum concentration measured increased. The majority of
concentrations measured in 2014, as indicated by the 5th and 95th percentiles, falls
into the tightest range among the years shown. This range expands considerably for
5-36
-------
2015, resembling the concentration profile for 2011. The concentration profile for
2016 resembles the concentration profile for 2012, with the exception of the
maximum concentration.
• The 1-year average concentration increases steadily between 2009 and 2011, then
decreases through 2014, with the 1-year average concentration falling to less than
0.2 |ig/m3 for the first time in 2014. The 1-year average concentration increases in
2015 and again in 2016. However, the 1-year average concentrations vary by only
0.1 |ig/m3, ranging from 0.19 |ig/m3 (2014) to 0.29 |ig/m3 (2011), and confidence
intervals calculated indicate these changes are not statistically significant.
Figure 5-26. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at SPAZ
i
T
2008 2009
2010 2011 2012 2013
Year
2014 2015
O 5th Percentile
— Minimum
Maximum O 95th Percentile Averc
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-26 for carbon tetrachloride concentrations measured at SPAZ
include the following:
• Two concentrations of carbon tetrachloride greater than 1.0 |ig/m3 have been
measured at SPAZ since the onset of sampling. One was measured in 2008 and one
was measured in 2011 (although another concentration just less than 1 |ig/m3 was also
measured in 2011). Conversely, two non-detects of carbon tetrachloride have been
measured at SPAZ, one in 2009 and one in 2011.
• The box and whisker plots for this pollutant appear "inverted" for several years, with
the minimum concentration extending farther away from the majority of
5-37
-------
measurements for several years rather than the maximum (see benzene or
1,3-butadiene as examples), which is more common.
• With the exception of 2012, the 1-year average concentration exhibits a slight
decreasing trend over most of the years shown, reaching a minimum for 2014
(0.60 |ig/m3). However, the overall change for this period is less than 0.12 |ig/m3.
Slight increases are shown for 2015 and 2016.
• The range of concentrations measured is at a minimum for 2015, with less than
0.25 |ig/m3 separating the minimum and maximum concentrations measured.
Figure 5-27. Yearly Statistical Metrics for /7-Dichlorobenzene Concentrations
Measured at SPAZ
¦o
I
2016
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-27 for p-dichlorobenzene concentrations measured at SPAZ
include the following:
• The widest range of p-dichlorobenzene concentrations measured is shown for 2008
(ranging from a single non-detect to 0.90 |ig/m3), while the range of concentrations
measured the following year is roughly half as large. A review of the data shows that
the number of p-dichlorobenzene concentrations greater than 0.3 |ig/m3 decreased by
half from 2008 (8) to 2009 (4). All of the statistical metrics exhibit increases from
2009 to 2010, with the number of p-dichlorobenzene concentrations greater than
0.3 |ig/m3 increasing nearly four-fold (15).
f V
2008 2009
2010 2011 2012 2013
Year
o*
UjU
T
2014 2015
5-38
-------
• The 1-year average concentration decreased from 2008 to 2009, increased for 2010,
then decreased slightly each year between 2011 and 2014. The 1-year average
concentration exhibits an increase for 2015 and again for 2016, increasing by nearly
50 percent between 2014 and 2016, and reaching a maximum of 0.30 |ig/m3.
However, confidence intervals calculated for these averages indicate that the changes
are not statistically significant.
Figure 5-28. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at SPAZ
5th Percentile
95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-28 for 1,2-dichloroethane concentrations measured at SPAZ
include the following:
• There were no measured detections of 1,2-dichloroethane in 2007 and only one
measured in 2008. The number of measured detections increased slightly each year
through 2011, then increased substantially in 2012, with measured detections
accounting for nearly 87 percent of the measurements. Between 2012 and 2016,
measured detections account for between 61 (2013) and 94 percent (2015) of the
measurements.
• The median concentration is zero for each year until 2012, indicating that at least
50 percent of the measurements were non-detects. As the number of measured
detections increase, so do the corresponding central tendency statistics shown in
Figure 5-28.
5-39
-------
The median concentration is greater than the 1-year average concentration for each
year beginning with 2012. This is because the non-detects (or zeros) factored into
each 1-year average concentration are pulling the average down in the same manner
that a maximum or outlier concentration can drive the average upward. These two
central tendency statistics are closest for 2015 when there were only two non-detects.
Between 2012 and 2016, the 1-year average concentrations vary by less than
0.025 |ig/m3, despite the apparent fluctuations shown in Figure 5-28. Confidence
intervals calculated for the last five years of sampling indicate that the changes shown
in the 1-year average concentrations are not statistically significant due to the
variability in the concentrations measured.
Figure 5-29. Yearly Statistical Metrics for Ethylbenzene Concentrations
Measured at SPAZ
o.
20071 2008 2009
2011 2012
Year
2013 2014
O 5th Percentile — Minimum — Median
Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2007.
Observations from Figure 5-29 for ethylbenzene concentrations measured at SPAZ
include the following:
• Two concentrations of ethylbenzene greater than 3.0 |ig/m3 have been measured at
SPAZ (one in 2007 and one in 2011). All 10 concentrations greater than 2.0 |ig/m3
were measured in either 2007 or 2011 (five each year).
• The median concentration is at a maximum for 2007, then decreases by half for 2008.
(2007 includes only half a year's worth of samples). Further decreases are shown for
2009, followed by an increase in 2010 and again in 2011. The median concentration
5-40
-------
decreases for 2012, with additional decreases for each year through 2015. The median
concentration for 2016 increases for the first time since 2011.
• The 1-year average concentrations have a similar pattern as the median concentration
through 2014. A significant increasing trend is shown between 2009 and 2011, which
is followed by a significant decreasing trend through 2014. The 1-year average
concentration exhibits an increase for 2015 and again for 2016. These patterns are
similar to the patterns shown for 1,3-butadiene in Figure-5-25.
• The only non-detects of ethylbenzene were measured during the first two full-years of
sampling at SPAZ.
5.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Arizona monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
5.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Arizona monitoring sites, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 5-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
5-41
-------
Table 5-4. Risk Approximations for the Arizona Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Phoenix, Arizona - PXSS
Acetaldehyde
0.0000022
0.009
33/33
NA
NA
NA
58/58
2.75
±0.28
6.04
0.31
Benzene
0.0000078
0.03
58/58
1.04
±0.21
8.13
0.03
60/60
1.13
±0.19
8.83
0.04
1.3 -Butadiene
0.00003
0.002
58/58
0.20
±0.05
5.86
0.10
60/60
0.22
±0.05
6.64
0.11
Carbon Tetrachloride
0.000006
0.1
58/58
0.63
±0.02
3.76
0.01
60/60
0.61
±0.03
3.65
0.01
p-Dichlorobenzene
0.000011
0.8
53/58
0.15
±0.03
1.65
<0.01
59/60
0.20
±0.03
2.21
<0.01
1,2-Dichloroethane
0.000026
2.4
55/58
0.08
±0.01
2.09
<0.01
55/60
0.09
±0.01
2.36
<0.01
Ethylbenzene
0.0000025
1
58/58
0.54
±0.12
1.35
<0.01
60/60
0.65
±0.11
1.63
<0.01
Formaldehyde
0.000013
0.0098
33/33
NA
NA
NA
58/58
3.80
±0.27
49.36
0.39
Arsenic (PMio)
0.0043
0.000015
59/59
0.57
±0.09
2.45
0.04
61/61
0.71
±0.16
3.07
0.05
Naphthalene
0.000034
0.003
55/55
74.36
± 15.68
2.53
0.02
51/51
NA
NA
NA
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
-------
Table 5-4. Risk Approximations for the Arizona Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
South Phoenix, Arizona - SPAZ
Benzene
0.0000078
0.03
32/32
1.28
±0.31
10.00
0.04
31/31
1.33
±0.32
10.36
0.04
1.3 -Butadiene
0.00003
0.002
32/32
0.23
±0.08
7.02
0.12
30/31
0.27
±0.10
8.25
0.14
Carbon Tetrachloride
0.000006
0.1
32/32
0.60
±0.02
3.63
0.01
31/31
0.62
±0.02
3.73
0.01
p-Dichlorobenzene
0.000011
0.8
32/32
0.25
±0.06
2.80
<0.01
30/31
0.30
±0.06
3.33
<0.01
1,2-Dichloroethane
0.000026
2.4
30/32
0.07
±0.01
1.87
<0.01
26/31
0.07
±0.01
1.90
<0.01
Ethylbenzene
0.0000025
1
32/32
0.66
±0.15
1.64
<0.01
31/31
0.74
±0.14
1.84
<0.01
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
-------
Observations for PXSS from Table 5-4 include the following:
• The pollutants of interest with the highest annual average concentrations for 2015
(without the carbonyl compounds) are different than those for 2016 (with the
carbonyl compounds). The pollutants of interest with the highest annual average
concentrations for 2015 are benzene, carbon tetrachloride, and ethylbenzene. The
pollutants of interest with the highest annual average concentrations for 2016 are
formaldehyde, acetaldehyde, and benzene. For the VOCs, the annual averages for
2015 are similar to the annual averages for 2016.
• Based on the annual averages for 2015 and cancer UREs, all of the cancer risk
approximations are less than 10 in-a-million; benzene has the highest cancer risk
approximation (8.13 in-a-million). For 2016, the pollutant with the highest cancer risk
approximation is formaldehyde (49.36 in-a-million).
• None of the pollutants of interest for PXSS have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for 2015 is benzene (0.10). The pollutant with the highest noncancer
hazard approximation for 2016 is formaldehyde (0.39).
Observations for SPAZ from Table 5-4 include the following:
• The pollutants with the highest annual average concentrations for SPAZ are the same
for both years: benzene, ethylbenzene, and carbon tetrachloride. Only benzene's
annual average concentrations are greater than 1 |ig/m3.
• Based on the annual averages and cancer UREs, benzene has the highest cancer risk
approximations for both years, followed by 1,3-butadiene, and carbon tetrachloride.
The cancer risk approximations for benzene (10.00 in-a-million for 2015 and
10.36 in-a-million for 2016) are the only ones greater than 10 in-a-million for SPAZ
and are the highest cancer risk approximations calculated across the program for this
pollutant.
• None of the pollutants of interest for SPAZ have noncancer hazard approximations
greater than 1.0, indicating no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for SPAZ is 1,3-butadiene (0.12 for 2015 and 0.14 for 2016).
5-44
-------
5.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 5-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 5-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.3.4. Lastly,
Table 5-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 5-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 5-5. Table 5-6
presents similar information but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more
in-depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 5.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
5-45
-------
Table 5-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Arizona Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Phoenix, Arizona (Maricopa County) - PXSS
Formaldehyde
944.58
Formaldehyde
1.23E-02
Formaldehyde
49.36
Benzene
897.31
Naphthalene
8.70E-03
Benzene
8.83
Ethylbenzene
558.42
Benzene
7.00E-03
Benzene
8.13
Acetaldehyde
514.05
1,3-Butadiene
3.82E-03
1,3-Butadiene
6.64
Naphthalene
255.84
POM, Group 2b
1.72E-03
Acetaldehyde
6.04
Bis(2-ethylhexyl) phthalate
149.76
Ethylbenzene
1.40E-03
1,3-Butadiene
5.86
1.3 -Butadiene
127.30
POM, Group 2d
1.18E-03
Carbon Tetrachloride
3.76
POM, Group 2b
19.51
Acetaldehyde
1.13E-03
Carbon Tetrachloride
3.65
POM, Group 2d
13.42
POM, Group 5a
1.07E-03
Arsenic
3.07
2,4-Toluene diisocyanate
12.83
Arsenic, PM
3.95E-04
Naphthalene
2.53
South Phoenix, Arizona (Maricopa County) - SPAZ
Formaldehyde
944.58
Formaldehyde
1.23E-02
Benzene
10.36
Benzene
897.31
Naphthalene
8.70E-03
Benzene
10.00
Ethylbenzene
558.42
Benzene
7.00E-03
1,3-Butadiene
8.25
Acetaldehyde
514.05
1,3-Butadiene
3.82E-03
1,3-Butadiene
7.02
Naphthalene
255.84
POM, Group 2b
1.72E-03
Carbon Tetrachloride
3.73
Bis(2-ethylhexyl) phthalate
149.76
Ethylbenzene
1.40E-03
Carbon Tetrachloride
3.63
1,3-Butadiene
127.30
POM, Group 2d
1.18E-03
/?-Dichlorobcnzcnc
3.33
POM, Group 2b
19.51
Acetaldehyde
1.13E-03
/?-Dichlorobcnzcnc
2.80
POM, Group 2d
13.42
POM, Group 5a
1.07E-03
1,2-Dichloroethane
1.90
2,4-Toluene diisocyanate
12.83
Arsenic, PM
3.95E-04
1,2-Dichloroethane
1.87
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 5-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Arizona Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Noncancer
Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Phoenix, Arizona (Maricopa County) - PXSS
Toluene
3,935.87
Acrolein
3,491,697.55
Formaldehyde
0.39
Xylenes
2,003.31
2,4-Toluene diisocyanate
183,229.29
Acetaldehyde
0.31
Methanol
1,807.93
Formaldehyde
96,385.87
1,3-Butadiene
0.11
Formaldehyde
944.58
Naphthalene
85,280.03
1,3-Butadiene
0.10
Benzene
897.31
1,3-Butadiene
63,648.01
Arsenic
0.05
Hexane
789.32
Acetaldehyde
57,116.24
Arsenic
0.04
Ethylbenzene
558.42
Lead, PM
30,072.62
Benzene
0.04
Acetaldehyde
514.05
Benzene
29,910.26
Benzene
0.03
Naphthalene
255.84
Xylenes
20,033.14
Naphthalene
0.02
Ethylene glycol
206.28
Bis(2-ethylhexyl) phthalate, gas
14,976.35
Carbon Tetrachloride
0.01
South Phoenix, Arizona (Marico
)a County) - SPAZ
Toluene
3,935.87
Acrolein
3,491,697.55
1,3-Butadiene
0.14
Xylenes
2,003.31
2,4-Toluene diisocyanate
183,229.29
1,3-Butadiene
0.12
Methanol
1,807.93
Formaldehyde
96,385.87
Benzene
0.04
Formaldehyde
944.58
Naphthalene
85,280.03
Benzene
0.04
Benzene
897.31
1,3-Butadiene
63,648.01
Carbon Tetrachloride
0.01
Hexane
789.32
Acetaldehyde
57,116.24
Carbon Tetrachloride
0.01
Ethylbenzene
558.42
Lead, PM
30,072.62
Ethylbenzene
<0.01
Acetaldehyde
514.05
Benzene
29,910.26
Ethylbenzene
<0.01
Naphthalene
255.84
Xylenes
20,033.14
p-Dichlorobenzene
<0.01
Ethylene glycol
206.28
Bis(2-ethylhexyl) phthalate, gas
14,976.35
/?-Dichlorobcnzcnc
<0.01
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 5-5 include the following:
• Formaldehyde, benzene, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Maricopa County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, naphthalene, and benzene.
• Eight of the highest emitted pollutants in Maricopa County also have the highest
toxicity-weighted emissions.
• Formaldehyde (2016) has the highest cancer risk approximation for PXSS.
Formaldehyde has also the highest emissions and the highest toxicity-weighted
emissions for Maricopa County. Acetaldehyde (2016) also appears on all three lists.
Carbonyl compounds were not sampled for at SPAZ, thus, cancer risk approximations
are not available for this pollutant for SPAZ.
• Among the VOCs, benzene, 1,3-butadiene, and carbon tetrachloride have the highest
cancer risk approximations for PXSS and SPAZ. The cancer risk approximations for
benzene and 1,3-butadiene for SPAZ are slightly higher than those for PXSS, but the
cancer risk approximations for carbon tetrachloride are very similar between the two
sites. While benzene and 1,3-butadiene both appear among the pollutants with the
highest emissions and highest toxicity-weighted emissions for Maricopa County,
carbon tetrachloride does not appear on either list, ranking 25th for quantity emitted
and 32nd for it toxicity-weighted emissions.
• Naphthalene is among the highest emitted pollutants (fifth), has the second highest
toxicity-weighted emissions, and has the 10th highest cancer risk approximations for
PXSS (2015). POM, Groups 2b and 2d are among the highest emitted "pollutants" in
Maricopa County and rank among those with the highest toxicity-weighted emissions.
POM, Group 2b includes several PAHs sampled for at PXSS including acenaphthene,
benzo(e)pyrene, fluoranthene, and perylene. None of the PAHs included in POM,
Group 2b were identified as pollutants of interest for PXSS (or failed any screens).
POM, Group 2d does not include any pollutants sampled for at PXSS.
• Arsenic (2016) has the ninth highest cancer risk approximation among the pollutants
of interest for PXSS. This pollutant ranks tenth for its toxicity-weighted emissions but
does not appear among the highest emitted pollutants in Maricopa County (it ranks
24th).
Observations from Table 5-6 include the following:
• Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in Maricopa County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 2,4-toluene diisocyanate, and formaldehyde.
• Five of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Maricopa County.
5-48
-------
• Acrolein has the highest toxicity-weighted emissions for Maricopa County. Although
acrolein was sampled for at both sites, this pollutant was excluded from the pollutants
of interest designation, and thus subsequent risk-based screening evaluations, due to
questions about the consistency and reliability of the measurements, as discussed in
Section 3.2. The emissions for acrolein rank 15th for Maricopa County.
• All of the noncancer hazard approximations calculated for PXSS and SPAZ are less
than an HQ of 1.0. Formaldehyde and acetaldehyde have the highest noncancer
hazard approximations for PXSS, both of which appear among those pollutants with
the highest emissions and toxicity-weighted emissions for Maricopa County.
• 1,3-Butadiene has the highest noncancer hazard approximations among the VOCs for
both PXSS and SPAZ. This pollutant ranks fifth for its toxicity-weighted emissions
but does not appear among the highest emitted in Maricopa County (it ranks 12th).
• For PXSS, noncancer risk approximations for arsenic, benzene, naphthalene (2015),
and carbon tetrachloride (2015) also appear in Table 5-6. Benzene ranks fifth for its
total emissions and eighth for its toxicity-weighted emissions. Naphthalene also
appears on both emissions-based lists, ranking ninth for its total emissions and fourth
for its toxicity-weighted emissions. Arsenic and carbon tetrachloride do not appear on
either emissions-based list.
• In addition to 1,3-butadiene, benzene, carbon tetrachloride, ethylbenzene, and
/;-dichlorobenzene have the highest noncancer hazard approximations for SPAZ. For
each pollutant, noncancer risk approximations for both years are shown in Table 5-6
and are similar to each other. Benzene appears on both emissions-based list;
ethylbenzene ranks seventh for its total emissions in Maricopa County, but does not
appear among those with the highest toxicity-weighted emissions; and carbon
tetrachloride and />dichlorobenzene appear on neither emissions-based list.
Summary of the 2015-2016 Monitoring Data for PXSS and SPAZ
Results from several of the data analyses described in this section include the following:
~~~ Twenty pollutants failed screens for PXSS; eight pollutants failed screens for SPAZ.
~~~ Of the site-specific pollutants of interest for PXSS, formaldehyde had the highest
annual average concentration, though 2016 was the only year for which an annual
average could be calculated. For SPAZ, benzene had the highest annual average
concentration (both years) among this site's pollutants of interest.
~~~ Concentrations of several VOCs, particularly benzene and 1,3-butadiene, tended to
be higher during the colder months of the year. This was also reflected in the
concentration data from previous years of sampling.
~~~ SPAZ has the highest annual average concentrations of benzene, p-dichlorobenzene,
and ethylbenzene among NMP sites sampling VOCs, as well as the second and third
highest annual average concentrations of 1,3-butadiene. Annual average
concentrations of these pollutant for PXSS also rank among the highest calculated
across the program.
5-49
-------
The most significant trends shown for the pollutants of interest for PXSS are for
1,2-dichloroethane; the detection rate of 1,2-dichloroethane increased significantly
during the later years of sampling, with the 1-year average concentration at a
maximum for 2016. This increase in the detection rate also occurred at SPAZ. For
SPAZ, the maximum 1,3-butadiene concentration measured since the onset of
sampling at this site (10 years) was measured in 2016.
Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for PXSS; benzene has the highest cancer risk approximation among the
pollutants of interest for SPAZ. None of the pollutants of interest for either site have
noncancer hazard approximations greater than an HQ of 1.0.
5-50
-------
6.0 Sites in California
This section examines those data from samples collected at three NATTS sites in
California and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and
2016 monitoring efforts. This section also examines the
Sections 1 through 4 for detailed discussions and
definitions regarding the various data analyses presented below.
6.1 Site Characterization
This section characterizes the California monitoring sites by providing a description of
the nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
Three NATTS monitoring sites are located in California. Two are located in Southern
California, specifically in Los Angeles (CELA) and Rubidoux (RUCA), and a third monitoring
site is located in Northern California, in San Jose (SJJCA). Figure 6-1 presents a composite
satellite image retrieved from ArcGIS Explorer showing the CELA monitoring site and its
immediate surroundings. Figure 6-2 identifies nearby point source emissions locations by source
category, as reported in the 2014 NEI for point sources, version 1. Note that only sources within
10 miles of CELA are included in the facility counts provided in Figure 6-2. A 10-mile boundary
was chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference but have been grayed out to emphasize
emissions sources within the boundary. Figures 6-3 through 6-6 present the composite satellite
images and emissions maps for the Rubidoux and San Jose monitoring sites. Table 6-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
IXiUi ijcncmlcd h> sources oilier
lluin LR(.i. LIJ.\'s conliiicl
kiliomioiA Tortile \\ll\ cue noi
included in I lie d^ila aiuil> ses
conliiined in llns report
6-1
-------
Figure 6-1. Los Angeles, California (CELA) Monitoring Site
; I | i -
i.A-1 L
iManitou Ave
-------
Figure 6-2. NEI Point Sources Located Within 10 Miles of CELA
Source Category Group (No. of Facilities)
Los Angeles
' County
1
e
Pacific \ I A
Ocean v
l\
t m \A?
* fF ar
118°25'0" W 118C20'0"W 118°15'0"W 118;10'0"W 118C5'0"W 118WW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
~ CELA NATTS site
10 mile radius
County boundary
* Aerospace/Aircraft Manufacturing Facility (2)
"t* Airport/Airline/Airport Support Operations (60)
£ Asphalt Production/Hot Mix Asphalt Plant (2)
^ Auto Body Shop/Painters/Automotive Stores (7)
@ Automotive/RV Dealership (1)
$£ Battery Manufacturing Facility (2)
£=~ Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminal/Bulk Plant (3)
C Chemical Manufacturing Facility (7)
A Concrete Batch Plant (2)
0 Dry Cleaning Facility (1)
G Electrical Equipment Manufacturing Facility (5)
f Electricity Generation Facility (5)
P Electroplating, Plating, Polishing, Anodizing, and
Coloring Facility (15)
F Food Processing/Agriculture Facility (10)
1 Foundry, Iron and Steel (4)
A Foundry, non-ferrous (4)
|f Gasoline/Diesel Service Station (1)
Glass Plant (1)
> Hotels/Motels/Lodging (1)
Industrial Machinery or Equipment Plant (2)
O Institution (school, hospital, prison, etc.) (23)
m Landfill (3)
, j-, Metal Can, Box, and Other Metal Container
v Manufacturing Facility (1)
A Metal Coating, Engraving, and Allied Services to
Manufacturers (4)
Metals Processing/Fabrication Facility (11)
X Mine/Quarry/Mineral Processing Facility (3)
0 Miscellaneous Commercial/Industrial Facility (32)
[Ml Municipal Waste Combustor (1)
• Oil and/or Gas Production (6)
Paint and Coating Manufacturing Facility (3)
-§)§¦ Petroleum Products Manufacturing Facility (2)
A Petroleum Refinery (1)
CD Pharmaceutical Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (6)
^ Printing, Coating and Dyeing of Fabrics Facility
T (1)
D Printing/Publishing/Paper Product Manufacturing
P Facility (18)
E3 Pulp and Paper Plant (2)
X Rail Yard/Rail Line Operations (4)
T Textile, Yarn, or Carpet Plant (3)
Truck/Bus/Transportation Operations (2)
I Wastewater Treatment Facility (5)
4 V\teter Treatment Facility (2)
... Woodwork, Furniture, Millwork and Wood
Preserving Facility (7)
6-3
-------
Figure 6-3. Rubidoux, California (RUCA) Monitoring Site
L/ ^ i 9 ¦
| I 'Slf 0,0
iM-r y W* i«
Kth.St'
Jensen-Alvarado Historic-Ranch Park
\ Source: USGS^?/ f -
Source: NASA, N G A , USGS
£ 2 008 Microsoft Coir p..
geonarhes.org
6 2d f,t.
—"' r
On
-k
-------
Figure 6-4. NEI Point Sources Located Within 10 Miles of RUCA
County
Riverside
County
Legend
tiy^S'O-W 117U20'0" W 117"15'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
RUCA NATTS site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
1^4 Aerospace/Aircraft Manufacturing Facility (1)
Airport/Airline/Airport Support Operations (11)
A Animal Feedlot or Farm (5)
£ Asphalt Production/Hot Mix Asphalt Plant (3)
Auto Body Shop/Painters/Automotive Stores (2)
« Automobile/Truck Manufacturing Facility (3)
I Brick, Structural Clay, or Clay Ceramics Plant (1)
f Building/Construction (1)
B Bulk Terminal/Bulk Plant (2)
C Chemical Manufacturing Facility (2)
i Compressor Station (1)
6 Electrical Equipment Manufacturing Facility (3)
f Electricity Generation Facility (3)
Ethanol Biorefinery (1)
Food Processing/Agriculture Facility (4)
Foundry, Iron and Steel (1)
Gasoline/Diesel Service Station (1)
Glass Plant (1)
Industrial Machinery or Equipment Plant (2)
Institution (school, hospital, prison, etc.) (8)
Landfill (5)
Metal Can, Box, and Other Metal Container
Manufacturing Facility (1)
Metal Coating, Engraving, and Allied Services to
Manufacturers (2)
Metals Processing/Fabrication Facility (13)
Military Base/National Security Facility (1)
Mine/Quarry/Mineral Processing Facility (6)
Miscellaneous Commercial/Industrial Facility (15)
Paint and Coating Manufacturing Facility (3)
Plastic, Resin, or Rubber Products Plant (3)
Portland Cement Manufacturing Facility (2)
Printing/Publishing/Paper Product Manufacturing
Facility (4)
Pulp and Paper Plant (1)
Rail Yard/Rail Line Operations (4)
Ship/Boat Manufacturing or Repair Facility (1)
Steel Mill (1)
Truck/Bus/Transportation Operations (1)
Wastewater Treatment Facility (3)
Water Treatment Facility (2)
Wbodwork, Furniture, Millwork and Wood
Preserving Facility (5)
I
San Bernardino I
A
~ ~
^ aa
6-5
-------
Figure 6-5. San Jose, California (SJJCA) Monitoring Site
-------
Figure 6-6. NEI Point Sources Located Within 10 Miles of SJJCA
122 0 0 W
Alameda
County
m
Santa Clara
County
1
O c t>
Santa Cruz
County
122"5'0"W 122°0'0"W 121°55'0"W 12r'50'0"W 121°45'0"W 121°40'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
SJJCA NATTS site Q 10 mile radius
County boundary
Source Category Group (No. of
* Aerospace/Aircraft Manufacturing Facility (3)
"1" Airport/Airline/Airport Support Operations (19)
£ Asphalt Production/Hot Mix Asphalt Plant (2)
0 Auto Body Shop/Painters/Automotive Stores (160)
« Automobile/Truck Manufacturing Facility (3)
(Q.' Automotive/RV Dealership (4)
1 1 Brick, Structural Clay, or Clay Ceramics Plant (2)
f Building/Construction (16)
B Bulk Terminal/Bulk Plant (4)
C Chemical Manufacturing Facility (14)
I Compressor Station (2)
A Concrete Batch Plant (6)
tX] Crematory (5)
® Dry Cleaning Facility (9)
G Electrical Equipment Manufacturing Facility (213)
f Electricity Generation Facility (51)
P Electroplating, Plating, Polishing, Anodizing, and
Coloring Facility (16)
V Fertilizer Plant (1)
Facilities)
F Food Processing/Agriculture Facility (20)
A Foundry, non-ferrous (1)
If Gasoline/Diesel Service Station (5)
> Hotels/Motels/Lodging (8)
Industrial Machinery or Equipment Plant (24)
O Institution (school, hospital, prison, etc.) (147)
A Landfill (4)
^ Leather and Leather Products Facility (1)
-gv Metal Can, Box, and Other Metal Container
™ Manufacturing Facility (2)
. Metal Coating, Engraving, and Allied Services to
^ Manufacturers (24)
0 Metals Processing/Fabrication Facility (29)
X Mine/Quarry/Mineral Processing Facility (4)
i]t Mineral Processing Plant (1)
? Miscellaneous Commercial/Industrial Facility (295)
• Oil and/or Gas Production (1)
jQ Paint and Coating Manufacturing Facility (3)
Petroleum Products Manufacturing Facility (2)
CD Pharmaceutical Manufacturing Facility (8)
R Plastic, Resin, or Rubber Products Plant (4)
Port and Harbor Operations (1)
'Is Printing, Coating and Dyeing of Fabrics Facility (4)
p Printing/Publishing/Paper Product Manufacturing
K Facility (30)
H Pulp and Paper Plant (3)
X Rail Yard/Rail Line Operations (2)
Ship/Boat Manufacturing or Repair Facility (1)
TT Telecommunications/Radio Facility (90)
© Testing Laboratory (3)
T Textile, Yarn, or Carpet Plant (2)
M Tobacco Manufacturing Facility (7)
Truck/Bus/Transportation Operations (6)
Utilities/Pipeline Construction (1)
* Wastewater Treatment Facility (9)
4 Water Treatment Facility (24)
... Woodwork, Furniture, Millworkand Wood
*v Preserving Facility (30)
6-7
-------
Table 6-1. Geographical Information for the California Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
CELA
06-037-1103
Los
Angeles
Los
Angeles
Los Angeles-Long
Beach-Anaheim, CA
34.066590,
-118.226880
Residential
Urban/City
Center
231,000
1-5 between Main St. and Broadway
(exit 136 and 137)
RUCA
06-065-8001
Rubidoux
Riverside
Riverside-San
Bernardino-Ontario, CA
33.999580,
-117.416010
Residential
Suburban
166,000
Rte 60 (Mission Blvd) between
Rubidoux Blvd and Valley Way
SJJCA
06-085-0005
San Jose
Santa
Clara
San Jose-Sunnyvale-
Santa Clara, CA
37.348497,
-121.894898
Commercial
Urban/City
Center
126,000
Rte 87 (Guadalupe Pkwy) between
Julian St and W Taylor St
lAADT reflects 2015 data (CA DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
On
00
-------
CELA is located on the rooftop of a two-story building northeast of downtown Los
Angeles, just southeast of Dodgers' Stadium and Los Angeles State Historic Park, which are
prominent features in Figure 6-1. CELA is surrounded by major freeways, including 1-5 and
Route 110. Highway 101 is located farther south. Although the area is classified as residential, a
freight yard is located to the south of the site. The Los Angeles River runs north-south just east
of the site. This monitoring site was originally set up as an emergency response monitoring site.
Figure 6-2 shows that CELA is situated among numerous point sources. The source
category with the greatest number of emissions sources near this monitoring site is the airport
source category, which includes airports and related operations as well as small runways and
heliports, such as those associated with hospitals or television stations. Other source categories
with a relatively large number of emissions sources within 10 miles of CELA include institutions
such as schools, hospitals, and/or prisons; printing, publishing, and paper product manufacturing;
and electroplating, plating, polishing, anodizing, and coloring facilities. A high-density cluster of
emissions sources is located just to the west and southwest of CELA. The sources closest to
CELA are a mineral processing facility, a facility involved in oil/gas production, and a heliport at
a detention center.
RUCA is located just north of Riverside, in a residential area in the town of Rubidoux.
RUCA is adjacent to a power substation next to a storage facility and apartment building near the
intersection of Mission Boulevard and Riverview Drive. Residential areas surround RUCA,
including three schools: a middle school north of Mission Boulevard, an elementary school south
of Riverview Drive, and a high school to the west of Pacific Avenue, the football and baseball
fields of which are prominent features in Figure 6-3. Highway 60 runs east-west to the north of
the site. Flabob Airport is located approximately three-quarters of a mile to the southeast of the
site. RUCA is located approximately 47 miles east of CELA.
Figure 6-4 shows that fewer emissions sources surround RUCA than CELA. Most of the
emissions sources within 10 miles of RUCA are located to the northeast, north, and northwest of
the site, in San Bernardino County. The point source located closest to RUCA is Flabob Airport
and is the only source located within a half-mile of the site. Although the emissions source
categories are varied, the emissions source categories with the greatest number of sources within
10 miles of RUCA include metals processing and fabrication; airport operations; institutions
such as schools, hospitals, and/or prisons; and mines, quarries, and mineral processing facilities.
6-9
-------
SJJCA is located in central San Jose. Figure 6-5 shows that SJJCA is located in a
commercial area surrounded by residential areas. A railroad is shown east of the monitoring site,
running north-south in Figure 6-5. Guadalupe Parkway (Route 87) intersects with 1-880
approximately 1 mile northwest of the monitoring site. San Jose International Airport is just on
the other side of this intersection. The Guadalupe River runs along the eastern boundary of the
airport and runs parallel to the Guadalupe Parkway, as does the Guadalupe River Park and
Gardens, a park and trail system which can be seen on the bottom left of Figure 6-5. Figure 6-6
shows that the density of point sources is significantly higher near SJJCA than the other
California monitoring sites. The emissions source categories with the greatest number of sources
surrounding SJJCA include electrical equipment manufacturing; auto body, paint, and
automotive shops; institutions such as schools, hospitals, and/or prisons; and
telecommunications/radio facility. Sources within a half-mile of SJJCA include a food
processing facility, an auto body shop, and two sources in the miscellaneous source category.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 6-1 also contains traffic volume information for each site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. CELA
experiences a higher traffic volume compared to the other California sites, although the traffic
volumes near each of these sites are all greater than 100,000. Compared to other NMP sites,
CELA has the second highest traffic volume, RUCA ranks sixth, and SJJCA ranks ninth highest.
The traffic volumes for CELA, RUCA and SJJCA were obtained from heavily traveled
highways.
6-10
-------
6.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
California site to identify site-specific "pollutants of interest," which allows analysts and readers
to focus on a subset of pollutants through the context of risk. For each site, each pollutant's
preprocessed daily measurement was compared to its associated risk screening value. If the
concentration was greater than the risk screening value, then the concentration "failed the
screen." The site-specific results of this risk-based screening process are presented in Table 6-2
and incorporate measurements from both 2015 and 2016. Pollutants of interest are those for
which the individual pollutant's total failed screens contribute to the top 95 percent of the site's
total failed screens and are shaded in gray in Table 6-2. It is important to note which pollutants
were sampled for at each site when reviewing the results of this analysis. PAHs were sampled for
at all three California sites; in addition, metals (PMio) were also sampled for at SJJCA.
Table 6-2. 2015-2016 Risk-Based Screening Results for the California Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Los Angeles, California - CELA
Naphthalene
0.029
115
118
97.46
98.29
98.29
Benzo(a)pyrene
0.00057
1
96
1.04
0.85
99.15
Fluorene
0.011
1
101
0.99
0.85
100.00
Total
117
315
37.14
Rubidoux, California - RUCA
Naphthalene
0.029
83
118
70.34
97.65
97.65
Benzo(a)pyrene
0.00057
2
87
2.30
2.35
100.00
Total
85
205
41.46
San Jose, California - SJJCA
Naphthalene
0.029
88
119
73.95
47.31
47.31
Arsenic (PMio)
0.00023
84
116
72.41
45.16
92.47
Nickel (PMio)
0.0021
11
116
9.48
5.91
98.39
Benzo(a)pyrene
0.00057
1
63
1.59
0.54
98.92
Cadmium (PMio)
0.00056
1
116
0.86
0.54
99.46
Lead (PMio)
0.015
1
116
0.86
0.54
100.00
Total
186
646
28.79
6-11
-------
Observations from Table 6-2 include the following:
• Concentrations of naphthalene failed the majority of screens for CELA, accounting
for 115 of the 117 total failed screens for this site, while benzo(a)pyrene and fluorene
concentrations failed a single screen each. Thus, naphthalene is the only pollutant
identified as a pollutant of interest for CELA.
• Similarly, concentrations of naphthalene failed the majority of screens for RUCA,
accounting for 83 of the 85 total failed screens for this site, while two concentrations
of benzo(a)pyrene also failed screens. Thus, naphthalene is also the only pollutant
identified as a pollutant of interest for RUCA. Note that the percentage of screens
failed for naphthalene is higher for CELA (97 percent) than for RUCA (70 percent).
• Metals (PMio) were also sampled for at SJJCA, in addition to PAHs. For SJJCA,
concentrations of naphthalene also account for the most failed screens (88 of 186, or
47 percent), although arsenic concentrations contributed a similar number of failed
screens (84). Together, these two pollutants account for more than 92 percent of
SJJCA's total failed screens. Concentrations of nickel account for another 6 percent
of the total failed screens for this site (11). Together, these three pollutants contribute
to more than 95 percent of failed screens for SJJCA and were therefore identified as
pollutants of interest for this site. Benzo(a)pyrene, cadmium, and lead also failed a
single screen each for SJCCA but were not identified as pollutants of interest.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
6.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the California monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics from 2015, 2016, and previous
years of monitoring are presented in order to characterize concentration trends at each
site.
6-12
-------
Each data analysis is performed where the applicable criteria is met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at the California
monitoring sites are provided in Appendices N and O.
6.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each California site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the California
monitoring sites are presented in Table 6-3, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
6-13
-------
Table 6-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the California Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Los Angeles, California - CELA
Naphthalene
57/57/57
91.18
± 14.56
NA
59.11
± 12.38
92.13
±20.41
76.85
±9.42
61/61/61
74.98
± 15.21
61.14
± 13.29
65.44
± 16.02
111.65
± 19.08
78.40
±9.06
Rubidoux, California - RUCA
Naphthalene
59/59/59
56.35
± 16.22
32.58
± 10.08
32.09
±7.79
72.68
± 17.91
48.70
±7.87
59/59/59
58.67
± 15.61
39.37
± 10.62
52.96
± 16.50
101.27
±29.84
63.56
± 11.03
San Jose, California - SJJCA
Arsenic (PMio)
59/59/59
0.79
±0.27
0.42
±0.18
0.36
±0.16
0.48
±0.10
0.52
±0.10
57/57/57
0.52
±0.20
0.36
±0.10
0.51
±0.22
0.43
±0.19
0.45
±0.08
Naphthalene
58/58/58
78.13
± 20.60
36.93
±9.63
44.92
± 12.89
97.84
±23.16
65.13
± 10.58
61/61/61
69.11
±21.63
34.99
± 12.05
35.73
± 10.30
77.09
±27.50
54.48
± 10.29
Nickel (PMio)
59/59/59
1.97
±0.87
0.95
±0.16
1.32
±0.32
1.17
±0.35
1.36
±0.26
57/57/57
1.11
±0.34
1.31
±0.35
1.48
±0.35
1.25
±0.41
1.28
±0.17
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
-------
Observations for the California monitoring sites from Table 6-3 include the following:
• Naphthalene was identified as a pollutant of interest for all three sites. The annual
average concentrations of naphthalene range from 48.70 ± 7.87 ng/m3 (RUCA, 2015)
to 78.40 ± 9.06 ng/m3 (CELA, 2016); CELA has the highest annual average
concentration of naphthalene for both years.
• For each site, naphthalene concentrations appear highest during the fourth quarters of
each year, particularly for 2016, based on the quarterly averages shown. However, the
confidence intervals calculated for each of these quarterly averages are relatively
large, indicating that there is considerable variability in the measurements. Quarterly
average concentrations vary considerably, varying the most for RUCA, which range
from 32.09 ± 7.79 ng/m3 (third quarter 2015) to 101.27 ± 29.84 ng/m3 (fourth quarter
2016). CELA does not have a second quarter average concentration for naphthalene
for 2015, as shown in Table 6-4. This is a result of issues with the collection system
experienced throughout most of June 2015.
• Naphthalene concentrations measured at CELA across both years range from
10.9 ng/m3 to 170 ng/m3, with a median concentration of 73.00 ng/m3. Naphthalene
concentrations measured at RUCA range from 7.78 ng/m3 to 181 ng/m3, with a
median concentration of 48.20 ng/m3. Naphthalene concentrations measured at
SJJCA range from 13.4 ng/m3 to 205 ng/m3, with a median concentration of
45.00 ng/m3. Despite having the widest range of naphthalene concentrations
measured and the only measurements (2) greater than 200 ng/m3 among the
California sites, SJJCA has the lowest median concentration of this pollutant. While
the median concentrations for RUCA and SJJCA are fairly similar to each other, the
median for CELA is considerably higher. The number of naphthalene concentrations
greater than 100 ng/m3 measured at CELA (33) is nearly twice the number measured
at RUCA (17), with the number measured at SJJCA in between (21).
• At each California site, the highest concentrations of naphthalene were measured
during the first or fourth quarters of either year, predominantly in January, February,
November, or December. However, some of the lowest concentrations measured at
CELA and RUCA, including the minimum concentrations, were also measured in
December, helping to explain the large confidence intervals shown for these sites'
fourth quarter naphthalene concentrations.
• Arsenic and nickel are also pollutants of interest for SJJCA. The quarterly and annual
average concentrations of these pollutants are significantly less than those shown for
naphthalene.
• Arsenic concentrations measured across both years at SJJCA range from 0.078 ng/m3
to 2.02 ng/m3. With the exception of the first quarter of 2015, the quarterly average
concentrations do not vary considerably. The first quarter average for 2015
(0.79 ± 0.27 ng/m3) is more than 0.25 ng/m3 greater than the next highest quarterly
average concentration for SJJCA. The first quarter of 2015 has the most arsenic
measurements greater than 1 ng/m3 (4), including the maximum concentration
measured at this site. However, the confidence interval associated with this quarterly
6-15
-------
average is the largest confidence interval shown in Table 6-3, indicating a relatively
high level of variability in the measurements.
• Concentrations of nickel measured across both years at SJJCA range from
0.274 ng/m3 to 6.10 ng/m3. A review of the quarterly average concentrations shows
that the first quarter average concentration for 2015 (1.97 ± 0.87 ng/m3) is
considerably higher than the others and has a relatively large confidence interval
associated with it. The two highest nickel concentrations measured at SJJCA
(6.10 ng/m3 and 5.43 ng/m3) were both measured in January 2015.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the
California sites from those tables include the following:
• CELA is the only California site to appear in Table 4-12 for naphthalene, with its
annual average concentrations ranking eighth (2016) and tenth (2015).
• SJJCA does not appear in Table 4-13 for PMio metals.
6.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 6-3 for CELA, RUCA, and SJJCA. Figures 6-7 through 6-9 overlay the sites' minimum,
annual average, and maximum concentrations for each year onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
Figure 6-7. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
P
1
2
3 4 5
Concentration (ng/m3)
6
7
Progra m:
1st Qua rtile
2nd Quartile 3rd Quartile
4th Quartile
Ave ra ge
¦
~
~
~
i
Site:
2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
6-16
-------
Figure 6-7 presents the box plot for arsenic (PMio) for SJJCA and shows the following:
• The maximum arsenic concentrations measured each year at SJJCA are considerably
less than the maximum concentration measured across the program.
• The range of arsenic concentrations measured at SJJCA in 2015 is larger than the
range of concentrations measured in 2016.
• Both annual average arsenic concentrations calculated for SJJCA are less than the
program-level average concentration (0.70 ng/m3) and just less than the program-
level median concentration of arsenic (0.55 ng/m3).
• Non-detects of arsenic were not measured at SJJCA.
Figure 6-8. Program vs. Site-Specific Average Naphthalene Concentrations
m
CELA
RUCA
J
150 200 250
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
Figure 6-8 presents the box plots for naphthalene for all three sites and shows the
following:
• Among the California sites, the highest naphthalene concentrations were measured at
SJJCA. However, all of the naphthalene concentrations measured at these sites are
considerably less than the maximum concentration measured across the program.
• CELA is the only California site for which both years' annual average concentrations
are greater than the program-level average (61.23 ng/m3); both annual averages are
just less than the program-level third quartile (82.15 ng/m3). Note that the minimum
6-17
-------
concentration measured at CELA in 2016 is greater than the program-level first
quartile.
• For RUC A and SJJC A, one year's annual average concentration is just greater than
the program-level average and one year's annual average is less than the program-
level average.
• There were no non-detects of naphthalene measured at CELA, RUCA, SJJCA, or
across the program.
Figure 6-9. Program vs. Site-Specific Average Nickel (PMio) Concentrations
E "
Program MaxConcentration
69.5 ng/m3
i
0 3 6 9 12 15 18 21 24
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~ ~ ~ 1
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 6-9 presents the box plot for nickel for SJJCA and shows the following:
• The program-level maximum nickel concentration (69.5 ng/m3) is not shown directly
on the box plot in Figure 6-9 because the scale of the box plot would be too large to
readily observe data points at the lower end of the concentration range. Thus, the
scale of the box plot has been reduced to 24 ng/m3.
• The maximum nickel concentration measured across the program is more than 10
times greater than the maximum nickel concentration measured at SJJCA.
• The range of nickel concentrations measured at SJJCA in 2015 is approximately
twice the range measured in 2016.
• Both of SJJCA's annual average nickel concentrations are greater than the program-
level average concentration (1.09 ng/m3), with the 2015 annual average concentration
just greater than the program-level third quartile.
• Non-detects of nickel were not measured at SJJCA.
6.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
Both CELA and RUCA began sampling PAHs under the NMP in 2007. SJJCA began sampling
6-18
-------
PAHs and metals under the NMP in 2008. Thus, Figures 6-10 through 6-14 present the 1-year
statistical metrics for each of the pollutants of interest first for CELA, then for RUC A, and
finally for SJJCA. The statistical metrics presented for assessing trends include the substitution
of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of sampling is
required for inclusion in the trends analysis; in these cases, a 1-year average concentration is not
provided, although the range and percentiles are still presented.
Figure 6-10. Yearly Statistical Metrics for Naphthalene Concentrations Measured at CELA
o
20071 2008
• • O-
¦o-
2011 2012
Year
-r
O 5th Percentile
— Minimum
Maximum O 95th Percentile Averc
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2007.
Observations from Figure 6-10 for naphthalene concentrations measured at CELA
include the following:
• CELA began sampling PAHs under the NMP in April 2007. Because a full year's
worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided.
• The minimum concentration measured at CELA was measured in 2007 (1.30 ng/m3);
2007 is the only year in which concentrations less than 10 ng/m3 were measured (five
in total). The range of naphthalene measurements increased considerably from 2007
to 2008 and again in 2009, when the maximum naphthalene concentration was
measured (736 ng/m3 on October 16, 2009). Concentrations greater than 500 ng/m3
were also measured in 2008 and 2010. The maximum naphthalene concentration
decreases steadily after 2009, with the smallest range of concentrations measured in
2016.
6-19
-------
• All of the statistical parameters shown in Figure 6-10 exhibit an increase from 2011
to 2012 except the maximum concentration. The increase in the 1-year average
concentration from 2011 to 2012 is significant, even though the range of
concentrations measured in 2012 is the smallest since the onset of sampling. The
number of naphthalene concentrations greater than 200 ng/m3 increased from nine in
2011 to 24 for 2012, which is the most for any year of sampling at CELA.
• Each of the statistical metrics exhibits a decrease from 2012 to 2013, with the 1-year
average concentration decreasing significantly (by nearly 40 percent). This decreasing
trend continues through 2015, with little change shown for 2016. 2015 is the first year
that a naphthalene concentration greater than 200 ng/m3 was not measured at CELA.
Figure 6-11. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
RUCA
T
[ _
o
•o
T
o
rh
o
-0
•o.
I
1
1
I
L2r
£-
2007 1 2008 2009
2011 2012
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2007.
Observations from Figure 6-11 for naphthalene concentrations measured at RUCA
include the following:
• RUCA began sampling PAHs under the NMP in May 2007. Because a full year's
worth of data is not available, a 1-year average concentration for 2007 is not
presented, although the range of measurements is provided.
• The range of naphthalene concentrations measured increased through the early years
of sampling at RUCA, in a similar manner to those measured at CELA. The
maximum naphthalene concentration was measured at RUCA in 2009 (406 ng/m3),
although a concentration of similar magnitude was also measured at RUCA in 2013.
6-20
-------
These are the only two naphthalene concentrations greater than 400 ng/m3 measured
at RUCA. Excluding 2009, the maximum concentration increases steadily between
2007 and 2013.
• The 1-year average concentration increases by more than 20 ng/m3 from 2008 to
2009, changes little for 2010, then continues to increase slightly through 2012,
reaching a maximum of nearly 100 ng/m3. After 2012, 1-year average concentration
begins to decrease, with the most significant decrease shown for 2015, when the
1-year average concentration is less than 50 ng/m3 for the first (and only) time. The
median concentration exhibits a similar pattern. All of the statistical parameters
exhibit an increase from 2015 to 2016.
• Concentrations measured at RUCA in 2015 exhibit the least variability among the
years of sampling. But the concentrations measured most years reflect a relatively
high level of variability. For 2009, 2012, and 2013, the maximum concentration
measured is twice the 95th percentile. For these years, more than 100 ng/m3 separates
the maximum concentration and the next highest concentration measured. In addition,
concentrations less than 10 ng/m3 have been measured during five year of sampling,
including two measurements less than or equal to 1 ng/m3.
Figure 6-12. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SJJCA
3.5
2008 2009 2010 2011 2012 2013 2014 2015 2016
Title
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
6-21
-------
Observations from Figure 6-12 for arsenic concentrations measured at SJJCA include the
following:
• The maximum concentration of arsenic (3.09 ng/m3) was measured on the first day of
sampling at SJJCA (January 1, 2008), though an arsenic concentration of similar
magnitude was also measured in 2013 (2.92 ng/m3). Only one other arsenic
concentration greater than 2 ng/m3 has been measured at SJJCA (2015).
• The 1-year average arsenic concentration decreased from 2008 to 2009. Although this
is due in part to the magnitude of the maximum concentration measured in 2008, all
of the statistical parameters exhibit a decrease from 2008 to 2009, indicating that the
decrease is not only due to the difference in the maximum concentrations. The
number of concentrations at the lower end of the concentration range increased for
2009. Seven arsenic concentrations less than 0.1 ng/m3 were measured in 2009,
compared to only two in 2008; in addition, two non-detects were measured at SJJCA
in 2009, compared to none in 2008.
• Between 2010 and 2012, the range of arsenic concentrations measured changed little
and the 1-year average concentration varied between 0.37 ng/m3 for 2010 to
0.39 ng/m3 for 2011 and 2012. With the exception of the minimum and 5th percentile
(which did not change), each of the statistical metrics exhibit an increase for 2013,
with the 1-year average concentration increasing to 0.52 ng/m3. Along with the
second highest concentration measured since the onset of sampling, the number of
arsenic concentrations greater than 0.75 ng/m3 measured at SJJCA increased to 16 for
2013, the most for any year of sampling thus far (none of the previous years had more
than six).
• Between 2013 and 2016, the statistical parameters have a slight undulating pattern,
with years of higher concentrations (and thus, higher 1-year average concentrations)
following a year of lower concentrations. The median concentration for each of these
years, though, varies by less than 0.04 ng/m3.
6-22
-------
Figure 6-13. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
SJJCA
o-.
2012
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
Observations from Figure 6-13 for naphthalene concentrations measured at SJJCA
include the following:
• SJJCA began sampling PAHs under the NMP in May 2008. Because a full year's
worth of data is not available, a 1-year average concentration for 2008 is not
presented, although the range of measurements is provided.
• The maximum concentration of naphthalene was measured at SJJCA in 2009
(496 ng/m3). No additional naphthalene concentrations greater than 400 ng/m3 have
been measured at SJJCA and few greater than 300 ng/m3 have been measured at this
site.
• There is very little change among the minimum concentrations and 5th percentiles
across the years of sampling while there are considerable fluctuations in the statistical
parameters representing the upper end of the concentration range, particularly
between 2009 and 2014.
• The median concentration changed little over the years through 2012, ranging from
43.00 ng/m3 (2010) to 49.90 ng/m3 (2011). 2013 is the first year with a median
concentration greater than 50 ng/m3 (57.70 ng/m3). For each year between 2013 and
2016, a median greater than 50 ng/m3 is followed by a median less than 40 ng/m3.
6-23
-------
• The 1-year average concentration exhibits more variability, having an undulating
pattern through 2014, ranging from 59.73 ng/m3 (2014) to 94.13 ng/m3 (2013) during
this period. Between 2014 and 2016, the year-to-year changes are smaller
• Both the 1-year average and median concentrations are at a minimum for 2016.
Figure 6-14. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at
SJJCA
O 5th Percentile
O 95th Percentile
Observations from Figure 6-14 for nickel concentrations measured at SJJCA include the
following:
• The maximum concentration of nickel measured at SJJCA has a steady increasing
trend after 2009, reaching a maximum of 9.73 ng/m3 in 2014. The maximum
concentration of nickel measured in 2014 is considerably higher than the next highest
nickel concentration measured at this site (6.10 ng/m3 measured in 2015).
• Both the 1-year average and median concentrations have a significant decreasing
trend between 2008 and 2010, when both statistical parameters are at a minimum for
the period of sampling. This is followed by a significant increase for 2011. The
number of nickel concentrations greater than 1 ng/m3 more than doubled from 2010
(16) to 2011 (37).
• The changes in the 1-year average and median concentrations have been more subtle
in more recent years. After a slight decrease for 2012, both central tendency
parameters increased for 2013. Between 2013 and 2016, the median concentration has
varied by only 0.10 ng/m3 and the 1-year average concentration has varied by less
6-24
-------
than 0.15 ng/m3, despite the changes in the maximum concentration and 95th
percentile shown.
6.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each California monitoring site. Refer to Sections 3.2,
3.4.2.3, and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time
frames, and calculations associated with these risk-based screenings.
6.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the California monitoring sites, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 6-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the California sites from Table 6-4 include the following:
• Annual average concentrations of naphthalene range from 48.70 ± 7.87 ng/m3
(RUCA, 2015) to 78.40 ± 9.06 ng/m3 (CELA, 2016). All of the cancer risk
approximations for naphthalene are less than 3 in-a-million, ranging from 1.66 in-a-
million (RUCA, 2015) to 2.67 in-a-million (CELA, 2016). All of the noncancer
hazard approximations for naphthalene are considerably less than 1.0 (all are less than
an HQ of 0.05).
• SJJCA is the only California site with pollutants of interest other than naphthalene.
The cancer risk approximations for arsenic for both years are similar to the cancer
risk approximations calculated for naphthalene each year for SJJCA, while the cancer
risk approximations for nickel are both less than 1 in-a-million.
• The noncancer hazard approximations for arsenic and nickel for SJJCA are also
considerably less than 1.0, similar to those calculated for naphthalene, indicating that
no adverse noncancer health effects are expected from these individual pollutants.
6-25
-------
Table 6-4. Risk Approximations for the California Monitoring Sites
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Los Angeles, California - CELA
Naphthalene
0.000034
0.003
57/57
76.85
±9.42
2.61
0.03
61/61
78.40
±9.06
2.67
0.03
Rubidoux, California - RUCA
Naphthalene
0.000034
0.003
59/59
48.70
±7.87
1.66
0.02
59/59
63.56
± 11.03
2.16
0.02
San Jose, California - SJJC
Arsenic (PMio)
0.0043
0.000015
59/59
0.52
±0.10
2.21
0.03
57/57
0.45
±0.08
1.95
0.03
Naphthalene
0.000034
0.003
58/58
65.13
± 10.58
2.21
0.02
61/61
54.48
± 10.29
1.85
0.02
Nickel (PMio)
0.00048
0.00009
59/59
1.36
±0.26
0.65
0.02
57/57
1.28
±0.17
0.61
0.01
— = A Cancer URE or Noncancer RfC is not available.
-------
6.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 6-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 6-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 6-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 6-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 6-5. Table 6-6
presents similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 6.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 6-5 include the following:
• Formaldehyde and benzene are the highest emitted pollutants with cancer UREs in
Los Angeles, Riverside, and Santa Clara Counties. The quantity of emissions for the
pollutants shown is considerably greater for Los Angeles County than Riverside and
Santa Clara Counties.
• Formaldehyde has the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Los Angeles County, followed by POM, Group la, and hexavalent
chromium. These same pollutants have the highest toxicity-weighted emissions for
Riverside and Santa Clara Counties but the order varies, with hexavalent chromium
ranking first for both counties.
• Six of the highest emitted pollutants in Los Angeles County also have the highest
toxicity-weighted emissions, while there are eight in common for Riverside County
and seven in common for Santa Clara County. Despite the relatively high ranking for
hexavalent chromium for each county's toxicity-weighted emissions, this pollutant
does not appear among the highest pollutants emitted in each county.
6-27
-------
• Naphthalene, which is a pollutant of interest for all three sites, appears on both
emissions-based lists for all three counties.
• While arsenic and nickel do not appear among the highest emitted pollutants in Santa
Clara County (they rank lower than tenth), they rank seventh and ninth, respectively,
for their toxicity-weighted emissions.
Observations from Table 6-6 include the following:
• Toluene is the highest emitted pollutant with a noncancer RfC in all three California
counties. The quantity emitted is significantly higher for Los Angeles County than
Riverside and Santa Clara Counties.
• Acrolein and chlorine are the pollutants with the highest toxicity-weighted emissions
(of the pollutants with noncancer RfCs) for all three counties. Although these two
pollutants rank highest for toxicity-weighted emissions for each county, neither
pollutant appears among the highest emitted.
• Three of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Los Angeles and Riverside Counties, while only two of the highest
emitted pollutants also have the highest toxicity-weighted emissions for Santa Clara
County.
• Naphthalene is the only pollutant of interest for CELA. Naphthalene does not appear
among the highest emitted pollutants (of those with a noncancer RfC) for Los
Angeles County, although it does rank tenth for its toxicity-weighted emissions. A
similar observation can be made for RUCA and SJJCA, where naphthalene's toxicity-
weighted emissions rank among the highest for both Riverside and Santa Clara
Counties (ranking seventh highest for both counties) but this pollutant does not
appear among the highest emitted.
• Arsenic and nickel are also pollutants of interest for SJJCA. Similar to naphthalene,
these two pollutants appear among those with the highest toxicity-weighted emissions
for Santa Clara County, but not among the highest emitted. Lead, which was sampled
for at SJJCA and but was not identified as a pollutant of interest, also appears among
those with the highest toxicity-weighted emissions in Table 6-6. Concentrations of
lead failed a single screen for SJJCA.
6-28
-------
Table 6-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the California Monitoring Sites
Top 10 Total Emissions for Pollutants with
Top 10 Cancer Toxicity-Weighted
Top 10 Cancer Risk Approximations Based
Cancer UREs
Emissions
on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Los
Angeles, California (Los Angeles County) - CELA
Formaldehyde
1,720.49
Formaldehyde
2.24E-02
Naphthalene
2.67
Benzene
1,565.85
POM, Group la
1.89E-02
Naphthalene
2.61
Dichloromethane
1,379.64
Hexavalent Chromium PM
1.61E-02
Acetaldehyde
700.59
Ethylene oxide
1.26E-02
Ethylbenzene
622.36
Benzene
1.22E-02
1.3 -Butadiene
274.82
1,3-Butadiene
8.24E-03
p-Dichlorobenzene
238.30
Naphthalene
4.64E-03
POM, Group la
215.10
p-Dichlorobcnzcnc
2.62E-03
Naphthalene
136.42
Nickel, PM
2.48E-03
T etrachloroethylene
57.81
POM, Group 5a
1.85E-03
Rubidoux, California (Riverside County) - RUCA
Formaldehyde
390.30
Hexavalent Chromium PM
9.95E-03
Naphthalene
2.16
Benzene
314.48
Formaldehyde
5.07E-03
Naphthalene
1.66
Dichloromethane
172.55
POM, Group la
4.02E-03
Acetaldehyde
156.68
Benzene
2.45E-03
Ethylbenzene
136.34
1,3-Butadiene
1.98E-03
1.3 -Butadiene
66.12
Naphthalene
1.31E-03
p-Dichlorobenzene
52.30
p-Dichlorobcnzcnc
5.75E-04
POM, Group la
45.69
Nickel, PM
5.67E-04
Naphthalene
38.45
Acetaldehyde
3.45E-04
1,3 -Dichloropropene
17.12
Ethylbenzene
3.41E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 6-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the California Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Top 10 Cancer Toxicity-Weighted
Top 10 Cancer Risk Approximations Based
Cancer UREs
Emissions
on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
San Jose, California (Santa Clara County) - SJJCA
Formaldehyde
243.18
Hexavalent Chromium, PM
3.74E-03
Arsenic
2.21
Benzene
227.20
POM, Group la
3.25E-03
Naphthalene
2.21
Acetaldehyde
151.12
Formaldehyde
3.16E-03
Arsenic
1.95
Ethylbenzene
106.49
Benzene
1.77E-03
Naphthalene
1.85
Dichloromethane
90.81
Naphthalene
1.66E-03
Nickel
0.65
Bis(2-ethylhexyl) phthalate, gas
69.83
1,3-Butadiene
1.54E-03
Nickel
0.61
1.3 -Butadiene
51.26
Arsenic, PM
5.36E-04
Naphthalene
48.73
p-Dichlorobcnzcnc
4.83E-04
p-Dichlorobenzene
43.93
Nickel, PM
3.91E-04
POM, Group la
36.95
Acetaldehyde
3.32E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 6-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the California Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Los Angeles, California (Los Angeles County) - CELA
Toluene
5,286.05
Acrolein
4,082,251.24
Naphthalene
0.03
1,1,1 -T richloroethane
3,400.41
Chlorine
642,408.04
Naphthalene
0.03
Xylenes
2,683.77
Cyanide Compounds, PM
199,875.63
Methanol
2,187.37
Formaldehyde
175,560.45
Formaldehyde
1,720.49
1,3-Butadiene
137,409.43
Benzene
1,565.85
Manganese, PM
103,247.72
Dichloromethane
1,379.64
Acetaldehyde
77,843.69
Hexane
1,167.71
Nickel, PM
57,307.96
Acetaldehyde
700.59
Benzene
52,195.08
Ethylbenzene
622.36
Naphthalene
45,474.14
Rubidoux, California (Riverside County) - RUCA
Toluene
1,072.19
Acrolein
782,148.33
Naphthalene
0.02
Xylenes
574.60
Chlorine
96,879.61
Naphthalene
0.02
1,1,1 -T richloroethane
407.00
Formaldehyde
39,826.03
Formaldehyde
390.30
1,3-Butadiene
33,058.33
Methanol
375.60
Acetaldehyde
17,409.28
Benzene
314.48
Nickel, PM
13,127.70
Hexane
288.62
Naphthalene
12,816.64
Dichloromethane
172.55
Lead, PM
11,794.83
Styrene
169.01
Benzene
10,482.56
Acetaldehyde
156.68
Hexavalent Chromium. PM
8,289.97
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 6-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the California Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
San Jose, California (Santa Clara County) - SJJCA
Toluene
628.93
Acrolein
684,804.86
Arsenic
0.03
Xylenes
409.19
Chlorine
118,671.89
Arsenic
0.03
Methanol
293.38
2,4-Toluene diisocyanate
85,429.29
Naphthalene
0.02
Formaldehyde
243.18
1.3 -Butadiene
25,631.72
Naphthalene
0.02
Benzene
227.20
Formaldehyde
24,814.38
Nickel
0.02
Hexane
173.95
Acetaldehyde
16,791.50
Nickel
0.01
Ethylene glycol
165.72
Naphthalene
16,243.44
Acetaldehyde
151.12
Nickel, PM
9,059.86
1,1,1 -T richloroethane
118.67
Lead, PM
8,632.37
Ethylbenzene
106.49
Arsenic, PM
8,310.91
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
6.5 Summary of the 2015-2016 Monitoring Data for the California Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Naphthalene failed the most screens for all three California sites and thus, was
identified as a pollutant of interest for each of them. Two additional PMio metals
were identified as pollutants of interest for SJJCA, the only California site at which
PMwmetals were sampled.
~~~ CELA has the highest annual average concentrations of naphthalene among the
California monitoring sites. CELA 's annual averages of naphthalene are among the
highest calculated for NMP sites sampling PAHs.
~~~ Naphthalene concentrations have a decreasing trend at CELA in recent years. This is
also true for RUCA through 2015, but concentrations exhibit an increase for 2016.
~~~ None of the pollutants of interest for the California sites have cancer risk
approximations greater than 3 in-a-million; none of the pollutants of interest for the
California sites have noncancer hazard approximations greater than an HQ of 1.0.
6-33
-------
7.0 Sites in Colorado
This section summarizes those data from samples collected at the NATTS and UATMP
sites in Colorado and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines the
context of risk. Readers are encouraged to refer to Sections conumed in ihis ivpori
1 through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
7.1 Site Characterization
This section characterizes the Colorado monitoring sites by providing a description of
the nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient monitoring measurements.
The Colorado NATTS site is in Grand Junction (GPCO) while the six UATMP sites are
in neighboring Garfield County, in the towns of Battlement Mesa (BMCO), Silt (BRCO),
Parachute (PACO), Carbondale (RFCO), Glenwood Springs (GSCO), and Rifle (RICO).
Figure 7-1 for GPCO presents a composite satellite image retrieved from ArcGIS Explorer
showing the monitoring site and its immediate surroundings. Figure 7-2 identifies nearby point
source emissions locations by source category, as reported in the 2014 NEI for point sources,
version 1. Note that only sources within 10 miles of GPCO are included in the facility counts
provided in Figure 7-2. A 10-mile boundary was chosen to give the reader an indication of which
emissions sources and emissions source categories could potentially have a direct effect on the
air quality at the monitoring site. Further, this boundary provides both the proximity of emissions
sources to the monitoring site as well as the quantity of such sources within a given distance of
the site. Sources outside the 10-mile boundary are still visible on the map for reference but have
been grayed out to emphasize emissions sources within the boundary. Figures 7-3 through 7-10
are the composite satellite maps and emissions sources maps for the Garfield County sites.
Table 7-1 provides supplemental geographical information such as land use, location setting, and
locational coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
IXiln jjcncmlcd h> sources oilier
lluin I !R( i. LIJ.\'s conliiicl
kiliomioiA lor iIk- \\ll\ arc noi
included in I lie d^ila ;iiuil\ses
7-1
-------
Figure 7-1. Grand Junction, Colorado (GPCO) Monitoring Site
Ouray-Ave
Quray Ave
Grand Ave
Grand Ave
-iT-r&sL:
-------
Figure 7-2. NEI Point Sources Located Within 10 Miles of GPCO
Legend
~
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
GPCO NATTS site
D 10 mile radius
County boundary
Source Category Group (No. of Facilities)
r
Airport/Airline/Airport Support Operations (9)
O
Institution (school, hospital, prison, etc.) (4)
£
Asphalt Production/Hot Mix Asphalt Plant (2)
M
Landfill (1)
0
Auto Body Shop/Painters/Automotive Stores (4)
A
Metal Coating, Engraving, and Allied Services to
®
Automotive/RV Dealership (1)
Manufacturers (1)
~
Brick, Structural Clay, or Clay Ceramics Plant (2)
0
Metals Processing/Fabrication Facility (3)
B
Bulk Terminal/Bulk Plant (3)
X
Mine/Quarry/Mineral Processing Facility (23)
C
Chemical Manufacturing Facility (2)
?
Miscellaneous Commercial/Industrial Facility (1)
A
Concrete Batch Plant (3)
•
Oil and/or Gas Production (3)
M
Crematory (5)
R
Plastic, Resin, or Rubber Products Plant (2)
e
Electrical Equipment Manufacturing Facility (1)
X
Rail Yard/Rail Line Operations (2)
£
Electroplating, Plating, Polishing, Anodizing, and Coloring
T
Textile, Yarn, or Carpet Plant (1)
Facility (1)
I
Wastewater Treatment Facility (1)
IT
Gasoline/Diesel Service Station (41)
W
Woodwork, Furniture, Millwork and Wood Preserving
*
Industrial Machinery or Equipment Plant (2)
Facility (1)
Mesa County '
* i
7-3
-------
Figure 7-3. Battlement Mesa, Colorado (BMCO) Monitoring Site
¦W-Battlement-Pkwvj
4^attlpmcnt-Pkvvs
-------
Figure 7-4. Silt, Colorado (BRCO) Monitoring Site
-------
Figure 7-5. Parachute, Colorado (PACO) Monitoring Site
-------
Figure 7-6. Rifle, Colorado (RICO) Monitoring Site
-------
Figure 7-7. NEI Point Sources Located Within 10 Miles of BMCO, BRCO, PACO, and
RICO
108°15'0"W
10'0"W 108"5'0"W 108"0'0"W 107°55'0"W 107°50'0"W 107"45'0"W 107J40'0"W 107U35'0"W
Rio Blanco
County
Garfie d
County
f.
Mesa
County
108C10'0"W 108C5'0"W
3°0'0"W 107C55'0"W 107C50'0"W 107a45'0"W 107°40'0"W 107°35,0"W 107C30'0"W 107C25'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
tV BMCO UATMP site ^ BRCO UATMP site PACO UATMP site ~ RICO UATMP site
County boundary
Source Category Group (No. of Facilities)
t
Airport/Airline/Airport Support Operations (7)
r
Gasoline/Diesel Service Station (19)
£
Asphalt Production/Hot Mix Asphalt Plant (1)
¦
Landfill (1)
B
Bulk Terminal/Bulk Plant (1)
X
Mine/Quarry/Mineral Processing Facility (10)
i
Compressor Station (18)
Miscellaneous Commercial/Industrial Facility (1)
ffi
Dry Cleaning Facility (1)
•
Oil and/or Gas Production (870)
f
Electricity Generation Facility (1)
X
Rail Yard/Rail Line Operations (1)
®
Gas Plant (1)
TT
Telecommunications/Radio Facility (1)
7-8
-------
Figure 7-8. Glenwood Springs, Colorado (GSCO) Monitoring Site
• 4 ,
-------
Figure 7-9. Carbondale, Colorado (RFCO) Monitoring Site
-------
Figure 7-10. NET Point Sources Located Within 10 Miles of GSCO and RFCO
Eagle
County
' Mesa
I County
Pitkin
County
Miles
Legend
107"35'0"W 107"30'0"W 107n25'0"W 107"20,0"W 107n15,0"W 107"10'0"W 107"5'0"W 107WW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
RFCO UATMP site
GSCO UATMP site Q 10 mile radius
County boundary
I Garfield |
County j
Source Category Group (No. of Facilities)
r
Airport/Airline/Airport Support Operations (6)
a
Gasoline/Diesel Service Station (17)
«
Building/Construction (1)
O
Institution (school, hospital, prison, etc.) (1)
B
Bulk Terminal/Bulk Plant (1)
X
Mine/Quarry/Mineral Processing Facility (3)
i
Compressor Station (1)
•
Oil and/or Gas Production (3)
w
Crematory (2)
i
Wastewater Treatment Facility (1)
7-11
-------
Table 7-1. Geographical Information for the Colorado Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
GPCO
08-077-0017
08-077-0018
Grand
Junction
Mesa
Grand Junction,
CO
39.064289,
-108.561550
Commercial
Urban/City
Center
12,000
Bus-70 (Pitkin Ave) just E of 7th St
BMCO
08-045-0019
Battlement
Mesa
Garfield
Glenwood Springs,
CO
39.438060,
-108.026110
Commercial
Suburban
1,880
S Battlement Pkwy
BRCO
08-045-0009
Silt
Garfield
Glenwood Springs,
CO
39.487755,
-107.659685
Agricultural
Rural
1,182
Dry Hollow Rd
PACO
08-045-0005
Parachute
Garfield
Glenwood Springs,
CO
39.453654,
-108.053259
Residential
Urban/City
Center
17,000
1-70 near exit 75
RICO
08-045-0007
Rifle
Garfield
Glenwood Springs,
CO
39.531813,
-107.782298
Commercial
Urban/City
Center
16,000
Rte 13 connecting US-6 and 1-70
GSCO
08-045-0020
Glenwood
Springs
Garfield
Glenwood Springs,
CO
39.5464,
-107.3286
Commercial
Suburban
27,000
Hwy 82/Grand Ave, south of 6th St
RFCO
08-045-0018
Carbondale
Garfield
Glenwood Springs,
CO
39.412278,
-107.230397
Residential
Rural
18,000
Rte 133 just south of Hwy 82
1AADT reflects 2015 data for GPCO, PACO, GSCO RFCO, and RICO (CO DOT, 2015) and 2014 data for BMCO and BRCO (GCRBD, 2014)
BOLD ITALICS = EPA-designated NATTS Site
-------
The GPCO monitoring site is comprised of two locations. The first location is a small
1-story shelter that houses the VOC and carbonyl compound collection systems, with the PAH
collection system located just outside the shelter. The second location, which is on the roof of an
adjacent 2-story building, is comprised of the metals collection systems. As a result, two AQS
codes are provided in Table 7-1. Figure 7-1 shows that the area surrounding GPCO is of mixed
usage, with commercial businesses to the west, northwest, and north; residential areas to the
northeast and east; and industrial areas to the southeast, south, and southwest. This site's location
is next to a major east-west roadway (1-70 Business) in central Grand Junction. A rail line runs
northeast-southwest a few blocks to the south of the GPCO monitoring site, and merges with
another rail line to the southwest of the site. The Colorado River can be seen in the bottom left-
hand corner of Figure 7-1 near the junction with the Gunnison River. Grand Junction is located
in the Grand Valley, which lies north and northeast of the Colorado National Monument.
As Figure 7-2 shows, GPCO is located within 10 miles of numerous emissions sources.
Many of the sources are located along a diagonal line running roughly northwest to southeast
along Highways 6 and 50 and Business-70 and oriented along the mountain valley. Many of the
point sources near GPCO fall into the gasoline/diesel service station or the mine/quarry/mineral
processing source categories. The sources closest to GPCO (within a half-mile) are a bulk
terminal/bulk plant, a gasoline/diesel service station, and an auto body shop.
There are six monitoring sites located in the eastern half of Garfield County; four of the
six Garfield County monitoring sites are situated in towns located within a river valley along the
Colorado River and paralleling 1-70. The BMCO monitoring site is located in Battlement Mesa, a
rural community located to the east and southeast of Parachute. The monitoring site is located on
the roof of the Grand Valley Fire Protection District facility, near the intersection of Stone
Quarry Road and West Battlement Parkway, as shown in Figure 7-3. Developed property around
the site is primarily residential subdivisions, although a gas station is located immediately to the
north of the site and a cemetery is located to the south.
The BRCO monitoring site is located on Bell/Melton Ranch, off Owens Drive,
approximately 4 miles south of the town of Silt. The site is both rural and agricultural in nature.
As shown in Figure 7-4, the closest major roadway is County Road 331, Dry Hollow Road.
PACO is located on the roof of the old Parachute High School building, which is
presently operating as an early education facility. This location is in the center of the town of
7-13
-------
Parachute. The surrounding area is considered residential. Interstate-70 is less than a quarter of a
mile from the monitoring site, as shown in Figure 7-5. PACO is located approximately 1.8 miles
from BMCO; these are the two sites in Garfield County that are the closest to each other.
RICO is located on the roof of the Henry Annex Building in downtown Rifle. This
location is near the crossroads of several major roadways through town, as shown in Figure 7-6.
Highway 13 and US-6/24 intersect just south of RICO and 1-70 is just over a half-mile south of
the monitoring site, across the Colorado River. The surrounding area is commercial in nature.
These four Garfield County sites are located along a line running roughly east-west and
spanning approximately 20 miles; hence, they are shown together in Figure 7-7. These four sites
lie within an area with high oil and gas related activity. There are nearly 900 petroleum or natural
gas wells (collectively shown as the oil and/or gas production source category) within 10 miles
of these sites. Garfield County is collecting SNMOC samples to characterize the effects these
wells may have on the air quality in the surrounding areas (GCPH, 2015). There are also
numerous gasoline/diesel service stations, mine/quarry/mineral processing facilities, and
compressor stations within 10 miles of these sites.
The instrumentation at BMCO was moved to a new location, GSCO, in February 2015.
The GSCO site is located in Glenwood Springs, which is one of the easternmost towns in
Garfield County along the 1-70 corridor. GSCO is located at Vogelaar Park, adjacent to
Glenwood Springs Elementary School. This monitoring site is in a commercial area, with town
government buildings to the northeast, a decommissioned wastewater treatment facility to the
north (GCPH, 2017) and a church to the west. This location is also near the confluence of the
Colorado River with the Roaring Fork River, which are prominent features along the top and
center of Figure 7-8. 1-70 is located about a quarter-mile north of GSCO. After 13 months of
sampling, the instrumentation returned to the BMCO location.
Approximately 11 miles south of Glenwood Springs is Carbondale, where RFCO is
located. The RFCO monitoring site is the only site in Garfield County not located along the 1-70
corridor. The town of Carbondale resides in the Roaring Fork Valley (GCA, 2016), between the
Roaring Fork and Crystal Rivers, north of Mt. Sopris (Carbondale, 2017). The RFCO monitoring
site is located near the boathouse of the Rocky Mountain School on the bank of the Crystal River
in the northern part of town. The surrounding area is residential and rural in nature. Highway 82,
which runs southward from Glenwood Springs and separates Carbondale from the base of Red
7-14
-------
Hill, is just over one-third of a mile north of RFCO and is visible in the top right-hand corner of
Figure 7-9.
The emissions sources surrounding the sites in the Roaring Fork Valley are provided in a
separate map in Figure 7-10. This figure shows that GSCO and RFCO are located outside the oil
and gas fields of Garfield County. The emissions source category with the most sources within
10 miles of these sites is the gasoline and/or diesel service stations category. Within a half-mile
of GSCO are a crematory, a waste water treatment facility, a bulk terminal/bulk plant, and two
gasoline/diesel service stations. There are no emissions sources located within a half-mile of
RFCO; the closest emissions source is a gasoline/diesel service station.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 7-1 also contains traffic volume information for each site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. Among the
Colorado sites, the traffic volume for BRCO is the lowest while the traffic volume is highest near
GSCO. The traffic volumes near RICO, RFCO, PACO, GSCO, and GPCO are considerably
higher than the traffic volumes near BMCO and BRCO, which have some of the lowest traffic
volumes among NMP sites. Yet, the traffic volumes for all seven Colorado sites rank in the
bottom half compared to the traffic volumes for other NMP sites.
7.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
monitoring site to identify site-specific "pollutants of interest," which allows analysts and
readers to focus on a subset of pollutants through the context of risk. For each site, each
pollutant's preprocessed daily measurement was compared to its associated risk screening value.
If the concentration was greater than the risk screening value, then the concentration "failed the
screen." The site-specific results of this risk-based screening process are presented in Table 7-2
and incorporate measurements from both 2015 and 2016, where applicable. Pollutants of interest
are those for which the individual pollutant's total failed screens contribute to the top 95 percent
of the site's total failed screens and are shaded in gray in Table 7-2. It is important to note which
pollutants were sampled for at each site when reviewing the results of this analysis. VOCs,
carbonyl compounds, PMio metals, and PAHs were sampled for at GPCO while SNMOCs and
carbonyl compounds were sampled for at each of the Garfield County sites except RFCO.
Between January and September 2015, canister samples collected at RFCO were analyzed for
7-15
-------
both VOCs and SNMOCs, after which only SNMOCs were analyzed for the rest of 2015, as well
as throughout 2016.
Table 7-2. 2015-2016 Risk-Based Screening Results for the Colorado Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Grand Junction, Colorado - GPCO
Acetaldehyde
0.45
114
114
100.00
11.94
11.94
Formaldehyde
0.077
114
114
100.00
11.94
23.87
Naphthalene
0.029
114
116
98.28
11.94
35.81
Benzene
0.13
112
112
100.00
11.73
47.54
1.3 -Butadiene
0.03
111
112
99.11
11.62
59.16
Carbon Tetrachloride
0.17
111
112
99.11
11.62
70.79
1,2-Dichloroethane
0.038
98
99
98.99
10.26
81.05
Arsenic (PMio)
0.00023
65
117
55.56
6.81
87.85
Ethylbenzene
0.4
29
112
25.89
3.04
90.89
Acenaphthene
0.011
17
113
15.04
1.78
92.67
Dichloromethane
60
14
112
12.50
1.47
94.14
Fluoranthene
0.011
14
116
12.07
1.47
95.60
Hexachloro-1,3 -butadiene
0.045
10
14
71.43
1.05
96.65
Benzo(a)pyrene
0.00057
9
111
8.11
0.94
97.59
Fluorene
0.011
8
102
7.84
0.84
98.43
p-Dichlorobenzene
0.091
5
45
11.11
0.52
98.95
1,2-Dibromoethane
0.0017
4
4
100.00
0.42
99.37
Bro mo methane
0.5
3
112
2.68
0.31
99.69
Tricliloroethylene
0.2
2
19
10.53
0.21
99.90
Acenaphthylene
0.011
1
93
1.08
0.10
100.00
Total
955
1849
51.65
Battlement Mesa, Colorado - BMCO
Benzene
0.13
54
54
100.00
58.06
58.06
Formaldehyde
0.077
27
27
100.00
29.03
87.10
Acetaldehyde
0.45
12
27
44.44
12.90
100.00
Total
93
108
86.11
Silt, Colorado - BRCO
Benzene
0.13
109
109
100.00
60.89
60.89
Formaldehyde
0.077
49
54
90.74
27.37
88.27
Acetaldehyde
0.45
19
54
35.19
10.61
98.88
1.3 -Butadiene
0.03
2
2
100.00
1.12
100.00
Total
179
219
81.74
7-16
-------
Table 7-2. 2015-2016 Risk-Based Screening Results for the Colorado Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Glenwood Springs, Colorado - GSCO
Benzene
0.13
64
64
100.00
46.72
46.72
Formaldehyde
0.077
31
31
100.00
22.63
69.34
1.3 -Butadiene
0.03
25
27
92.59
18.25
87.59
Acetaldehyde
0.45
17
31
54.84
12.41
100.00
Total
137
153
89.54
Parachute, Colorado - PACO
Benzene
0.13
108
108
100.00
55.10
55.10
Formaldehyde
0.077
46
47
97.87
23.47
78.57
Acetaldehyde
0.45
32
46
69.57
16.33
94.90
1,3-Butadiene
0.03
9
9
100.00
4.59
99.49
Ethylbenzene
0.4
1
106
0.94
0.51
100.00
Total
196
316
62.03
Carbondale, Colorado - RFCO
Benzene
0.13
51
51
100.00
53.13
53.13
Carbon Tetrachloride
0.17
20
20
100.00
20.83
73.96
1,2-Dichloroethane
0.038
18
19
94.74
18.75
92.71
1.3 -Butadiene
0.03
7
7
100.00
7.29
100.00
Total
96
97
98.97
Rifle, Colorado - RICO
Benzene
0.13
106
106
100.00
34.42
34.42
1,3-Butadiene
0.03
77
77
100.00
25.00
59.42
Formaldehyde
0.077
56
56
100.00
18.18
77.60
Acetaldehyde
0.45
46
55
83.64
14.94
92.53
Ethylbenzene
0.4
23
107
21.50
7.47
100.00
Total
308
401
76.81
Observations from Table 7-2 include the following:
• The number of pollutants failing screens varied significantly between GPCO and the
Garfield County monitoring sites; this is expected given the difference in pollutants
measured at the sites.
• Concentrations of 20 pollutants failed at least one screen for GPCO; 52 percent of the
concentrations for these 20 pollutants were greater than their associated risk screening
value (or failed screens). GPCO ranks third for the number of pollutants failing
screens, behind only NBIL (23 pollutants), BTUT and S4MO (22 each), and tying
PXSS with 20.
7-17
-------
Twelve pollutants contributed to 95 percent of failed screens for GPCO and therefore
were identified as pollutants of interest for GPCO. These 12 include two carbonyl
compounds, six VOCs, three PAHs, and one PMio metal.
• The number of pollutants failing screens for the Garfield County sites range from
three (BMCO) to five (PACO and RICO). Benzene failed screens for each Garfield
County site. 1,3-Butadiene also failed screens at four of the six sites (BMCO and
BRCO are the exceptions). Formaldehyde and acetaldehyde failed screens for each
site sampling carbonyl compounds.
• For four of the six Garfield County sites (BMCO, GSCO, RFCO, RICO) all of the
pollutants that failed screens were also identified as site-specific pollutants of interest.
For BRCO and PACO, one pollutant was excluded from this designation.
• Benzene failed 100 percent of screens for all seven Colorado sites.
• Carbonyl compound samples were collected on a l-in-12 day sampling schedule at
BMCO, BRCO, GSCO, PACO, and RICO, while SNMOC samples were collected on
a l-in-6 day sampling schedule; thus, the number of carbonyl compound samples
collected at these sites were often less than half the number of SNMOC samples
collected. At RFCO, the concurrent VOC and SNMOC sampling frequency appears
more variable because site operators were collecting make-up samples for a number
of invalid samples, after which, a l-in-12 day sampling schedule resumed.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
7.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Colorado monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
7-18
-------
However, site-specific statistical summaries for all pollutants sampled for at the Colorado
monitoring sites are provided in Appendices J, K, M, N, and O.
7.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Colorado monitoring site, as described in Section 3.1. The
quarterly average concentration of a particular pollutant is simply the average concentration of
the preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated for a given year and where method completeness was greater than or equal to
85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
pollutants of interest for the Colorado monitoring sites are presented in Table 7-3, where
applicable. Note that concentrations of the PAHs and metals for GPCO are presented in ng/m3
for ease of viewing. Also note that if a pollutant was not detected in a given calendar quarter, the
quarterly average simply reflects "0" because only zeros substituted for non-detects were
factored into the quarterly average concentration.
7-19
-------
Table 7-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Colorado Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
frig/m3)
Q4
Avg
Oig/m3)
Annual
Average
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Grand Junction, Colorado - GPCO
Acetaldehyde
55/55/55
NA
1.54
±0.38
1.51
±0.18
1.67
±0.27
1.63
±0.17
59/59/59
1.52
±0.20
1.72
±0.31
2.06
±0.33
1.95
±0.38
1.81
±0.15
Benzene
50/50/50
NA
0.60
±0.06
0.91
±0.39
NA
NA
62/62/62
1.00
±0.23
0.53
±0.06
0.51
±0.11
0.86
±0.14
0.71
±0.09
1.3 -Butadiene
50/49/50
NA
0.09
±0.01
0.10
±0.02
NA
NA
62/49/62
0.17
±0.04
0.07
±0.01
0.07
±0.02
0.13
±0.03
0.11
±0.02
Carbon Tetrachloride
50/50/50
NA
0.60
±0.02
0.64
±0.03
NA
NA
62/62/62
0.55
±0.06
0.68
±0.04
0.59
±0.07
0.56
±0.05
0.60
±0.03
1,2-Dichloroethane
47/41/50
NA
0.07
± <0.01
0.04
±0.01
NA
NA
52/50/62
0.09
±0.01
0.09
±0.01
0.03
±0.02
0.06
±0.02
0.07
±0.01
Dichloromethane
50/50/50
NA
108.73
± 167.61
90.69
± 60.23
NA
NA
62/62/62
14.81
± 18.74
0.68
±0.22
1.14
±0.65
0.65
±0.15
4.19
±4.52
Ethylbenzene
50/49/50
NA
0.29
±0.04
0.40
±0.08
NA
NA
62/62/62
0.33
±0.06
0.24
±0.05
0.30
±0.09
0.38
±0.09
0.30
±0.04
Formaldehyde
55/55/55
NA
3.12
± 1.45
3.32
±0.36
2.99
±0.39
3.16
±0.51
59/59/59
2.85
±0.25
2.88
±0.49
3.11
±0.33
2.37
±0.28
2.80
±0.18
Acenaphthene3
55/55/55
6.06
±3.45
7.79
±3.03
17.80
± 15.96
2.46
±0.89
8.29
±3.97
58/58/61
1.64
±0.60
6.29
±2.56
10.46
±2.25
4.56
± 1.55
5.67
± 1.20
Arsenic (PMi0)a
57/57/57
0.35
±0.08
0.27
±0.06
0.22
±0.04
0.27
±0.06
0.28
±0.03
60/60/60
0.37
±0.12
0.24
±0.06
0.22
±0.05
0.44
±0.18
0.32
±0.06
Fluoranthene
55/55/55
2.11
±0.95
2.74
±0.82
15.75
±6.79
2.46
±0.51
5.59
±2.16
61/61/61
1.81
±0.29
6.79
±3.32
6.00
±2.17
2.07
±0.48
4.13
± 1.09
Naphthalene3
55/55/55
98.50
±25.20
79.14
± 17.27
95.14
±42.55
92.01
± 20.60
91.01
± 12.89
61/61/61
69.10
± 15.96
55.13
±8.28
70.75
± 16.45
102.89
±26.59
74.38
±9.54
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Table 7-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Colorado Monitoring Sites (Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
frig/m3)
Q4
Avg
Oig/m3)
Annual
Average
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Battlement Mesa, Colorado - BMCO
Acetaldehyde
3/3/3
NA
NS
NS
NS
NA
24/24/24
NS
0.45
±0.25
0.62
±0.17
0.42
±0.12
0.48
±0.10
Benzene
6/6/6
NA
NS
NS
NS
NA
48/48/48
NS
0.70
±0.08
0.75
±0.08
0.97
±0.14
0.79
±0.06
Formaldehyde
3/3/3
NA
NS
NS
NS
NA
24/24/24
NS
0.78
±0.31
1.34
±0.27
0.78
±0.18
0.93
±0.18
Silt, Colorado - BRCO
Acetaldehyde
28/28/28
0.32
±0.15
0.69
±0.34
0.58
±0.18
0.43
±0.19
0.51
±0.11
26/26/26
NA
0.39
±0.26
0.11
±0.06
0.16
±0.09
0.22
±0.08
Benzene
52/52/52
0.88
±0.36
0.55
±0.11
NA
0.66
±0.14
0.66
±0.12
57/57/58
1.09
±0.23
0.46
±0.11
0.60
±0.20
0.64
±0.21
0.71
±0.11
Formaldehyde
28/28/28
0.47
±0.25
1.08
±0.44
1.05
±0.31
0.67
±0.27
0.82
±0.17
26/26/26
NA
0.67
±0.41
0.16
±0.12
0.22
±0.15
0.37
±0.14
Glenwood S
jrings, Colorado - GSCO
Acetaldehyde
26/26/26
NA
0.65
±0.26
0.48
±0.19
0.52
±0.18
0.53
±0.10
5/5/5
NA
NA
NS
NS
NA
Benzene
52/52/52
NA
0.42
±0.05
0.47
±0.07
0.59
±0.12
0.49
±0.04
12/12/12
0.83
±0.12
NA
NS
NS
NA
1.3 -Butadiene
19/6/52
NA
0.01
±0.01
0
0.05
±0.02
0.02
±0.01
8/6/12
0.05
±0.03
NA
NS
NS
NA
Formaldehyde
26/26/26
NA
0.92
±0.34
0.93
±0.26
0.69
±0.18
0.80
±0.14
5/5/5
NA
NA
NS
NS
NA
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Table 7-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Colorado Monitoring Sites (Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
frig/m3)
Oig/m3)
frig/m3)
Oig/m3)
frig/m3)
# Samples
frig/m3)
frig/m3)
frig/m3)
frig/m3)
frig/m3)
Parachute, Colorado - PACO
0.57
0.45
0.38
0.61
0.71
0.64
0.59
Acetaldehyde
17/17/18
NA
NA
±0.22
±0.28
NA
29/29/29
±0.30
±0.25
±0.12
±0.19
±0.11
1.56
0.92
1.05
1.21
1.18
1.22
1.12
1.20
Benzene
54/54/54
±0.35
±0.15
±0.15
NA
±0.16
54/54/54
±0.41
NA
±0.18
±0.23
±0.14
0.02
<0.01
0.01
<0.01
<0.01
<0.01
1.3 -Butadiene
7/3/54
±0.02
±0.01
0
NA
±0.01
2/0/54
±0.01
NA
0
±0.01
±<0.01
1.41
0.97
0.66
1.21
1.59
1.19
1.18
Formaldehyde
18/18/18
NA
NA
±0.32
±0.65
NA
29/29/29
±0.47
±0.37
±0.16
±0.25
±0.19
Carbondale, Colorado - RFCO
0.46
0.42
0.42
0.92
0.59
0.34
0.43
0.45
Benzene
24/24/26
NA
±0.18
±0.05
±0.13
±0.68
27/27/27
±0.20
±0.07
±0.18
NA
±0.08
0.02
0.02
0.01
0.02
<0.01
1,3-Butadiene
5/4/26
NA
±0.03
0
±0.03
±0.01
2/1/27
±0.02
0
0
NA
±0.01
0.60
0.64
Carbon Tetrachloride
20/20/20
NA
±0.05
±0.06
NS
NA
NS
NS
NS
NS
NS
NS
0.07
0.04
1,2-Dichloroethane
19/14/20
NA
±0.01
±0.01
NS
NA
NS
NS
NS
NS
NS
NS
Rifle
, Colorado - RICO
0.66
1.01
0.90
0.57
0.90
1.04
0.86
Acetaldehyde
25/25/26
±0.24
NA
NA
±0.24
NA
30/30/30
±0.44
±0.15
±0.29
±0.24
±0.15
1.25
1.27
1.71
0.71
0.76
1.13
1.10
Benzene
44/44/46
±0.24
NA
NA
±0.22
NA
62/62/62
±0.32
±0.07
±0.05
±0.20
±0.14
0.10
0.12
0.14
0.02
0.03
0.10
0.07
1.3 -Butadiene
38/32/46
±0.04
NA
NA
±0.02
NA
39/27/62
±0.04
±0.01
±0.03
±0.03
±0.02
0.36
0.34
0.36
0.22
0.27
0.30
0.29
Ethylbenzene
45/44/46
±0.10
NA
NA
±0.07
NA
62/62/62
±0.07
±0.04
±0.03
±0.05
±0.03
0.96
1.17
1.44
1.24
1.20
1.04
1.52
1.47
1.31
Formaldehyde
26/26/26
±0.33
±0.72
NA
±0.37
±0.21
30/30/30
±0.46
±0.22
±0.19
±0.31
±0.16
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Observations for GPCO from Table 7-3 include the following:
• A number of VOC and carbonyl compound samples that did not run long enough
were collected in February and March 2015; thus, these pollutants do not have first
quarter average concentrations provided. This combined with additional invalid
samples throughout the year, particularly during the fourth quarter, results in a
completeness less than 85 percent for the VOCs and thus, fourth quarter and annual
averages could not be calculated either. However, statistical summaries for 2015 are
provided in Appendix J for the valid samples collected.
• Excluding the VOCs, formaldehyde is the pollutant with the highest annual average
concentration for GPCO in 2015 (3.16 ± 0.51 |ig/m3). While the available 2015
quarterly average concentrations do not vary significantly, the confidence interval for
the second quarter is three to four times greater than the other confidence intervals
shown. A review of the data shows that the maximum formaldehyde concentration
(15.1 |ig/m3) was measured at GPCO on May 3, 2015; the next highest formaldehyde
(5.12 |ig/m3) is one-third as high. GPCO is one of only six NMP sites at which a
formaldehyde concentration greater than 15 |ig/m3 was measured during 2015 and
2016. Formaldehyde concentrations measured at GPCO over the two years of
sampling range from 1.01 |ig/m3 to 15.1 |ig/m3. Based on the averages shown in
Table 7-3, formaldehyde concentrations appear higher in 2015 than in 2016, though
the difference is not statistically significant. All but one of the 10 formaldehyde
concentrations greater than 4 |ig/m3 were measured at GPCO in 2015.
• The maximum acetaldehyde concentration was also measured at GPCO on May 3,
2015 (3.95 |ig/m3), although the difference between this measurement and other
"higher" concentrations measured at this site are considerably less. Acetaldehyde
concentrations measured at GPCO range from 0.645 |ig/m3 to 3.95 |ig/m3, with a total
of six concentrations greater than or equal to 3 |ig/m3 measured over the two-year
period.
• The available quarterly average concentrations of dichloromethane exhibit
considerable variability, spanning three orders of magnitude. The second and third
quarter average concentrations for 2015 are both around 100 |ig/m3, with very large
confidence intervals associated with them. Prior to the April 2016, dichloromethane
concentrations measured at GPCO range from 0.397 |ig/m3 to 1,493 |ig/m3, including
seven concentrations greater than 100 |ig/m3 and 22 greater than 10 |ig/m3. After the
first quarter of 2016, both the magnitude and variability of the measurements
decrease significantly; concentrations greater than 5 |ig/m3 were not measured at
GPCO after the first quarter of 2016, with only eight concentrations greater than
1 |ig/m3 measured during this nine-month period.
• Concentrations of benzene and 1,3-butadiene appear highest during the colder months
of the year, based on the quarterly average concentrations for 2016 shown in
Table 7-3. A review of the data shows that all but one of the 11 benzene
concentrations greater than 1 |ig/m3 were measured during the first or fourth quarters
of 2016 (and the one exception was in mid-September). Conversely, all but three of
the 21 benzene concentrations less than 0.5 |ig/m3 were measured between April and
September, during the second or third quarters of 2016. While similar trends are also
7-23
-------
shown in the 2015 data, the maximum benzene concentration was measured in July
(3.39 |ig/m3), explaining the relatively large confidence interval for the third quarter
of 2015. Similarly, the 15 highest 1,3-butadiene concentrations were measured at
GPCO during the first or fourth quarters of 2016 while most of the lowest
concentrations were measured between April and September. Similar observations
were made in the 2014 NMP report.
All four quarterly average concentrations for both years are provided in Table 7-3 for
the PAH and metal pollutants of interest. Concentrations of arsenic measured across
the two years of sampling do not exhibit much variability; arsenic concentrations
measured at GPCO range from 0.070 ng/m3 to 1.40 ng/m3, with the bulk of the
measurements falling between 0.10 ng/m3 and 0.60 ng/m3. The four highest arsenic
concentrations (those greater than 0.7 ng/m3) were measured at GPCO during the first
and fourth quarters of 2016, contributing to the larger confidence intervals shown for
these quarterly average concentrations.
Among the PAHs, naphthalene has the highest annual average concentrations for both
years. Concentrations of naphthalene measured at GPCO range from 18.0 ng/m3 to
321 ng/m3; the maximum naphthalene concentration measured at GPCO is the fourth
highest among NMP sites sampling this pollutant. The second quarter average
concentrations of naphthalene for both years, but particularly for 2016, appear lower
than the other quarterly averages shown. The maximum concentration measured
during the second quarter of 2016 is 87.3 ng/m3; multiple concentrations greater than
87.3 ng/m3 were measured during every other calendar quarter, from as few as four
(first quarter 2016) to as many as 11 (fourth quarter 2016).
The third quarter average concentration of acenaphthene for 2015 is considerably
higher than the other quarterly average concentrations shown for both years, and its
associated confidence interval is nearly the same magnitude as the average itself
(17.80 ± 15.96 ng/m3). A review of the data shows that the maximum acenaphthene
concentration (108 ng/m3) was measured at GPCO on July 17, 2015; the next highest
concentration (25.5 ng/m3) was also measured in July 2015, but is one-quarter the
magnitude. GPCO's maximum acenaphthene concentration is the highest
concentration of this pollutant measured among NMP sites sampling this pollutant.
The highest concentrations of acenaphthene measured at GPCO tended to be
measured during the warmer months of the year. Of the 18 acenaphthene
concentrations greater than 10 ng/m3 measured at GPCO, 16 were measured between
June and September.
The highest concentrations of fluoranthene were also measured at GPCO during the
third quarter of 2015, including all four measurements greater than 25 ng/m3. There is
little variation in the quarterly averages for 2015, with the exception of the third
quarter. For 2016, the second and third quarter averages are significantly higher than
the first and fourth quarter averages. Similar to acenaphthene, the highest
fluoranthene concentrations tended to be measured during the warmer months of the
year. The 17 highest fluoranthene concentrations were measured at GPCO between
June and September of either year. Few fluoranthene concentrations greater than
5 ng/m3 were measured outside these months (one of 12 in 2015 and two in 2016, out
of a total of 25).
7-24
-------
Observations for the Garfield County sites from Table 7-3 include the following:
• Several quarterly average concentrations, and some annual averages are "missing" for
the Garfield County sites in Table 7-3. The reasons for this are varied, including:
- The instrumentation at BMCO was moved to GSCO in February 2015, then was
returned to the BMCO location in March 2016.
- Due to the instrument relocation from GSCO back to BMCO in March 2016 and
one invalid sample, the criteria for a quarterly average concentration to be
calculated was not met.
- A quarterly average concentration for the third quarter of 2015 could not be
calculated for BRCO due to flow controller issues with the SNMOC collection
system in September and October 2015. Quarterly averages for the carbonyl
compounds for the first quarter of 2016 could not be calculated for BRCO due to
collection system issues experienced in January and February.
- A series of carbonyl compound samples resembling field blanks was collected at
PACO during the first half of 2015, resulting in the invalidation of one-third of
the samples collected in 2015.
- Canisters were analyzed for both VOCs and SNMOCs concurrently at RFCO
between January and September 2015; thus, VOC pollutants of interest for 2015
not on the SNMOC analyte list would not have any averages for 2016. Also, a
number of invalid samples were collected in the first half of 2015 resulting in no
quarterly averages for the first quarter of 2015.
- A number of SNMOC samples did not run properly at RICO in June and first part
of July 2015; this combined with other invalid samples throughout the year
resulted in a completeness less than 85 percent for RICO in 2015.
- RICO has a quarterly average concentration for formaldehyde for the second
quarter of 2015 while acetaldehyde does not. This is due to co-elution for one
sample, for which an acetaldehyde concentration could not be resolved.
• Acetaldehyde and formaldehyde are pollutants of interest for each Garfield County
site that sampled carbonyl compounds (all sites except RFCO).
• Quarterly average concentrations of acetaldehyde range from 0.11 ± 0.06 |ig/m3
(BRCO, third quarter 2016) to 1.04 ± 0.37 |ig/m3 (RICO, fourth quarter 2016). BRCO
has four of the five lowest quarterly average concentrations among the Garfield
County sites as well as NMP sites. In fact, the Garfield County sites account for 10 of
the 11 quarterly averages of acetaldehyde less than 0.5 |ig/m3. RICO is the only site
for which a quarterly average concentration of acetaldehyde greater than 1 |ig/m3 was
calculated. BRCO and RICO also have the lowest (0.22 ± 0.08 |ig/m3 - BRCO, 2016)
and highest (0.86 ±0.15 |ig/m3 - RICO, 2016) annual average concentrations,
respectively, of acetaldehyde among the Garfield County sites. Excluding RICO,
Garfield County sites account for the five lowest annual average concentrations of
acetaldehyde among sites sampling this pollutant. BRCO is the only Garfield County
site for which an annual average concentration could be calculated for both years.
BRCO's annual average concentration for 2015 (0.51 ± 0.11 |ig/m3) is more than
twice the annual average concentration for 2016 (0.22 ± 0.08 |ig/m3).
7-25
-------
• Quarterly average concentrations of formaldehyde range from 0.16 ± 0.06 |ig/m3
(BRCO, third quarter 2016) to 1.59 ± 0.37 |ig/m3 (PACO, third quarter 2016). BRCO
has the two lowest quarterly average concentrations of formaldehyde among NMP
sites. The Garfield County sites account for 13 of the 21 quarterly averages of
formaldehyde less than 1 |ig/m3 (SEWA accounts for the others). RICO has the most
quarterly average concentrations of formaldehyde greater than 1 |ig/m3 (six) among
the Garfield County sites. BRCO and RICO also have the lowest (0.37 ± 0.14 |ig/m3 -
BRCO, 2016) and highest (1.31 ± 0.16 |ig/m3 - RICO, 2016) annual average
concentrations, respectively, of formaldehyde among the Garfield County sites.
BRCO's 2016 annual average concentration of formaldehyde is the lowest annual
average concentration of formaldehyde among sites sampling this pollutant. BRCO
and RICO are the only Garfield County sites for which an annual average
formaldehyde concentration could be calculated for both years. BRCO's annual
average concentration for 2015 (0.82 ± 0.17 |ig/m3) is also more than twice the
annual average concentration for 2016 (0.37 ± 0.14 |ig/m3), while RICO's annual
averages are similar to each other.
• Benzene was sampled for and is a pollutant of interest for all six Garfield County
sites. Quarterly average concentrations of benzene range from 0.34 ± 0.07 |ig/m3
(RFCO, second quarter 2016) to 1.71 ± 0.37 |ig/m3 (RICO, first quarter 2016). This
quarterly average concentration for RICO is among the highest quarterly averages of
benzene among NMP sites sampling this pollutant, while the quarterly average for
RFCO is among the lowest. PACO (5) and RICO (4) have the most quarterly average
concentrations of benzene greater than 1 |ig/m3. RFCO has the lowest annual average
concentration of benzene (0.45 ± 0.08 |ig/m3 - 2016) while PACO has the highest
(1.21 ± 0.16 |ig/m3 - 2015) among the Garfield County sites. Both of PACO's annual
average concentrations of benzene are among the highest annual averages for this
pollutant. BRCO, PACO, and RFCO are the Garfield County sites for which an
annual average benzene concentration could be calculated for both years. There is
little difference between the annual averages for BRCO and PACO, while the annual
average for 2015 for RFCO (0.92 ± 0.68 |ig/m3) is twice the annual average
concentration for 2016 (0.45 ± 0.08 |ig/m3). Note that the confidence interval for the
RFCO's 2015 annual average is considerably large, indicating the potential for
outliers. The confidence intervals shown for the available 2015 quarterly average
concentrations do not reflect this level of variability. A review of the data shows that
the four highest benzene concentrations measured at RFCO were measured in
February and March and range from 0.902 |ig/m3 to 6.62 |ig/m3; the two highest
benzene concentrations measured at RFCO are the second and third highest
concentrations of benzene measured across the program. Note that for pollutants on
both the VOC and SNMOC analyte list, such as benzene, the measurements provided
by the SNMOC method were used for RFCO.
• 1,3-Butadiene is also a pollutant of interest for GSCO, PACO, RFCO, and RICO.
This pollutant is detected at these sites at a lower rate than the other pollutants of
interest shown in Table 7-3. In fact, several of the quarterly average concentrations
are zero, indicating that 1,3-butadiene was not detected during that calendar quarter.
RICO is the only Garfield County site with a quarterly average concentration of
1,3-butadiene greater than 0.1 |ig/m3. Annual average concentrations of 1,3-butadiene
7-26
-------
range from <0.01 ± <0.01 |ig/m3 (PACO, 2016) to 0.07 ± 0.02 |ig/m3 (RICO, 2016)
among the Garfield County sites.
• Carbon tetrachloride and 1,2-dichloroethane were also identified as pollutants of
interest for RFCO. These pollutants are analytes on the VOC list only; thus, few
quarterly average concentrations are shown in Table 7-3.
• Ethylbenzene is also a pollutant of interest for RICO. The quarterly average
concentrations of ethylbenzene do not vary significantly among the available
quarterly averages shown in Table 7-3.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Colorado
sites from those tables include the following:
• Annual average concentrations for the Colorado sites appear in Tables 4-10 through
4-13 a total of seven times, with GPCO appearing the most (4).
• PACO has the second (2015) and third (2016) highest annual average concentrations
of benzene among all NMP sites sampling this pollutant, as indicated above, with
RICO's 2016 annual average ranking sixth highest.
• None of the Colorado sites appear in Table 4-11 for the carbonyl compounds.
• GPCO has the fifth highest annual concentration of naphthalene (2015) among NMP
sites sampling PAHs, as shown in Table 4-12. This site's annual average naphthalene
concentration for 2016 ranks 11th highest. GPCO's annual averages of acenaphthene
rank 7th (2015) and 10th (2016) highest among sites sampling these pollutants.
GPCO's 2015 annual average of fluorene also ranks 7th highest among sites sampling
these pollutants, with this site's 2016 annual average ranking 12th.
• GPCO does not appear in Table 4-13 for arsenic.
7.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for each of the pollutants
listed in Table 7-3 for each site. Note that the box plots for benzene, 1,3-butadiene, and
ethylbenzene were split into separate figures, one for samples collected and analyzed with
Method TO-15 (GPCO, RFCO) and one for samples collected and analyzed with the SNMOC
method (the Garfield County sites), where annual averages could be calculated. Figures 7-11
through 7-22 overlay the sites' minimum, annual average, and maximum concentrations for each
year onto the program-level minimum, first quartile, median, average, third quartile, and
maximum concentrations, as described in Section 3.4.2.1, and are discussed below. If an annual
7-27
-------
average concentration could not be calculated, the range of concentrations is still provided in the
figures that follow.
Figure 7-11. Program vs. Site-Specific Average Acenaphthene Concentrations
! Proeram Max Concentration = 108 ne/m3 !
A L
1
**
; GPCO 2016 Max Concentration = 108 ng/m3 !
0 15 30 45 60 75
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
~ ~ ~ ~ i
Site: 2015Average 2016 Avereage Concentration Range, 2015&2016
o o
Figure 7-11 presents the box plot for acenaphthene for GPCO and shows the following:
• The program-level maximum acenaphthene concentration (108 ng/m3) is not shown
directly on the box plot in Figure 7-11 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale has been reduced.
• The maximum acenaphthene concentration measured at GPCO in 2015 is the
maximum acenaphthene concentration measured across the program, as discussed in
the previous section. Although the range of concentrations measured at GPCO in
2015 appears considerably larger than the range measured in 2016, if the maximum
concentration is excluded, the ranges are more similar (the second highest
concentration measured in 2015 is 25.5 ng/m3).
• The minimum concentration measured in 2015 is greater than the program-level 25th
percentile. There were three non-detects measured in 2016 (compared to none in
2015).
• Both annual average concentrations of acenaphthene for GPCO are greater than the
program-level average concentration as well as the program-level third quartile. As
discussed in the previous section, GPCO's annual averages are among the highest
calculated for this pollutant.
7-28
-------
Figure 7-12. Program vs. Site-Specific Average Acetaldehyde Concentrations
GPCO
BMCO
BRCO
GSCO
PACO
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 7-12 presents the box plots for acetaldehyde and shows the following:
• The box plots show that the range of acetaldehyde concentrations measured at GPCO
is considerably larger than the range of concentrations measured at the Garfield
County sites. The entire range of acetaldehyde concentrations measured at most of the
Garfield County sites (RICO being the exception) is less than the program-level
average concentration. However, the range of concentrations measured at all of the
Colorado sites is relatively small compared to the range measured across the program.
=
a-
91
7-29
-------
Note that the range of measurements for BMCO in 2015 is very small but represents
only one month of sampling.
• GPCO has the highest annual average acetaldehyde concentrations among the
Colorado sites, where they could be calculated. The 2015 annual average
concentration for GPCO is similar to the program-level average concentration and the
2016 annual average is just greater than the program average. By comparison, the
annual average concentrations for most of the Garfield County sites are less than the
program-level first quartile.
• Carbonyl compounds were not sampled for at RFCO.
Figure 7-13. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
r
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~ ~ ~ 1
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 7-13 presents the box plot for arsenic for GPCO and shows the following:
• The maximum arsenic concentration measured at GPCO in 2016 is more than twice
the maximum arsenic concentration measured in 2015. The entire range of
concentrations measured in 2015 is less than the program-level average concentration
(0.703 ng/m3). For 2016, only the maximum concentration measured is greater than
the program-level third quartile; all but the three highest concentrations measured at
GPCO in 2016 are less than the program-level average concentration.
• The 2015 and 2016 annual average concentrations for GPCO are similar to each other
and both are just less than the program-level first quartile. This site has the lowest
annual average concentrations of arsenic among NMP sites sampling arsenic.
7-30
-------
Figure 7-14a. Program vs. Site-Specific Average Benzene (Method TO-15) Concentrations
GPCO
Bi
t
Concentration ([ig/m3]
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 7-14a presents the box plots for benzene (TO-15) and shows the following:
• Figure 7-14a presents the minimum, maximum, and annual average concentrations of
benzene, where available, for GPCO and RFCO compared to the benzene
concentrations measured across the program for NMP sites sampling VOCs with
Method TO-15. As previously discussed, canisters collected at RFCO were analyzed
with Method TO-15 between January and September 2015.
• The two highest benzene concentrations measured across the program were measured
at RFCO (7.33 |ig/m3 and 7.30 |ig/m3). The next highest benzene concentration
measured at RFCO is also among the highest across the program (4.42 |ig/m3). These
three concentrations were measured at RFCO on three consecutive sample days in
March 2015; the remaining measurements are less than 1 |ig/m3, ranging from
0.198 |ig/m3 to 0.567 |ig/m3.
• Although the maximum benzene concentration measured at GPCO is roughly half the
magnitude as the maximum concentration measured at RFCO, it is still among the
higher benzene concentrations measured across the program. The range of benzene
concentrations measured at GPCO in 2015 is larger than the range measured in 2016.
• The annual average concentration for GPCO for 2016 is similar to the program-level
average concentration. (An annual average could not be calculated for 2015.)
7-31
-------
Figure 7-14b. Program vs. Site-Specific Average Benzene (SNMOC) Concentrations
BMCO
BRCO
PACO
RFCO
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Average
Concentration Range, 2015 & 2016
a
o
Figure 7-14b presents the box plots for benzene (SNMOC) and shows the following:
• Figure 7-14b presents the minimum, maximum, and annual average benzene
concentrations for the Garfield County sites compared to the benzene concentrations
measured across the program for NMP sites sampling SNMOCs. Note that the scales
vary slightly between Figures 7-14a and 7-14b.
• The higher benzene concentrations measured at RFCO with Method TO-15 are
reflected in the benzene concentrations measured with the concurrent SNMOC
7-32
-------
method. The range of benzene concentrations measured across the Garfield County
sites is highly variable.
• Of the Garfield County sites, PACO has the highest annual average concentrations of
benzene, followed by RICO (2016) and RFCO (2015). The available annual average
concentrations for the remaining sites (and RFCO's 2016 annual average) are less
than the program-level average concentration, with several also less than the
program-level median concentration.
• For the sites with two available annual averages, the difference between the two is
largest for RFCO.
Figure 7-15a. Program vs. Site-Specific Average 1,3-Butadiene (Method TO-15)
Concentrations
GPCO
1
Yf. '
i
J Program Max Concentration
= 3.90 |ig/m3
\
1
It
1
j Program Max Concentration
= 3.90 |ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 7-15a presents the box plots for 1,3-butadiene (TO-15) and shows the following:
• Figure 7-15a presents the minimum, maximum, and annual average concentrations of
1,3-butadiene, where available, for GPCO and RFCO compared to the 1,3-butadiene
concentrations measured across the program for NMP sites sampling VOCs with
Method TO-15. As previously discussed, canisters collected at RFCO were analyzed
with Method TO-15 between January and September 2015.
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plots in Figure 7-15a because the scale of the box plots has been
reduced to allow for the observation of data points at the lower end of the
concentration range.
• Figure 7-15a shows that the maximum 1,3-butadiene concentration measured at
GPCO is an order of magnitude less than the program-level maximum concentration.
The range of 1,3-butadiene concentrations measured at GPCO in 2016 is just slightly
less than the range measured in 2015. For RFCO, all but one 1,3-butadiene
7-33
-------
concentration are less than 0.1 |ig/m3. Non-detects of 1,3-butaidene were not
measured at GPCO, as the minimum 1,3-butadiene concentration for both years are
similar to the program-level first quartile. Two non-detects were measured at RFCO.
• GPCO's annual average concentration of 1,3-butadiene for 2016 is greater than both
the program-level average concentration and the program-level third quartile.
Figure 7-15b. Program vs. Site-Specific Average 1,3-Butadiene (SNMOC) Concentrations
GSCO
PACO
0.05
0.1
0.15 0.2
Concentration (jug/m3)
0.25
0.3
0.35
Program: 1st Quartile
¦
Site: 2015 Average
2nd Quartile 3rd Quartile 4th Quartile Avera§
~ ~ ~
2016 Average Concentration Range, 2015 & 2016
o —
Figure 7-15b presents the box plots for 1,3-butadiene (SNMOC) and shows the
following:
• Figure 7-15b includes a box plot for GSCO, PACO, RFCO, and RICO, the Garfield
County sites for which 1,3-butadiene is a pollutant of interest. The program-level
first, second, and third quartiles are zero for sites sampling SNMOCs, and thus, not
visible in Figure 7-15b, indicating that at least 75 percent of the 1,3-butadiene
concentrations measured by sites sampling SNMOCs were non-detects. The box plots
show that non-detects were measured at each of the Garfield County sites shown.
7-34
-------
• The maximum 1,3-butadiene concentration measured at RICO (0.311 |ig/m3) is
roughly three to five times the maximum concentrations measured among the
remaining Garfield County sites.
• Of the Garfield County sites shown, RICO has the highest annual average
concentration of 1,3-butadiene (2016) and is the only one greater than the program-
level average 1,3-butadiene concentration.
Figure 7-16. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
GPCO
RFCO
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 7-16 presents the box plots for carbon tetrachloride for GPCO and RFCO and
shows the following:
• The range of carbon tetrachloride concentrations measured at these two sites is
considerably smaller than the range measured across the program, particularly for
RFCO.
• The program-level median and average concentrations are nearly identical and thus,
plotted nearly on top of each other in Figure 7-16.
• The annual average carbon tetrachloride concentration for GPCO for 2016 is greater
than the program-level first quartile but less than the program-level median and
average concentrations. GPCO's 2016 annual average of carbon tetrachloride is
among the lowest annual averages among NMP sites sampling this pollutant.
However, the variability in annual averages of carbon tetrachloride among NMP sites
is rather small, with 0.1 |ig/m3 separating most sites' annual averages.
7-35
-------
Figure 7-17. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1
GPCO
s
Program Max Concentration =45.8 (.ig/m3
1
1
1
"
Program Max Concentration = 45.8 (.ig/m3 !
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 7-17 presents the box plots for 1,2-dichloroethane for GPCO and RFCO and
shows the following:
• The program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is not
shown directly on the box plots in Figure 7-17 as the program-level maximum
concentration is considerably greater than the majority of concentrations measured
across the program.
• All of the concentrations of 1,2-dichloroethane measured at GPCO and RFCO are
less than the program-level average concentration of 0.30 |ig/m3. The maximum
concentration of 1,2-dichloroethane measured at GPCO (0.146 |ig/m3) is half the
magnitude of the program-level average concentration and RFCO's maximum
concentration (0.105 |ig/m3) is one-third the magnitude.
• The annual average concentration for GPCO for 2016 falls between program-level
first quartile and median concentration. The program-level average concentration is
being driven by the measurements at the upper end of the concentration range.
7-36
-------
Figure 7-18. Program vs. Site-Specific Average Dichloromethane Concentrations
Program Max Concentration 1,493 ug/m3
1
¦
GPCO 2015 Max= 1,493 |ig/m3; 2016 Max= 104 ^ig/m3
012345678
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 7-18 presents the box plot for dichloromethane for GPCO and shows the
following:
• The program-level maximum dichloromethane concentration (1,493 |ig/m3) is not
shown directly on the box plot in Figure 7-18 as the program-level maximum
concentration is considerably greater than the majority of concentrations measured
across the program.
• The program-level average concentration of dichloromethane is also being driven by
the measurements at the upper end of the concentration range. While the maximum
dichloromethane concentration measured at GPCO in 2015 is the maximum
concentration measured across the program, GPCO is not the only site at which a
dichloromethane concentration greater than 1000 |ig/m3 was measured. The
maximum dichloromethane concentration measured at GPCO in 2016 is an order of
magnitude less than the maximum concentration measured in 2015, though both of
these measurements exceed the scale in Figure 7-18.
Figure 7-19a. Program vs. Site-Specific Average Ethylbenzene (Method TO-15)
Concentrations
¦
>
0 0.5 1 1.5 2 2.5 3
Concentration (|jg/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
7-37
-------
Figure 7-19a presents the box plot for ethylbenzene (TO-15) for GPCO and shows the
following:
• Figure 7-19a presents the minimum, maximum, and annual average concentrations of
ethylbenzene, where available, for GPCO compared to the ethylbenzene
concentrations measured across the program for NMP sites sampling VOCs with
Method TO-15. A box plot is not provided for RFCO because ethylbenzene is not a
pollutant of interest for this site.
• Six concentrations measured at GPCO in 2015 are greater than the maximum
concentration measured in 2016, as the maximum ethylbenzene concentration
measured at GPCO in 2015 is nearly twice the magnitude of the maximum
concentration measured in 2016.
• Although a few non-detects of ethylbenzene were measured across the program, none
were measured at GPCO.
• GPCO's annual average concentration of ethylbenzene for 2016 falls between the
program-level average concentration and third quartile.
Figure 7-19b. Program vs. Site-Specific Average Ethylbenzene (SNMOC) Concentrations
mc° Ei ~j 1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Average
Concentration Range, 2015 & 2016
a
o
Figure 7-19b present the box plot for ethylbenzene (SNMOC) for RICO and shows the
following:
• Figure 7-19b includes a box plot for RICO, the only Garfield County site for which
ethylbenzene is a pollutant of interest.
• Although maximum concentrations measured at RICO each year are fairly similar,
the minimum concentrations are considerably different. A single non-detect was
measured at RICO in 2015; if this non-detect was excluded from the dataset, the
minimum concentrations for each year would be more similar.
• RICO's annual average concentration of ethylbenzene for 2016 is greater than the
program-level average concentration and third quartile for ethylbenzene
concentrations measured at sites sampling SNMOCs.
7-38
-------
Figure 7-20. Program vs. Site-Specific Average Fluoranthene Concentrations
i
1
i i
r
Program Max Concentration = 57.3 ng/rrr i
0 5 10 15 20 25 30 35 40
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 7-20 presents the box plot for fluoranthene for GPCO and shows the following:
• The program-level maximum fluoranthene concentration (57.3 ng/m3) is not shown
directly on the box plot in Figure 7-20 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range; thus,
the scale has been reduced.
• GPCO is one of only two NMP sites sampling PAHs where a fluoranthene
concentration greater than 25 ng/m3 was measured, though concentrations measured
at GPCO account for only four of the 17 fluoranthene measurements of this
magnitude.
• Both annual average concentrations of fluoranthene for GPCO are greater than the
program-level average concentration and third quartile. GPCO's 2015 annual average
concentration ranks fourth highest among sites sampling this pollutant. Because
fluoranthene is not a program-level pollutant of interest, this pollutant is not shown in
annual average comparison table (Table 4-12).
7-39
-------
Figure 7-21. Program vs. Site-Specific Average Formaldehyde Concentrations
BMCO
BRCO
GSCO
PACO
RICO
0 3 6 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
o
Figure 7-21 presents the box plots for formaldehyde and shows the following:
• The box plots show that the range of formaldehyde concentrations measured at GPCO
is considerably larger than the range of concentrations measured at the Garfield
County sites, particularly for 2015. The entire range of formaldehyde concentrations
measured at the Garfield County sites is less than the program-level average
concentration (and in many cases, also less than the program-level median
concentration). Note that the range of measurements for BMCO in 2015 is very small
but represents only one month of sampling.
¦
t |
C"g
E
1
7-40
-------
• GPCO has the highest annual average formaldehyde concentrations among the
Colorado sites, where they could be calculated. The annual averages for GPCO fall
on either side of the program-level average concentration. By comparison, the
available annual average concentrations for the Garfield County sites are all less than
the program-level first quartile.
• Carbonyl compounds were not sampled for at RFCO.
Figure 7-22. Program vs. Site-Specific Average Naphthalene Concentrations
Fr^ ¦
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o
Figure 7-22 presents the box plot for naphthalene for GPCO and shows the following:
• Although the maximum naphthalene concentration measured at GPCO is not the
maximum concentration measured across the program, it is among one of the highest
measured.
• The minimum concentration of naphthalene measured at GPCO in 2015 is greater
than the program-level first quartile, indicating that this measurement is higher than
25 percent of the naphthalene concentrations measured over the two years of
sampling.
• GPCO's 2015 annual average naphthalene concentration is greater than the program-
level average concentration and third quartile, while GPCO's 2016 annual average
naphthalene concentration falls between the two. As previously discussed, GPCO's
annual average concentration for 2015 is the fifth highest annual average naphthalene
concentration among NMP sites sampling PAHs.
7.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
GPCO has sampled carbonyl compounds and VOCs under the NMP since 2004, PAHs since
2008, and metals since 2014; thus, Figures 7-23 through 7-33 present the 1-year statistical
metrics for each of the pollutants of interest for GPCO except arsenic, since metals have not been
sampled at this site for the minimum of 5 consecutive years. BRCO, PACO, and RICO began
7-41
-------
sampling SNMOCs and carbonyl compounds under the NMP in 2008 and RFCO began sampling
SNMOCs under the NMP in 2012. Thus, Figures 7-34 through 7-47 present the 1-year statistical
metrics for each of the pollutants of interest for these sites.
The statistical metrics presented for assessing trends include the substitution of zeros for
non-detects. If sampling began mid-year, a minimum of 6 months of sampling is required for
inclusion in the trends analysis; in these cases, a 1-year average concentration is not provided,
although the range and percentiles are still presented. While sampling at BMCO has occurred
since late 2010, sampling has not been performed consecutively (with the 1-year temporary
relocation to GSCO) and thus, a trends analysis was not conducted for BMCO (or GSCO).
Figure 7-23. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
GPCO
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 7-23 for acenaphthene concentrations measured at GPCO
include the following:
• Sampling for PAHs at GPCO began under the NMP in April 2008. Because a full
year's worth of data is not available for 2008, a 1-year average is not presented,
although the range of measurements is provided.
• The maximum acenaphthene concentration (182 ng/m3) was measured during the
spring of 2012. The next highest concentration was measured during the summer of
7-42
-------
2015 (108 ng/m3). Another acenaphthene concentration greater than 100 ng/m3
measured at GPCO was also measured in 2012; five of the seven acenaphthene
concentrations greater than 50 ng/m3 were measured at GPCO in March and April of
2012.
• Concentrations of acenaphthene decreased significantly from 2009 to 2010, based on
the 1-year averages, after which a steady increasing trend is shown through 2012.
Concentrations measured in 2012 were higher overall compared to other years; nine
of the 17 acenaphthene concentrations greater than 30 ng/m3 were measured in 2012
while only one or two were measured in each of the other years of sampling (except
2014 and 2016 when none were measured). Even if the two highest concentrations
measured in 2012 were removed from the dataset, the 1-year average concentration
for acenaphthene for 2012 would still represent more than a 50 percent increase from
2011.
• All of the statistical metrics shown in Figure 7-23 exhibit a decrease for 2013. Both
the 1-year average and median concentrations decreased by more than half from 2012
to 2013. Each of the statistical parameters exhibit additional decreases for 2014. The
slight increase in the 1-year average concentration for 2015 is a result of the
maximum concentration measured; if this data point is removed from the dataset, the
1-year average concentration would exhibit additional decreases. Even with the
outlier, the 5th percentile and median concentration exhibit continued decreases for
2015.
• With the exception of the 95th percentile, all of the statistical parameters are at
minimum for 2016; 2016 is the first year that an acenaphthene concentration greater
than 20 ng/m3 was not measured at GPCO.
7-43
-------
Figure 7-24. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
GPCO
Maxi mum Concentration
for 2004 is 93.0 ^ig/m3
T
T
•o~
n
•o-
2
rXn
£
LjJ
¦o
T
X
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 7-24 for acetaldehyde concentrations measured at GPCO
include the following:
• The two highest acetaldehyde concentrations (93.0 |ig/m3 and 54.9 |ig/m3) were both
measured at GPCO in 2004. The third highest acetaldehyde concentration
(17.2 |ig/m3) was measured in 2005, after which acetaldehyde concentrations greater
than 7 |ig/m3 were not measured again until 2013. In 2013, six concentrations ranging
from 7.00 |ig/m3 to 10.7 |ig/m3 were measured.
• Between 2005 and 2012, the 1-year average concentrations vary by less than 1 |ig/m3,
ranging from 2.00 |ig/m3 (2010) to 3.00 |ig/m3 (2005). The largest year-to-year
change shown is from 2009 to 2010. The 1-year average concentration increases
significantly between 2010 and 2013, with the 1-year average at its highest since
2004. The median concentration exhibits a similar pattern.
• Concentrations measured in 2014 return to levels near those shown for 2012, with all
of the statistical parameters exhibiting a decrease from 2013 to 2014. Additional
decreases are shown for most of the parameters for 2015, when the 1-year average
concentration is less than 2 |ig/m3 for the first time (1.63 |ig/m3).
• Even though the 1-year average and median concentrations increased slightly for
2016, the smallest range of acetaldehyde concentrations was measured at GPCO in
2016. Acetaldehyde concentrations greater than 4 |ig/m3 were not measured in 2015
7-44
-------
or 2016. The 1-year average concentration for 2016 is also less than 2 |ig/m3
(1.81 |ig/m3).
Figure 7-25. Yearly Statistical Metrics for Benzene Concentrations Measured at GPCO
o
r-J-i
1
2
**
r^i
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-25 for benzene concentrations measured at GPCO include
the following:
• The maximum benzene concentration (10.6 |ig/m3) was measured on June 8, 2011.
Three additional benzene concentrations greater than 5 |ig/m3 have been measured at
GPCO, two in 2004 and one in 2009.
• Concentrations of benzene have a decreasing trend between 2004 and 2007, based on
the 1-year average concentration. After a period of increasing for 2008 and 2009, a
significant decrease is shown for 2010. This decreasing trend continues through 2016,
when the 1-year average concentration (and several of the other statistical metrics) is
at a minimum. This is also true for the median concentration, except that the median
increases slightly for 2014, before continuing to decrease.
• Although the range of benzene concentrations measured is at a minimum for 2014,
the difference between the 5th and 95th percentiles is at a minimum for 2016 and is
less than 1 |ig/m3 for the first time. This indicates that the majority of concentrations
fell within the tightest range in 2016 (with the majority, or 90 percent, of benzene
concentrations falling between 0.33 |ig/m3 and 1.24 |ig/m3).
7-45
-------
Figure 7-26. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
GPCO
A
1
1
T
. o
2008 2009
2010 2011
Year
2
2013
2
20151 2016
O 5th Percentile
O 95th Percentile Averz
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-26 for 1,3-butadiene concentrations measured at GPCO
include the following:
• The only 1,3-butadiene concentration greater than 1 |ig/m3 measured at GPCO was
measured on December 11, 2004. The second highest concentration was also
measured in 2004 (0.75 |ig/m3), although a similar concentration was measured in
2009 (0.71 |ig/m3).
• The 1-year average concentrations have an undulating pattern and vary by less than
0.09 |ig/m3 over the years of sampling, ranging from 0.110 |ig/m3 (2016) to
0.197 |ig/m3 (2006). The increase in the 1-year average concentration from 2011 to
2012 represents the largest year-to-year change (approximately 0.05 |ig/m3). The
median also increased by this much from 2011 to 2012. Not only are the
measurements at the upper end of the concentration range higher for 2012 compared
to 2011 (three 1,3-butadiene concentrations greater than 0.3 |ig/m3 were measured in
2011 compared to nine in 2012), but there were also no non-detects reported for 2012,
while there were seven reported for 2011. 2011 is the last year non-detects of
1,3-butadiene were reported at GPCO.
• Most of the statistical parameters exhibit decreases from 2014 to 2015 and again for
2016. The smallest range of 1,3-butadiene concentrations was measured at GPCO in
2016, when both the 1-year average and median concentrations are at a minimum.
7-46
-------
Figure 7-27. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations Measured
at GPCO
1.4
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 20151 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-27 for carbon tetrachloride concentrations measured at
GPCO include the following:
• Seven concentrations of carbon tetrachloride greater than 1 |ig/m3 have been
measured at GPCO (four in 2008, and one each in 2006, 2009, and 2014).
Conversely, 16 non-detects have been measured (nine in 2004, five in 2005, and one
each in 2006 and 2013).
• The year with the least variability is 2012, with a difference of 0.38 |ig/m3 between
the minimum and maximum concentrations and a difference of 0.24 |ig/m3 between
the 5th and 95th percentiles. However, the year with the highest 1-year average and
median concentrations (0.67 |ig/m3 and 0.68 |ig/m3, respectively) is also 2012. Note
that the 5th percentile for 2012 is greater than or similar to the 1-year average and/or
median concentrations for several of the previous years of sampling.
• For most of the years of sampling, the median concentration is slightly higher than
the 1-year average concentration. This indicates that the concentrations at the lower
end of the concentration range are pulling down the 1-year average in the same
manner than an outlier can drive an average upward. However, the difference
between the 1-year average and median concentrations for most years is less than
0.05 |ig/m3.
7-47
-------
• There is a significant increase in the 1-year average concentrations from 2007 to 2008
as the range of concentrations measured doubled from one year to the next. After
2008, a decreasing trend is shown through 2010, with little change in the 1-year
average from 2010 to 2011. These statistical parameters increased significantly from
2011 to 2012, and are at a maximum for the period of sampling. All of the statistical
metrics exhibit a decrease from 2012 to 2013, primarily as a result of the higher
number of concentrations at the lower end of the concentration range. The number of
carbon tetrachloride concentrations less than 0.5 |ig/m3 increased from one in 2012 to
12 in 2013.
• Between 2013 and 2016, the central tendency statistics change little, with both the
1-year average and median concentrations varying by about 0.02 |ig/m3 across these
four years.
Figure 7-28. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured at
GPCO
^ ^
2004 2005 2006 2007 2008 2009
o
o
I
2010 2011 2012 2013 2014 2015
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-28 for 1,2-dichloroethane concentrations measured at GPCO
include the following:
• Between 2004 and 2008, there were only three measured detections of
1,2-dichloroethane measured at GPCO. The median concentration is zero for all years
through 2011, indicating that at least 50 percent of the measurements were non-
detects prior to 2012. The number of measured detections began to increase in 2009,
from 12 percent for 2009 and 2010, to 27 percent in 2011, and 90 percent for 2012.
7-48
-------
Since 2012, measured detections have accounted for the majority of measurements,
ranging from 74 percent in 2013 to 94 percent in 2015.
• As the number of measured detections increases, so do each of the corresponding
statistical metrics shown in Figure 7-28. The percentage of measured detections
increased by 64 percent from 2011 to 2012, thus, the 1-year average and median
concentrations exhibit considerable increases.
• The median concentration is greater than the 1-year average concentration for each
year after 2011. This is because there are still non-detects (or zeros) factoring into the
1-year average concentration for each year, which pull the average down in the same
manner than an outlier drives an average upward. Excluding non-detects, the
minimum concentration for each year between 2012 and 2016 would range from
0.03 |ig/m3 to 0.05 |ig/m3.
Figure 7-29. Yearly Statistical Metrics for Dichloromethane Concentrations Measured at
GPCO
Me
Maximum Concentration
for 2010 is 5,256 ^g/m3
2004 2005 2006 2007 2003 2009 2010 2011 2012 2013 2014 2015 2016
I
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 20151 2016
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile Averc
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-29 for dichloromethane concentrations measured at GPCO
include the following:
• Due to the range of concentrations measured, Figure 7-29 contains an inset showing
how the statistical metrics look at the lower end of the scale.
• The maximum dichloromethane concentration was measured at GPCO in 2010
(5,256 |ig/m3); an additional concentration greater than 1,000 |ig/m3 was also
7-49
-------
measured in 2015 (1,493 |ig/m3). In total, 21 concentrations of dichloromethane
greater than 100 |ig/m3 have been measured at this site, all of which were measured in
2010 or later.
• The central tendency statistics vary little during the early years of sampling, as shown
in the inset, with less than 0.1 |ig/m3 separating the 1-year average concentrations.
Between 2008 and 2016, the 1-year average concentrations range from as little as
1.32 |ig/m3 (2011) to as high as 91.79 |ig/m3 (2010).
• The median concentration holds steady between 2004 and 2008, ranging from
0.31 |ig/m3 (2004) to 0.40 |ig/m3 (2008). Between 2009 and 2012, the median
concentrations vary slightly more, ranging from 0.49 |ig/m3 (2011) to 0.65 |ig/m3
(2012). The median increases considerably, in a similar manner as the 1-year average
concentration, for the years between 2013 and 2015. For 2016, the median decreases
back to level similar 2008-2012 levels.
Figure 7-30. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
GPCO
J
o.
2005 2006
I
2008 2009
1
LjJ
2010
Year
r-J-i
2
2011 2012
O 5th Percentile
— Minimum — Media
Maximum O 95th Percentile Averc
1 A 1-year average is not presented for 2015 due to low completeness.
Observations from Figure 7-30 for ethylbenzene concentrations measured at GPCO
include the following:
• The maximum ethylbenzene concentration was measured at GPCO in 2005
(5.31 |ig/m3), as was the second highest concentration (3.96 |ig/m3). Three additional
concentrations greater than 3 |ig/m3 have been measured at GPCO, two in 2004 and
7-50
-------
one in 2012. All but two of the 18 ethylbenzene measurements greater than 2 |ig/m3
were measured during these three years.
• The 1-year average concentration increased from 2004 to 2005, although there is a
relatively high level of variability in the measurements. A significant decrease in all
the statistical parameters is shown from 2005 to 2006, with a slight decreasing trend
continuing through 2008.
• Although the maximum concentration measured increased by more than 1 |ig/m3
from 2008 to 2009, only a slight change in the 1-year and median concentrations is
exhibited for 2009. The range of concentrations measured in 2010 is similar to the
range of concentrations measured in 2008. An increasing trend in the 1-year average
concentration is shown from 2010 through 2012. The median concentration begins
increasing in 2009 and continues through 2012.
• All of the statistical parameters exhibit a decrease from 2012 to 2013, with the
maximum concentration decreasing by more than half. Between 2013 and 2016, the
95th percentile decreased at least slightly each year, indicating that the range within
which the majority of concentrations fall is getting tighter. With the exception of the
minimum concentration, all of the statistical parameters are at a minimum for 2016.
Figure 7-31. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at
GPCO
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
7-51
-------
Observations from Figure 7-31 for fluoranthene concentrations measured at GPCO
include the following:
• There is little change is the fluoranthene concentrations measured between 2009 (the
first full year of sampling) and 2013. The 1-year average concentration varies by less
than 1 ng/m3 during this time, ranging from 3.30 ng/m3 in 2010 to 3.98 ng/m3 in
2009.
• A decrease in most of the statistical parameters is shown for 2014, when the smallest
range of fluoranthene concentrations was measured at GPCO.
• The range of fluoranthene concentrations increased dramatically in 2015, which is the
first year that a fluoranthene concentration greater than 15 ng/m3 was measured at
GPCO. The 1-year average concentration for 2015 is greater than the 95th percentile
for 2014. However, the median concentration for 2015 is the lowest median since the
first full year of sampling. This indicates that the higher fluoranthene concentrations
are driving the 1-year average concentration for 2015, while concentrations on the
lower end of the concentration range account for about half of the measurements for
GPCO in 2015. The number of fluoranthene concentrations less than 2.5 ng/m3 is the
same for 2014 as it is for 2015 (29).
• The statistical parameters representing the concentrations on the upper end of the
range exhibit decreases for 2016 and thus, the 1-year average concentration decreased
for 2016. But the median concentration for 2016 is similar to the median
concentration for 2015, and the number of concentrations less than 2.5 ng/m3 is at a
maximum (33).
7-52
-------
Figure 7-32. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
GPCO
Maximum Concentration |
for 2004 is 40.5 ng/m3
-I -
-.i^i
T
n5-|
£
5
O-
5
.0.
o
-t-'
rh
iOn..
T
LjJ
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 7-32 for formaldehyde concentrations measured at GPCO
include the following:
• The maximum formaldehyde concentration (40.5 |ig/m3) was measured in 2004 and
is significantly higher than the maximum concentrations measured in subsequent
years. The second highest concentration was also measured in 2004 (23.5 |ig/m3);
these two concentrations of formaldehyde were measured on the same back-to-back
days in 2004 as the two highest acetaldehyde concentrations. The next seven highest
formaldehyde concentrations were measured at GPCO in 2013 and range from
15.8 |ig/m3 to 21.9 |ig/m3. The only other formaldehyde concentration greater than
15 |ig/m3 was measured at GPCO in 2015.
• Even with decreasing maximum concentrations, the 1-year average concentrations
have an increasing trend through 2006. The 1-year average concentration is
approximately 4 |ig/m3 for each year between 2006 and 2009. A significant decrease
in all of the statistical parameters is shown for 2010. Although an even smaller range
of concentrations was measured in 2011, there is little change in the 1-year average
concentration between 2010 and 2011. With a few higher concentrations measured in
2012, the 1-year average calculated for 2012 is slightly higher than the 1-year average
concentrations for the previous two years, although the increase is not statistically
significant.
• All of the statistical parameters exhibit increases for 2013, particularly those
representing concentrations at the upper end of the concentration range. The 1-year
7-53
-------
average concentration for 2013 is greater than the maximum concentrations measured
in several of the previous years and is greater than the 95th percentile for each of the
previous years. Even the median concentration, which is less affected by outlier
concentrations, increased by more than 70 percent from 2012 to 2013. Ten
formaldehyde concentrations measured in 2013 are greater than the maximum
concentration measured in 2012; further, the number of formaldehyde concentrations
greater than 5 |ig/m3 increased from two in 2012 to 26 in 2013.
• All of the statistical metrics for 2014 exhibit a decrease from 2013 levels, although
the 1-year average and median concentrations are still higher than they were in the
three years prior to 2013. The 1-year average formaldehyde concentration continues
to decrease each year through 2016.
• For 2016, the difference between the 5th and 95th percentiles is at a minimum,
indicating that the majority of concentrations fell into the tightest range in 2016.
Figure 7-33. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
GPCO
O:
o
2009 2010
2012
Year
2
2013
O 5th Percentile — Minimum — Median
Maximum o 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 7-33 for naphthalene concentrations measured at GPCO
include the following:
• The maximum naphthalene concentration measured at GPCO was measured in 2012
(822 ng/m3). Concentrations of 400 ng/m3 or higher have been measured in four of
the years of sampling and concentrations greater than 250 ng/m3 have been measured
in all years of sampling except 2014 and 2016.
7-54
-------
• The trends graph for naphthalene resembles the trends graphs for acenaphthene
shown in Figure 7-23. The 1-year average concentration for naphthalene decreased
significantly from 2009 to 2010. A slight increase from 2010 to 2011 is followed by
an additional increase for 2012. All of the statistical parameters increased from 2011
to 2012 and are at a maximum across the years of sampling.
• A significant decreasing trend is shown after 2012. Both the 1-year average and
median concentrations are at a minimum for 2016, with the 1-year average
concentration decreasing by nearly half between 2012 and 2016.
Figure 7-34. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
BRCO
2.5
pl-
E
1
c
O
c
m
i
-
—
2~~
—2—
T
T
I
Tr
2~
o
2008 2009 2010 1 20111 2012 2013 2014 2 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 There was a gap in sampling between October 2010 and September 2011.
2 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-34 for acetaldehyde concentrations measured at BRCO
include the following:
• BRCO began sampling carbonyl compounds under the NMP in February 2008. A
1-year average concentration is not presented for 2010 and statistical metrics are not
provided for 2011 because sampling was discontinued in October 2010 and did not
begin again until September 2011. In addition, the completeness criteria was not met
for 2014, and thus, a 1-year average concentration is not provided for 2014. Note that
carbonyl compounds are sampled on a l-in-12 sampling schedule at BRCO.
• The maximum acetaldehyde concentration (1.97 |ig/m3) was measured on the second
day of sampling, February 12, 2008. In total, only 28 acetaldehyde concentrations
7-55
-------
greater than 1 |ig/m3 have been measured at BRCO since the onset of sampling, with
none measured in 2016.
• Concentrations of acetaldehyde measured at BRCO have a decreasing trend across
the years of sampling, and all of the statistical parameters are at a minimum for 2016.
Figure 7-35. Yearly Statistical Metrics for Benzene Concentrations Measured at BRCO
,.o
Maximum Concentrati on
for 2008 is 13.7^g/m3
2012
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-35 for benzene concentrations measured at BRCO include
the following:
• BRCO began sampling benzene under the NMP in January 2008. Similar to
acetaldehyde, a 1-year average concentration is not provided for benzene for 2014 as
the completeness criteria was not met.
• The maximum benzene concentration (13.7 |ig/m3) was measured on July 29, 2008
and is three times greater than the next highest concentration (4.55 |ig/m3, measured
on January 7, 2009). Two additional benzene concentrations greater than 4 |ig/m3
have been measured at BRCO, another in 2009 and one in 2010.
• Most of the statistical parameters for benzene exhibit a steady decreasing trend over
the years of sampling at BRCO between 2009 and 2012. Prior to 2013, the 1-year
average concentration decreased by roughly half, from a maximum of 1.39 |ig/m3 in
2009 to a minimum of 0.68 |ig/m3 in 2012. The median concentration also has a
decreasing trend, and has decreased each year between 2008 and 2012, from
1.05 |ig/m3 in 2008 to 0.65 |ig/m3 in 2012.
7-56
-------
• All of the statistical metrics exhibit an increase from 2012 to 2013, returning to
concentration levels similar to 2010. The fewest benzene concentrations less than
0.5 |ig/m3 were measured in 2013 (two), compared to 18, or about one-third of the
measurements, in 2012. The increases for 2013 are followed by a return to 2012
levels for 2014, based on the available statistical metrics. Additional slight decreases
are shown for several of the statistical parameters for 2015.
• Even though the 1-year average concentration for 2016 exhibits a slight increase, the
median concentration continues to decrease, and is at a minimum for 2016 over the
period of sampling; 2016 is the first year for which the median concentration is less
than 0.5 |ig/m3.
Figure 7-36. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
BRCO
Maximum Concentration
for 2009 is 10.2 pg/m3
...o
o.
o
<>
2012
Year
O 5th Percentile
O 95th Percentile
1 There was a gap in sampling between October 2010 and September 2011.
2 A 1-year average is not presented due to low method completeness in 2014.
Observations from Figure 7-36 for formaldehyde concentrations measured at BRCO
include the following:
• The maximum formaldehyde concentration (10.2 |ig/m3) was measured at BRCO on
January 7, 2009, the same day as the second highest benzene concentration. This
formaldehyde measurement is more than three times higher than the next highest
concentration measured at this site (3.11 |ig/m3, measured on August 31, 2012). Four
additional formaldehyde concentrations greater than 2.0 |ig/m3 have been measured at
BRCO.
7-57
-------
• The increase in the 1-year average concentration shown from 2008 to 2009 results
primarily from the maximum concentration measured in 2009. The median
concentrations are similar to each other for these two years (1.02 |ig/m3 and
1.03 |ig/m3) and, if the maximum concentration for 2009 was removed from the
dataset, the 1-year average concentrations would also be similar to each other.
• Several of the statistical parameters exhibit increases for 2010, although these
parameters do not include measurements for an entire year.
• Between 2012 and 2016 there is an overall decreasing trend in the formaldehyde
concentrations measured at BRCO. Nearly all of the statistical parameters are at a
minimum for 2016, when only one formaldehyde concentration greater than 1 |ig/m3
was measured.
Figure 7-37. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
PACO
2012
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile Averc
1 A 1-year average is not presented due to low method completeness in 2011, 2014, and 2015.
Observations from Figure 7-37 for acetaldehyde concentrations measured at PACO
include the following:
• PACO began sampling acetaldehyde under the NMP in February 2008. A 1-year
average concentration is not presented for 2011 due to low method completeness.
This is also true for 2014 and 2015. Note that carbonyl compounds are sampled on a
l-in-12 sampling schedule at PACO.
7-58
-------
• The maximum acetaldehyde concentration (2.04 |ig/m3) was measured at PACO on
January 13, 2009 and is the only acetaldehyde concentration greater than 2 |ig/m3
measured at this site.
• Acetaldehyde concentrations measured at PACO have an overall decreasing trend
through 2016 (although two of the years do not follow this pattern, and are discussed
in the bullets that follow). The 1-year average concentration is at a minimum for
2016, though several of the years do not have 1-year averages presented. The median
concentration, which is available for each year of sampling, is at a minimum for
2014, although the median concentrations for 2015 and 2016 are similar.
• Concentrations greater than 1 |ig/m3 make up a higher percentage of the
measurements in 2011, compared to 2010 and 2012, resulting in a higher median
concentration. In addition, the minimum concentration measured in 2011 was higher
than most other years of sampling; 2011 also has fewer concentrations less than
0.5 |ig/m3 than most other years.
• For 2013, both the 1-year average and median concentrations exhibit an increase. The
largest range of acetaldehyde concentrations was measured in 2013, and the range
within which the majority of the measurements fall, indicated by the 5th and 95th
percentiles, is at a maximum for 2013 over the years of sampling.
Figure 7-38. Yearly Statistical Metrics for Benzene Concentrations Measured at PACO
X
2008 2009 2010 2011
2012
Year
2013 2014 2015 2016
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented due to low method completeness in 2012.
7-59
-------
Observations from Figure 7-38 for benzene concentrations measured at PACO include
the following:
• PACO began sampling SNMOCs under the NMP in January 2008. A 1-year average
concentration is not presented for 2012 due to low method completeness resulting
from issues with the collection system.
• The maximum benzene concentration (11.1 |ig/m3) was measured at PACO on
October 15, 2008. The next highest measurement (10.1 |ig/m3) was measured
3 months later on January 7, 2009. The third highest concentration was measured on
the next sample day in 2009 (7.52 |ig/m3). In total, 12 benzene concentrations greater
than 5.0 |ig/m3 have been measured at PACO, with three measured in 2008, eight
measured in 2009, and one in 2013.
• Even though the maximum concentration decreased some from 2008 to 2009,
benzene concentrations increased overall from 2008 to 2009, as indicated by the
increases in the 1-year average, median, and 95th percentile.
• Concentrations of benzene decreased significantly between 2009 and 2010, when the
maximum and 95th percentile decreased by nearly half. This decreasing trend
continued into 2011 and 2012. Benzene concentrations greater than 3 |ig/m3 were not
measured in 2012.
• All of the statistical parameters shown exhibit considerable increases from 2012 to
2013. The range within which the majority of the measurements fall, indicated by the
5th and 95th percentiles, more than doubled and is at its largest since 2009. Nine
benzene concentrations greater than the maximum concentration for 2012
(2.97 |ig/m3) were measured in 2013.
• The increases shown for 2013 were followed by significant decreases over the last
three years of sampling. The 1-year average concentration is at a minimum for 2016
while the median concentration is at a minimum for 2015.
7-60
-------
Figure 7-39. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
PACO
.1 0.5
Maximum Concentration for
2009 is 3.15 jjg/m3
pl-
1-
T
~
—
i. £ I
O 5th Percentile
Minimum — Median — Maximum O 95th Percentile •••~•••Avera
1 A 1-year average is not presented due to low method completeness in 2012.
Observations from Figure 7-39 for 1,3-butadiene concentrations measured at PACO
include the following:
• The maximum 1,3-butadiene concentration (3.15 |ig/m3) was measured on
December 27, 2009 and is the only 1,3-butadiene measurement greater than 1 |ig/m3
measured at this site.
• The increase in the 1-year average concentration from 2008 to 2009 is a result of this
outlier concentration measured in 2009. The second highest concentration measured
in 2009 is substantially less (0.188 |ig/m3). Excluding the maximum concentration for
2009 would result is a 1-year average concentration of only 0.028 |ig/m3 (rather than
0.088 |ig/m3), and thus, a decrease in the 1-year average concentration by almost half
from 2008 to 2009. Note that the median 1,3-butadiene concentration for 2009 is
zero, indicating that at least half of the measurements for 2009 are non-detects.
• The second, third, fourth, and fifth highest 1,3-butadiene concentrations measured at
PACO were all measured in December 2010 and range from 0.39 |ig/m3 to
0.66 |ig/m3. The next two highest concentrations for this year was also measured in
December but were considerably less (0.16 |ig/m3 and 0.12 |ig/m3). The 95th
percentile for 2010 is greater than the maximum concentration measured for all other
years except 2009, when the outlier was measured. Even though half of the
measurements in 2010 were non-detects, the December measurements for 2010 are
driving the top-end statistical parameters upward.
7-61
-------
• Nearly all of the statistical parameters decreased from 2010 to 2011, except the
minimum and 5th percentile, which are both zero for these years.
• Prior to 2012, the percentage of non-detects of 1,3-butadiene measured at PACO
ranged from 47 percent (2008) to 58 percent (2009 and 2011). This explains why the
median concentration is at or near zero for these years. For 2012, the number of non-
detects is at a minimum (29 percent) and explains why the median increased
considerably, although the range of measurements did not change much from 2011
and 2012.
• For 2013, the median concentration returned to zero as the number of non-detects
increased from 29 percent in 2012 to 83 percent for 2013. Between 2013 and 2016,
the percentage of non-detects is greater than 75 percent for each year, with the
percentage of non-detects reaching a maximum of 96 percent in 2016. With only two
measured detections, and 52 zeros factoring in for non-detects, it's not surprising that
the 1-year average concentration for 2016 is considerably lower than the other 1-year
average concentrations shown in Figure 7-39.
Figure 7-40. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
PACO
2012
Year
O 5th Percentile — Minimum — Median
Maximum O 95th Percentile Averc
1 A 1-year average is not presented due to low method completeness in 2011, 2014, and 2015.
Observations from Figure 7-40 for formaldehyde concentrations measured at PACO
include the following:
• Four formaldehyde concentrations greater than 3 |ig/m3 have been measured at PACO
(one in 2008, two in 2009, and one in 2010).
7-62
-------
• The 1-year average concentration changed little between 2008 and 2009. The
decreases in the minimum and maximum concentrations for 2009 are countered by an
increase in the number of measurements at the higher end of the concentration range,
as indicated by the increases in the median and 95th percentile.
• The data distribution statistics for 2010 resemble those for 2008, although the 1-year
average and median concentrations both exhibit decreases. The number of
formaldehyde concentrations greater than 2 |ig/m3 decreased by more than half from
2009 to 2010, while the number of concentrations less than 1 |ig/m3 more than
doubled.
• Although the maximum concentration decreased for 2011, all of the other statistical
parameters that could be calculated exhibit increases from 2010 to 2011.
• All of the statistical parameters exhibit decreases from 2011 to 2012, particularly at
the lower end of the concentration range, as the 5th percentile decreased from just
less than 1 |ig/m3 to just greater than 0.1 |ig/m3. The concentration profiles for the
four years following 2012 more resemble 2012 than the four years prior. Less than
0.25 |ig/m3 separates the median concentrations for 2012 through 2016, and
approximately 0.10 |ig/m3 separates the available 1-year average concentrations for
these years.
Figure 7-41. Yearly Statistical Metrics for Benzene Concentrations Measured at RFCO
7.0
T
^ 40
3
c
O
c
m
o
u
; a - ± _
—I—
2012 1 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling did not begin until June 2012.
7-63
-------
Observations from Figure 7-41 for benzene concentrations measured at RFCO include
the following:
• Sampling for SNMOCs at RFCO began under the NMP in June 2012. Because a full
year's worth of data is not available for 2012, a 1-year average is not presented,
although the range of measurements is provided.
• With the exception of 2015, the range of benzene concentrations has not varied much
across the years of sampling, with roughly 1 |ig/m3 separating the minimum and
maximum concentrations measured.
• The two highest benzene concentrations measured at RFCO were measured on back-
to-back sample days in March 2015 (6.62 |ig/m3 and 5.87 |ig/m3). The next highest
concentration measured in 2015 is considerably less (0.74 |ig/m3). If the two highest
benzene concentrations were removed from the dataset, 2015 would have the smallest
range of measurements and the 1-year average concentration would decrease by half,
making it the lowest 1-year average. The median, however, would change very little.
In fact, the median concentrations shown in Figure 7-41 vary by only 0.11 |ig/m3.
Figure 7-42. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
RFCO
0.35
c
O
c
m
o
u
—L
I
o
0.00
2012 1
2013
2014
Year
2015
20
16
O
5th Percentile
- Mi
limum
— Median
-
Maximum
O 95th Percentile
Aver a
e
1 A 1-year average is not presented because sampling did not begin until June 2012.
7-64
-------
Observations from Figure 7-42 for 1,3-butadiene concentrations measured at RFCO
include the following:
• The six highest 1,3-butadiene concentrations, ranging from 0.23 |ig/m3 to 0.31 |ig/m3,
were measured at RFCO in 2012. Concentrations of 1,3-butadiene greater than
0.15 |ig/m3 were not measured during any other year of sampling at RFCO and
concentrations greater than 0.10 |ig/m3 were not measured at RFCO in 2015 or 2016.
• The median concentration is zero for the years after 2012, indicating that at least half
of the 1,3-butadiene measurements were non-detects. There were five non-detects in
2012, accounting for nearly 30 percent of the measurements (note that sampling
began in June in 2012, such that the concentration profile represents seven months of
sampling). For subsequent years, the detection rate decreased considerably, with non-
detects accounting for between 75 percent (2014) and 93 percent (2016) of
measurements.
Figure 7-43. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
RICO
2012
Year
2r
2013
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
'A
2 A
1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
1-year average is not presented for 2015 due to low method completeness combined with coelution.
Observations from Figure 7-43 for acetaldehyde concentrations measured at RICO
include the following:
• RICO began sampling carbonyl compounds under the NMP in February 2008. A
1-year average concentration is not presented for 2010, 2011, or 2013 due to low
method completeness; a 1-year average is not presented for 2015 due to low
7-65
-------
completeness combined with coelution, as indicated in Section 7.3.1. However, the
range of measurements is provided for each of these years.
• The maximum acetaldehyde concentration (2.91 |ig/m3) was measured at RICO in
July 2008, although a similar concentration was also measured on the sample day
prior.
• Few 1-year average concentrations could be calculated for RICO. However, the
measurements have a decreasing trend through 2014, based on the decreases shown
for nearly all of the other statistical parameters, many of which are at a minimum for
2014.
• All of the available statistical parameters exhibit increases for 2015, and several
exhibit additional increases for 2016 (the median concentration is the exception).
Figure 7-44. Yearly Statistical Metrics for Benzene Concentrations Measured at RICO
o..
2012
Year
O 5th Percentile
O 95th Percentile ....... Averz
1 A 1-year average is not presented for 2015 due to low method completeness.
Observations from Figure 7-44 for benzene concentrations measured at RICO include the
following:
• RICO began sampling SNMOCs under the NMP in January 2008.
• The maximum benzene concentration (6.67 |ig/m3) was measured in January 2009.
The six highest benzene concentrations measured at RICO were all measured in 2009,
with the four highest measured in January.
7-66
-------
All of the statistical parameters exhibit increases from 2008 to 2009, particularly the
maximum concentration and the 95th percentile, after which a steady decreasing trend
is shown through 2012. The number of measurements greater than 2 |ig/m3 increased
from 19 to 25 from 2008 to 2009, then decreased by half for 2010 and continued to
decrease, reaching a minimum of two for 2012. This helps explain the increase in the
statistical parameters shown from 2008 to 2009 as well as the subsequent decreases in
the years that follow. The median concentration is 0.96 |ig/m3 for 2012, indicating
that at least half of the measurements are less than 1 |ig/m3. The 1-year average
concentration is also less than 1 |ig/m3 for 2012.
All of the statistical parameters exhibit increases for 2013 as benzene concentrations
were higher overall in 2013. The number of concentrations greater than 2 |ig/m3
increased six-fold from 2012 (2) to 2013 (13). Five concentrations measured in 2013
are greater than the maximum concentration measured in 2012, while 11
concentrations measured in 2012 are less than the minimum concentration measured
in 2013.
The increases shown for 2013 were followed by significant decreases for 2014,
although not quite returning to 2012 levels. The statistical parameters shown for
RICO's benzene concentrations resemble the ones shown for benzene concentrations
measured at PACO (and to a lesser extent BRCO), as all three sites exhibit a
decreasing trend through 2012 followed by a considerable increase for 2013 and
additional decreases for 2014.
The smallest range of benzene concentrations was measured in 2015. The range of
measurements increases for 2016, such that the concentration profile for 2016 is
similar to that shown for 2014.
7-67
-------
Figure 7-45. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
RICO
1.2
E
3
c
° 06
C
m
c
O
u
JL
0.2
0.0
.<
pL
rh
"O"
2008 2009 2010 2011 2012 2013 2014 2015 1 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented for 2015 due to low method completeness.
Observations from Figure 7-45 for 1,3-butadiene concentrations measured at RICO
include the following:
• The five highest 1,3-butadiene concentrations were measured at RICO in December
2010 and ranged from 0.57 |ig/m3 to 0.98 |ig/m3 (although a measurement of 0.57
|ig/m3 was also measured in 2012). Higher 1,3-butadiene concentrations were also
measured at PACO during December 2010.
• The minimum concentration and 5th percentile are both zero for each year of
sampling; this indicates that at least 5 percent of the measurements were non-detects
each year. The percentage of non-detects has varied from 7 percent (2012) to
39 percent (2014).
• With the exception of the maximum concentration, the range of concentrations
measured in 2008 and 2009 were similar to each other, as indicated by most of the
statistical parameters shown. This was followed by an increase in the magnitude of
the concentrations measured in 2010. Even though the maximum concentration and
95th percentile more than doubled and the 1-year average concentration increased by
more than 50 percent, the median concentration changed very little for 2010. This
indicates that there are roughly the same number of measurements at the lower end of
the concentration range while the measurements at the higher end of the concentration
range are driving the 1-year average concentration upward.
7-68
-------
Although the range of concentrations measured varies between 2010 and 2012, the
1-year average concentration decreases slightly while the median concentration
increases slightly.
• Most of the statistical parameters exhibit decreases from 2012 to 2013 (the minimum
and 5th percentile both stay the same), with the median concentration decreasing by
almost half. This indicates that the 1,3-butadiene concentrations measured were lower
in 2013.
• Although little change is shown in the 1-year average concentration for 2014, five
concentrations measured in 2014 are greater than the maximum concentration
measured in 2013. On the lower end of the scale, the number of non-detects increased
four-fold, from five measured in 2013 to 21 measured in 2014.
• The smallest range of 1,3-butadiene concentrations was measured in 2015; all
1,3-butadiene concentrations measured at RICO in 2015 are less than 0.2 |ig/m3.
• Despite the slightly larger range of concentrations measured, the central tendency
statistics are both at minimum for 2016.
Figure 7-46. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
RICO
| MaximumConcentraion
1 for 2010 is25.7 |^/m3
2008 2009 2010 2011 2012 2013 2014 2015 1 2016
! © 5th Percentile — Minimum ~ Median - Maximum © 95th Percentile Average
1 A 1-year average is not presented for 2015 due to low method completeness.
7-69
-------
Observations from Figure 7-46 for ethylbenzene concentrations measured at RICO
include the following:
• The maximum ethylbenzene concentration measured at RICO was measured on
August 18, 2010 (25.7 |ig/m3). The next highest concentration was also measured in
2010 but is considerably less (6.86 |ig/m3). No other ethylbenzene concentrations
greater than 2 |ig/m3 have been measured at RICO and only nine concentrations
greater than 1 |ig/m3 have been measured at this site. This explains why the 1-year
average concentration is greater than the 95th percentile for 2010, it is skewed by the
outlier. Excluding the maximum concentration from the 1-year average calculation
for 2010 would result in a 1-year average concentration similar to that shown for
2009.
• Excluding the outlier, there is a decreasing trend in most of the statistical parameters
shown between 2009 and 2012, with most of the statistical parameters at a minimum
for 2012.
• Each of the statistical parameters shown in Figure 7-46 increased from 2012 to 2013,
with several of them returning to levels similar to those calculated for 2011.
• The range of ethylbenzene concentrations measured decreases slightly each year
between 2013 and 2016, with only slight changes in the 1-year average and median
concentrations.
• Three non-detects of ethylbenzene have been measured at RICO, two in 2014 and
another in 2015.
7-70
-------
Figure 7-47. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
RICO
I
2012 2013 2014
Year
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented due to low method completeness in 2010, 2011, and 2013.
Observations from Figure 7-47 for formaldehyde concentrations measured at RICO
include the following:
• The maximum formaldehyde concentration (4.82 |ig/m3) was measured at RICO in
November 2008. The only other formaldehyde concentration greater than 4 |ig/m3
was measured in August 2013 (4.38 |ig/m3). Three additional formaldehyde
concentrations measured at RICO are greater than 3 |ig/m3 (one each in 2008, 2010,
and 2011).
• Formaldehyde concentrations measured at RICO have an overall decreasing trend
between 2010 and 2014, despite a few higher concentrations measured, based on the
decreases shown for several of the other statistical parameters. The median
concentration decreases by 64 percent during this time, from 1.88 |ig/m3 in 2010 to
0.68 |ig/m3 for 2014. All of the statistical parameters except the minimum
concentration are at a minimum for 2014.
• A 1-year average concentration is available for each year from 2014 through 2016. A
significant increase in this parameter is shown from 2014 to 2015, with a slight
increase for 2016. The median concentration has a similar pattern. The 1-year average
and median concentrations for 2015 and 2016 are just slightly less than the 95th
percentile for 2014. This results from changes at both the upper and lower ends of the
concentration range. The number of formaldehyde concentrations greater than
1 |ig/m3 nearly tripled between 2014 (8) and 2016 (22) while the number of
7-71
-------
formaldehyde concentrations less than 0.5 |ig/m3 decreased from seven (2014) to
none (2016).
7.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Colorado monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
7.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Colorado monitoring sites, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 7-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for GPCO from Table 7-4 include the following:
• Dichloromethane (2016 only), formaldehyde, and acetaldehyde have the highest
annual average concentrations among GPCO's pollutants of interest.
• Formaldehyde has the highest cancer risk approximations for this site (41.07 in-a-
million for 2015 and 36.35 in-a-million for 2015). The remaining cancer risk
approximations, where they could be calculated, are all less than 10 in-a-million, with
several less than 1 in-a-million.
• None of the pollutants of interest for GPCO have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. Acetaldehyde and formaldehyde have the highest
noncancer hazard approximations (the only ones greater than an HQ of 0.10) among
the pollutants of interest for GPCO.
7-72
-------
Table 7-4. Risk Approximations for the Colorado Monitoring Sites
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HO)
Grand Junction, Colorado - GPCO
Acetaldehyde
0.0000022
0.009
55/55
1.63
±0.17
3.58
0.18
59/59
1.81
±0.15
3.97
0.20
Benzene
0.0000078
0.03
50/50
NA
NA
NA
62/62
0.71
±0.09
5.56
0.02
1.3 -Butadiene
0.00003
0.002
50/50
NA
NA
NA
62/62
0.11
±0.02
3.31
0.06
Carbon Tetrachloride
0.000006
0.1
50/50
NA
NA
NA
62/62
0.60
±0.03
3.60
0.01
1,2-Dichloroethane
0.000026
2.4
47/50
NA
NA
NA
52/62
0.07
±0.01
1.78
<0.01
Dichloromethane
0.000000016
0.6
50/50
NA
NA
NA
62/62
4.19
±4.52
0.07
0.01
Ethylbenzene
0.0000025
1
50/50
NA
NA
NA
62/62
0.30
±0.04
0.76
<0.01
Formaldehyde
0.000013
0.0098
55/55
3.16
±0.51
41.07
0.32
59/59
2.80
±0.18
36.35
0.29
Acenaphthene3
0.000088
55/55
8.29
±3.97
0.73
58/61
5.67
± 1.20
0.50
Arsenic (PMi0)a
0.0043
0.000015
57/57
0.28
±0.03
1.19
0.02
60/60
0.32
±0.06
1.36
0.02
Fluoranthene3
0.000088
55/55
5.59
±2.16
0.49
61/61
4.13
± 1.09
0.36
Naphthalene3
0.000034
0.003
55/55
91.01
± 12.89
3.09
0.03
61/61
74.38
±9.54
2.53
0.02
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Table 7-4. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Battlement Mesa, Colorado - BMCO
Acetaldehyde
0.0000022
0.009
3/3
NA
NA
NA
24/24
0.48
±0.10
1.05
0.05
Benzene
0.0000078
0.03
6/6
NA
NA
NA
48/48
0.79
±0.06
6.17
0.03
Formaldehyde
0.000013
0.0098
3/3
NA
NA
NA
24/24
0.93
±0.18
12.10
0.09
Silt, Colorado - BRCO
Acetaldehyde
0.0000022
0.009
28/28
0.51
±0.11
1.12
0.06
26/26
0.22
±0.08
0.49
0.02
Benzene
0.0000078
0.03
52/52
0.66
±0.12
5.16
0.02
57/58
0.71
±0.11
5.51
0.02
Formaldehyde
0.000013
0.0098
28/28
0.82
±0.17
10.71
0.08
26/26
0.37
±0.14
4.81
0.04
Glenwood Springs, Colorado - GSCO
Acetaldehyde
0.0000022
0.009
26/26
0.53
±0.10
1.16
0.06
5/5
NS
NS
NS
Benzene
0.0000078
0.03
52/52
0.49
±0.04
3.85
0.02
12/12
NS
NS
NS
1.3 -Butadiene
0.00003
0.002
19/52
0.02
±0.01
0.63
0.01
8/12
NS
NS
NS
Formaldehyde
0.000013
0.0098
26/26
0.80
±0.14
10.42
0.08
5/5
NS
NS
NS
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Table 7-4. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Parachute, Colorado - PACO
Acetaldehyde
0.0000022
0.009
17/18
NA
NA
NA
29/29
0.59
±0.11
1.29
0.07
Benzene
0.0000078
0.03
54/54
1.21
±0.16
9.46
0.04
54/54
1.20
±0.14
9.39
0.04
1.3 -Butadiene
0.00003
0.002
7/54
0.01
±0.01
0.24
<0.01
2/54
<0.00
± <0.00
0.06
<0.01
Formaldehyde
0.000013
0.0098
18/18
NA
NA
NA
29/29
1.18
±0.19
15.32
0.12
Carbondale, Colorado - RFCO
Benzene
0.0000078
0.03
24/26
0.92
±0.68
7.18
0.03
27/27
0.45
±0.08
3.52
0.02
1,3-Butadiene
0.00003
0.002
5/26
0.01
±0.01
0.41
0.01
2/27
<0.01
±0.01
0.15
<0.01
Carbon Tetrachloride
0.000006
0.1
20/20
NA
NA
NA
NA
NA
NA
NA
1,2-Dichloroethane
0.000026
2.4
19/20
NA
NA
NA
NA
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Table 7-4. Risk Approximations for the Colorado Monitoring Sites (Continued)
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HO)
Rifle, Colorado - RICO
Acetaldehyde
0.0000022
0.009
25/26
NA
NA
NA
30/30
0.86
±0.15
1.89
0.10
Benzene
0.0000078
0.03
44/46
NA
NA
NA
62/62
1.10
±0.14
8.56
0.04
1.3 -Butadiene
0.00003
0.002
38/46
NA
NA
NA
39/62
0.07
±0.02
2.23
0.04
Ethylbenzene
0.0000025
1
45/46
NA
NA
NA
62/62
0.29
±0.03
0.72
<0.01
Formaldehyde
0.000013
0.0098
26/26
1.24
±0.21
16.06
0.13
30/30
1.31
±0.16
17.04
0.13
— = A Cancer URE or Noncancer RfC is not available.
NA = Not available due to the criteria for calculating an annual average.
NS = Sampling was not conducted during this time.
a Average concentrations provided for the pollutants below the blue line for GPCO are presented in ng/m3 for ease of viewing.
-------
Observations for the Garfield County sites from Table 7-4 include the following:
• Benzene, acetaldehyde, and formaldehyde were identified as pollutants of interest for
most of the Garfield County sites. RICO and PACO have the highest annual average
concentrations of these pollutants among the Garfield County sites (where they could
be calculated).
• The cancer risk approximations for formaldehyde for these sites range from 4.81 in-a-
million (BRCO, 2016) to 17.04 in-a-million (RICO, 2016). The noncancer hazard
approximations calculated for formaldehyde for the Garfield County sites with
available annual average concentrations are considerably less than 1.0 (all are less
than an HQ of 0.15). This indicates that no adverse noncancer health effects are
expected from this individual pollutant.
• The cancer risk approximations for acetaldehyde for these sites range from 0.49 in-a-
million (BRCO, 2016) to 1.89 in-a-million (RICO, 2016). The noncancer hazard
approximations calculated for acetaldehyde for the Garfield County sites with
available annual average concentrations are considerably less than 1.0 (all are 0.10 or
less). This indicates that no adverse noncancer health effects are expected from this
individual pollutant.
• Cancer risk approximations for benzene for these sites range from 3.52 in-a-million
(RFCO, 2016) to 9.46 in-a-million (PACO, 2015). The noncancer hazard
approximations calculated for benzene for the Garfield County sites with available
annual average concentrations are considerably less than 1.0 (all are less than or equal
to an HQ of 0.04). This indicates that no adverse noncancer health effects are
expected from this individual pollutant.
• 1,3-Butadiene was identified as a pollutant of interest for GSCO, PACO, RFCO, and
RICO. The cancer risk approximations for these sites for 1,3-butadiene range from
0.06 in-a-million (PACO, 2016) to 2.23 in-a-million (RICO, 2016). The noncancer
hazard approximations calculated for 1,3-butadiene for these Garfield County sites
are less than 0.05.
7-77
-------
7.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 7-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 7-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 7-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 7-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 7-5. Table 7-6
presents similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more
in-depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 7.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
7-78
-------
Table 7-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Colorado Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grand Junction, Colorado (Mesa County) - GPCO
Benzene
124.55
Formaldehyde
1.34E-03
Formaldehyde
41.07
Formaldehyde
103.42
Benzene
9.72E-04
Formaldehyde
36.35
Acetaldehyde
39.94
Ethylene oxide
3.25E-04
Benzene
5.56
Ethylbenzene
34.04
1,3-Butadiene
2.88E-04
Acetaldehyde
3.97
1.3 -Butadiene
9.59
Naphthalene
2.34E-04
Carbon Tetrachloride
3.60
Naphthalene
6.87
POM, Group 2b
9.59E-05
Acetaldehyde
3.58
Bis(2-ethylhexyl) phthalate, gas
6.86
Acetaldehyde
8.79E-05
1,3-Butadiene
3.31
POM, Group 2b
1.09
Ethylbenzene
8.51E-05
Naphthalene
3.09
POM, Group 2d
0.82
POM, Group 2d
7.19E-05
Naphthalene
2.53
2,4-Toluene diisocyanate
0.59
POM, Group 5a
5.46E-05
1,2-Dichloroethane
1.78
Battlement Mesa, Colorado (Garfield County) - BMCO
Benzene
831.13
Formaldehyde
9.09E-03
Formaldehyde
12.10
Formaldehyde
699.31
Benzene
6.48E-03
Benzene
6.17
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
Acetaldehyde
1.05
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
1,3-Butadiene
13.58
Naphthalene
1.84E-04
Naphthalene
5.41
Ethylbenzene
1.61E-04
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Silt, Colorado (Garfield County) - BRCO
Benzene
831.13
Formaldehyde
9.09E-03
Formaldehyde
10.71
Formaldehyde
699.31
Benzene
6.48E-03
Benzene
5.51
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
Benzene
5.16
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
Formaldehyde
4.81
1.3 -Butadiene
13.58
Naphthalene
1.84E-04
Acetaldehyde
1.12
Naphthalene
5.41
Ethylbenzene
1.61E-04
Acetaldehyde
0.49
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
Glenwood Springs, Colorado (Garfield County) - GSCO
Benzene
831.13
Formaldehyde
9.09E-03
Formaldehyde
10.42
Formaldehyde
699.31
Benzene
6.48E-03
Benzene
3.85
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
Acetaldehyde
1.16
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
1,3-Butadiene
0.63
1,3-Butadiene
13.58
Naphthalene
1.84E-04
Naphthalene
5.41
Ethylbenzene
1.61E-04
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Parachute, Colorado (Garfield County) - PACO
Benzene
831.13
Formaldehyde
9.09E-03
Formaldehyde
15.32
Formaldehyde
699.31
Benzene
6.48E-03
Benzene
9.46
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
Benzene
9.39
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
Acetaldehyde
1.29
1.3 -Butadiene
13.58
Naphthalene
1.84E-04
1,3-Butadiene
0.24
Naphthalene
5.41
Ethylbenzene
1.61E-04
1,3-Butadiene
0.06
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
Carbondale, Colorado (Garfield County) - RFCO
Benzene
831.13
Formaldehyde
9.09E-03
Benzene
7.18
Formaldehyde
699.31
Benzene
6.48E-03
Benzene
3.52
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
1,3-Butadiene
0.41
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
1,3-Butadiene
0.15
1,3-Butadiene
13.58
Naphthalene
1.84E-04
Naphthalene
5.41
Ethylbenzene
1.61E-04
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
Top 10 Cancer Toxicity-Weighted
Emissions
Top 10 Cancer Risk Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Cancer
Cancer Risk
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(in-a-million)
Rifle, Colorado (Garfield County) - RICO
Benzene
831.13
Formaldehyde
9.09E-03
Formaldehyde
17.04
Formaldehyde
699.31
Benzene
6.48E-03
Formaldehyde
16.06
Acetaldehyde
141.90
1,3-Butadiene
4.07E-04
Benzene
8.56
Ethylbenzene
64.35
Acetaldehyde
3.12E-04
1,3-Butadiene
2.23
1.3 -Butadiene
13.58
Naphthalene
1.84E-04
Acetaldehyde
1.89
Naphthalene
5.41
Ethylbenzene
1.61E-04
Ethylbenzene
0.72
Bis(2-ethylhexyl) phthalate, gas
2.65
1,2-Dibromoethane
1.26E-04
POM, Group la
1.07
POM, Group la
9.43E-05
Dichloro methane
0.62
POM, Group 2b
5.41E-05
POM, Group 2b
0.62
POM, Group 2d
4.36E-05
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Colorado Monitoring Sites
Top 10 Total Emissions for Pollutants
Top 10 Noncancer Toxicity-Weighted
Top 10 Noncancer Hazard Approximations
with Noncancer RfCs
Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Noncancer
Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Grand Junction, Colorado (Mesa County) - GPCO
Toluene
245.90
Acrolein
521,714.21
Formaldehyde
0.32
Xylenes
209.88
Formaldehyde
10,553.12
Formaldehyde
0.29
Benzene
124.55
2,4-Toluene diisocyanate
8,393.50
Acetaldehyde
0.20
Formaldehyde
103.42
1.3 -Butadiene
4,793.06
Acetaldehyde
0.18
Methanol
75.05
Acetaldehyde
4,437.71
1,3-Butadiene
0.06
Hexane
56.39
Benzene
4,151.73
Naphthalene
0.03
Acetaldehyde
39.94
Naphthalene
2,289.23
Naphthalene
0.02
Ethylbenzene
34.04
Xylenes
2,098.80
Benzene
0.02
Styrene
11.62
Lead, PM
1,291.29
Arsenic
0.02
Acrolein
10.43
Antimony, PM
1,051.02
Arsenic
0.02
Battlement Mesa, Colorado (Garfield County) - BMCO
Xylenes
1,453.50
Acrolein
3,865,686.15
Formaldehyde
0.09
Benzene
831.13
Formaldehyde
71,358.40
Acetaldehyde
0.05
Toluene
751.51
Benzene
27,704.24
Benzene
0.03
Formaldehyde
699.31
Acetaldehyde
15,766.16
Methanol
495.03
Xylenes
14,534.96
Hexane
159.66
1,3-Butadiene
6,788.90
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1.3 -Butadiene
13.58
Lead, PM
570.68
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
Top 10 Noncancer Toxicity-Weighted
Top 10 Noncancer Hazard Approximations
with Noncancer RfCs
Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Noncancer
Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Silt, Colorado (Garfield County) - BRCO
Xylenes
1,453.50
Acrolein
3,865,686.15
Formaldehyde
0.08
Benzene
831.13
Formaldehyde
71,358.40
Acetaldehyde
0.06
Toluene
751.51
Benzene
27,704.24
Formaldehyde
0.04
Formaldehyde
699.31
Acetaldehyde
15,766.16
Acetaldehyde
0.02
Methanol
495.03
Xylenes
14,534.96
Benzene
0.02
Hexane
159.66
1,3-Butadiene
6,788.90
Benzene
0.02
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1.3 -Butadiene
13.58
Lead, PM
570.68
Glenwood Springs, Colorado (Garfield County) -
GSCO
Xylenes
1,453.50
Acrolein
3,865,686.15
Formaldehyde
0.08
Benzene
831.13
Formaldehyde
71,358.40
Acetaldehyde
0.06
Toluene
751.51
Benzene
27,704.24
Benzene
0.02
Formaldehyde
699.31
Acetaldehyde
15,766.16
1,3-Butadiene
0.01
Methanol
495.03
Xylenes
14,534.96
Hexane
159.66
1,3-Butadiene
6,788.90
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1,3-Butadiene
13.58
Lead, PM
570.68
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
Top 10 Noncancer Toxicity-Weighted
Top 10 Noncancer Hazard Approximations
with Noncancer RfCs
Emissions
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Noncancer
Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Parachute, Colorado (Garfield County) - PACO
Xylenes
1,453.50
Acrolein
3,865,686.15
Formaldehyde
0.12
Benzene
831.13
Formaldehyde
71,358.40
Acetaldehyde
0.07
Toluene
751.51
Benzene
27,704.24
Benzene
0.04
Formaldehyde
699.31
Acetaldehyde
15,766.16
Benzene
0.04
Methanol
495.03
Xylenes
14,534.96
1,3-Butadiene
<0.01
Hexane
159.66
1,3-Butadiene
6,788.90
1,3-Butadiene
<0.01
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1.3 -Butadiene
13.58
Lead, PM
570.68
Carbondale, Colorado (Garfield County) - RFCO
Xylenes
1,453.50
Acrolein
3,865,686.15
Benzene
0.03
Benzene
831.13
Formaldehyde
71,358.40
Benzene
0.02
Toluene
751.51
Benzene
27,704.24
1,3-Butadiene
0.01
Formaldehyde
699.31
Acetaldehyde
15,766.16
1,3-Butadiene
<0.01
Methanol
495.03
Xylenes
14,534.96
Hexane
159.66
1,3-Butadiene
6,788.90
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1,3-Butadiene
13.58
Lead, PM
570.68
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 7-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Colorado Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Rifle, Colorado (Garfield County) - RICO
Xylenes
1,453.50
Acrolein
3,865,686.15
Formaldehyde
0.13
Benzene
831.13
Formaldehyde
71,358.40
Formaldehyde
0.13
Toluene
751.51
Benzene
27,704.24
Acetaldehyde
0.10
Formaldehyde
699.31
Acetaldehyde
15,766.16
1,3-Butadiene
0.04
Methanol
495.03
Xylenes
14,534.96
Benzene
0.04
Hexane
159.66
1.3 -Butadiene
6,788.90
Ethylbenzene
<0.01
Acetaldehyde
141.90
2,4-Toluene diisocyanate
3,245.03
Acrolein
77.31
Naphthalene
1,802.56
Ethylbenzene
64.35
Cadmium, PM
761.59
1.3 -Butadiene
13.58
Lead, PM
570.68
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 7-5 include the following:
• The five highest emitted pollutants with cancer UREs in Mesa County are also the
five highest emitted pollutants in Garfield County, although the emissions are higher
in Garfield County. In total, there are eight pollutants in common for Garfield County
and Mesa County.
• The two pollutants with the highest toxicity-weighted emissions (of the pollutants
with cancer UREs) are formaldehyde and benzene for both Mesa and Garfield
Counties. These two counties have eight pollutants in common among the pollutants
with the highest toxicity-weighted emissions.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Mesa County; eight pollutants have the highest emitted pollutants and
highest toxicity-weighted emissions for Garfield County.
• Formaldehyde has the highest cancer risk approximations for GPCO, followed by
benzene (2016 only). Formaldehyde and benzene appear at the top of all three lists in
Table 7-5. Acetaldehyde, 1,3-butadiene, and naphthalene also appear on all three lists.
• Each of the pollutants of interest identified for the Garfield County sites also appear
on both emissions-based lists for Garfield County in Table 7-5.
Observations from Table 7-6 include the following:
• Toluene, xylenes, and benzene are the highest emitted pollutants with a noncancer
RfC in Mesa County; these same pollutants are also the highest emitted in Garfield
County, although the order is different. Note that the emissions of these pollutants,
particularly for xylenes, are considerably higher in Garfield County. These two
counties have an additional six pollutants in common on their lists of highest emitted
pollutants with noncancer RfCs.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein. Although acrolein was sampled for at
GPCO, this pollutant was excluded from the pollutants of interest designation, and
thus subsequent risk-based screening evaluations, due to questions about the
consistency and reliability of the measurements, as discussed in Section 3.2. Acrolein
is not a target analyte for the SNMOC method. Although acrolein has the highest
toxicity-weighted emissions for all but two counties with an NMP site, it does not
often appear among the highest emitted pollutants. Mesa and Garfield Counties are
two of only five counties with an NMP site for which acrolein ranks among the 10
highest emitted. The acrolein emissions for Garfield County are the third highest
among counties with NMP sites (behind only Cook County, Illinois and Los Angeles
County, California). A similar observation was made in previous NMP reports.
• Five of the highest emitted pollutants in Mesa County (including acrolein) also have
the highest toxicity-weighted emissions. Six of the 10 highest emitted pollutants in
Garfield County (including acrolein) also have the highest toxicity-weighted
emissions. Toluene, the highest emitted pollutant for Mesa County and third highest
emitted pollutant in Garfield County, is not among those pollutants with the highest
7-87
-------
toxicity-weighted emissions. Several metals appear near the bottom of each toxicity-
weighted emissions list for each county but do not appear among the highest emitted.
• Formaldehyde, acetaldehyde, and benzene are pollutants of interest for GPCO that
appear on all three lists in Table 7-6. Naphthalene and 1,3-butadiene appear among
the pollutants with the highest noncancer hazard approximations and highest toxicity-
weighted emissions but are not among the highest emitted pollutants with a noncancer
RfC in Mesa County. This is also true for arsenic.
• Each of the pollutants of interest identified for the Garfield County sites appear on
both emissions-based lists in Table 7-6, with one exception. Ethylbenzene is a
pollutant of interest for RICO. Ethylbenzene appears among the pollutants with the
highest emissions in Garfield County, but is not among those with the 10 highest
toxicity-weighted emissions.
Summary of the 2015-2016 Monitoring Data for the Colorado Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Twenty pollutants failed screens for GPCO. The number ofpollutants failing screens
for the Garfield County sites ranged from three to five.
~~~ Among GPCO's pollutants of interest, dichlorome thane had the highest
concentrations measured, particularly for 2015. Dichloromethane, formaldehyde,
and acetaldehyde are the only pollutants with annual average concentrations greater
than 1 ng/m3.
~~~ Among the Garfield County sites, RICO andPACO had the highest annual average
concentrations of benzene andformaldehyde, the only pollutants of interest for these
sites with annual average concentrations greater than 1 ng/m3. PACO had the second
(2015) and third (2016) highest annual average concentrations of benzene among all
NMP sites sampling this pollutant.
~~~ GPCO andfour Garfield County sites have sampled under the NMP for at least
5 years. Notable trends for these sites include: The 1-year average concentrations for
several pollutants for GPCO are at a minimum for 2016, including acenaphthene,
benzene, 1,3-butadiene, ethylbenzene, and naphthalene. The significant increase in
the detection rate of 1,2-dichloroethane beginning in 2012 continues through 2016.
The highest fluoranthene concentrations measured since the onset of sampling, were
measured at GPCO in 2015 and 2016. Concentrations of acetaldehyde and
formaldehyde decreased considerably at BRCO in 2016. Concentrations of
acetaldehyde appear to have a decreasing trend at PACO. The detection rate of
1,3-butadiene has decreased considerably at PACO andRFCO, particularly in 2016.
~~~ Formaldehyde had the highest cancer risk approximations among the pollutants of
interest for GPCO. Benzene andformaldehyde have the highest cancer risk
approximations for the Garfield County sites, depending on the year and whether
annual average concentrations could be calculated. None of the pollutants of interest
for the Colorado monitoring sites have noncancer hazard approximations greater
than an HQ of 1.0.
7-88
-------
8.0 Site in the District of Columbia
This section summarizes those data from samples collected at the NATTS site in
Washington, D.C. and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines the
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to Sections
1 through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
8.1 Site Characterization
This section characterizes the Washington, D.C. monitoring site by providing a
description of the nearby area surrounding the monitoring site; plotting emissions sources
surrounding the monitoring site; and presenting traffic data and other characterizing information
for the site. This information is provided to give the reader insight regarding factors that may
influence the air quality near the site and assist in the interpretation of the ambient
measurements.
Figure 8-1 presents a composite satellite image retrieved from ArcGIS Explorer showing
the monitoring site and its immediate surroundings. Figure 8-2 identifies nearby point source
emissions locations by source category, as reported in the 2014 NEI for point sources, version 1.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 8-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference but have been
grayed out to emphasize emissions sources within the boundary. Table 8-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. r
8-1
-------
Figure 8-1. Washington, D.C. (WADC) Monitoring Site
00
to
•Swiimcmy'tn'
z\- £ B^umbia M ™
NW .£ - •' . -. i •
IfjwMMi lT^55SiB@4*g-gg|
WBBSmwz
NW KXc^9an
lair IK
M'ich>9ani
Franklin St NE1
ii?montStNW
'Evarts-StNE
Hazard UmvmrtV
,upt.|»w
Douglas St~N£.
McMillan Reservoir
. i
£ Howaf
,EdgeW°(
iGhanninq*St -
<51 Howard University H@rta £
i2-4f m I
U St NW *
r-,i -j
ifKBHii'S
¦ imk« i . ^
^Thomas SfcNW-
i /**wj* i •
luiiuinb^
HT St NW-
LTcdd.Pl.NE.
AStNW
i «*
ft
*T iStNW
Hyde ,
leadership
Charter' i
CrtGwjk
iWevtminster|St'NW-i
-------
Figure 8-2. NEI Point Sources Located Within 10 Miles of W A DC
DISTRICT OF
COLUMBIA
MARYLAND
VIRGINIA
Potomac
[ River
WADC NATTS site O 10 mile radius
Source Category Group (No. of Facilities)
County boundary
*
Airport/Airline/Airport Support Operations (23)
X
Mine/Quarry/Mineral Processing Facility (2)
'i
Asphalt Production/Hot Mix Asphalt Plant (4)
?
Miscellaneous Commercial/Industrial Facility (7)
B
Bulk Terminal/Bulk Plant (1)
a
Paint and Coating Manufacturing Facility (1)
*
Electricity Generation Facility (1)
p
Printing/Publishing/Paper Product
Manufacturing Facility (3)
>
Hotels/Motels/Lodging (5)
X
Rail Yard/Rail Line Operations (3)
O
Institution (school, hospital, prison, etc.) (21)
»
Water Treatment Facility (2)
A
Military Base/National Security Facility (10)
Legend
77"5'0"W 77n0'0"W 76"55'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
8-3
-------
Table 8-1. Geographical Information for the Washington, D.C. Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
WADC
11-001-0043
Washington
District Of
Columbia
Washington-
Arlington-Alexandria,
DC-VA-MD-WV
38.921847,
-77.013178
Commercial
Urban/City
Center
3,600
Bryant St NW at First St
lAADT reflects 2014 data (DC DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
00
-L
-------
Figure 8-1 shows that the WADC monitoring site is located in an open field at the
southeast end of the McMillan Water Reservoir in Washington, D.C. It is also located within a
short distance of several heavily traveled roadways. The site is located in a commercial area, and
is surrounded by a hospital, a cemetery, and a university. The First Street Tunnel Project, a
construction project which commenced to reduce sewer flooding in nearby neighborhoods, was
completed in October 2016 (DC WSA, 2017).
As Figure 8-2 shows, WADC is surrounded by a number of emissions sources, many of
which are included in three sources categories: 1) the airport and airport support operations
source category, which includes airports and related operations as well as small runways and
heliports, such as those associated with hospitals or televisions stations; 2) the institution source
category, which includes hospitals, schools, and prisons; and 3) the military bases and national
security facilities source category. The closest sources to WADC are a wastewater treatment
facility, hospitals, and heliports at hospitals.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 8-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near WADC is less than 4,000 vehicles and is among the 10 lowest
compared to other NMP sites. The traffic volume provided is for Bryant Street NW at First Street
NW, the closest intersection east of the monitoring site. The tunnel project may have affected
typical traffic patterns in the area during construction.
8.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 8-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 8-2. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. PAHs were sampled for at WADC in 2015 and 2016.
8-5
-------
Table 8-2. 2015-2016 Risk-Based Screening Results for the Washington, D.C. Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Washington, D.C. - WADC
Naphthalene
0.029
107
118
90.68
98.17
98.17
Benzo(a)pyrene
0.00057
2
99
2.02
1.83
100.00
Total
109
217
50.23
Observations from Table 8-2 include the following:
• Concentrations of two PAHs failed screens for WADC: naphthalene, and
benzo(a)pyrene.
• Concentrations of naphthalene failed nearly 91 percent of screens, while
concentrations of benzo(a)pyrene failed 2 percent of screens.
• Naphthalene accounted for more than 98 percent of the total failed screens for
WADC; thus, naphthalene is WADC's only pollutant of interest.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutant(s) of interest identified via the risk-based screening process, as described in
Section 3.4.2.
8.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Washington, D.C. monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
Site-specific statistical summaries for all pollutants sampled for at WADC are provided in
Appendix N.
8-6
-------
8.3.1 2015 and 2016 Concentration Averages
Quarterly and annual average concentrations for 2015 and 2016 were calculated for the
pollutants of interest for the Washington, D.C. monitoring site, as described in Section 3.1. The
quarterly average concentration of a particular pollutant is simply the average concentration of
the preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual average concentrations were calculated for pollutants where three valid
quarterly averages could be calculated for a given year and where method completeness was
greater than or equal to 85 percent, as presented in Section 2.4. Quarterly and annual average
concentrations for the pollutant of interest for WADC are presented in Table 8-3, where
applicable. Note that if a pollutant was not detected in a given calendar quarter, the quarterly
average simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
Observations for WADC from Table 8-3 include the following:
• Naphthalene was detected in every valid PAH sample collected at WADC.
• Concentrations of naphthalene measured at WADC range from 16.0 ng/m3 to
257 ng/m3.
• The annual average concentration of naphthalene for 2015 is similar the annual
average concentration for 2016.
• A comparison of both years' quarterly average concentrations shows that the fourth
quarter average is the highest quarterly average for both years and has the largest
confidence intervals. For 2015, both the minimum and maximum concentrations of
naphthalene were measured during the fourth quarter of the year, including all six
naphthalene concentrations greater than 100 ng/m3. For 2016, nine naphthalene
concentrations greater than 100 ng/m3 were measured, four each during the first and
fourth quarters, with an additional one measured during the second quarter. Again,
the minimum concentration was measured during the fourth quarter. Even though the
maximum concentration of naphthalene was not measured during the fourth quarter in
2016, the second, third, and fourth highest naphthalene concentrations measured in
2016 were measured during the fourth quarter.
8-7
-------
Table 8-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Washington, D.C. Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Washington, D.C. - WADC
Naphthalene
60/60/60
58.61
± 10.31
47.73
±5.07
50.68
±9.30
82.09
±27.37
60.56
±8.81
58/58/58
73.84
± 32.90
52.85
± 13.68
45.79
±9.53
83.97
± 30.00
65.23
± 12.52
00
00
-------
8.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, a box plot was created for naphthalene for WADC.
Figure 8-3 overlays the site's minimum, annual average, and maximum naphthalene
concentrations for each year onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.2.1, and are discussed
below. If an annual average concentration could not be calculated, the range of concentrations is
still provided in the figure(s) that follow.
Figure 8-3. Program vs. Site-Specific Average Naphthalene Concentrations
U4 1 I I ' I
WADC T
'
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
» o
Figure 8-3 presents the box plot for naphthalene for WADC and shows the following:
• The range of naphthalene concentrations measured at WADC in 2016 is slightly
larger than the range measured in 2015. The maximum naphthalene concentrations
measured at WADC in both years are considerably less than the program-level
maximum concentration (403 ng/m3).
• The annual average concentration of naphthalene for 2015 is similar to the program-
level average concentration (61.23 ng/m3), while the annual average for 2016 is just
slightly higher.
• There were no non-detects of naphthalene measured at WADC, or across the
program.
8.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
WADC has sampled PAHs under the NMP since mid-2008. Thus, Figure 8-4 presents the 1-year
statistical metrics for naphthalene for WADC. The statistical metrics presented for assessing
8-9
-------
trends include the substitution of zeros for non-detects. If sampling began mid-year, a minimum
of 6 months of sampling is required for inclusion in the trends analysis; in these cases, a 1-year
average concentration is not provided, although the range and percentiles are still presented.
Figure 8-4. Yearly Statistical Metrics for Naphthalene Concentrations Measured at WADC
5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until late June 2008.
Observations from Figure 8-4 for naphthalene concentrations measured at WADC
include the following:
• WADC began sampling PAHs under the NMP in late June 2008. Because a full
year's worth of data is not available, a 1-year average concentration for 2008 is not
presented, although the range of measurements is provided.
• The maximum naphthalene concentration was measured in 2009 and is the only
concentration greater than 500 ng/m3 measured at this site (553 ng/m3).
Concentrations greater than 400 ng/m3 were measured in each year of sampling
between 2009 and 2012, after which concentrations greater than 300 ng/m3 were not
measured.
• The 1-year average concentrations exhibit a significant decreasing trend over the
years of sampling, decreasing by more than half over the period, from a maximum of
128.63 ng/m3 in 2009 to a minimum of 60.56 ng/m3 in 2015. A slight uptick is shown
in the 1-year average concentration for 2012 before the decreasing trend resumes; less
than 2 ng/m3 separates the 2011 and 2012 averages. A similar observation can be
made for 2016.
8-10
-------
• The median concentration exhibits a similar pattern as the 1-year average
concentration, though the increase from 2011 to 2012 is larger. The median
concentration is less than 100 ng/m3 for each year shown in Figure 8-4, and is less
than 50 ng/m3 for 2015 and 2016.
8.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the WADC monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
8.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutant(s) of interest for WADC, risk was examined by calculating cancer risk
and noncancer hazard approximations for each year annual average concentrations could be
calculated. These approximations can be used as risk estimates for cancer and noncancer effects
attributable to the pollutants of interest. Although the use of these approximations is limited, they
may help identify where policy-makers want to shift their air monitoring priorities. Refer to
Section 3.4.2.3 for an explanation of how cancer risk and noncancer hazard approximations are
calculated and what limitations are associated with them. Annual averages, cancer UREs and/or
noncancer RfCs, and cancer risk and noncancer hazard approximations are presented in
Table 8-4, where applicable. Cancer risk approximations are presented as probabilities while the
noncancer hazard approximations are ratios and thus, unitless values.
Observations for WADC from Table 8-4 include the following:
• The annual average concentrations of naphthalene for WADC fall in the middle of the
range compared to annual average concentrations for other NMP sites sampling this
pollutant.
• The cancer risk approximations for naphthalene are 2.06 in-a-million for 2015 and
2.22 in-a-million for 2016.
• The noncancer hazard approximations for naphthalene are significantly less than 1.0
(0.02 for both years), indicating that no adverse noncancer health effects are expected
from this individual pollutant.
8-11
-------
Table 8-4. Risk Approximations for the Washington, D.C. Monitoring Site
2015
2016
Risk Approximations
Risk Approximations
# of
# of
Cancer
Noncancer
Measured
Annual
Measured
Annual
URE
RfC
Detections vs.
Average
Cancer
Noncancer
Detections vs.
Average
Cancer
Noncancer
Pollutant
(Ug/m3)1
(mg/m3)
# of Samples
(ng/m3)
(in-a-million)
(HQ)
# of Samples
(ng/m3)
(in-a-million)
(HQ)
Washington, D.C. - WADC
60.56
65.23
Naphthalene
0.000034
0.003
60/60
±8.81
2.06
0.02
58/58
± 12.52
2.22
0.02
-------
8.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 8-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 8-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 8-5 provides the cancer risk approximations (in-a-million) for the pollutant of interest for
WADC, as presented in Table 8-4. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 8-5. Table 8-6 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 8.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
8-13
-------
Table 8-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Washington, D.C. Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Washington, D.C. (District of Columbia) - WADC
Benzene
92.37
Formaldehyde
1.20E-03
Naphthalene
2.22
Formaldehyde
91.97
Benzene
7.20E-04
Naphthalene
2.06
Ethylbenzene
49.50
1,3-Butadiene
3.88E-04
Acetaldehyde
48.19
Naphthalene
2.58E-04
1.3 -Butadiene
12.93
POM, Group 2b
1.50E-04
Naphthalene
7.58
POM, Group 2d
1.28E-04
POM, Group 2b
1.71
Ethylbenzene
1.24E-04
POM, Group 2d
1.45
Acetaldehyde
1.06E-04
Trichloroethylene
0.18
POM, Group 5a
9.87E-05
POM, Group 6
0.17
Arsenic, PM
8.95E-05
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 8-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Washington, D.C. Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
Top 10 Noncancer Toxicity-Weighted
Emissions
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(County-Level)
(County-Level)
(Site-Specific)1
Noncancer
Noncancer Hazard
Pollutant
Emissions
(tpy)
Pollutant
Toxicity
Weight
Pollutant
Approximation
(HQ)
Washington, D.C. (District of Columbia) - WADC
Toluene
335.28
Acrolein
309,020.44
Naphthalene
0.02
Methanol
294.90
Formaldehyde
9,384.59
Naphthalene
0.02
Xylenes
179.06
1.3 -Butadiene
6,464.65
Ethylene glycol
101.49
Acetaldehyde
5,354.92
Benzene
92.37
Benzene
3,079.01
Formaldehyde
91.97
Naphthalene
2,526.39
Hexane
67.97
Xylenes
1,790.63
Ethylbenzene
49.50
Arsenic, PM
1,387.77
Acetaldehyde
48.19
Cadmium, PM
1,106.59
Glycol ethers, gas
13.59
Propionaldehyde
842.80
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 8-5 include the following:
• Benzene and formaldehyde are the highest emitted pollutants with cancer UREs in the
District of Columbia. Formaldehyde and benzene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs).
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
• Naphthalene is the only pollutant of interest for WADC. This pollutant appears on
both emissions-based lists. Naphthalene is the sixth highest emitted pollutant with a
cancer URE in the District of Columbia and has the fourth highest toxicity-weighted
emissions (of the pollutants with cancer UREs).
• Several POM Groups are among the highest emitted "pollutants" in the District
and/or rank among the pollutants with the highest toxicity-weighted emissions. POM,
Groups 2b and 6 includes several PAHs sampled for at WADC, although none of
these failed screens. POM, Group 5a includes benzo(a)pyrene, which failed two
screens but was not identified as a pollutant of interest for WADC. POM, Group 2d
does not include any PAHs sampled for at WADC.
Observations from Table 8-6 include the following:
• Toluene and methanol are the highest emitted pollutants with noncancer RfCs in the
District of Columbia.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
• Four of the highest emitted pollutants in the District of Columbia also have the
highest toxicity-weighted emissions.
• Naphthalene has the sixth highest toxicity-weighted emissions but is not one of the
10 highest emitted pollutants (of the pollutants with noncancer RfCs).
• None of the other pollutants sampled for at WADC under the NMP appear in
Table 8-6.
Summary of the 2015-2016 Monitoring Data for WADC
Results from several of the data analyses described in this section include the following:
~~~ Concentrations of two PAHs failed screens, with naphthalene failing the majority of
screens and therefore is the only pollutant of interest identified via the risk screening
process.
~~~ The annual average concentration of naphthalene for 2015 is similar the annual
average concentration for 2016 and both fall in the middle of the range compared to
annual average concentrations for other NMP sites sampling this pollutant.
~~~ Concentrations of naphthalene have an overall decreasing trend at WADC.
8-16
-------
~~~ Both cancer risk approximations for naphthalene are approximately 2 in-a-million;
the noncancer hazard approximations for naphthalene are significantly less than an
HQ of 1.0.
8-17
-------
9.0 Sites in Florida
This section summarizes those data from samples collected at the NATTS and UATMP
sites in Florida and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines the
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to Sections 1
through 4 for detailed discussions and definitions regarding
the various data analyses presented below.
9.1 Site Characterization
This section characterizes the Florida monitoring sites by providing a description of the
nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
The five Florida sites are located in two separate urban areas; AZFL, SKFL, and SYFL
are located in the Tampa-St. Petersburg-Clearwater, Florida CBSA, ORFL and PAFL are located
in the Orlando-Kissimmee-Sanford, Florida CBSA. Figures 9-1 and 9-2 present composite
satellite images retrieved from ArcGIS Explorer showing the St. Petersburg area monitoring sites
and their immediate surroundings. Figure 9-3 identifies nearby point source emissions locations
that surround these two sites by source category, as reported in the 2014 NEI for point sources,
version 1. Note that only sources within 10 miles of the sites are included in the facility counts
provided in Figure 9-3. A 10-mile boundary was chosen to give the reader an indication of which
emissions sources and emissions source categories could potentially have a direct effect on the
air quality at the monitoring sites. Further, this boundary provides both the proximity of
emissions sources to the monitoring sites as well as the quantity of such sources within a given
distance of the sites. Sources outside the 10-mile boundaries are still visible on the map for
reference but have been grayed out to emphasize emissions sources within the boundaries.
Figures 9-4 through 9-8 present the composite satellite images and emissions sources maps for
the Tampa site and the two sites in the Orlando area. Table 9-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates. Each
figure and table is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. ^
9-1
-------
Figure 9-1. St. Petersburg, Florida (AZFL) Monitoring Site
Xountry.Glub.Rd.N,
•16th Terrace N
u16th,AveNi
15th.Ave.N-
»13thfAveNi
«12 th .Terrace.N —5
4* jS**' $1
«t2tlnTerrace>N-_c
.8th. Ave N,
-------
Figure 9-2. Pinellas Park, Florida (SKFL) Monitoring Site
-------
Figure 9-3. NEI Point Sources Located Within 10 Miles of AZFL and SKFL
Pinellas
County
I,
Tampa Bay \
82°50'0"W 82°45'0"W 82°40'0"W 82°35'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
82°35'0"W
—b.
Hillsborough ^
County ^
~ AZFL UATMP site ~ SKFL NATTS site
Source Category Group (No. of Facilities)
10 mile radius
County boundary
*
Aerospace/Aircraft Manufacturing Facility (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
T
Airport/Airline/Airport Support Operations (8)
Metals Processing/Fabrication Facility (7)
Asphalt Production/Hot Mix Asphalt Plant (1)
?
Miscellaneous Commercial/Industrial Facility (7)
«
Automobile/Truck Manufacturing Facility (1)
H
Municipal Waste Combustor (1)
B
Bulk Terminal/Bulk Plant (1)
~
Paint and Coating Manufacturing Facility (2)
e
Electrical Equipment Manufacturing Facility (4)
o
Pharmaceutical Manufacturing Facility (1)
f
Electricity Generation Facility (2)
R
Plastic, Resin, or Rubber Products Plant (5)
F
Food Processing/Agriculture Facility (1)
P
Printing/Publishing/Paper Product Manufacturing Facility (9)
*
Industrial Machinery or Equipment Plant (3)
A
Ship/Boat Manufacturing or Repair Facility (6)
O
Institution (school, hospital, prison, etc.) (1)
I
Wastewater Treatment Facility (2)
¦
Landfill (2)
w
Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
Guff of
Legend
9-4
-------
vo
iydrvey.Rd.
Figure 9-4. Valrico, Florida (SYFL) Monitoring Site
-------
Figure 9-5. NEI Point Sources Located Within 10 Miles of SYFL
Hillsborough
County
Source Category Group (No. of Facilities)
*
Aerospace/Aircraft Manufacturing Facility (1)
A
Metal Coating, Engraving, and Allied Services to Manufacturers (1)
T
Airport/Airline/Airport Support Operations (7)
<•>
Metals Processing/Fabrication Facility (4)
i
Asphalt Production/Hot Mix Asphalt Plant (3)
X
Mine/Quarry/Mineral Processing Facility (1)
0
Auto Body Shop/Painters/Automotive Stores (1)
?
Miscellaneous Commercial/Industrial Facility (4)
»
Automobile/Truck Manufacturing Facility (1)
El
Municipal Waste Combustor (1)
~
Brick, Structural Clay, or Clay Ceramics Plant (1)
<
Pesticide Manufacturing Plant (1)
i
Compressor Station (1)
Petroleum Refinery (1)
©
Dry Cleaning Facility (1)
R
Plastic, Resin, or Rubber Products Plant (1)
e
Electrical Equipment Manufacturing Facility (2)
P
Printing/Publishing/Paper Product Manufacturing Facility (2)
V
Fertilizer Plant (1)
X
Rail Yard/Rail Line Operations (2)
F
Food Processing/Agriculture Facility (3)
w
Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
SYFL NATTS site
O 10 mile radius
County boundary
Tampa '
Bay |
Legend
82°15'0"W 82°10'0"W 82°5'C
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
9-6
-------
Figure 9-6. Winter Park, Florida (ORFL) Monitoring Site
-Whipple-Ave'
.W-SwoopeAve-
rE Cnfllahd'Ave"
/Fairview-Avp -
Dallas Ave<
Fairbanks Ave WQJ
Kentucky Ave
WlesbwAwfl-
-------
Figure 9-7. Orlando, Florida (PAFL) Monitoring Site
•E.Rark'.Uke.StLS.
Colonial Landing
MtVefnonSU
lAmelia-Su
- Stanley
•Valencia Rd
-------
Figure 9-8. NEI Point Sources Located Within 10 Miles of QRFL and PAFL
Orange
County
Legend
~ ORFL UATMP site
81°25'0"W 81C20'0"W 81°15'0"W 81'10'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
PAFL UATMP site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
"1" Airport/Airline/Airport Support Operations (23) •
'it Asphalt Production/Hot Mix Asphalt Plant (4) jj)
0 Auto Body Shop/Painters/Automotive Stores (1) A
« Automobile/Truck Manufacturing Facility (4) <•>
B Bulk Terminal/Bulk Plant (1) x
1 Compressor Station (1) ?
e Electrical Equipment Manufacturing Facility (2) ^
f Electricity Generation Facility (2) P
F Food Processing/Agriculture Facility (4) X
Industrial Machinery or Equipment Plant (2)
o Institution (school, hospital, prison, etc.) (6)
Landfill (1)
Metal Can, Box, and Other Metal Container Manufacturing Facility (1)
Metal Coating, Engraving, and Allied Services to Manufacturers (2)
Metals Processing/Fabrication Facility (2)
Mine/Quarry/Mineral Processing Facility (1)
Miscellaneous Commercial/Industrial Facility (5)
Paint and Coating Manufacturing Facility (3)
Printing/Publishing/Paper Product Manufacturing Facility (4)
Rail Yard/Rail Line Operations (2)
Ship/Boat Manufacturing or Repair Facility (1)
Seminole
County
9-9
-------
Table 9-1. Geographical Information for the Florida Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
AZFL
12-103-0018
St.
Petersburg
Pinellas
Tampa-St.
Petersburg-
Clearwater, FL
27.785866,
-82.739875
Residential
Suburban
3,1002
9th Ave N, W of Park St N
SKFL
12-103-0026
Pinellas
Park
Pinellas
Tampa-St.
Petersburg-
Clearwater, FL
27.850348,
-82.714465
Residential
Suburban
4,0002
60th St N, N of 82 Ave N
SYFL
12-057-3002
Valrico
Hillsborough
Tampa-St.
Petersburg-
Clearwater, FL
27.965650,
-82.230400
Residential
Rural
3,900
Sydney Road, W of S Forbes Rd
ORFL
12-095-2002
Winter
Park
Orange
Orlando-
Kissimmee-
Sanford, FL
28.596389,
-81.362500
Commercial
Urban/City
Center
33,000
Orlando Ave, N of Morse Blvd
PAFL
12-095-1004
Orlando
Orange
Orlando-
Kissimmee-
Sanford, FL
28.550833,
-81.345556
Commercial
Suburban
50,000
Colonial/MLK Blvd, b/w
Primrose Rd & Bumby Ave
1AADT reflects 2016 data (FL DOT, 2016)
2 Previous traffic count location moved to closer location; reflects a large decrease.
BOLD ITALICS = EPA-designated NATTS Site
-------
AZFL is located at Azalea Park in St. Petersburg. Figure 9-1 shows that the area
surrounding AZFL consists of mixed land use, including residential, commercial, and industrial
properties. The industrial property separated from Azalea Park by 72nd Street North is a former
electronics manufacturer and a permanently closed facility, and was purchased in 2015 by a
commercial redevelopment company (Girardi, 2015). Heavily traveled roadways are located less
than 1 mile from the monitoring site. AZFL is located less than 1 mile east of Boca Ciega Bay,
the edge of which can be seen in the bottom-left corner of Figure 9-1.
SKFL is located in Pinellas Park, north of St. Petersburg. This site is located on the
property of Skyview Elementary School, at the corner of 86th Avenue North and 60th Street
North. Figure 9-2 shows that SKFL is located in a primarily residential area. A rail line intersects
the Pinellas Park Ditch near a construction company on the left-hand side of Figure 9-2.
Population exposure is the purpose behind monitoring at this location. This site is the Pinellas
County NATTS site.
Figure 9-3 shows the location of the St. Petersburg sites in relation to each other. AZFL is
located approximately 5 miles south-southwest of SKFL. Most of the emissions sources on the
Tampa Bay Peninsula are located north of SKFL. A small cluster of point sources is also located
southeast of SKFL. The source categories with the greatest number of emissions sources in the
St. Petersburg area (based on the areas covered by the 10-mile boundaries) include printing,
publishing, and paper product manufacturing; the airport source category, which includes
airports and related operations as well as small runways and heliports, such as those associated
with hospitals or television stations; metals processing and fabrication; and ship/boat
manufacturing or repair. The emissions source closest to AZFL is a plastic, resin, or rubber
products plant, and is the only source within 2 miles of the site. While the emissions source
closest to SKFL falls into the miscellaneous commercial/industrial facility source category, a
plastic, resin, or rubber products plant, two metals processing/fabrication facilities, two ship/boat
manufacturing or repair facilities, an industrial machinery or equipment plant, and two additional
miscellaneous facilities are also located within 2 miles of SKFL; many of these facilities are in
the cluster of sources to the northwest of SKFL.
SYFL is located in Valrico, which is also part of the Tampa-St. Petersburg-Clearwater,
Florida CBSA, although it is on the eastern outskirts of the area. The SYFL monitoring site is
located in a rural area, although, as Figure 9-4 shows, a residential community and country club
lie just to the west of the site. Located to the south of the site (and shown in the bottom-center
9-11
-------
portion of Figure 9-4) are tanks that are part of the local water treatment facility. This site serves
as a background site, although the effects of increased development in the area are likely being
captured by the monitoring site. This site is the Tampa NATTS site.
Figure 9-5 shows that most of the emissions sources surrounding SYFL are greater than
5 miles away from the site. The point sources shown are included in a variety of sources
categories. The airport source category and metals processing and fabrication are among the
source categories with the greatest number of emissions sources near SYFL. The closest source
to SYFL with reportable air emissions in the 2014 NEI is a food processing facility.
ORFL is located in Winter Park, north of Orlando. Figure 9-6 shows that ORFL is
located near Lake Mendsen, just behind Community Playground. The site is east of Lake
Killarney and south of Winter Park Village. This site lies in a commercial area and is a
population exposure site.
PAFL is located in northeast Orlando, on the northwestern edge of the Orlando Executive
Airport property, as shown in Figure 9-7. The area is commercial in nature and experiences
heavy traffic. The airport is bordered by Colonial Drive to the north and the East-West
Expressway (Toll Road 408) to the south (although not shown in Figure 9-7). A large shopping
complex is located to the northeast of the site, just north of the airport, between Colonial Drive
and Maguire Boulevard. Interstate-4 runs north-south approximately 2 miles to the west of the
monitoring site.
ORFL is located 3.3 miles north-northwest of PAFL. Most of the point sources
surrounding these sites are located on the western side of the 10-mile boundaries, as shown in
Figure 9-8. Although the emissions sources surrounding ORFL and PAFL are involved in a
variety of industries and processes, the airport and airport support operations source category has
the greatest number of emissions sources within 10 miles of these sites. The closest emissions
source to PAFL is Orlando Executive Airport, which is located under the star symbol for PAFL
in Figure 9-8. The closest emissions sources to ORFL are located to the south of the site: a
hospital, which falls into the institutions category, and the heliport located at the hospital, which
falls into the airport source category.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 9-1 also contains traffic volume information for each site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
9-12
-------
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volumes in Table 9-1 are higher near the Orlando sites than the Tampa/St. Petersburg sites.
Different traffic locations for AZFL and SKFL were chosen for this NMP report compared to
previous reports. The traffic volume for PAFL ranks 20th highest among other NMP sites, with
ORFL ranking 24th. The traffic volumes near AZFL, SKFL, and SYFL are in the bottom third
compared to other NMP sites.
9.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 9-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 9-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. Carbonyl compounds were sampled for at
AZFL, SYFL, and ORFL. PAHs were sampled for in addition to carbonyl compounds at SKFL.
PMio metals were sampled for at PAFL.
Observations from Table 9-2 include the following:
• For AZFL, SYFL, and ORFL, the sites sampling only carbonyl compounds,
acetaldehyde and formaldehyde were the only two pollutants to fail screens. Among
the carbonyl compounds, only acetaldehyde, formaldehyde, and propionaldehyde
have risk screening values. Propionaldehyde did not fail any screens for these three
sites.
• For SYFL, formaldehyde and acetaldehyde failed the same number of screens and
contributed equally to the total number of failed screens. For AZFL and ORFL,
concentrations of acetaldehyde failed a few less screens than formaldehyde. For all
three sites, formaldehyde failed 100 percent of screens.
• Concentrations of five pollutants failed at least one screen for SKFL (two carbonyl
compounds and three PAHs). Acetaldehyde and formaldehyde failed the most
screens, followed by naphthalene. Together, these three pollutants account for
99 percent of failed screens for SKFL, and thus were identified as pollutants of
interest for this site.
9-13
-------
• Arsenic and nickel are the only PMio metals to fail screens for PAFL, with arsenic
accounting for just less than 95 percent of the failed screens. Thus, both arsenic and
nickel are pollutants of interest for PAFL.
Table 9-2. 2015-2016 Risk-Based Screening Results for the Florida Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Petersburg
, Florida - AZFL
Formaldehyde
0.077
119
119
100.00
51.52
51.52
Acetaldehyde
0.45
112
119
94.12
48.48
100.00
Total
231
238
97.06
Pinellas Park, Florida - SKFL
Formaldehyde
0.077
117
117
100.00
37.03
37.03
Acetaldehyde
0.45
116
117
99.15
36.71
73.73
Naphthalene
0.029
80
118
67.80
25.32
99.05
Fluorene
0.011
2
95
2.11
0.63
99.68
Acenaphthene
0.011
1
110
0.91
0.32
100.00
Total
316
557
56.73
Valrico, Florida - SYFL
Acetaldehyde
0.45
113
113
100.00
50.00
50.00
Formaldehyde
0.077
113
113
100.00
50.00
100.00
Total
226
226
100.00
Winter Park, Florida - ORFL
Formaldehyde
0.077
99
99
100.00
50.51
50.51
Acetaldehyde
0.45
97
99
97.98
49.49
100.00
Total
196
198
98.99
Orlando, Florida - PAFL
Arsenic (PMio)
0.00023
52
53
98.11
94.55
94.55
Nickel (PMio)
0.0021
3
54
5.56
5.45
100.00
Total
55
107
51.40
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
9-14
-------
9.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Florida monitoring sites. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections below) and is limited to the site-specific pollutants of interest. However,
site-specific statistical summaries for all pollutants sampled for at the Florida monitoring sites
are provided in Appendices M, N, and O.
9.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Florida monitoring site, as described in Section 3.1. The quarterly
average concentration of a particular pollutant is simply the average concentration of the
preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated for a given year and where method completeness was greater than or equal to
85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
pollutants of interest for the Florida monitoring sites are presented in Table 9-3, where
applicable. Note that concentrations of the PAHs and metals for SKFL and PAFL are presented
in ng/m3 for ease of viewing. Also note that if a pollutant was not detected in a given calendar
quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects
were factored into the quarterly average concentration.
9-15
-------
Table 9-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Florida Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
Oig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
frig/m3)
Q4
Avg
Oig/m3)
Annual
Average
Oig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
St. Petersburg, Florida - AZFL
Acetaldehyde
58/58/58
0.99
±0.19
0.71
±0.09
NA
0.78
±0.16
0.80
±0.09
61/61/61
1.17
±0.28
1.56
±0.41
1.58
±0.24
1.35
±0.68
1.41
±0.21
Formaldehyde
58/58/58
1.52
±0.26
2.00
±0.28
NA
1.77
±0.37
1.79
±0.18
61/61/61
1.96
±0.33
10.81
±5.01
13.30
±2.55
3.53
±2.15
7.31
± 1.85
Pinellas Park, Florida - SKFL
Acetaldehyde
58/58/58
1.28
±0.22
1.25
±0.13
1.11
±0.26
1.09
±0.13
1.19
±0.09
59/59/59
1.30
±0.27
1.14
±0.12
1.01
±0.16
1.39
±0.28
1.21
±0.11
Formaldehyde
58/58/58
1.80
±0.34
3.18
±0.54
2.63
±0.72
5.90
± 1.32
3.39
±0.56
59/59/59
3.41
±2.31
3.46
±0.97
3.66
±0.62
8.61
±4.37
4.72
± 1.29
Naphthalene1
59/59/59
41.13
± 13.52
32.49
±5.51
39.01
±7.58
33.01
±6.35
36.05
±3.91
59/59/59
59.45
± 25.06
49.89
± 14.32
45.02
± 10.76
41.98
±7.34
49.34
±7.81
Valrico, Florida - SYFL
Acetaldehyde
57/57/57
1.36
±0.29
1.39
±0.22
1.31
±0.19
0.66
±0.11
1.19
±0.13
56/56/56
1.11
±0.21
1.14
±0.12
NA
1.02
±0.15
1.04
±0.09
Formaldehyde
57/57/57
1.58
±0.41
3.08
±0.45
3.35
± 1.09
1.41
±0.24
2.37
±0.37
56/56/56
2.09
±0.68
2.97
±0.40
NA
1.94
±0.24
2.28
±0.25
Winter Park, Florida - ORFL
Acetaldehyde
58/58/58
2.21
±0.36
1.26
±0.21
1.43
±0.21
0.96
±0.21
1.45
±0.17
41/41/41
1.53
±0.30
1.29
±0.17
1.25
±0.21
NS
1.36
±0.14
Formaldehyde
58/58/58
1.82
±0.92
3.00
±0.47
3.01
±0.39
1.74
±0.41
2.37
±0.32
41/41/41
1.95
±0.35
3.08
±0.44
3.20
±0.57
NS
2.71
±0.31
Orlando, Florida - PAFL
Arsenic (PMi0)a
29/29/30
0.43
±0.24
0.44
±0.24
1.03
±0.56
0.54
±0.10
0.62
±0.18
24/23/24
0.56
±0.18
0.49
±0.15
0.51
±0.30
NS
0.52
±0.12
Nickel (PMi,;,)a
30/28/30
4.00
±4.97
0.59
±0.24
0.54
±0.22
0.46
±0.16
1.34
± 1.14
24/23/24
0.52
±0.16
2.81
±4.69
0.75
±0.33
NS
1.36
± 1.47
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Observations for AZFL from Table 9-3 include the following:
• Both acetaldehyde and formaldehyde were detected in every valid sample collected at
AZFL.
• Quarterly average concentrations are not presented for the third quarter of 2015 due
to one too many invalid samples.
• Concentrations of acetaldehyde measured at AZFL range from 0.168 |ig/m3 to
5.72 |ig/m3. Both of these concentrations were measured during the fourth quarter of
2016, explaining, at least in part, the relatively large confidence interval associated
with this quarterly average. The quarterly average concentrations of acetaldehyde for
2016 are greater than the quarterly averages for 2015; none of the quarterly averages
for 2015 are greater than 1 |ig/m3 while all of them for 2016 are. Further, the annual
average for 2016 (1.41 ± 0.21 |ig/m3) is significantly higher than the annual average
for 2015 (0.80 ± 0.09 |ig/m3). A review of the acetaldehyde data collected at AZFL
shows that all but two of the 25 acetaldehyde concentrations greater than 1.5 |ig/m3
were measured during 2016. The difference between the two year's measurements is
even more pronounced in the formaldehyde data.
• Concentrations of formaldehyde measured at AZFL range from 0.341 |ig/m3 to
22.2 |ig/m3. Although not the maximum concentration measured across the program
during the 2015 and 2016 monitoring efforts, the maximum formaldehyde
concentration measured at AZFL ranks third highest. Concentrations of formaldehyde
measured at AZFL account for half of the 30 measurements greater than 15 |ig/m3.
Based on the quarterly averages in Table 9-3, a significant increase in the magnitude
of formaldehyde concentrations occurs during the second quarter of 2016. The second
quarter average concentration for 2016 (10.81 ± 5.01 |ig/m3) is more than five times
the previous quarterly averages. The concentrations remain elevated through the third
quarter, but decrease somewhat for the fourth quarter. Between May 24, 2016 and
October 15, 2016, all but two of the 26 formaldehyde measurements are greater than
10 |ig/m3. A new collection system was installed on October 27, 2016. If the
concentrations from the first three sample days in October 2016 were excluded from
the quarterly average, the fourth quarter average would decrease by more than half.
Observations for SKFL from Table 9-3 include the following:
• Acetaldehyde, formaldehyde, and naphthalene were detected in every valid sample
collected at SKFL.
• Concentrations of acetaldehyde measured at SKFL range from 0.0542 |ig/m3 to
2.24 |ig/m3. The minimum acetaldehyde concentration measured at SKFL is an order
of magnitude less than the next lowest acetaldehyde concentration measured at this
site and among only a few less than 0.1 |ig/m3 measured across the program. Less
than 0.4 |ig/m3 separates the quarterly average concentrations of acetaldehyde for
SKFL and there is little difference between the 2015 (1.19 ± 0.09 |ig/m3) and 2016
(1.21 ±0.11 |ig/m3) annual average concentrations of acetaldehyde.
• Concentrations of formaldehyde measured at SKFL range from 0.0948 |ig/m3 to
19.1 |ig/m3, with the minimum formaldehyde concentration measured on the same
9-17
-------
day as the minimum acetaldehyde concentration (September 27, 2016). For 2015,
quarterly average concentrations are highly variable, with the highest quarterly
average concentration (fourth quarter, 5.90 ± 1.32 |ig/m3) three times higher than the
lowest quarterly average concentration (first quarter, 1.80 ± 0.34 |ig/m3). For 2015,
10 of the 11 formaldehyde concentrations greater than 6 |ig/m3 were measured
between late October and the end of the year. For 2016, most of the higher
formaldehyde concentrations were also measured during the fourth quarter of the
year. Six of the seven formaldehyde concentrations greater than 10 |ig/m3 were
measured in November or December 2016. All seven formaldehyde concentrations
greater than 10 |ig/m3 measured at SKFL were measured in 2016, one in February
and the other six in November or December. The variability in the formaldehyde
concentrations measured is reflected the confidence intervals shown, particularly for
the first and fourth quarters of 2016. This can also be seen in the annual averages.
• Concentrations of naphthalene measured at SKFL range from 10.7 ng/m3 to
190 ng/m3. The 2016 annual average naphthalene concentration is significantly higher
than the 2015 annual average concentration; the quarterly average concentrations for
these years reflect this as well. All four naphthalene concentrations greater than
100 ng/m3 were measured in 2016, three during the first quarter and one during the
second; further, 15 of the 16 highest naphthalene concentrations measured at SKFL
(those greater than 70 ng/m3) were measured in 2016. The quarterly average
naphthalene concentrations reflect considerable variability, as indicated in the
confidence intervals, particularly those for the first quarter of 2015, and most of those
for 2016.
Observations for SYFL from Table 9-3 include the following:
• Acetaldehyde and formaldehyde were detected in every valid sample collected at
SYFL.
• Concentrations of acetaldehyde measured at SYFL range from 0.468 |ig/m3 to
2.44 |ig/m3. For 2015, the first three quarterly average concentrations of acetaldehyde
are similar to each other, but the fourth quarter average is significantly lower than the
others; this quarterly average is also significantly lower than the available quarterly
averages for 2016. Twelve (of the 13) acetaldehyde concentrations measured during
the fourth quarter of 2015 are less than 1 |ig/m3; between two (third quarter) and four
(first quarter) acetaldehyde concentrations less than 1 |ig/m3 were measured during
the other calendar quarters for 2015. For 2016, the available quarterly average
concentrations are fairly similar to each other. Excluding the fourth quarter average
for 2015, the quarterly averages and annual average for 2016 are slightly lower than
those calculated for 2015, although the differences are not statistically significant.
• Concentrations of formaldehyde measured at SYFL range from 0.769 |ig/m3 to
9.70 |ig/m3. For 2015, the second and third quarter average concentrations of
formaldehyde are significantly higher than the first and fourth quarter averages. All
but one of the 16 formaldehyde concentrations greater than 3 |ig/m3 measured in 2015
were measured between April and September, including the maximum concentration,
which is considerably higher than the next highest concentration measured at SYFL
(5.51 |ig/m3). Fewer concentrations greater than 3 |ig/m3 were measured at SYFL in
9-18
-------
2016 (seven), with most (four) of these measured during the second quarter of 2016.
Similar to acetaldehyde, the annual average concentration of formaldehyde for 2016
is slightly lower than the annual average calculated for 2015, although the difference
is not statistically significant.
• Quarterly average concentrations for the third quarter of 2016 could not be calculated
due to a series of invalid samples at the end of July and beginning of August 2016.
Observations for ORFL from Table 9-3 include the following:
• Acetaldehyde and formaldehyde were detected in every valid sample collected at
ORFL.
• Concentrations of acetaldehyde measured at ORFL span an order of magnitude,
ranging from 0.356 |ig/m3 to 3.72 |ig/m3. The four highest acetaldehyde
concentrations measured at ORFL, those greater than 2.75 |ig/m3, were all measured
in January, with three in 2015 and one in 2016. Twelve of the 15 acetaldehyde
concentrations greater than 2 |ig/m3 were measured during the first quarter of either
year, with nine measured during the first quarter of 2015 and three measured during
the first quarter of 2016. This is reflected in the quarterly averages shown for the first
quarter, particularly for 2015. The annual average concentrations of acetaldehyde for
ORFL vary by less than 0.1 |ig/m3.
• Concentrations of formaldehyde measured at ORFL range from 0.582 |ig/m3 to
7.43 |ig/m3. For 2015, the first and fourth quarter average concentrations are less than
the second and third quarter averages, although the confidence interval for the first
quarter is more than twice the magnitude of the other confidence intervals. All seven
formaldehyde concentrations less than 1 |ig/m3 measured over the 2 years of
sampling were measured during either the first (4) or fourth (3) quarters of 2015. The
maximum formaldehyde concentration was also measured during the first quarter of
2015; the combination of the lower concentrations and the maximum concentration
explains the relatively large confidence interval shown for the first quarter of 2015.
• Carbonyl compound sampling at ORFL was discontinued after the third quarter of
2016.
Observations for PAFL from Table 9-3 include the following:
• PAFL is the only Florida monitoring site that did not sample carbonyl compounds.
• PMio metals were sampled for at PAFL on a l-in-12 day schedule, while the other
Florida sites sampled on a l-in-6 day schedule, thus, yielding roughly half the number
of samples as the remaining sites. Sampling at PAFL was discontinued after the third
quarter of 2016.
• Concentrations of arsenic measured at PAFL range from 0.019 ng/m3 to 2.77 ng/m3.
With the exception of the third quarter of 2015, the quarterly average concentrations
of arsenic do not vary considerably; the third quarter average for 2015 is roughly
twice the other quarterly averages shown in Table 9-3. A review of the data shows
that the maximum arsenic concentration measured at PAFL was measured during this
9-19
-------
calendar quarter (on August 22, 2015). The third quarter of 2015 is the only calendar
quarter in which more than one arsenic concentration greater than 1 ng/m3 was
measured at PAFL.
• Concentrations of nickel measured at PAFL span two orders of magnitude, ranging
0.170 ng/m3 to 18.0 ng/m3, including three nickel measurements greater than
10 ng/m3 (the most of any NMP site). The first quarter average for 2015 and the
second quarter average for 2016 are considerably higher than the other quarterly
averages, each with a confidence interval greater than the average itself. Two
relatively high nickel concentrations (14.2 ng/m3 and 11.2 ng/m3) were measured
during the first quarter of 2015, and another was measured during the second quarter
of 2016 (18.0 ng/m3). The next highest nickel concentration measured at PAFL is an
order of magnitude less. The effects of these higher measurements are also reflected
in the annual averages for each year, with confidence intervals similar or greater in
magnitude than the averages themselves.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Florida
sites from those tables include the following:
• None of the Florida monitoring sites appear in Table 4-11 for acetaldehyde. Among
the Florida sites, annual averages of acetaldehyde range from 0.80 ± 0.09 |ig/m3
(AZFL, 2015) to 1.45 ± 0.17 |ig/m3 (ORFL, 2015).
• AZFL annual average concentration for 2016 is the second highest annual average of
formaldehyde among NMP sites sampling carbonyl compounds; SKFL's annual
average for 2016 ranks fifth highest. Interestingly, AZFL has both the lowest and
highest annual averages of this pollutant among the Florida sites: 1.79 ± 0.18 |ig/m3
for 2015 and 7.31 ± 1.85 |ig/m3 for 2016.
• SKFL is not among the sites with the highest annual average concentrations of
naphthalene among NMP sites sampling this pollutant, as shown in Table 4-12.
• PAFL is not among the sites with the highest annual average concentrations of
arsenic among NMP sites sampling this pollutant, as shown in Table 4-13.
9.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 9-3 for each of the Florida monitoring sites. Figures 9-9 through 9-13 overlay the sites'
minimum, annual average, and maximum concentrations for each year onto the program-level
minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.4.2.1, and are discussed below. If an annual average concentration could
not be calculated, the range of concentrations is still provided in the figures that follow.
9-20
-------
Figure 9-9. Program vs. Site-Specific Average Acetaldehyde Concentrations
m
m-
n
| ,
H 1 1 1 1 1 1 1 1
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
0
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 9-9 presents the box plots for acetaldehyde for the four Florida sites sampling
carbonyl compounds and shows the following:
• Among the Florida sites, the maximum acetaldehyde concentration was measured at
AZFL (2016), but this measurement is still about one-third the maximum
concentration measured across the program.
• All of the annual average concentrations of acetaldehyde for the Florida sites are less
than the program-level average concentration of 1.67 |ig/m3.
• The acetaldehyde concentrations measured at AZFL exhibit the largest differences
between the two years. For 2015, the entire range of acetaldehyde concentrations
measured is less than the program-level average concentration, with an annual
average less than the first quartile across the program. For 2016, the range of
measurements is much larger, with the annual average similar to the program-level
median.
• The minimum acetaldehyde concentration measured at SKFL in 2016 is the minimum
concentration measured among the Florida sites. However, if this concentration is
9-21
-------
excluded, the range of concentrations measured at SKFL is similar across both years.
A similar range of concentrations was measured at SYFL across both years. The
annual averages for these two sites fall between the program-level first quartile and
median concentrations.
• A slightly larger range of acetaldehyde concentrations was measured at ORFL in
2015 compared to 2016; these years' annual average concentrations fall on either side
of the program-level median concentration (1.43 |ig/m3).
Figure 9-10. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
¦ ¦ a
S
,
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 9-10 presents the box plot for arsenic for PAFL and shows the following:
• The range of arsenic concentrations measured at PAFL in 2015 is approximately
twice the range measured in 2016. The maximum concentration measured at PAFL is
considerably less than the maximum concentration measured across the program.
• Although the minimum arsenic concentration measured each year at PAFL appears
similar, the minimum concentration for 2015 is a non-detect while the minimum
concentration for 2016 is 0.019 ng/m3; the minimum concentration measured at PAFL
in 2016 is among one of the lower arsenic concentrations measured across the
program.
• Both annual average concentrations of arsenic for PAFL are less than the program-
level average concentration of 0.70 ng/m3. The annual average concentration for 2016
is also just less than the program-level median concentration of 0.55 ng/m3.
9-22
-------
Figure 9-11. Program vs. Site-Specific Average Formaldehyde Concentrations
^ ' I
N ¦
N—
0 3 6 9 12 15 18 21 24 27
Concentration (jug/rn3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
9
o
Figure 9-11 presents the box plots for the four Florida sites sampling carbonyl
compounds and shows the following:
• While the maximum formaldehyde concentrations measured at AZFL and SKFL in
2016 are considerably higher than the maximum concentrations measured at the other
Florida sites, they are still not the highest concentrations measured across the
program (two concentrations greater than 25 |ig/m3 were measured at BTUT).
• AZFL has the largest disparity in the formaldehyde concentrations measured between
the two years of sampling. The annual average concentration for 2016 is more than
four times greater than the annual average for 2015. The annual average
formaldehyde concentration for AZFL for 2015 is just greater than the program-level
first quartile, while the annual average for 2016 is nearly twice the program-level
third quartile.
• Formaldehyde concentrations measured at SKFL also vary considerably between
2015 and 2016. While more than 1 |ig/m3 separates this site's annual averages, both
are greater than the program-level average concentration of 3.05 |ig/m3.
9-23
-------
• For SYFL and ORFL, the range of formaldehyde concentrations measured in 2015 is
greater than the range measured in 2016. SYFL's annual average concentrations of
formaldehyde are both just less than the program-level median concentration. Despite
the larger range of concentrations measure at ORFL in 2015, the annual average for
2015 is less than the annual average for 2016, with both less than the program-level
average concentration.
Figure 9-12. Program vs. Site-Specific Average Naphthalene Concentrations
200 250
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4thQuartile Avera
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
e o
Figure 9-12 presents the box plot for naphthalene for SKFL and shows the following:
• The range of naphthalene concentrations measured at SKFL in 2016 is more than
twice the range measured in 2015. The maximum concentration measured at SKFL
across both years of sampling is considerably less than the maximum concentration
measured across the program (403 ng/m3).
• Both annual average concentrations of naphthalene for SKFL are less than the
program-level average concentration (61.2 ng/m3); the annual average for 2016 is
similar to the program-level median concentration (48.9 ng/m3), while the annual
average for 2015 falls between the program-level first quartile and median
concentration.
Figure 9-13. Program vs. Site-Specific Average Nickel (PMio) Concentrations
L
r — _j
,1
Program Max Concentration
= 69.5 ng/m3
! r
r
0 3 6 9 12 15 18 21 24
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
9-24
-------
Figure 9-13 presents the box plot for nickel for PAFL and shows the following:
• The program-level maximum nickel (PMio) concentration (69.5 ng/m3) is not shown
directly on the box plot in Figure 9-13 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plot has been reduced to 24 ng/m3.
• The maximum nickel concentration measured at PAFL in 2016 (18.0 ng/m3) is
somewhat higher than the maximum concentration measured in 2015 (14.2 ng/m3).
• The annual average nickel concentrations for PAFL are similar to each other, both of
which are greater than the program-level average concentration and third quartile.
• Non-detects of nickel were not measured at PAFL.
9.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
AZFL, ORFL, SKFL, and SYFL have sampled carbonyl compounds under the NMP for at least
5 consecutive years; in addition, sampling for PAHs at SKFL and PMio metals at PAFL began in
2008. Thus, Figures 9-14 through 9-24 present the 1-year statistical metrics for each of the
pollutants of interest for the Florida monitoring sites. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began or ended
mid-year, a minimum of 6 months of sampling is required for inclusion in the trends analysis; in
these cases, a 1-year average concentration is not provided, although the range and percentiles
are still presented.
9-25
-------
Figure 9-14. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
AZFL
j
L
l A A
<
¦
L
7
? *
J 1
^ u
<
r i
*
a
>-
i.—
"¦<
<
i
A
>
'¦<
..<
>•••• .
1
- x ft A
>
H
r
J 1
-l
H
r L
r
J '""a;""
^ J]
r
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
Observations from Figure 9-14 for acetaldehyde concentrations measured at AZFL
include the following:
• Carbonyl compounds have been measured at AZFL under the NMP since 2001,
making this site one of the longest running NMP sites.
• The maximum acetaldehyde concentration was measured in 2010 (8.09 |ig/m3),
although similar concentrations were also measured in 2003 (8.00 |ig/m3) and 2009
(7.74 |ig/m3).
• The 1-year average and median concentrations did not change significantly during the
first 2 years of sampling, although the range of measurements is twice as large for
2001 compared to 2002. The 1-year average and median concentrations increased
significantly from 2002 to 2003, remained elevated through 2004, then began
decreasing significantly, a trend that continued through 2008.
• The 1-year average and median concentrations began to increase again in 2009.
Nearly all of the statistical parameters exhibit this increase and the trend continued
into 2010. The 95th percentile more than doubled from 2008 to 2009, and the 1-year
average and median concentrations exhibit increases slightly less in magnitude.
• A significant decrease is shown for 2011 and continued into 2012, despite the
increase in the maximum concentration measured in 2012. Slight increases in the
central tendency statistics are shown for 2013, with a return to 2012 levels for 2014.
9-26
-------
Additional decreases are shown for 2015, when all of the statistical parameters
(except the minimum) are at a minimum over the years of sampling; 2015 is the only
year in which both central tendency parameters are less than 1 |ig/m3.
• A significant increase in the 1-year average and median concentrations is shown for
2016, returning to near 2014 levels.
Figure 9-15. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
AZFL
I
I fy
Oi
* ?
I
1/
T
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
O 95th Percentile
Observations from Figure 9-15 for formaldehyde concentrations measured at AZFL
include the following:
• The maximum formaldehyde concentration (22.2 |ig/m3) was measured in 2016,
along with three other measurements greater than 20 |ig/m3; the 12 highest
formaldehyde concentrations were all measured in 2016. Twenty-five formaldehyde
concentrations greater than 10 |ig/m3 have been measured at AZFL, one each in 2001
and 2014, and 23 in 2016.
• Slight changes in the 1-year average and median formaldehyde concentrations are
shown between 2001 and 2002, despite the difference in the range of concentrations
measured. These central tendency parameters decreased significantly from 2002 to
2003. The decreasing trend continued through 2004, after which an increasing trend is
shown, which lasted through 2008. A second significant decrease is shown from 2008
to 2009 and into 2010 (although the median concentration increased for 2010). Little
change is shown for the next 3 years of sampling.
9-27
-------
• Each of the statistical parameters exhibits at least a slight increase from 2013 to 2014,
though the increase in the maximum concentration measured in 2014 is the most
apparent. If the maximum concentration was removed from the calculation, the 1-year
average concentration would decrease only slightly, and the other parameters would
change little. Five concentrations measured in 2013 were less than the minimum
concentration measured in 2014. On the other end of the concentration range, the
number of formaldehyde concentrations greater than 2.5 |ig/m3 doubled from 2013
(8) to 2014 (16). Thus, the slight increases shown for 2014 were not solely
attributable to the maximum concentration.
• Even though the majority of concentrations measured in 2015 fell into a similar range
as those measured in 2014, as indicated by the 5th and 95th percentiles, both central
tendency parameters exhibit slight decreases. This would be true even if the
maximum concentration measured in 2014 was excluded from the calculations.
• With the exception of the minimum concentration, all of the statistical parameters
exhibit increases for 2016. The maximum concentration and 95th percentile exhibit
five- and six-fold increases, respectively, and the 1-year average concentration
increased more than four times from 2015 to 2016. As discussed in the previous
section, a new collection system was installed at this site in the latter part of 2016,
after which higher formaldehyde concentrations were not measured.
Figure 9-16. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SKFL
I
A.
T
o
ri-,
£
n
6
LjJ
V
2007 2008
2010 2011
Year
20141 2015 2016
O 5th Percentile
— Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because some samples collected from July to Sept 2014 were
invalidated.
9-28
-------
Observations from Figure 9-16 for acetaldehyde concentrations measured at SKFL
include the following:
• Sampling for carbonyl compounds under the NMP began at SKFL in late July 2004.
Because this represents less than half of the sampling year, Figure 9-16 excludes data
from 2004.
• The maximum acetaldehyde concentration shown was measured at SKFL in
2010 (10.3 |ig/m3), as were the third, fourth, and fifth highest concentrations of
acetaldehyde. Of the 18 acetaldehyde concentrations greater than 5 |ig/m3, 11 were
measured in 2010.
• Even though the range of concentrations measured decreased by more than half from
2005 to 2006, the change in the 1-year average concentration is not statistically
significant. After 2006, the 1-year average acetaldehyde concentration increased
steadily each year, reaching a maximum in 2010. Most of the statistical parameters
are at a maximum for 2010. A significant decrease is shown for 2011 and continued
into 2012. Although the range of concentrations measured decreased by half for 2013,
the 1-year average concentration changed little.
• Although an annual average concentration could not be calculated for 2014, the
median concentration shown for 2014 is at a minimum for 2014, and is less than
1 |ig/m3 for the first time since the first full year of sampling.
• The smallest range of acetaldehyde concentrations was measured in 2015 and 1-year
average concentration is at a minimum over the period of sampling. Little change is
shown in this parameter for 2016, although the lowest acetaldehyde concentration
measured at SKFL since the onset of sampling was measured in 2016. This
measurement is an order of magnitude less than the minimum concentration measured
in 2015.
9-29
-------
Figure 9-17. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SKFL
Maximum
Concentration for
2005 is 91.7 (J.g/m3
LjJ
X
a &
x
X
o
2006 2007 2008 2009
2010 2011 2012 2013
Year
20141 2015
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because some samples collected from July to Sept 2014 were
invalidated.
Observations from Figure 9-17 for formaldehyde concentrations measured at SKFL
include the following:
• The maximum formaldehyde concentration was measured at SKFL on July 9, 2005
(91.7 |ig/m3). The next seven highest formaldehyde concentrations were measured at
SKFL in 2016, and range from roughly 15 |ig/m3 to 20 |ig/m3. Only one other
formaldehyde concentration greater than 10 |ig/m3 has been measured at SKFL
(11.0 |ig/m3 in 2012).
• For 2005, the 1-year average concentration is greater than the 95th percentile,
reflecting the effect that an outlier can have on statistical measurements. The second
highest concentration measured in 2005 was 4.07 |ig/m3.
• The 1-year average concentration exhibits an overall decreasing trend through 2010.
The range of measurements is at a minimum for 2010 and the 1-year average and
median concentrations are nearly equivalent, reflecting little variability in the
measurements.
• All of the statistical parameters increased from 2010 to 2011 and again for 2012. The
5th percentile for 2012 is greater than several of the central tendency statistics for
several of the previous years.
9-30
-------
• All of the statistical parameters exhibit a decrease for 2013, with the concentration
profile returning to near 2011 levels.
• Although a 1-year average could not be calculated for 2014, the median concentration
exhibits a slight increase for 2014, despite the smaller range of formaldehyde
measurements measured in 2014.
• Significant increases are shown for 2015 and 2016, particularly in the statistical
metrics representing the upper end of the concentration range. The 1-year average
concentration increased three-fold from 2013 to 2016. The median concentration also
exhibits an increasing trend between 2013 and 2016, and is at a maximum for 2016.
Figure 9-18. Yearly Statistical Metrics for Naphthalene Concentrations Measured at SKFL
o
•o...
2012
Year
I
O 5th Percentile
— Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 9-18 for naphthalene concentrations measured at SKFL include
the following:
• Sampling for PAHs began at SKFL under the NMP on March 1, 2008.
• The maximum naphthalene concentration was measured at SKFL in 2012
(435 ng/m3). Three additional measurements greater than 300 ng/m3 have been
measured at SKFL (one each in 2008, 2010, and 2013).
• The range within which the majority of naphthalene concentrations fall changed little
through 2011. An increase is shown for 2012 as this year has the greatest number of
measurements greater than 200 ng/m3 (seven). This increase is followed by a
9-31
-------
considerable decrease for 2013, which has the fewest measurements greater than
200 ng/m3 (one) since the onset of sampling PAHs at SKFL.
• A decreasing trend in naphthalene concentrations is shown after 2012, with both the
1-year average and median concentrations at a minimum for 2015. Concentrations of
naphthalene greater than 200 ng/m3 were not measured at SKFL after 2013, and
concentrations greater than 100 ng/m3 were not measured in 2015, the only year for
which this is true.
• A significant increase in concentrations is shown for 2016, which has a range of
measurements similar to 2014. Although four concentrations measured in 2016 are
greater than the maximum concentration measured in 2015, and both the 1-year
average and median concentrations increased by at least 10 ng/m3 from 2015 to 2016,
the 1-year average concentration is still less than 50 ng/m3 for only the second time
since the onset of sampling of PAHs at SKFL.
Figure 9-19. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SYFL
„ T
L2J
r-5-, /
£
SL
<>¦¦¦¦.
rh
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
Observations from Figure 9-19 for acetaldehyde concentrations measured at SYFL
include the following:
• Carbonyl compounds have been measured at SYFL under the NMP since January
2004.
• Two acetaldehyde concentrations greater than 10 |ig/m3 were measured at SYFL in
January 2007 (15.3 |ig/m3 and 12.6 |ig/m3). The next highest concentration, measured
9-32
-------
in 2008, is roughly half as high (7.29 |ig/m3). Only one additional acetaldehyde
concentration greater than 6 |ig/m3 has been measured at SYFL (6.81 |ig/m3 in 2004).
• After a significant decreasing trend through 2006, all of the statistical parameters
increased for 2007. Even if the two measurements of acetaldehyde discussed above
were removed from the calculation, the 1-year average concentration for 2007 would
still be more than 1 |ig/m3 greater than the next highest 1-year average concentration.
2007 has the greatest number of acetaldehyde concentrations greater than 3 |ig/m3
(16), while every other year of sampling has three or less. Thus, it is not just the two
highest measurements driving this 1-year average concentration.
• All of the statistical parameters exhibit a decrease from 2007 to 2008; the 1-year
average concentration decreased significantly for 2008. Little change is shown in this
parameter for 2009, followed by an undulating pattern through 2013 and a steady,
albeit slight, decreasing trend through 2016, when the range of acetaldehyde
concentrations measured is at a minimum.
• Between 2008 and 2016, the 1-year average concentration has fluctuated between
1 |ig/m3 and 1.5 |ig/m3. With the exception of 2007, less than 0.6 |ig/m3 separates all
of the 1-year averages shown over the period of sampling.
Figure 9-20. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SYFL
Maximum
Concentration for
2005 is 32.5 (J.g/m3
An
1
o-
H
L2-
4L
i-J-i
I
¦o
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
9-33
-------
Observations from Figure 9-20 for formaldehyde concentrations measured at SYFL
include the following:
• The maximum formaldehyde concentration was measured at SYFL in 2005
(32.5 |ig/m3) and is nearly twice the next highest concentrations (17.8 |ig/m3,
measured in 2007, and 17.1 |ig/m3, measured in 2008). In all, seven formaldehyde
concentrations greater than 10 |ig/m3 have been measured at SYFL, four in 2007 and
one each in 2005, 2008, and 2010.
• Even though the maximum concentration was measured in 2005, the second highest
concentration measured that year is considerably less (4.17 |ig/m3). The 1-year
average concentration exhibits a slight increase from 2004 to 2005 while the median
concentration decreased slightly; if the outlier was excluded from the calculation, the
1-year average concentration would exhibit a decrease for 2005 while the median
would change little.
• The slight decrease in most of the statistical parameters for 2006 is followed by large
increases for 2007. In particular, the 95th percentile increased five-fold and the 1-year
average doubled from 2006 to 2007. As discussed in the previous bullet, 2007 is the
only year in which multiple formaldehyde concentrations greater than 10 |ig/m3 were
measured.
• Although similar maximum concentrations were measured in 2007 and 2008, the
change in the 95th percentile between these two years stands out. The second highest
concentration measured in 2008 (4.17 |ig/m3) is considerably less than the maximum
concentration measured (17.1 |ig/m3).
• The 1-year average concentrations of formaldehyde vary little during the five years
between 2008 and 2012, ranging from 2.23 |ig/m3 (2012) to 2.75 |ig/m3 (2010).
• The range of formaldehyde concentrations measured is at a minimum for 2013, and
the 1-year average concentration for 2013 is the lowest since 2006; 2006 and 2013 are
the only two years with 1-year average concentrations less than 2 |ig/m3.
• The 1-year average concentrations for 2014, 2015, and 2016 fall into a similar range
as those calculated for the years prior to 2013.
9-34
-------
Figure 9-21. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
ORFL
12.0
E
3
c
.2 g Q
T
c
m
c
o
u
4.0
1
T
2.0
£
£
n
•o
1
&
'"2"
o
-2
—
LI"
i 1
LjJ
1
-
o
o
1
^
r
L2J
o
'V
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 20162
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because sampling under the NMP stopped in September 2016.
Observations from Figure 9-21 for acetaldehyde concentrations measured at ORFL
include the following:
• Sampling for carbonyl compounds under the NMP began at ORFL in April 2003. A
1-year average concentration is not presented for 2003 because a full year's worth of
data is not available, although the range of measurements is provided. This is also
true for 2016, when sampling was discontinued at the end of September (the 10-
month criteria discussed in Section 3.4.2.2 was not met).
• The maximum acetaldehyde concentration was measured in 2006 (9.55 |ig/m3); two
additional acetaldehyde concentrations greater than 6 |ig/m3 were measured in 2007
and 2008. Acetaldehyde concentrations of at least 5 |ig/m3 were measured in half the
years of sampling, including each year between 2004 and 2009.
• Between 2004 and 2007, the 1-year average concentrations have an undulating
pattern, with a higher year followed by a lower year. Between 2007 and 2009, little
change is shown in the 1-year average concentrations, when these averages varied by
about 0.1 |ig/m3, despite the considerable increase in the median shown for 2009. The
undulating pattern returns between 2009 and 2012.
9-35
-------
• The 1-year average concentration is at a minimum for 2012 (1.08 |ig/m3) and
represents a significant decrease from 2011 and most of the previous years of
sampling. The median concentration decreased by almost half from 2011 to 2012.
Only one concentration less than 1 |ig/m3 was measured in 2011 compared to 38 for
2012.
• All of the statistical metrics exhibit increases for 2013, with most exhibiting
additional increases for 2014. The 1-year average concentration shown for 2014 is the
highest average since 2006, when the maximum concentration was measured.
• All of the statistical metrics exhibit decreases for 2015. The range of concentrations
measured decreases further for 2016, with the smallest range of measurements since
the onset of sampling. Sampling was discontinued at the end of September 2016, and
thus, the concentration profile shown represents only 9 months of sampling.
Figure 9-22. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
ORFL
E
io.o
f
T
r5n
LjJ
rS",
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile Aver e
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because sampling under the NMP stopped in September 2016.
Observations from Figure 9-22 for formaldehyde concentrations measured at ORFL
include the following:
• The maximum formaldehyde concentration was measured in 2007 (16.1 |ig/m3), on
the same day as the second highest acetaldehyde concentration (September 21, 2007).
Formaldehyde concentrations greater than 10 |ig/m3 were also measured in 2005
(two) and 2008 (one).
9-36
-------
• The 1-year average concentrations exhibit an overall decreasing trend through 2011,
starting at 3.26 |ig/m3 for 2004 and decreasing to 1.89 |ig/m3 by 2011. The statistical
metrics for 2007 are the exception to this pattern. However, if the maximum
concentration measured in 2007 was excluded from the calculation, the 1-year
average concentration would exhibit a consistent decreasing trend across the years
through 2011. The 1-year average concentrations hover around 2 |ig/m3 for several
years of sampling at ORFL.
• The 1-year average concentration exhibits slight increases for 2014 and 2015.
Although a 1-year average is not presented for 2016, the median concentration
calculated for 2016 is at its highest in 12 years.
Figure 9-23. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
PAFL
_L
o
2—
—©—
~
O"
o
2012
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP stopped in September 2016.
Observations from Figure 9-23 for arsenic concentrations measured at PAFL include the
following:
• Sampling for PMio metals under the NMP began at PAFL in January 2008; metals
sampling occurred on a l-in-12 day sampling schedule at this site.
• Two of three highest arsenic concentrations, those greater than 3 ng/m3, were
measured at PAFL in 2012, and the other was measured in 2013. Seven arsenic
concentrations greater than 2 ng/m3 have been measured at PAFL, four in 2012 and
one each in 2013, 2014, and 2015.
9-37
-------
• The 1-year average concentration decreases through 2010, increases slightly for 2011,
and increases even more for 2012, when the largest range of arsenic concentrations
was measured. The 1-year average concentration for 2012 is the only one greater than
1 ng/m3.
• The range of concentrations measured decreases for 2013 and again for 2014. Despite
this decrease, both the 1-year average and median concentrations exhibit increase for
2014. Fewer concentrations less than 0.5 ng/m3 were measured in 2014, which
accounted for about half of the measurements in 2013 and about a quarter of the
measurements for 2014. The difference between the 1-year average and median
concentrations is at minimum for 2014, indicating reduced variability in the central
tendency for 2014.
• With the exception of the maximum concentration, most of the statistical parameters
decreased from 2014 to 2015. The only non-detect of arsenic measured at PAFL was
measured in 2015.
• The smallest range of arsenic concentrations since 2010 was measured in 2016. The
median concentration exhibits a decrease for 2016 and is less than 0.5 ng/m3 for only
the second time since 2011.
Figure 9-24. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at
PAFL
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP stopped in September 2016.
9-38
-------
Observations from Figure 9-24 for nickel concentrations measured at PAFL include the
following:
• The maximum nickel concentration was measured at PAFL in 2016 (18.0 ng/m3),
with three additional concentrations greater than 10 ng/m3 also measured in 2008 (1)
and 2015 (two). Nickel concentrations greater than 3.5 ng/m3 were not measured at
this site during any other years of sampling. Less than 1.5 ng/m3 separates the
minimum and maximum concentrations measured in 2009 and between 2011 and
2014.
• The 1-year average concentration decreased by almost half from 2008 to 2009, as
concentrations greater than 2 ng/m3 were not measured in 2009. The 1-year average
concentration increased slightly for 2010 as a few concentrations greater than 2 ng/m3
were again measured. The 1-year average concentration decreased again for 2011, as
an even smaller concentration range was measured. The concentration profiles for
2011, 2012, and 2013 are similar to each other. A slight increase in the 1-year average
concentration is shown for 2014.
• The range of concentrations increased by an order of magnitude for 2015, at both
ends of the concentration range. The two highest nickel concentrations (14.2 ng/m3
and 11.2 ng/m3) are an order of magnitude greater than the next two highest
concentrations (1.07 ng/m3 and 1.01 ng/m3) measured in 2015. Eighteen
concentrations measured in 2015 are less than the minimum concentration measured
in 2014. This represents 75 percent of the nickel measurements, explaining the
50 percent decrease shown in the median concentration, despite the larger
concentration range for 2015. The difference between the 1-year average and median
concentrations for 2015 is at a maximum, indicating a high level of variability in the
measurements for this year.
• The range of concentrations expands further for 2016, although this is primarily a
result of the maximum concentration measured. Two orders of magnitude separate the
minimum (0.17 ng/m3) and maximum (18.0 ng/m3) concentration measured in 2016;
this maximum concentration is the only one greater than 2 ng/m3 measured in 2016.
9.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Florida monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
9.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Florida sites, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
9-39
-------
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 9-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the Florida sites from Table 9-4 include the following:
• For the four sites at which carbonyl compounds were sampled for, the annual average
concentration for each year for formaldehyde is greater than the annual average for
acetaldehyde.
• Formaldehyde has the highest cancer risk approximations among the various
pollutants of interest for the Florida sites. These cancer risk approximations range
from 23.28 in-a-million (AZFL, 2015) to 95.04 in-a-million (AZFL, 2016).
• The cancer risk approximations for acetaldehyde are an order of magnitude less than
the cancer risk approximations for formaldehyde, ranging from 1.76 in-a-million
(AZFL, 2015) to 3.18 in-a-million (ORFL, 2015).
• For SKFL, the naphthalene cancer risk approximations for 2015 and 2016 are 1.23 in-
a-million and 1.68 in-a-million, respectively.
• For PAFL, cancer risk approximations for arsenic (2.66 in-a-million for 2015 and
2.24 in-a-million for 2016) are greater than the cancer risk approximations for nickel
(0.64 in-a-million for 2015 and 0.65 in-a-million for 2016).
• All of the noncancer hazard approximations for the site-specific pollutants of interest
are less than 1.0, indicating that no adverse noncancer health effects are expected
from these individual pollutants. The highest noncancer hazard approximation was
calculated for formaldehyde (0.75), based on the 2016 annual average concentration
for AZFL. This is the second highest noncancer hazard approximation calculated
across the program.
9-40
-------
Table 9-4. Risk Approximations for the Florida Monitoring Sites
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(Ug/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HO)
St. Petersburg, Florida - AZFL
Acetaldehyde
0.0000022
0.009
58/58
0.80
±0.09
1.76
0.09
61/61
1.41
±0.21
3.11
0.16
Formaldehyde
0.000013
0.0098
58/58
1.79
±0.18
23.28
0.18
61/61
7.31
± 1.85
95.04
0.75
Pinellas Park, Florida - SKFL
Acetaldehyde
0.0000022
0.009
58/58
1.19
±0.09
2.61
0.13
59/59
1.21
±0.11
2.66
0.13
Formaldehyde
0.000013
0.0098
58/58
3.39
±0.56
44.04
0.35
59/59
4.72
± 1.29
61.38
0.48
Naphthalene3
0.000034
0.003
59/59
36.05
±3.91
1.23
0.01
59/59
49.34
±7.81
1.68
0.02
Valrico, Florida - SYFL
Acetaldehyde
0.0000022
0.009
57/57
1.19
±0.13
2.63
0.13
56/56
1.04
±0.09
2.28
0.12
Formaldehyde
0.000013
0.0098
57/57
2.37
±0.37
30.81
0.24
56/56
2.28
±0.25
29.70
0.23
Winter Park, Florida - ORFL
Acetaldehyde
0.0000022
0.009
58/58
1.45
±0.17
3.18
0.16
41/41
1.36
±0.14
3.00
0.15
Formaldehyde
0.000013
0.0098
58/58
2.37
±0.32
30.81
0.24
41/41
2.71
±0.31
35.19
0.28
Orlando, Florida - PAFL
Arsenic (PMi0)a
0.0043
0.000015
29/30
0.62
±0.18
2.66
0.04
24/24
0.52
±0.12
2.24
0.03
Nickel (PMio)a
0.00048
0.00009
30/30
1.34
± 1.14
0.64
0.01
24/24
1.36
± 1.47
0.65
0.02
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
-------
As an extension of this analysis, pollution roses were created for each of the site-specific
pollutants of interest that have a cancer risk approximation greater than 75 in-a-million and/or a
noncancer hazard approximation greater than 1.0, where applicable. Thus, a pollution rose was
created for AZFL's formaldehyde measurements. A pollution rose is a plot of the ambient
concentration versus the wind direction; the magnitude of the concentration is indicated using
different colored dots and are shown in relation to the average wind direction oriented about a
16-point compass. Thus, high concentrations may be shown in relation to the direction of
potential emissions sources. Hourly wind observations collected at the NWS station at St.
Petersburg/Whitted Airport and obtained from NOAA are used in this analysis and were
averaged (using vector averaging techniques) to compute daily wind direction averages for
comparison to the 24-hour concentration data. This analysis is intended to help identify the
geographical area where the emissions sources of these pollutants may have originated.
Additional information regarding this analysis is also presented in Section 3.4.2.3. Figure 9-25
presents the pollution rose for all 119 formaldehyde concentrations measured at AZFL over the
two-year sampling period.
Figure 9-25. Pollution Rose for Formaldehyde Concentrations Measured at AZFL
360/0
180
O < 1 |ig/m3 O 1-10 |ig/m3 © > 10 |ig/m3
9-42
-------
Observations from Figure 9-25 include the following:
• Most of the formaldehyde concentrations measured at AZFL are less than 5 |ig/m3,
and are shown in pink. These concentrations account for 81 percent of the
measurements shown. The higher measurements collected between May and October
2016 are greater than 10 |ig/m3 and are those shown in blue. Concentrations between
5 |ig/m3 and 10 |ig/m3 were not measured at AZFL.
• Formaldehyde concentrations are shown in relation to a variety of average wind
directions, including the higher measurements, indicating that the samples were
collected on sample days with varying average wind directions. The higher
concentrations are shown in relation to all wind directions, except the south-southeast
(between 135° and 180°).
• The facility map in Figure 9-3 shows that the only point sources within 2.5 miles of
AZFL is located about three-quarters of a mile north of the site.
• The distribution of formaldehyde concentrations on the pollution rose, including the
higher concentrations, indicates that the concentrations measured at AZFL do not
reflect emissions from any one particular source.
9.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 9-5 and 9-6 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 9-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 9-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 9-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 9-4. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 9-5. Table 9-6 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
9-43
-------
noncancer hazard approximations provided in Section 9.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 9-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Pinellas, Hillsborough, and Orange Counties.
• Formaldehyde, ethylene oxide, benzene have the highest toxicity-weighted emissions
for Pinellas County. Formaldehyde, benzene, and 1,3-butadiene have the highest
toxicity-weighted emissions for Hillsborough County. Hexavalent chromium,
formaldehyde, and benzene have the highest toxicity-weighted emissions for Orange
County.
• Seven of the highest emitted pollutants in Pinellas County also have the highest
toxicity-weighted emissions; this is also true for Orange County. Eight of the highest
emitted pollutants in Hillsborough County also have the highest toxicity-weighted
emissions.
• Formaldehyde, which has the highest cancer risk approximations for each of the sites
sampling carbonyl compounds, is one of the highest emitted pollutants in each county
and has one of the highest toxicity-weighted emissions for each county. This is also
true for acetaldehyde for Hillsborough County, but acetaldehyde does not appear
among those pollutants with the highest toxicity-weighted emissions for Pinellas or
Orange Counties (it ranks 11th for both counties).
• Naphthalene, which is a pollutant of interest for SFKL, is one of the highest emitted
pollutants in all three counties and has one of the highest toxicity-weighted emissions
for each county.
• POM, Groups 2b and 2d are also among the highest emitted "pollutants" in all three
counties and appear among the pollutants with the highest toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at SKFL, including
fluorene and acenaphthene, both of which failed screens for SKFL but were not
identified as pollutants of interest.
• Arsenic and nickel are the pollutants of interest for PAFL. Neither metal appears
among the highest emitted in Orange County. Arsenic does not appear among the
pollutants with the highest toxicity-weighted emissions for Orange County, while
nickel ranks ninth.
9-44
-------
Table 9-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Petersburg, Florida (Pinellas County) - AZFL
Benzene
251.18
Formaldehyde
2.45E-03
Formaldehyde
95.04
Formaldehyde
188.23
Ethylene oxide
2.00E-03
Formaldehyde
23.28
Ethylbenzene
165.34
Benzene
1.96E-03
Acetaldehyde
3.11
Acetaldehyde
107.84
1,3-Butadiene
1.20E-03
Acetaldehyde
1.76
1.3 -Butadiene
39.94
Hexavalent Chromium. PM
8.14E-04
Naphthalene
21.80
Naphthalene
7.41E-04
T richloroethylene
4.79
Ethylbenzene
4.13E-04
POM, Group 2b
4.37
POM, Group 2b
3.85E-04
Dichloromethane
3.77
Nickel, PM
3.66E-04
POM, Group 2d
3.42
POM, Group 2d
3.01E-04
Pinellas Park, Florida (Pinellas County) - SKFL
Benzene
251.18
Formaldehyde
2.45E-03
Formaldehyde
61.38
Formaldehyde
188.23
Ethylene oxide
2.00E-03
Formaldehyde
44.04
Ethylbenzene
165.34
Benzene
1.96E-03
Acetaldehyde
2.66
Acetaldehyde
107.84
1,3-Butadiene
1.20E-03
Acetaldehyde
2.61
1,3-Butadiene
39.94
Hexavalent Chromium. PM
8.14E-04
Naphthalene
1.68
Naphthalene
21.80
Naphthalene
7.41E-04
Naphthalene
1.23
T richloroethylene
4.79
Ethylbenzene
4.13E-04
POM, Group 2b
4.37
POM, Group 2b
3.85E-04
Dichloromethane
3.77
Nickel, PM
3.66E-04
POM, Group 2d
3.42
POM, Group 2d
3.01E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 9-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Valrico, Florida (Hillsborough County) - SYFL
Benzene
355.47
Formaldehyde
4.31E-03
Formaldehyde
30.81
Formaldehyde
331.57
Benzene
2.77E-03
Formaldehyde
29.70
Ethylbenzene
233.29
1,3-Butadiene
1.75E-03
Acetaldehyde
2.63
Acetaldehyde
176.93
Arsenic, PM
1.22E-03
Acetaldehyde
2.28
1.3 -Butadiene
58.39
Naphthalene
1.13E-03
Naphthalene
33.25
POM, Group 2b
6.58E-04
POM, Group 2b
7.47
Ethylbenzene
5.83E-04
Methyl tert butyl ether
6.86
Hexavalent Chromium. PM
5.15E-04
POM, Group 2d
5.07
POM, Group 2d
4.46E-04
T richloroethylene
3.84
Acetaldehyde
3.89E-04
Winter Park, Florida (Orange County) - ORFL
Benzene
343.02
Hexavalent Chromium. PM
1.24E-02
Formaldehyde
35.19
Formaldehyde
333.48
Formaldehyde
4.34E-03
Formaldehyde
30.81
Ethylbenzene
232.56
Benzene
2.68E-03
Acetaldehyde
3.18
Acetaldehyde
160.40
1,3-Butadiene
1.63E-03
Acetaldehyde
3.00
1,3-Butadiene
54.32
Naphthalene
1.11E-03
Naphthalene
32.58
Ethylbenzene
5.81E-04
POM, Group 2b
6.32
POM, Group 2b
5.56E-04
T etrachloroethylene
5.45
POM, Group 2d
4.27E-04
T richloroethylene
4.94
Nickel, PM
4.10E-04
POM, Group 2d
4.86
POM, Group 5a
3.81E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 9-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Orlando, Florida (Orange County) - PAFL
Benzene
343.02
Hexavalent Chromium, PM
1.24E-02
Arsenic
2.66
Formaldehyde
333.48
Formaldehyde
4.34E-03
Arsenic
2.24
Ethylbenzene
232.56
Benzene
2.68E-03
Nickel
0.65
Acetaldehyde
160.40
1,3-Butadiene
1.63E-03
Nickel
0.64
1.3 -Butadiene
54.32
Naphthalene
1.11E-03
Naphthalene
32.58
Ethylbenzene
5.81E-04
POM, Group 2b
6.32
POM, Group 2b
5.56E-04
T etrachloroethylene
5.45
POM, Group 2d
4.27E-04
T richloroethylene
4.94
Nickel, PM
4.10E-04
POM, Group 2d
4.86
POM, Group 5a
3.81E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 9-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Florida Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Petersburg, Florida (Pinellas County) - AZFL
Toluene
1,120.26
Acrolein
683,946.60
Formaldehyde
0.75
Xylenes
665.70
1,3-Butadiene
19,971.07
Formaldehyde
0.18
Methanol
658.85
Formaldehyde
19,207.50
Acetaldehyde
0.16
Benzene
251.18
Acetaldehyde
11,982.55
Acetaldehyde
0.09
Ethylene glycol
242.14
Nickel, PM
8,472.67
Hexane
228.78
Benzene
8,372.75
Formaldehyde
188.23
Naphthalene
7,267.81
Ethylbenzene
165.34
Xylenes
6,656.96
Acetaldehyde
107.84
Lead, PM
4,922.16
Styrene
80.67
Antimony, PM
3,500.85
Pinellas Park, Florida (Pinellas County) - SKFL
Toluene
1,120.26
Acrolein
683,946.60
Formaldehyde
0.48
Xylenes
665.70
1,3-Butadiene
19,971.07
Formaldehyde
0.35
Methanol
658.85
Formaldehyde
19,207.50
Acetaldehyde
0.13
Benzene
251.18
Acetaldehyde
11,982.55
Acetaldehyde
0.13
Ethylene glycol
242.14
Nickel, PM
8,472.67
Naphthalene
0.02
Hexane
228.78
Benzene
8,372.75
Naphthalene
0.01
Formaldehyde
188.23
Naphthalene
7,267.81
Ethylbenzene
165.34
Xylenes
6,656.96
Acetaldehyde
107.84
Lead, PM
4,922.16
Styrene
80.67
Antimony, PM
3,500.85
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 9-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Valrico, Florida (Hillsborough County) - SYFL
Hydrochloric acid
2,725.32
Acrolein
1,502,391.89
Formaldehyde
0.24
Toluene
1,474.94
Hydrochloric acid
136,266.15
Formaldehyde
0.23
Xylenes
923.76
Formaldehyde
33,833.20
Acetaldehyde
0.13
Methanol
885.69
Hydrofluoric acid
29,839.55
Acetaldehyde
0.12
Hydrofluoric acid
417.75
1,3-Butadiene
29,196.33
Benzene
355.47
Bromomethane
24,759.22
Ethylene glycol
339.03
Acetaldehyde
19,658.53
Formaldehyde
331.57
Arsenic, PM
18,968.49
Hexane
330.23
Lead, PM
13,955.68
Ethylbenzene
233.29
Benzene
11,849.02
Winter Park, Florida (Orange County) - ORFL
Toluene
1,537.19
Acrolein
1,246,277.62
Formaldehyde
0.28
Xylenes
921.43
Formaldehyde
34,028.31
Formaldehyde
0.24
Methanol
822.73
1,3-Butadiene
27,160.32
Acetaldehyde
0.16
Benzene
343.02
Chlorine
26,988.93
Acetaldehyde
0.15
Formaldehyde
333.48
Hexamethylene-l,6-diisocyanate, gas
18,000.00
Ethylene glycol
322.93
Acetaldehyde
17,821.97
Hexane
314.56
Benzene
11,434.08
Ethylbenzene
232.56
Naphthalene
10,861.47
Acetaldehyde
160.40
Hexavalent Chromium PM
10,313.82
Styrene
150.70
Nickel, PM
9,484.79
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 9-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Florida Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard
Approximations Based on Annual
Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Orlando, Florida (Orange County) - PAFL
Toluene
1,537.19
Acrolein
1,246,277.62
Arsenic
0.04
Xylenes
921.43
Formaldehyde
34,028.31
Arsenic
0.03
Methanol
822.73
1.3 -Butadiene
27,160.32
Nickel
0.02
Benzene
343.02
Chlorine
26,988.93
Nickel
0.01
Formaldehyde
333.48
Hexamethylene-l,6-diisocyanate, gas
18,000.00
Ethylene glycol
322.93
Acetaldehyde
17,821.97
Hexane
314.56
Benzene
11,434.08
Ethylbenzene
232.56
Naphthalene
10,861.47
Acetaldehyde
160.40
Hexavalent Chromium PM
10,313.82
Styrene
150.70
Nickel, PM
9,484.79
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 9-6 include the following:
• Toluene, xylenes, methanol, and benzene are the highest emitted pollutants with
noncancer RfCs in Pinellas and Orange Counties. Hydrochloric acid is the highest
emitted pollutant in Hillsborough County, followed by toluene, xylenes, and
methanol.
• Acrolein has the highest toxicity-weighted emissions of the pollutants with noncancer
RfCs for each county, but is not among the highest emitted pollutants in the three
Florida counties.
• Four of the highest emitted pollutants in Pinellas and Hillsborough Counties also have
the highest toxicity-weighted emissions. Three of the highest emitted pollutants in
Orange County also have the highest toxicity-weighted emissions.
• Formaldehyde appears on both emissions-based lists for each county. Acetaldehyde
appears on both lists for Pinellas and Orange Counties; while acetaldehyde is not
among the highest emitted in Hillsborough County, it is among those with the highest
toxicity-weighted emissions.
• Naphthalene is among the pollutants with the highest toxicity-weighted emissions for
two of the three counties (except Hillsborough County), but is not among the highest
emitted pollutants (with a noncancer RfC) in any of these counties.
• Arsenic does not appear among the highest emitted pollutants in Orange County, or
among those with the highest toxicity-weighted emissions for this county. Nickel also
does not appear among the pollutants with the highest emissions for Orange County,
but does rank tenth for its toxicity-weighted emissions.
• Several metals appear among those with the highest toxicity-weighted emissions for
Pinellas and Hillsborough Counties, ranking highest for Hillsborough County, but
none of these metals are among the highest emitted.
Summary of the 2015-2016 Monitoring Data for the Florida Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Acetaldehyde andformaldehyde failed screens for AZFL, SYFL, and ORFL, where
only carbonyl compounds were sampled. Formaldehyde, acetaldehyde, and
naphthalene failed screens for SKFL. Arsenic and nickel failed screens for PAFL.
~~~ The 25 highest formaldehyde concentrations measured across the program in 2016
were measured at AZFL and SKFL. The 2016 annual averages for these sites are the
second andfifth highest annual average concentrations of formaldehyde,
respectively.
~~~ Concentrations of naphthalene have a decreasing trend at SKFL through 2015 but
increased somewhat for 2016. Acetaldehyde concentrations have a slight decreasing
trend at SYFL.
9-51
-------
Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for each Florida site, where carbonyl compounds were sampled. None of the
pollutants of interest have noncancer hazard approximations greater than an HQ of
1.0.
9-52
-------
10.0 Sites in Illinois
This section summarizes those data from samples collected at the NATTS and UATMP
sites in Illinois and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines the
context of risk. Readers are encouraged to refer to Sections 1 coiiinmcd in this ivpori
through 4 for detailed discussions and definitions regarding
the various data analyses presented below.
10.1 Site Characterization
This section characterizes the Illinois monitoring sites by providing a description of the
nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
Two monitoring sites are located in northwestern suburbs of Greater Chicago; NBIL is
located in Northbrook and SPIL is located in Schiller Park. A third site (ROIL) is located in
Roxana, just north of the St. Louis CBS A. Figures 10-1 and 10-2 present composite satellite
images retrieved from ArcGIS Explorer showing the Chicago monitoring sites and their
immediate surroundings. Figure 10-3 identifies the nearby point source emissions locations by
source category, as reported in the 2014 NEI for point sources, version 1, for NBIL and SPIL.
Note that only sources within 10 miles of the sites are included in the facility counts provided in
Figure 10-3. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring sites. Further, this boundary provides both the proximity of emissions sources to
the monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile boundaries are still visible on the map for reference but have been
grayed out to emphasize emissions sources within the boundaries. Figures 10-4 and 10-5 present
the composite satellite image and facility map for ROIL, respectively. Table 10-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates for each site. Each figure and table is discussed in detail in the paragraphs that
follow.
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
IXiUi jjcncr;ilct.l h> sources oilier
I hail LR(.i. LIJ.\'s conlriicl
kiliomloiA lor llic \\ll\ ;iiv nol
included in I lie dala ;uuil\ses
10-1
-------
Figure 10-1. Northbrook, Illinois (Mill,) Monitoring Site
© Edens Expy Spur E
5 —-RosemaryAn.
Chicago Botanic Garden
Green Acres Country
Dundee Rd
Dundee Rd
O \Dundee Rd
¦Ridgewood.Dr-
Source: USGS 1
irce: NASA, NGA, USGS
,2008 Microsoft CoirjiW
o
to
-------
Figure 10-2. Schiller Park, Illinois (SPIL) Monitoring Site
Lawrenc
•Lawrence
.Eastwood Ave
-------
Figure 10-3. NEI Point Sources Located Within 10 Miles of NBIL and SPIL
Lake
County
Lake
Michigan
Cook
County
NBIL NATTS site SPIL UATMP site 0 10 mile radius | | County boundary
Source Category Group (No. of Fi
"f Airport/Airline/Airport Support Operations (31)
£ Asphalt Production/Hot Mix Asphalt Plant (9)
0 Auto Body Shop/Painters/Automotive Stores (1)
» Automobile/Truck Manufacturing Facility (5)
® Automotive/RV Dealership (1)
Brick, Structural Clay, or Clay Ceramics Plant
(1)
A Building/Construction (7)
B Bulk Terminal/Bulk Plant (8)
C Chemical Manufacturing Facility (34)
1 Compressor Station (7)
A Concrete Batch Plant (13)
[X]Crematory (20)
0 Dry Cleaning Facility (57)
© Electrical Equipment Manufacturing Facility (38)
f Electricity Generation Facility (7)
_ Electroplating, Plating, Polishing, Anodizing,
and Coloring Facility (63)
F Food Processing/Agriculture Facility (44)
I Foundry, Iron and Steel (2)
A Foundry, non-ferrous (16)
If Gasoline/Diesel Service Station (1)
Glass Plant (3)
> Hotels/Motels/Lodging (2)
Industrial Machinery or Equipment Plant (29)
O Institution (school, hospital, prison, etc.) (42)
A Landfill (6)
Metal Can, Box, and Other Metal Container
^ Manufacturing Facility (1)
. Metal Coating, Engraving, and Allied Services
' to Manufacturers (25)
(•} Metals Processing/Fabrication Facility (70)
X Mine/Quarry/Mineral Processing Facility (29)
r\ Miscellaneous Commercial/Industrial Facility
¦ (93)
• Oil and/or Gas Production (5)
Paint and Coating Manufacturing Facility (9)
^ Petroleum Products Manufacturing Facility (1)
Q Pharmaceutical Manufacturing Facility (7)
R Plastic, Resin, or Rubber Products Plant (29)
^ Printing, Coating and Dyeing of Fabrics Facility
T (2)
p Printing/Publishing/Paper Product
Manufacturing Facility (76)
H Pulp and Paper Plant (1)
X Rail Yard/Rail Line Operations (8)
W Steel Mill (3)
TT Telecommunications/Radio Facility (21)
^ Testing Laboratory (1)
M Tobacco Manufacturing Facility (1)
[0 Utilities/Pipeline Construction (2)
* V\festewater Treatment Facility (6)
6 Vteter Treatment Facility (4)
... V\foodwork, Furniture, Millwork and Wood
Preserving Facility (8)
XI
' A ffi
DuPage \
County |
,. - -far
is © 1
Lfr '
HP* - o
?TT
Legend
87°35'0"W
Note; Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
10-4
-------
Figure 10-4. Roxana, Illinois (ROIL) Monitoring Site
-------
Figure 10-5. NEI Point Sources Located Within 10 Miles of ROIL
Madison'
County I
Mississippi
River 1
I Missouri River
MISSOURI
St. Louis
CityjJ
I St. Charles
I County
I '
St. Louis 1
County I
Legend
~
ROIL UATMP site
St. Clair
I County
i 1
| I* B A jp Ai
j
*\ _ . &•- -L
90l10'0"W 90°5'0"W 90°0'0"W 89°55'0"W 89°50'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
S4MO NATTS site Q 10 mile radius
County boundary
Source Category Group (No. o
Airport/Airline/Airport Support
Operations (7)
« Asphalt Production/Hot Mix Asphalt
Plant (4)
_Brick, Structural Clay, or Clay
Ceramics Plant (1)
B Bulk Terminal/Bulk Plant (11)
C Chemical Manufacturing Facility (3)
I Compressor Station (7)
AConcrete Batch Plant (6)
^Crematory (3)
CD Dry Cleaning Facility (1)
e Electrical Equipment Manufacturing
Facility (1)
Facilities)
f Electricity Generation Facility (2)
F Food Processing/Agriculture Facility
(4)
AFoundry, non-ferrous (1)
^Industrial Machinery or Equipment
Plant (1)
Institution (school, hospital, prison,
etc.)(5)
• Landfill (2)
^ Metal Coating, Engraving, and Allied
' Services to Manufacturers (2)
0Metals Processing/Fabrication
Facility (3)
, Mine/Quarry/Mineral Processing
* Facility (7)
rf Miscellaneous Commercial/Industrial
' Facility (10)
< Pesticide Manufacturing Plant (1)
^Petroleum Products Manufacturing
Facility (1)
14 Petroleum Refinery (1)
• Port and Harbor Operations (2)
p Printing/Publishing/Paper Product
Manufacturing Facility (1)
S3 Pulp and Paper Plant (1)
XRail Yard/Rail Line Operations (1)
•Steel Mill (2)
1 Wastewater Treatment Facility (1)
10-6
-------
Table 10-1. Geographical Information for the Illinois Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
NBIL
17-031-4201
Northbrook
Cook
Chicago-
Naperville-Elgin
IL-IN-WI
42.139996,
-87.799227
Residential
Suburban
115,100
1-94 north of intersection with
Dundee Rd
SPIL
17-031-3103
Schiller
Park
Cook
Chicago-
Naperville-Elgin
IL-IN-WI
41.965193,
-87.876265
Mobile
Suburban
193,800
1-294, just south of Lawrence Ave
ROIL
17-119-9010
Roxana
Madison
St. Louis, MO-IL
38.848382,
-90.076413
Industrial
Suburban
6,850
S Central Ave at W Thomas St
1 AADT reflects 2013 data for SPIL, 2014 data for NBIL, and 2015 data for ROIL (IL DOT, 2017)
BOLD ITALICS = EPA-designated NATTS Site
-------
NBIL is located on the property of the Northbrook Water Filtration Station. Figure 10-1
shows that NBIL is located off State Highway 68 (Dundee Road), near Exit 30 on 1-94. A rail
line runs north-south next to the water filtration station, separating the municipal buildings and
nearby residential subdivision from a business complex to the east, and intersects Dundee Road
just south of the monitoring site. The surrounding area is suburban and residential. Commercial,
residential, and forested areas surround the site, along with a country club and golf course. The
NBIL monitoring site is the Chicago NATTS site.
SPIL is located on the eastern edge of the Chicago-O'Hare International Airport, between
Mannheim Road and 1-294, just north of the toll plaza. The nearest runway is less than one-
half mile from the site. The surrounding area is classified as suburban and mobile. Commercial
and residential areas are located to the east of the airport and 1-294. The rail yard located to the
east of 1-294 is an intermodal terminal/facility that has been closed (Podmolik, 2015).
NBIL and SPIL are located within 13 miles of each other. Each site is located within
10 miles of numerous point sources, although the quantity of emissions sources is higher near
SPIL than NBIL, as shown in Figure 10-3. The source categories with the largest number of
sources within 10 miles of NBIL and SPIL include printing/publishing/paper product
manufacturing; metals processing/fabrication; electroplating, plating, polishing, anodizing, and
coloring; dry cleaning; food processing/agriculture; and institutions (schools, hospitals, prisons,
etc.). Few point sources are located within 2 miles of NBIL, with most of the sources located
farther west or south. The closest source to NBIL is plotted under the symbol for the site in
Figure 10-3; this source is a dry cleaning facility. Besides the airport and related operations, the
closest point source to SPIL is involved in electroplating, plating, polishing, anodizing, and
coloring.
The ROIL monitoring site in Roxana is located at the fence line of a petroleum refinery.
Although this area is industrial, residential areas are wedged between the industrial properties, as
Figure 10-4 shows. Just north of the monitoring site are a junior high school and a high school,
whose track and tennis courts are shown across the street from the monitoring site. North of the
schools is a community park. Ambient monitoring data from this location will be used to assess
near-field concentrations in the neighboring community, with emphasis on comparing and
contrasting these data to the St. Louis NATTS site (S4MO), which is also pictured in Figure 10-5
(WUSTL, 2013 and 2016). The Mississippi River, which is the border between Missouri and
Illinois, is just over a mile and a half west of the ROIL monitoring site.
10-8
-------
In addition to showing the ROIL monitoring site's location relative to the S4MO
monitoring site, Figure 10-5 also shows the point sources within 10 miles of each site (although
only the facilities within 10 miles of ROIL are included in the facility counts below the map).
There are numerous emissions sources surrounding ROIL, most of which are located to the south
and northwest of the site. Many of the sources within 2 miles of ROIL are involved in or related
to the petroleum industry. A petroleum refinery, multiple compressor stations, and several bulk
terminals surround this site. Other nearby sources include a rail yard, an industrial
machinery/equipment facility, and several chemical manufacturers.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 10-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site. SPIL
experiences a higher traffic volume compared to NBIL, although the traffic volumes near these
two sites are both significantly greater than the traffic volume near ROIL. SPIL's traffic volume
is the third highest among NMP sites. The traffic volume for NBIL ranks tenth among NMP sites
while traffic volume near ROIL is in the bottom third. Note that the traffic volumes presented for
NBIL and SPIL are from interstate highways while the traffic volume for ROIL is not.
10.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 10-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 10-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. VOCs, carbonyl compounds, SNMOCs,
metals (PMio), and PAHs were sampled for at NBIL, while VOCs and carbonyl compounds were
sampled for at SPIL and ROIL. Note that sampling at ROIL was discontinued at the end of July
2015.
10-9
-------
Table 10-2. 2015-2016 Risk-Based Screening Results for the Illinois Monitoring Sites
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Northbrook, Illinois - NBIL
Acetaldehyde
0.45
116
116
100.00
11.34
11.34
Formaldehyde
0.077
116
116
100.00
11.34
22.68
Benzene
0.13
113
113
100.00
11.05
33.72
Carbon Tetrachloride
0.17
113
113
100.00
11.05
44.77
Arsenic (PMio)
0.00023
107
113
94.69
10.46
55.23
1,2 -Dichloroethane
0.038
103
103
100.00
10.07
65.30
Naphthalene
0.029
85
116
73.28
8.31
73.61
1.3 -Butadiene
0.03
67
107
62.62
6.55
80.16
Acenaphthene
0.011
49
110
44.55
4.79
84.95
Fluorene
0.011
47
104
45.19
4.59
89.54
Fluoranthene
0.011
37
116
31.90
3.62
93.16
p-Dichlorobenzene
0.091
21
50
42.00
2.05
95.21
Hexacliloro-1,3 -butadiene
0.045
11
11
100.00
1.08
96.29
Nickel (PMio)
0.0021
11
113
9.73
1.08
97.36
Clilorofonn
9.8
8
113
7.08
0.78
98.14
Ethylbenzene
0.4
6
112
5.36
0.59
98.73
Benzo(a)pyrene
0.00057
4
113
3.54
0.39
99.12
Manganese (PMio)
0.03
3
113
2.65
0.29
99.41
T ricliloroethylene
0.2
2
38
5.26
0.20
99.61
Bromomethane
0.5
1
111
0.90
0.10
99.71
1,2 -Dibromoethane
0.0017
1
1
100.00
0.10
99.80
Diclilorometliane
60
1
113
0.88
0.10
99.90
Lead (PMio)
0.015
1
113
0.88
0.10
100.00
Total
1,023
2,228
45.92
10-10
-------
Table 10-2. 2015-2016 Risk-Based Screening Results for the Illinois Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Schiller Park, Illinois - SPIL
Acetaldehyde
0.45
118
118
100.00
15.07
15.07
Benzene
0.13
118
118
100.00
15.07
30.14
Carbon Tetrachloride
0.17
118
118
100.00
15.07
45.21
Formaldehyde
0.077
118
118
100.00
15.07
60.28
1.3 -Butadiene
0.03
117
118
99.15
14.94
75.22
1,2 -Dichloroethane
0.038
112
114
98.25
14.30
89.53
T richloroethylene
0.2
36
90
40.00
4.60
94.13
p-Dichlorobenzene
0.091
14
50
28.00
1.79
95.91
Ethylbenzene
0.4
9
118
7.63
1.15
97.06
Hexachloro-1,3 -butadiene
0.045
9
10
90.00
1.15
98.21
Propionaldehyde
0.8
9
118
7.63
1.15
99.36
T etrachloroethylene
3.8
3
118
2.54
0.38
99.74
Bromomethane
0.5
1
116
0.86
0.13
99.87
1,2 -Dibromoethane
0.0017
1
1
100.00
0.13
100.00
Total
783
1,325
59.09
Roxana, Illinois - ROIL
Acetaldehyde
0.45
33
33
100.00
17.01
17.01
Formaldehyde
0.077
33
33
100.00
17.01
34.02
Benzene
0.13
32
32
100.00
16.49
50.52
Carbon Tetrachloride
0.17
32
32
100.00
16.49
67.01
1,2 -Dichloroethane
0.038
31
31
100.00
15.98
82.99
1.3 -Butadiene
0.03
25
30
83.33
12.89
95.88
Ethylbenzene
0.4
4
32
12.50
2.06
97.94
Propionaldehyde
0.8
2
31
6.45
1.03
98.97
1,2 -Dibromoethane
0.0017
1
1
100.00
0.52
99.48
Hexachloro-1,3 -butadiene
0.045
1
2
50.00
0.52
100.00
Total
194
257
75.49
Observations from Table 10-3 include the following:
• The number of pollutants failing screens for NBIL is higher than the other two
monitoring sites; this is expected given the difference in pollutants measured at each
site.
• Twenty-three pollutants failed at least one screen for NBIL; 46 percent of
concentrations for these 23 pollutants were greater than their associated risk screening
value (or failed screens).
10-11
-------
• Twelve pollutants contributed to 95 percent of failed screens for NBIL and therefore
were identified as pollutants of interest for this site. These 12 include two carbonyl
compounds, five VOCs, one PMio metal, and four PAHs.
• NBIL failed the third highest number of screens (1,023) among NMP sites, as shown
in Table 4-9, and had the highest number of pollutants whose concentrations failed
screens (23). However, the failure rate for NBIL, when incorporating all pollutants
with screening values, is relatively low, at nearly 20 percent. This is due primarily to
the relatively large number of pollutants sampled for at this site. NBIL is one of only
two NMP sites sampling five pollutant groups and one of only two sites to sample
with both the TO-15 and SNMOC methods. As described in Section 3.2, if a pollutant
was measured by both the TO-15 and SNMOC methods at the same site, the TO-15
results were used for the risk-based screening process. As NBIL sampled both VOCs
(TO-15) and SNMOCs, the TO-15 results were used for the 12 pollutants these
methods have in common.
• Fourteen pollutants failed screens for SPIL; 59 percent of concentrations for these 14
pollutants were greater than their associated risk screening value (or failed screens).
• Eight pollutants contributed to 95 percent of failed screens for SPIL and therefore
were identified as pollutants of interest for this site. These eight include two carbonyl
compounds and six VOCs. SPIL is the only NMP site with trichloroethylene as a
pollutant of interest. This was also true in the 2014 NMP report.
• Ten pollutants failed screens for ROIL; 75 percent of concentrations for these 10
pollutants were greater than their associated risk screening value (or failed screens).
• Six pollutants contributed to 95 percent of failed screens for ROIL and therefore were
identified as pollutants of interest for this site. These six include two carbonyl
compounds and four VOCs.
• The Illinois monitoring sites have six pollutants of interest in common: two carbonyl
compounds (acetaldehyde and formaldehyde) and four VOCs (benzene,
1,3-butadiene, carbon tetrachloride, 1,2-dichloroethane). Of these, acetaldehyde,
benzene, carbon tetrachloride, and formaldehyde failed 100 percent of screens for
each site.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
10-12
-------
10.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Illinois monitoring sites. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at NBIL, SPIL, and
ROIL are provided in Appendices J, K, and M through P.
10.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Illinois site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the Illinois
monitoring sites are presented in Table 10-3, where applicable. Note that concentrations of the
PAHs and metals for NBIL are presented in ng/m3 for ease of viewing. Also note that if a
pollutant was not detected in a given calendar quarter, the quarterly average simply reflects "0"
because only zeros substituted for non-detects were factored into the quarterly average
concentration.
10-13
-------
Table 10-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Illinois Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
Oig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Northbrook, Illinois - NBIL
Acetaldehyde
57/57/57
1.16
±0.20
1.17
±0.28
1.68
±0.26
1.71
±0.46
1.42
±0.16
59/59/59
2.18
±0.55
1.26
±0.35
1.49
±0.32
0.96
±0.29
1.47
±0.21
Benzene
54/54/54
0.59
±0.05
0.42
±0.10
0.61
±0.16
NA
0.52
±0.06
59/59/59
0.49
±0.09
0.36
±0.08
0.47
±0.09
0.43
±0.09
0.43
±0.04
1.3 -Butadiene
51/32/54
0.04
±0.01
0.03
±0.01
0.05
±0.02
NA
0.04
±0.01
56/9/59
0.03
±0.01
0.04
±0.01
0.04
±0.01
0.04
±0.01
0.04
±0.01
Carbon Tetrachloride
54/54/54
0.59
±0.08
0.63
±0.02
0.68
±0.04
NA
0.63
±0.03
59/59/59
0.51
±0.05
0.69
±0.05
0.50
±0.11
0.54
±0.06
0.56
±0.04
p-Dichlorobenzene
27/8/54
0.01
±0.01
0.04
±0.03
0.34
±0.39
NA
0.13
±0.11
23/3/59
<0.01
±0.01
0.05
±0.03
0.06
±0.03
0.02
±0.02
0.03
±0.01
1,2 -Dichloroethane
53/49/54
0.08
±<0.01
0.08
±0.01
0.06
±0.01
NA
0.07
± <0.01
50/50/59
0.08
±0.02
0.09
±0.01
0.06
±0.02
0.04
±0.02
0.07
±0.01
Formaldehyde
57/57/57
1.74
±0.27
1.87
±0.39
3.04
±0.50
1.38
±0.31
2.03
±0.25
59/59/59
1.41
±0.39
2.31
±0.72
3.53
±0.97
1.28
±0.23
2.16
±0.39
Acenaphthenea
55/55/60
3.20
±2.52
23.72
± 14.01
39.75
± 11.73
3.23
± 1.20
17.48
±5.87
55/55/56
2.82
± 1.07
19.49
± 11.80
46.29
± 11.43
5.03
±3.05
18.93
±6.15
Arsenic (PMi0)a
56/56/56
0.55
±0.12
0.83
±0.45
1.55
± 1.01
0.87
±0.25
0.94
±0.27
57/57/57
NA
1.08
±0.27
1.00
±0.21
0.75
±0.23
0.87
±0.13
Fluoranthene3
60/60/60
2.21
± 1.06
10.87
±6.30
20.29
±6.57
1.91
±0.59
8.82
±2.90
56/56/56
1.94
±0.54
13.65
±8.86
26.45
±7.57
2.51
± 1.12
11.38
±3.85
Fluorene3
53/53/60
3.56
±2.44
21.43
± 12.95
37.23
± 10.61
2.87
± 1.15
16.27
±5.41
51/51/56
2.29
± 1.14
20.78
± 12.35
46.67
± 12.88
4.98
±2.39
19.19
±6.42
Naphthalene1
60/60/60
43.62
± 14.02
97.32
±40.71
179.35
±54.50
36.97
±9.24
89.32
± 22.07
56/56/56
47.67
± 17.25
79.44
±33.74
145.74
±39.92
41.07
± 11.5
79.55
± 17.47
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 10-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Illinois Monitoring Sites (Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
Oig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Schiller Park, Illinois - SPIL
Acetaldehyde
61/61/61
2.49
±0.35
1.91
±0.35
2.54
±0.34
2.71
± 1.87
2.43
±0.52
57/57/57
4.07
± 1.78
1.71
±0.41
1.55
±0.27
2.17
± 1.04
2.45
±0.61
Benzene
60/60/60
0.78
±0.10
0.77
±0.35
0.68
±0.25
0.62
±0.12
0.71
±0.10
58/58/58
0.66
±0.09
0.58
±0.08
0.60
±0.12
0.68
±0.15
0.63
±0.05
1.3 -Butadiene
60/59/60
0.12
±0.04
0.14
±0.08
0.12
±0.04
0.11
±0.04
0.12
±0.02
58/53/58
0.12
±0.03
0.12
±0.02
0.10
±0.02
0.11
±0.03
0.11
±0.01
Carbon Tetrachloride
60/60/60
0.61
±0.05
0.62
±0.04
0.66
±0.04
0.59
±0.05
0.62
±0.02
58/58/58
0.58
±0.05
0.68
±0.04
0.61
±0.07
0.62
±0.05
0.62
±0.03
p-Dichlorobenzene
25/1/60
0.02
±0.02
0.04
±0.03
0.05
±0.03
0.02
±0.02
0.03
±0.01
25/3/58
0.02
±0.01
0.04
±0.02
0.06
±0.04
0.02
±0.03
0.03
±0.01
1,2 -Dichloroethane
60/54/60
0.08
±0.01
0.08
±0.01
0.06
±0.01
0.08
±0.01
0.08
± <0.01
54/53/58
0.09
±0.01
0.09
±0.01
0.06
±0.02
0.07
±0.01
0.07
±0.01
Formaldehyde
61/61/61
3.02
±0.33
3.35
±0.59
5.13
±0.79
3.85
± 1.23
3.85
±0.45
57/57/57
4.43
± 1.39
3.22
±0.58
3.48
±0.50
2.87
±0.73
3.54
±0.47
T ricliloroethylene
42/35/60
0.09
±0.06
0.31
±0.37
0.57
±0.57
0.19
±0.12
0.29
±0.16
48/36/58
0.20
±0.17
0.30
±0.19
0.60
±0.43
0.19
±0.17
0.32
±0.13
Roxana, Illinois - ROIL
Acetaldehyde
33/33/33
1.71
±0.41
2.21
±0.33
NA
NS
NS
NS
NS
NS
NS
NS
NS
Benzene
32/32/32
0.95
±0.18
1.48
±0.61
NA
NS
NS
NS
NS
NS
NS
NS
NS
1.3 -Butadiene
30/25/32
0.06
±0.02
0.04
±0.01
NA
NS
NS
NS
NS
NS
NS
NS
NS
Carbon Tetrachloride
32/32/32
0.61
±0.04
0.62
±0.05
NA
NS
NS
NS
NS
NS
NS
NS
NS
1,2 -Dichloroethane
31/30/32
0.09
±0.01
0.09
±0.01
NA
NS
NS
NS
NS
NS
NS
NS
NS
Formaldehyde
33/33/33
1.94
±0.39
4.15
± 1.04
NA
NS
NS
NS
NS
NS
NS
NS
NS
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Observations for NBIL from Table 10-3 include the following:
• The pollutants with the highest annual average concentrations for both years are
formaldehyde and acetaldehyde. The annual average concentrations for the remaining
pollutants of interest are less than 1 |ig/m3, with carbon tetrachloride as the next
highest.
• Formaldehyde concentrations measured at NBIL range from 0.406 |ig/m3 to
7.11 |ig/m3. For 2015, the third quarter average concentration of formaldehyde
(3.05 ± 0.50 |ig/m3) is significantly higher than the other quarterly averages, which
are all less than 2 |ig/m3. A review of the data shows that all but one of the 10
formaldehyde concentrations greater than or equal to 3 |ig/m3 in 2015 were measured
between July and September (with the exception measured in May). For 2016, the
third quarter average (3.53 ± 0.97 |ig/m3) is also the highest quarterly average
concentration, but the confidence interval associated with this average is relatively
large, indicating considerable variability in the measurements. This is also true for the
second quarter average for 2016 (2.31 ± 0.72 |ig/m3). Twelve formaldehyde
concentrations greater than or equal to 3 |ig/m3 were measured in 2016, with four
measured during the second quarter and eight measured during the third quarter of
2016.
• Acetaldehyde concentrations measured at NBIL range from 0.473 |ig/m3 to
3.95 |ig/m3. Quarterly average concentrations of acetaldehyde range from
0.96 ± 0.29 |ig/m3 (fourth quarter 2016) to 2.18 ± 0.55 |ig/m3 (first quarter 2016). The
three highest acetaldehyde concentrations were measured at NBIL during the first
quarter of 2016; further, six of the eight acetaldehyde concentrations greater than
2.5 |ig/m3 in 2016 were measured during the first quarter of the year (with one each
measured during the second and third quarters), explaining the relatively high
quarterly average acetaldehyde concentration shown in Table 10-3. By comparison,
three acetaldehyde concentrations greater than 2.5 |ig/m3 were measured in 2015.
• Fourth quarter average concentrations for 2015 for the VOCs in Table 10-3 could not
be calculated because there were too many invalid samples (due to varying sampling
issues) during this quarter to meet the 75 percent criterion. Carbon tetrachloride and
benzene have the highest annual average concentrations among the VOCs in
Table 10-3.
• Concentrations of carbon tetrachloride measured at NBIL range from 0.240 |ig/m3 to
0.832 |ig/m3. Even though six concentrations measured in 2016 are greater than the
maximum carbon tetrachloride concentration measured in 2015, including all four
concentrations greater than 0.8 |ig/m3, the annual average for 2015 is greater than the
annual average for 2016. This can be explained by reviewing the measurements at the
lower end of the concentration range. Three carbon tetrachloride concentrations less
than 0.5 |ig/m3 were measured in 2015 compared to 21 in 2016.
• Concentrations of benzene measured at NBIL range from 0.16 |ig/m3 to 1.20 |ig/m3.
Only two benzene concentrations greater than 1 |ig/m3 were measured at NBIL, both
of which were measured during the third quarter of 2015, which is the second fewest
10-16
-------
among NMP sites sampling benzene over the full 2-year period. NBIL's annual
average concentration of benzene for 2016 is the fifth lowest across the program.
• Concentrations of/>dichlorobenzene measured at NBIL range from 0.0361 |ig/m3 to
2.78 |ig/m3 plus 63 non-detects. The third quarter average concentration of
/;-dichlorobenzene for 2015 is an order of magnitude higher than the other quarterly
average shown and has a confidence interval greater than the average itself
(0.34 ± 0.39 |ig/m3). This indicates the likely presence of outlier(s). A review of the
data shows that the maximum p-dichlorobenzene concentration measured at NBIL on
September 21, 2015 (2.78 |ig/m3) is the maximum /;-dichlorobenzene concentration
measured across the program. The next highest />dichlorobenzene concentration
measured at NBIL is one-third the magnitude (0.837 |ig/m3), but even this
concentration is among the highest across the program. NBIL is one of five NMP
sites with at least one p-dichlorobenzene concentration greater than 0.5 |ig/m3 (NBIL
has four, second only to S4MO, which as 10), all of which were measured during the
third or fourth quarters of 2015.
• Arsenic is the only PMio metal identified as a pollutant of interest for NBIL. Arsenic
concentrations measured at NBIL range from 0.076 ng/m3 to 7.18 ng/m3. Although
the 2015 and 2016 annual average concentrations of arsenic are fairly similar to each
other, the confidence interval for 2015 is more than twice the confidence interval for
2016. The maximum arsenic concentration was measured at NBIL on July 5, 2015
and is more than twice the next highest concentration measured at this site, and the
second highest measured across the program. This is reflected in the third quarter
average concentration for 2015 (1.55 ± 1.01 ng/m3). A review of the data shows that
concentrations measured during the third quarter of 2015 were higher overall
compared to other calendar quarters. The third quarter of 2015 is the only calendar
quarter within which an arsenic concentration less than 0.5 ng/m3 was not measured;
among the other calendar quarters, the number of arsenic concentrations less than
0.5 ng/m3 ranged from as few as one (third quarter of 2016) to as many as seven (both
the second quarter of 2015 and first quarter of 2016). Even if the maximum
concentration was excluded, the third quarter average for 2015 would still be the
highest quarterly average concentration of arsenic. A quarterly average concentration
for first quarter of 2016 could not be calculated because there were too many invalid
samples (many for run time issues) to meet the 75 percent criterion.
• Of the PAHs, naphthalene has the highest annual average concentrations shown.
Naphthalene concentrations measured at NBIL span three orders of magnitude,
ranging from 0.446 |ig/m3 to 403 ng/m3. These are both the minimum and maximum
naphthalene concentrations measured across the program. Quarterly average
concentrations of naphthalene vary considerably, from 36.97 ± 9.24 ng/m3 (fourth
quarter of 2015) to 179.35 ± 54.50 ng/m3 (third quarter of 2015). The maximum
naphthalene concentration across the program has been measured at NBIL for the last
several years. Based on the quarterly averages shown, concentrations of naphthalene
appear higher during the warmer months of the year and exhibit the most variability.
All but one of the 17 naphthalene concentrations greater than 150 ng/m3 measured at
NBIL were measured between May and September of any given year (with the
exception measured in April). Conversely, all but two of the 23 naphthalene
concentrations less than 25 ng/m3 were measured outside these five months.
10-17
-------
• Some of the highest concentrations of acenaphthene, fluorene, and fluoranthene
measured across the program were also measured at NBIL. Concentrations of
acenaphthene measured at NBIL range from 0.338 ng/m3 to 103 ng/m3, plus six non-
detects, accounting for 12 of the 15 highest acenaphthene measurements across the
program (those greater than 50 ng/m3). Concentrations of fluorene measured at NBIL
range from 0.675 ng/m3 to 96 ng/m3, plus 12 non-detects; although the maximum
fluorene concentration across the program was not measured at NBIL, concentrations
measured at NBIL account for the next 15 highest measurements across the program.
Concentrations of fluoranthene range from 0.403 ng/m3 to 57.3 ng/m3, with
concentrations measured at NBIL accounting for all but one of the 11 fluoranthene
concentrations greater than 30 ng/m3 across the program. Concentrations of these
PAH pollutants of interest also tended to be measured during the warmer months of
the year and exhibit a relatively large amount of variability, based on the quarterly
average concentrations shown and their associated confidence intervals. Many of the
higher PAH concentrations were measured on the same sample days. For instance, the
highest fluorene and acenaphthene concentrations were measured at NBIL on June
23, 2015 and the second highest fluorene and acenaphthene concentrations were
measured at NBIL on August 10, 2016 (along with the third highest naphthalene
concentration and fourth highest fluoranthene concentration).
Observations for SPIL from Table 10-3 include the following:
• The pollutants with the highest annual average concentrations for SPIL are
formaldehyde and acetaldehyde (for both years). These are the only pollutants of
interest with annual average concentrations greater than 1 |ig/m3. Of the VOCs,
benzene and carbon tetrachloride have the highest annual average concentrations for
SPIL.
• Concentrations of formaldehyde measured at SPIL range from 1.45 |ig/m3 to
12.3 |ig/m3. The three highest formaldehyde concentrations were measured between
December 2015 and February 2016, explaining, at least in part, the relatively large
confidence intervals associated with the fourth quarter average of 2015 and the first
quarter average of 2016 (these quarterly averages have confidence intervals greater
than 1.00). Formaldehyde concentrations greater than 5 |ig/m3 were not measured at
SPIL prior to the third quarter of 2015. Of the 16 formaldehyde concentrations greater
than 5 |ig/m3 measured at SPIL, 13 were measured between July 2015 and February
2016, with one measured during each of the other calendar quarters in 2016. With the
exception of the third quarter average for 2015 and the first quarter average for 2016,
the quarterly average concentrations of formaldehyde vary by less than 1 |ig/m3.
• Concentrations of acetaldehyde measured at SPIL range from 0.793 |ig/m3 to
17.1 |ig/m3, which is the second highest acetaldehyde concentration measured across
the program. SPIL is one of three NMP sites at which an acetaldehyde concentration
greater than 10 |ig/m3 was measured; of the six acetaldehyde concentrations greater
than 10 |ig/m3, two were measured at SPIL. These two concentrations were measured
on December 2, 2015 and February 12, 2016, the same days the maximum
formaldehyde concentrations were measured (although the acetaldehyde
concentrations were higher). Acetaldehyde concentrations greater than 4 |ig/m3 were
measured between December 2015 and February 2016 (five) or in December 2016
10-18
-------
(two). The effect of these measurements can be seen in the confidence intervals for
the fourth quarter average of 2015 and the first and fourth quarter averages of 2016.
• Quarterly average concentrations for several of the VOCs, including 1,3-butadiene,
carbon tetrachloride,/?-dichlorobenzene, and 1,2-dichloroethane, did not vary much
across years.
• Concentrations of benzene appear higher at SPIL in 2015 than in 2016, particularly
for the first half of 2015, though the difference is not statistically significant. Twice
the number of benzene concentrations greater than 1 |ig/m3 were measured in 2015
(six) compared to 2016 (three), with the three highest measured in 2015.
• Although the two annual average concentrations of trichloroethylene are similar to
each other, the quarterly average concentrations vary significantly, from
0.09 ± 0.06 |ig/m3 (first quarter 2015) to 0.60 ± 0.43 |ig/m3 (third quarter 2016). Each
of the quarterly average concentrations shown has a relatively large confidence
interval, including two greater than or equal to the averages themselves. This is also
true for the annual average concentrations of this pollutant. A review of the data
shows that trichloroethylene was detected in 76 percent of the samples collected at
SPIL, with measured detections ranging from 0.0323 |ig/m3 to 4.02 |ig/m3. Eleven of
the 14 trichloroethylene concentrations greater than 1 |ig/m3 measured across the
program were measured at SPIL. Interestingly, NBIL is the only NMP site at which a
trichloroethylene concentration greater than the maximum concentration measured at
SPIL was measured. (Trichloroethylene was not identified as a pollutant of interest
for NBIL.) Nineteen of the 23 highest trichloroethylene concentrations (those greater
than 0.5 |ig/m3) across the program were measured at SPIL. SPIL is the only NMP
site for which trichloroethylene is a pollutant of interest. Similar observations were
also made in the 2011, 2012, 2013, and 2014 NMP reports.
• The annual averages for acetaldehyde and formaldehyde for SPIL for both years are
significantly higher than the annual averages for these compounds for NBIL. This is
also true for 1,3-butadiene and benzene.
Observations for ROIL from Table 10-3 include the following:
• Sampling at ROIL was discontinued at the end of July 2015, ending a four-year
monitoring effort at this location. Due to the discontinuation of sampling, quarterly
average concentration could be calculated for the first and second quarters of 2015
only and no annual averages are provided. However, statistical summaries for the
entire period of sampling in 2015 are provided in Appendices J and M.
• The pollutants with the highest quarterly average concentrations are formaldehyde,
acetaldehyde, and benzene. For each of these pollutants, the second quarter average
concentration is greater than the first quarter average, particularly for formaldehyde.
• The second quarter average concentration of formaldehyde is significantly higher
than the first quarter average. A review of the data shows that formaldehyde
concentrations measured at ROIL range from 1.20 |ig/m3 to 9.91 |ig/m3, with 17 of
the 18 formaldehyde concentrations greater than 3 |ig/m3 measured at ROIL after
April 1st (including all five concentrations measured in July). At the other end of the
10-19
-------
concentration range, none of the 10 formaldehyde concentrations less than 2 |ig/m3
were measured outside the first quarter of 2015.
• Concentrations of acetaldehyde also appear higher during the second quarter,
although the difference among the quarterly averages is considerably less. While
acetaldehyde concentrations greater than 2 |ig/m3 were measured during both
calendar quarters, most of them (10 of 13) were measured after April 1st, including
three in July. Conversely, the six lowest acetaldehyde concentrations were measured
at ROIL during the first quarter of 2015.
• Concentrations of benzene also appear higher during the second quarter and, similar
to formaldehyde, the confidence interval for the second quarter average concentration
is considerably larger than the confidence interval shown for the first quarter average.
Benzene concentrations measured at ROIL span an order of magnitude, ranging from
0.349 |ig/m3 to 3.94 |ig/m3, which is the 12th highest benzene concentration
measured at an NMP site in 2015 and 2016. The four highest benzene concentrations
(those greater than 2 |ig/m3) were measured at ROIL after the first quarter, with one
measured in April, two in May, and one in July. However, all three benzene
concentrations less than 0.5 |ig/m3 were also measured after the first quarter.
• For the remaining pollutants of interest, the first quarter average for 2015 is similar to
the second quarter average for 2015.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for NBIL and
SPIL from those tables include the following:
• The Chicago monitoring sites appear in Tables 4-10 through 4-13 a total of 15 times,
with NBIL appearing 10 times and SPIL appearing five times.
• Table 4-10 shows that annual average concentrations of 1,3-butadiene for SPIL rank
eighth (2015) and 10th (2016) among NMP sites sampling this pollutant. NBIL also
appears for its annual average concentration of />dichlorobenzene for 2015, ranking
seventh (the annual average for 2016 ranks considerably lower). NBIL's annual
average concentration of hexachloro-l,3-butadiene (for 2016) ranks 10th among
NMP sites sampling this pollutant, although the annual averages for this pollutant do
not vary considerably.
• SPIL appears in Table 4-11 for both of its annual average concentrations of
acetaldehyde, ranking eighth (2016) and ninth (2015). SPIL also ranks ninth for its
2015 annual average concentration of formaldehyde, with the annual average for
2016 outside the top 10.
• NBIL's annual average concentrations of acenaphthene and fluorene rank highest
among NMP sites sampling PAHs, as shown in Table 4-12. The confidence intervals
associated with NBIL's annual averages for these PAHs are the largest among the
averages shown, a reflection of the variability within the measurements. NBIL's
annual average concentrations of naphthalene also appear in Table 4-12, ranking sixth
10-20
-------
(2015) and seventh (2016) highest, and also have the largest confidence intervals
shown.
• The annual average concentrations of arsenic for NBIL rank fourth (2015) and eighth
(2016) highest among NMP sites sampling PMio metals, as shown in Table 4-13.
10.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 10-3 for NBIL, SPIL, and ROIL. Figures 10-6 through 10-18 overlay the sites' minimum,
annual average, and maximum concentrations for each year onto the program-level minimum,
first quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
Figure 10-6. Program vs. Site-Specific Average Acenaphthene Concentrations
Proeram Max Concentration = 108 ne/m3
.
...... .....
1
V
1
J NBIL 2015 Max = 103 ng/m3; 2016 Max = 92.4 ng/m3 ;
0 15 30 45 60 75
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
0 Q
Figure 10-6 presents the box plot for acenaphthene for NBIL and shows the following:
• NBIL is the only Illinois site to sample PAHs under the NMP in 2015 and 2016. The
program-level maximum concentration (108 ng/m3) of acenaphthene is not shown
directly on the box plot because the scale of the box plot would be too large to readily
observe data points at the lower end of the concentration range. Thus, the scale has
been reduced.
• The maximum acenaphthene concentrations measured each year at NBIL exceed the
scale of the box plot and thus, are also provided on the box plot. Two additional
acenaphthene concentrations measured at NBIL in 2015 exceed the scale of the box
plot. Four NMP sites have individual acenaphthene concentrations greater than
40 ng/m3, with concentrations measured at NBIL accounting for more than half of
them (19 out of 31).
10-21
-------
• NBIL's annual average concentrations of acenaphthene are both more than four times
the program-level average concentration. More than half of NBIL's individual
acenaphthene measurements are greater than the program-level average concentration
(4.36 ng/m3). Note that the program-level average is greater than the program-level
third quartile, an indication that the measurements at the upper end of the
concentration range are driving the program-level average upward.
Figure 10-7. Program vs. Site-Specific Average Acetaldehyde Concentrations
NBIL
ROIL
4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 10-7 presents the box plots for acetaldehyde for all three sites and shows the
following:
• The range of acetaldehyde concentrations measured is largest for SPIL and smallest
for NBIL. The maximum acetaldehyde concentration measured at SPIL in 2015 is
only slightly less than the maximum acetaldehyde concentration measured across the
program. The maximum acetaldehyde concentration measured at SPIL in 2016 is also
among the highest measured across the program. The minimum concentration
measured at SPIL in 2016 is greater than the program-level first quartile. By
comparison, only 11 percent of individual acetaldehyde concentrations measured at
NBIL are greater than the annual averages for SPIL.
• Both annual averages for SPIL greater than the program-level average and third
quartile. The annual average concentrations of acetaldehyde for NBIL are a full
1 |ig/m3 less than the annual averages for SPIL and are less than the program-level
average concentration.
10-22
-------
• The range of acetaldehyde concentrations measured at ROIL is more similar to the
range measured at NBIL than SPIL, although the minimum concentration measured at
ROIL is just less than the program-level first quartile.
• Sampling at ROIL under the NMP was discontinued at the end of July 2015.
Figure 10-8. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
1 1-
o
E !
>
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 10-8 presents the box plot for arsenic (PMio) for NBIL and shows the following:
• NBIL is the only Illinois site to sample PMio metals under the NMP in 2015 and
2016.
• The range of arsenic concentrations measured at NBIL in 2015 is considerably larger
than the range measured in 2016; the maximum arsenic concentration measured at
NBIL in 2015 is more than three times greater than the maximum arsenic
concentration measured at NBIL in 2016. This measurement is also the second
highest arsenic concentration measured across the program.
• Despite the differences in the range of concentrations, the annual average
concentrations of arsenic for NBIL are fairly similar; both annual averages are greater
than the program-level average (0.70 ng/m3), with the annual average for 2015 also
greater than the program-level third quartile.
• Non-detects of arsenic were not measured at NBIL, although a few were measured
across the program.
10-23
-------
Figure 10-9. Program vs. Site-Specific Average Benzene Concentrations
r
¦4-
E T
I
012345678
Concentration (|ig/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 10-9 presents the box plots for benzene for all three sites and shows the following:
• Despite the shortened sampling duration, the range of benzene concentrations
measured was larger for ROIL than the two Chicago sites. Twice the number of
benzene concentrations greater than 2 |ig/m3 were measured at ROIL (four) than
SPIL (two), with none measured at NBIL. Among the 27 benzene concentrations
greater than 1 |ig/m3 measured at these sites, only two were measured at NBIL,
compared to nine at SPIL and 16 at ROIL.
• Both of NBIL's annual average benzene concentrations are less than the program-
level median concentration, with the annual average for 2016 similar to the program-
level first quartile and the fifth lowest among NMP sites sampling this pollutant.
• SPIL's annual average benzene concentrations are both greater than the program-
level median concentration, with the annual average for 2015 similar to the program-
level average concentration.
10-24
-------
Figure 10-10. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
NBIL
p
i r ,
| Program Max Concentration = 3.90 (.ig/m3 i
1 1
SPIL
i
Program Max Concentration = 3.90 (.ig/m3 i
7
H
r*
" 1
¦ 1, r
1 Program Max Concentration =3.90 (.ig/m3 i
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (]ug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 10-10 presents the box plots for 1,3-butadiene for all three sites and shows the
following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plots as the scale has been reduced to allow for the observation of
data points at the lower end of the concentration range. Only one 1,3-butatdiene
concentration measured at the Illinois sites is greater than one-tenth of the program-
level maximum concentration.
• The range of 1,3-butadiene concentrations measured at NBIL is similar to the range
measured at ROIL, with a few non-detects measured at these two sites. The maximum
1,3-butadiene concentrations measured at these two sites are fairly similar to each
other.
• The range of 1,3-butadiene concentrations measured at SPIL in 2015 is more than
twice the range measured in 2016. Despite this, the annual average concentrations for
both years are similar to each other, with both greater than the program-level average
concentration and third quartile. The annual average 1,3-butadiene concentrations for
SPIL are more than twice the annual averages for NBIL.
10-25
-------
Figure 10-11. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
NBIL
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 10-11 presents the box plots for carbon tetrachloride for all three sites and shows
the following:
• Less than 0.6 |ig/m3 separates the minimum and maximum carbon tetrachloride
concentrations measured at the Illinois sites, with both the minimum and maximum
concentrations measured at NBIL.
• The program-level average and median concentrations are similar to each other and
are plotted nearly on top of one another in Figure 10-11. The annual average
concentrations of carbon tetrachloride for SPIL are similar for 2015 and 2016, both of
which are just less than the program-level median and average concentrations. For
NBIL, the 2015 annual average concentration is greater than the annual average for
2016, with the 2015 annual average similar to the program-level average and median
concentrations and the 2016 annual average just less than the program-level first
quartile.
10-26
-------
Figure 10-12. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
i
[ Program Max Concentration = 2.78 jxg/m3 ¦
NBIL 2015 Max Concentration = 2.78}ig/m3 |
SPIL
J
| Program MaxConcentration
= 2.78 |ig/m3
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (jug/rn3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 10-12 presents the box plots for p-dichlorobenzene for the Chicago sites and
shows the following:
• The program-level maximum /;-dichlorobenzene concentration (2.78 |ig/m3) is not
shown directly on the box plots as the scale has been reduced. Note that the program-
level first and second quartiles for p-dichlorobenzene are both zero, and therefore not
visible on the box plots. This pollutant has a 43 percent detection rate across the
program.
• The maximum concentration of p-dichlorobenzene measured at NBIL in 2015 is the
maximum concentration measured across the program. While this is the only
/;-dichlorobenzene concentration greater than 1 |ig/m3 measured at NBIL, six
additional p-dichlorobenzene concentrations measured in 2015 are greater than the
maximum concentration measured at this site in 2016. The maximum concentration
measured at NBIL in 2016 is similar to the maximum concentrations measured each
year at SPIL.
• NBIL's 2016 annual average concentration of p-dichlorobenzene is similar to both of
SPIL's annual average concentrations, all three of which are less than the program-
level average concentration. By comparison, NBIL's 2015 annual average
concentration of p-dichlorobenzene is nearly three times greater than the program-
level average concentration. As noted in the previous section, this annual average
concentration ranks seventh highest among NMP sites sampling this pollutant. Even
if the maximum concentration was excluded from the dataset, NBIL's ranking would
not change.
• A box plot for ROIL is not presented in Figure 10-12 because/?-dichlorobenzene was
not identified as a pollutant of interest for this site.
10-27
-------
Figure 10-13. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1
*
W
1 -j
| Program Max Concentration =45.8 (.ig/m3
1
1
fr
, ________ ,
Program Max Concentration = 45.8 }ig/m3 i
1
SPIL
¦
ROIL
Program Max Concentration = 45.8 p.g/m3 ¦
0.50 0.75 1.00
Concentration (jug/rn3)
Program: 1st Quartile 2nd Quartile 3rd Quartile 4thQuartile Ave rag
¦ ~ ~ ~ i
Site: 2015Average 2016Aveage Concentration Range, 2015 & 2016
o
Figure 10-13 presents the box plots for 1,2-dichloroethane for all three sites and shows
the following:
• The scale of the box plots for 1,2-dichloroethane has also been reduced to allow for
the observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably higher than the majority of measurements. This is an example of
measurements at the upper end of the concentration range driving the program-level
average concentration, as the program-level average is nearly three times greater than
the program-level third quartile.
• All of the 1,2-dichloroethane concentrations measured at the Illinois sites are less
than the program-level average concentration of 0.30 |ig/m3. In fact, all but one
measurement are less than 0.15 |ig/m3.
• Less than 0.01 |ig/m3 separates the annual average concentrations of
1,2-dichloroethane for NBIL and SPIL. Each of these fall between the program-level
first quartile and median concentrations.
10-28
-------
Figure 10-14. Program vs. Site-Specific Average Fluoranthene Concentrations
NBIL
Program Max Concentration = 57.3 ng/m3
NBIL 2015 Max= 51.6 ng/m3; 2016 Max = 57.3 ng/m3
5 10 15 20 25 30 35 40
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
~ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
O o
Figure 10-14 presents the box plot for fluoranthene for NBIL and shows the following:
• NBIL is the only Illinois site to sample PAHs under the NMP in 2015 and 2016. The
program-level maximum concentration (57.3 ng/m3) of fluoranthene is not shown
directly on the box plot because the scale has been reduced.
• The maximum fluoranthene concentrations measured each year at NBIL exceed the
scale of the box plot and thus, are provided directly on the box plot. In total, six
fluoranthene concentrations measured at NBIL exceed the scale of the box plot (three
each year). These are also the six highest fluoranthene concentrations measured
across the program.
• NBIL's annual average concentration of fluoranthene for 2016 is greater than the
annual average for 2015, although both are several times greater than the program-
level average of 2.39 ng/m3. More than half of NBIL's individual fluoranthene
measurements are greater than the program-level average concentration.
Figure 10-15. Program vs. Site-Specific Average Fluorene Concentrations
1
! Program Max Concentration = 105 ng/m3 |
f ,
, NBIL 2015 Max= 96.0 ng/m3; 2016 Max =91.7 ng/m3 I
30
Concentration (ng/m3]
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
~ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
e o
Figure 10-15 presents the box plot for fluorene for NBIL and shows the following:
• Fluorene is another PAIT pollutant of interest for NBIL. Similar to the box plots for
the other PAHs, the program-level maximum concentration (105 ng/m3) of fluorene is
10-29
-------
not shown directly on the box plot because the scale has been reduced. Note that the
program-level first quartile is zero and thus, not visible on the box plot.
• The maximum fluorene concentrations measured each year at NBIL also exceed the
scale of the box plot and thus, are also provided directly on the box plot. In total, nine
fluorene concentrations measured at NBIL exceed the scale of the box plot. Only two
NMP sites have individual fluorene concentrations greater than 45 ng/m3, with
concentrations measured at NBIL accounting for nearly all of them (15 out of 16).
• NBIL's annual average fluorene concentrations for 2015 and 2016 are nearly four and
five times greater than the program-level average concentration of 4.36 ng/m3; more
than half of NBIL's fluorene measurements are greater than the program-level
average concentration. A similar observation was made in the 2014 NMP report.
Figure 10-16. Program vs. Site-Specific Average Formaldehyde Concentrations
NBIL
SPIL
S 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
9
o
Figure 10-16 presents the box plots for formaldehyde for all three sites and shows the
following:
• The range of formaldehyde concentrations measured at these sites is smallest for
NBIL and largest for SPIL.
• The range of formaldehyde concentrations measured each year at SPIL are similar to
each other. At NBIL, the range of concentrations measured in 2016 is larger than the
range measured in 2015. Six formaldehyde concentrations measured in 2016, ranging
10-30
-------
from 4.80 |ig/m3 to 7.11 |ig/m3, are greater than the maximum formaldehyde
concentration measured in 2015 (4.30 |ig/m3).
Both annual average concentrations of formaldehyde for NBIL are less than the
program4evel median concentration of 2.47 |ig/m3. Both annual average
concentrations of formaldehyde for SPIL are greater than the program-level average
concentration of 3.05 |ig/m3, with the annual average for 2015 also greater than the
program-level third quartile.
Only one formaldehyde concentration measured at SPIL is less than the program-
level first quartile.
Figure 10-17. Program vs. Site-Specific Average Naphthalene Concentrations
200 250
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
e o
Figure 10-17 presents the box plot for naphthalene for NBIL and shows the following:
• The maximum naphthalene concentration measured at NBIL is the maximum
concentration measured across the program. This was also true in 2013 and 2014.
NBIL is the only NMP site at which a naphthalene concentration greater than
400 ng/m3 was measured.
• Both the minimum and maximum concentrations of naphthalene were measured at
NBIL in 2015. Note the difference between the minimum concentration measured at
NBIL in 2015 (0.446 ng/m3) and the minimum concentration measured at NBIL in
2016 (13.0 ng/m3).
• NBIL's annual average naphthalene concentrations are both greater than the program-
level average concentration of 61.23 ng/m3 and fall one either side of the program-
level third quartile.
10-31
-------
Figure 10-18. Program vs. Site-Specific Average Trichloroethylene Concentrations
1 A
'
0 1 2 3 4 5 6
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 10-18 presents the box plot for trichloroethylene for SPIL and shows the
following:
• SPIL is the only NMP site for which trichloroethylene is a pollutant of interest.
• The first, second, and third quartiles for trichloroethylene are zero due to the large
number of non-detects; thus, only the fourth quartile is visible in Figure 10-18. The
detection rate of trichloroethylene across the program in 2015 and 2016 is 21 percent.
• Although the maximum concentration of trichloroethylene across the program was
measured at NBIL, the next eight highest concentrations were measured at SPIL.
Among NMP sites sampling this pollutant, SPIL has the greatest number of measured
detections (90), with the next closest site at CSN (54). (NBIL is fourth with 38.)
• The annual average concentrations of trichloroethylene for SPIL are an order of
magnitude higher than the program-level average concentration (0.030 |ig/m3),
10.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
NBIL and SPIL have both sampled VOCs under the NMP since 2003. Both sites have also
sampled carbonyl compounds since 2005. NBIL has also sampled PMio metals since 2005 and
began sampling PAHs under the NMP in 2008. Thus, Figures 10-19 through 10-38 present the
1-year statistical metrics for each of the pollutants of interest first for NBIL, then for SPIL. The
statistical metrics presented for assessing trends include the substitution of zeros for non-detects.
If sampling began mid-year, a minimum of 6 months of sampling is required for inclusion in the
trends analysis; in these cases, a 1-year average concentration is not provided, although the range
and percentiles are still presented. Because sampling at ROIL began in 2012 and ended in 2015,
a trends analysis was not performed.
10-32
-------
Figure 10-19. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
NBIL
1
o
¦-o-
..<>•"'
2012
Year
..<>
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 10-19 for acenaphthene concentrations measured at NBIL
include the following:
• Although PAH sampling under the NMP began at NBIL in 2008, sampling did not
begin until June; because a full year's worth of data is not available for 2008, a 1-year
average is not presented, although the range of measurements is provided.
• The maximum acenaphthene concentration was measured at NBIL on August 9, 2014
(198 ng/m3), with three additional acenaphthene concentrations greater than
100 ng/m3 measured at NBIL (two in 2013 and one in 2015). All but one of the 11
acenaphthene concentrations greater than 75 ng/m3 were measured at NBIL in 2013
or later.
• The median concentration decreased significantly from 2008 to 2009. This is because
there are a greater number of concentrations at the lower end of the concentration
range in 2009. The number of acenaphthene concentrations less than 2 ng/m3
increased from seven in 2008 to 24 in 2009. As previously noted, 2008 does not
include a full year's worth of sampling. The median concentration increases steadily
between 2009 and 2012, after which the median doubles for 2013, and changes little
for 2014.
• The 1-year average concentration increases between 2009 and 2011, nearly doubling
over this time frame. However, confidence intervals calculated for these averages
10-33
-------
indicate that the increase is not statistically significant due to the relatively large
amount of variability in the measurements. The 1-year average decreased slightly for
2012, although the median continued to increase. For 2013, the 1-year average
concentration more than doubled, with similar increases for the median, 95th
percentile, and maximum concentration. Five acenaphthene concentrations measured
in 2013 are greater than the maximum concentrations measured in 2012.
Additionally, the number of acenaphthene concentrations greater than 50 ng/m3
measured at NBIL increased from one in 2012 to 11 in 2013, with no more than four
measured in any of the previous years.
• Even though the maximum concentration measured approached 200 ng/m3 in 2014,
the 95th percentile decreased considerably and the 1-year average exhibits a decrease
as well (while the median changed little). The number of acenaphthene concentrations
greater than 75 ng/m3 measured at NBIL decreased from five in 2013 to one in 2014.
Between 2013 and 2016, the median concentration decreases by half. A slight
decrease is also shown in the 1-year average concentration for 2015, followed by a
slight increase for 2016.
• Although difficult to discern, the only non-detects of acenaphthene measured at NBIL
were measured in 2015 (5) and 2016 (1).
Figure 10-20. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
NBIL
o
•o-
1 ,
—
9.
1
i£i
o
-2-
h 4,
o
L¥-
¦O
—Q-
I
U|J
20051 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2005.
10-34
-------
Observations from Figure 10-20 for acetaldehyde concentrations measured atNBIL
include the following:
• Carbonyl compound sampling at NBIL under the NMP began in March 2005;
because a full year's worth of data is not available for 2005, a 1-year average
concentration is not presented, although the range of measurements is provided.
• The maximum acetaldehyde concentration measured at NBIL (9.17 |ig/m3) was
measured in 2014, along with four of the five highest concentrations since the onset
of sampling. The 14 highest acetaldehyde concentrations were measured in 2013 and
2014 and all 44 acetaldehyde concentrations greater than 3 |ig/m3 measured at NBIL
were measured after 2009 (with the majority, 15 each, measured in 2013 and 2014).
• Prior to 2010, the 1-year average concentrations were all less than 1 |ig/m3,
fluctuating between 0.69 |ig/m3 (2009) and 0.98 |ig/m3 (2006). Acetaldehyde
concentrations measured at NBIL increased significantly after 2009. The steady
increasing trend continues through 2013. The increase in the 1-year average
concentration of acetaldehyde between 2009 and 2013 represents a 243 percent
increase.
• The range of acetaldehyde concentrations measured at NBIL expanded in 2014. Two
acetaldehyde concentrations greater than the maximum concentration for 2013 were
measured in 2014 while seven concentrations less than the minimum concentration
for 2013 were measured in 2014. Yet, little difference is shown in the 1-year average
concentration between these two years. The median concentration decreased slightly
for 2014, but is still greater than the 1-year average and median concentrations shown
for all years prior to 2013.
• A significant decrease in the acetaldehyde concentrations measured at NBIL is shown
for 2015, with the smallest range of concentrations measured since 2009, and a
40 percent decrease in the 1-year average concentration. While the range of
measurements increased somewhat for 2016, little change is shown in the 1-year
average concentration, and the median exhibits a slight decrease.
10-35
-------
Figure 10-21. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
NBIL
2005 2006 2007
•o
-2-j ¦—g-
2008 2009 2010 2011 2012 2013 2014
Year
o
2015 2016
O 5th Percentile
O 95th Percentile
Observations from Figure 10-21 for arsenic (PMio) concentrations measured at NBIL
include the following:
• Metals sampling under the NMP began at NBIL in January 2005.
• The maximum arsenic concentration was measured at NBIL on July 5, 2015
(7.18 ng/m3). Additional arsenic concentrations greater than 4 ng/m3 have not been
measured at this site. Excluding the maximum, five arsenic concentrations greater
than 3 ng/m3 have been measured at NBIL (one each in 2006, 2009, and 2015, and
two in 2010).
• Although the statistical parameters representing the upper end of the concentration
range have fluctuated somewhat each year, the 1-year average concentrations exhibit
relatively little significant change over most of the years sampling. Between 2005 and
2014, less than 0.25 ng/m3 separates the 1-year average concentrations, including a
5-year period where the 1-year average concentration hovered around 0.75 ng/m3.
• Most of the statistical parameters are at a minimum for 2013, with the 1-year average
concentration (0.62 ng/m3) at its lowest since the onset of sampling. Little change is
shown for 2014, after which most of the statistical parameters exhibit increases for
2015. This would be true even if the maximum concentration was excluded from the
dataset. The 1-year average concentration is at a maximum for 2015 (0.94 ng/m3).
10-36
-------
• While the maximum, 95th percentile, and 1-year average concentration exhibit
decrease for 2016, the minimum, 5th percentile, and median concentrations exhibit
increases. In fact, these latter three statistical parameters are at their highest since the
onset of sampling.
Figure 10-22. Yearly Statistical Metrics for Benzene Concentrations Measured at NBIL
i—5—i
1
o
i
A.
LyJ
I
i
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 through the
end of the year.
Observations from Figure 10-22 for benzene concentrations measured at NBIL include
the following:
• Although sampling for VOCs at NBIL under the NMP began in 2003, sampling did
not begin until April; because a full year's worth of data is not available, a 1-year
average is not presented, although the range of measurements is provided. In addition,
sampling for VOCs was discontinued in October 2004 through the end of the year.
Thus, a 1-year average is not presented for 2004 either.
• The maximum benzene concentration (4.51 |ig/m3) was measured on January 9, 2011
and is the only benzene measurement greater than 4 |ig/m3 measured at NBIL. Three
additional benzene concentrations greater than 3 |ig/m3 were measured in 2004 and
2005 and most of the measurements greater than 2 |ig/m3 were measured in 2004.
• A decreasing trend in the concentrations of benzene is shown through 2007. The
1-year average concentration decreased significantly from 2005 to 2006, with an
additional slight decrease for 2007. Between 2005 and 2007, the 1-year average
10-37
-------
concentration decreased by almost half, after which the 1-year average remained
steady through 2009.
• All of the statistical parameters exhibit increases from 2009 to 2010. Although the
maximum concentration nearly doubled from 2010 to 2011, the rest of the statistical
parameters decreased for 2011 (albeit slightly). This decreasing continued into 2012
(although the median concentration actually increased slightly) and 2013, which is the
first year the 1-year average concentration is less than 0.5 |ig/m3.
• Benzene concentrations greater than 1.5 |ig/m3 were not measured at NBIL during the
4-year period between 2013 and 2016 and benzene concentrations greater than
1 |ig/m3 were not measured in 2016. The entire range of concentrations spans less
than 0.75 |ig/m3 in 2016, when the 1-year average concentration is at a minimum
(0.43 |ig/m3).
Figure 10-23. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
NBIL
6
«-
X
Maximum concentration
for 2011 is 2.68 p.g/m3
o
I
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 through the
end of the year.
Observations from Figure 10-23 for 1,3-butadiene concentrations measured at NBIL
include the following:
• The maximum 1,3-butadiene concentration was measured at NBIL on the same day
as the maximum benzene concentration, January 9, 2011 (2.68 |ig/m3). Only three
1,3-butadiene concentrations greater than 1 |ig/m3 have been measured at NBIL (two
10-38
-------
in 2011 and one in 2010). All other concentrations of 1,3-butadiene measured at
NBIL are less than 0.35 |ig/m3.
For each year shown through 2014, the minimum and 5th percentile are zero,
indicating the presence of non-detects. For the first 2 years of sampling, the median
concentration is also zero, indicating that at least half of the measurements were non-
detects. The percentage of non-detects reported has fluctuated over the years of
sampling, from as high as 88 percent (2004) to as low as 5 percent (2016).
The 1-year average concentration decreased slightly between 2005 and 2009,
although the changes are not statistically significant. From 2009 to 2010, the 1-year
average doubled, and then nearly doubled again for 2011. There is, however, a
significant amount of variability associated with these measurements, based on the
confidence intervals associated with these averages. Even with the relatively high
concentrations measured in 2010 and 2011, the 95th percentile changed only slightly,
indicating that the majority of concentrations measured were within a similar range. If
the three outlier concentrations measured in 2010 and 2011 were excluded from the
calculations, the 1-year average concentrations for these years would still be greater
than the 1-year average for 2009, but they would be similar to the averages shown for
years prior.
Excluding the two years with outliers, the 1-year average and median concentrations
are highest for 2012. Although the range of concentrations measured is similar to
other years, 2012 has the highest number of 1,3-butadiene concentrations (13) greater
than 0.1 |ig/m3 than any other year of sampling.
The range within which the majority of concentrations fall, as determined by the 5th
and 95th percentiles, decreased considerably from 2012 to 2013, and remained fairly
steady through 2016.
The 1-year average concentration decreased significantly from 2012 to 2013, and
varies by about 0.01 |ig/m3 in the years that follow.
10-39
-------
Figure 10-24. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at NBIL
I
Maxi mum Concentration
for 2014 is 4.98 ng/m3
I
I
¦o
f
I
X
r5n
r-5-i
f
1
I
1
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 through the
end of the year.
Observations from Figure 10-24 for carbon tetrachloride concentrations measured at
NBIL include the following:
• The maximum concentration of carbon tetrachloride was measured in 2004
(4.98 |ig/m3). Only one additional measurement greater than 1.5 |ig/m3 has been
measured at NBIL (1.88 |ig/m3 in 2012).
• Five non-detects of carbon tetrachloride have been measured at NBIL. All of these
were measured during the first 2 years of sampling (two in 2003 and three in 2004).
• The statistical parameters for 2003 and 2004 have a different appearance than the
parameters shown for the years afterward, particularly for 2004, when the range of
measurements is at its largest and several non-detects were measured. However,
neither year represents a full year of sampling.
• After decreasing slightly between 2005 and 2007, the 1-year average concentration
increased significantly for 2008. This increase is followed by a significant decreasing
trend that continued through 2011, when the 1-year average returned to 2007 levels.
After exhibiting an increase for 2012, the 1-year average concentration decreased
again for 2013 and is less than 0.60 |ig/m3 for the first time. The slight increases
shown for 2014 and 2015 are followed by a decrease for 2016, when the 1-year
10-40
-------
average concentration is at a minimum (0.56 |ig/m3). However, less than 0.07 |ig/m3
separates the 1-year averages calculated between 2013 and 2016.
• The median carbon tetrachloride concentration has a similar pattern as the 1-year
average concentration and is also at a minimum for 2016 (0.56 |ig/m3).
Figure 10-25. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
NBIL
I ,
. a i
A
p Q $ £] dt] j$ "'1
J £i
20031 2004 2 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 through the
end of the year.
Observations from Figure 10-25 for /;-dichlorobenzene concentrations measured at NBIL
include the following:
• The maximum concentration of p-dichlorobenzene was measured in 2007
(3.00 |ig/m3), although a concentration of similar magnitude was also measured in
2015 (2.78 |ig/m3). In total, eight for /;-dichlorobenzene concentrations greater than
1 |ig/m3 have been measured at NBIL, the majority of which were measured in 2007
(five).
• For each year shown, the minimum and 5th percentile are zero, indicating the
presence of non-detects. The median concentration is also zero for several years,
indicating that at least half of the measurements were non-detects. The percentage of
non-detects reported has varied over the years of sampling, from 28 percent (2007) to
97 percent (2003).
10-41
-------
• In 2003 and 2004, non-detects account for nearly all of the measurements (only one
and two measured detections, respectively). The percentage of non-detects decreased
considerably each year through 2007, when the number of non-detects is at a
minimum and the magnitude of the measured detections is at a maximum. The 1-year
average concentration of p-dichlorobenzene increased six-fold between 2005 and
2007. The 1-year average concentration for 2007 is greater than the 95th percentile
for most other years of sampling.
• A significant decrease in concentrations of p-dichlorobenzene is shown after 2007
and through 2009, when the range of measurements is at its smallest since the first
year of sampling. Between 2010 and 2014, the 1-year average concentration has a
undulating pattern, where a slightly higher 1-year average is followed by a slightly
lower 1-year average. The median concentration returned to zero during this 5-year
period, indicating that at least half of the measurements are non-detects.
• An increase is shown for nearly all of the statistical parameters for 2015, when four
concentrations greater than 0.5 |ig/m3 were measured, the first year since 2010 and
the most since 2007. This increase is followed by a return to levels shown prior to
2015.
Figure 10-26. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at NBIL
•Vv Ur i-Qt"
I
X
.0
o
¦
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
2 A 1-year average is not presented because there was a gap in sampling from late October 2004 through the
end of the year.
10-42
-------
Observations from Figure 10-26 for 1,2-dichloroethane concentrations measured atNBIL
include the following:
• There were no measured detections of 1,2-dichloroethane in 2003, 2004, or 2008. The
percentage of non-detects between 2005 and 2007 was greater than 95 percent. Thus,
the minimum, 5th percentile, median, and in some cases the 1-year average
concentrations, were zero between 2003 and 2008. The median concentration
remained zero through 2011, indicating that at least half of the measurements are non-
detects.
• The number of non-detects began to decrease starting with 2009 and continued
through 2012. The median concentration is greater than zero for the first time for
2012 and is also greater than the 1-year average concentration. This is because the
eight non-detects (or zeros) factored into the 1-year average concentration are pulling
the average down (in the same manner that a maximum or outlier concentration can
drive the average up) but are not contributing to the majority of measurements for the
first time. This is also true for 2013, although the number of non-detects increased
slightly (from 8 to 10).
• The 5th percentile is greater than zero for the first time for 2014, when only two non-
detects of 1,2-dichloroethane were measured atNBIL. This is also true for 2015,
when a single non-detect was measured. The 5th percentile returned to zero for 2016,
when the number of non-detects increased considerably (nine).
• Between 2012 and 2016, the 1-year average concentrations ranged from 0.06 |ig/m3
(2013) to 0.07 |ig/m3 (2015), varying by only 0.01 |ig/m3. If 2013 is excluded, the
range varies by less than 0.001 |ig/m3. The decrease in the 1-year average for 2013
does not result solely based on the two additional non-detects measured, though they
are a factor. The 95th percentile for 2012 is the same as the maximum concentration
measured in 2013; seven concentrations greater than 0.1 |ig/m3 were measured in
2012 compared to only two in 2013. While this is also true for 2014, the low number
of 1,2-dichloroethane concentrations greater than 0.1 |ig/m3 measured in 2014 is
counterbalanced by the reduced number of non-detects, resulting in a 1-year average
concentration more similar to the one calculated for 2012.
10-43
-------
Figure 10-27. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at
NBIL
70.0
60.0
50.0
J> 40.0
c
0
c
01
£ 30.0
o
20.0
10.0
0.0
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 10-27 for fluoranthene concentrations measured at NBIL
include the following:
• Four fluoranthene concentrations greater than 50 ng/m3 have been measured at NBIL,
three in 2016 and one in 2015.
• The median concentration decreased by more than half from 2008 to 2009. This is
because there is a greater number of fluoranthene concentrations at the lower end of
the concentration range for 2009. The number of measurements less than 2 ng/m3
tripled from 2008 to 2009, increasing from nine in 2008 to 27 in 2009. Note,
however, that 2008 does not include a full year's worth of sampling. The median
fluoranthene concentrations shown between 2009 and 2014 vary by less than 2 ng/m3.
• An overall increasing trend is shown in the concentrations of fluoranthene measured
at NBIL. The 1-year average concentration has doubled since the first full year of
sampling. Confidence intervals calculated for these averages, though, indicate a
relatively large amount of variability in the measurements.
10-44
2008 2009 2010 2011
2012
Year
2013 2014 2015 2016
-------
Figure 10-28. Yearly Statistical Metrics for Fluorene Concentrations Measured at NBIL
-o- •
2012
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 10-28 for fluorene concentrations measured at NBIL include
the following:
• The statistical patterns for fluorene resemble the statistical patterns shown on the
trends graph for acenaphthene and, to a lesser extent, fluoranthene.
• The maximum fluorene concentration was measured at NBIL in 2014 (161 ng/m3),
and is the only concentration greater than 100 ng/m3 measured since the onset of
sampling (although one greater than 90 ng/m3 has been measured each year since
2012).
• The median concentration of fluorene also decreased significantly from 2008 to 2009
due to the number of fluorene concentrations at the lower end of the concentration
range for 2009. The number of measurements less than 3 ng/m3 increased three-fold
from 2008 to 2009, increasing from eight in 2008 to 29 in 2009. Note, however, that
2008 does not include a full year's worth of sampling.
• Like acenaphthene, the 1-year average concentration of fluorene increases between
2009 and 2011, then decreases slightly for 2012. The 1-year average concentration
then increases considerably for 2013, after which the 1-year average concentration
lies between 15 ng/m3 and 20 ng/m3 for each year. Confidence intervals calculated for
these averages indicate that there is a relatively large amount of variability in these
measurements.
10-45
-------
• Non-detects of fluorene were not measured at NBIL until 2013. Both the minimum
concentration and the 5th percentile are zero for 2014, 2015, and 2016, with between
five (2016) and seven (2015) non-detects measured each year during this period.
Figure 10-29. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
NBIL
Maximum Concentration
for 2006 is 91.7 M-g/m3
Maximum Concentration
for 2010 is 53.5 M-g/m3
2
2010 2011
Year
O 5th Percentile
— Minimum
O 95th Percentile Averz
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2005.
Observations from Figure 10-29 for formaldehyde concentrations measured at NBIL
include the following:
• The maximum formaldehyde concentration was measured on January 5, 2006
(91.7 |ig/m3). The next five highest concentrations, ranging from 14.4 |ig/m3 to
53.5 |ig/m3, were all measured in 2010. The only other formaldehyde concentration
greater than 10 |ig/m3 was measured in 2011 (13.7 |ig/m3).
• The maximum concentration measured in 2006 is 20 times higher than the next
highest concentration measured that year (4.46 |ig/m3). The magnitude of this outlier
explains why the 1-year average concentration for 2006 is greater than the 95th
percentile.
• The range of formaldehyde concentrations measured in 2007, 2008, and 2009, and
thus, the statistical parameters characterizing them, is considerably less than the
previous two years.
• The statistical metrics for 2010 are also affected by higher concentrations; however,
concentrations measured this year are higher overall, as indicated by seven-fold
10-46
-------
increase in the 95th percentile. The 1-year average concentration more than tripled
from 2009 to 2010 and the median increased by 50 percent. The concentrations
measured in 2011 were less than those measured in 2010, although still higher than
most years.
Although the maximum formaldehyde concentration measured in 2012 is less than
the 95th percentile for 2011, the 1-year average concentration changed little and the
median concentration increased. This is because the number of concentrations
between 2 |ig/m3 and 4 |ig/m3 nearly doubled from 2011 (15) to 2012 (29).
The range of concentrations measured at NBIL after 2010 has a decreasing trend
through 2014. The 1-year average concentration exhibits a 64 percent decrease
between 2010 and 2014, although those prior to 2010 are still lower.
Despite a smaller range of concentrations measured, the 1-year average concentration
increased significantly for 2015. Concentrations measured in 2015 were generally
higher than those measured in 2014. The number of formaldehyde concentrations
greater than 2 |ig/m3 increased three-fold, from eight in 2014 to 25 in 2015.
Additionally, the number of concentrations less than 1 |ig/m3 decreased from 21 in
2014 to seven in 2015; further, 11 concentrations measured in 2014 are less than the
minimum concentration measured in 2015.
The concentration profile for 2016 indicates that a few higher concentrations were
measured; six concentrations measured at NBIL in 2016 are greater than the
maximum concentration measured in 2015. Only slight changes in the central
tendency statistics are shown for 2016.
10-47
-------
Figure 10-30. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
NBIL
1000
900
800
700
JT" 600
c
o
4= 500
c
QJ
C
u 400
300
200
100
0
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 10-30 for naphthalene concentrations measured at NBIL
include the following:
• The maximum naphthalene concentration was measured on September 23, 2010
(869 ng/m3). Five additional naphthalene concentrations greater than 500 ng/m3 have
been measured at NBIL (one in 2011, three in 2013, and one in 2014).
• The 1-year average and median concentrations increase considerably from 2009 to
2010, when the maximum naphthalene concentration was measured. While this
measurement (869 ng/m3) is more than twice the next highest concentration measured
in 2010 (363 ng/m3), the increases are not solely a result of this outlier. Four
concentrations measured in 2010 are greater than the maximum concentration
measured in 2009, and the number of naphthalene concentrations greater than or
equal to 100 ng/m3 increased from 14 to 22. The concentration profile for 2011
resembles the profile for 2010, with only slight decreases shown for most of the
statistical parameters.
• After additional decreases for 2012, naphthalene concentrations exhibit significant
increases for 2013, when the 1-year average concentration doubled, and is at a
maximum for the period of sampling (155.98 ng/m3). The highest number of
naphthalene concentrations greater than 300 ng/m3 was measured in 2013 (11), with
no other year having more than three.
~Qr
2011
2012
Year
10-48
-------
• After 2013, a significant decreasing trend is shown in the concentrations of
naphthalene measured at NBIL.
Figure 10-31. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SPIL
r-5-i
o
L-^J
r-S-i
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 10-31 for acetaldehyde concentrations measured at SPIL
include the following:
• Although the first carbonyl compound sample was collected at SPIL in February
2005, consistent sampling did not begin until March 2005; because a full year's worth
of data is not available for 2005, a 1-year average is not presented, although the range
of measurements is provided.
• The maximum acetaldehyde concentration was measured at SPIL on
November 17, 2012 (20.4 |ig/m3). Thirty-six of the 38 concentrations of acetaldehyde
greater than 5 |ig/m3 were measured after 2010 (eight each in 2011, 2012, and 2014,
seven in 2013, one in 2015, and four in 2016); the other two were measured in 2006.
• The 1-year average concentration decreased significantly from 2006 to 2007, as did
most of the other statistical parameters. Between 2007 and 2009, the 1-year average
concentration changed little, hovering between 1.25 |ig/m3 and 1.45 |ig/m3. The
1-year average concentration increased slightly in 2010 then increased significantly in
2011. All of the statistical metrics increased for 2011, particularly the maximum and
95th percentile, indicating that the increases shown are not attributable to a few of
10-49
-------
outliers. As an illustration, the number of measurements greater than 2 |ig/m3
increased from three in 2009 to 15 for 2010 to 40 in 2011.
• The concentration profiles for 2012 through 2016 are more similar to 2011 than other
years of sampling. Yet, these measurements reflect considerable variability, based on
the range of concentrations measured and spread of the central tendency statistics.
Even though the 95th percentile for 2015 decreased to less than 5 |ig/m3 for the first
times since 2010, the 1-year average concentration changed little and the median
concentration exhibits an increase. The second highest acetaldehyde concentration
was measured at SPIL in 2015 (17.1 |ig/m3) but is more than four times higher than
the next highest measurement that year (3.66 |ig/m3). Even though this disparity is
large, it balances out with other years where the spread of concentrations higher in
magnitude is less. In addition, 2015 has the most concentrations greater than 2 |ig/m3
since 2011, which explains the increase in the median concentration. These are the
only two years in which the median concentration is greater than 2 |ig/m3.
Figure 10-32. Yearly Statistical Metrics for Benzene Concentrations Measured at SPIL
y
T
T
y
o
T
¦0
1
,
o
Lr
1
LfJ
4
t
£
<
4
£
X
X
<
I
~
I
_0
1
O
11 i
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 10-32 for benzene concentrations measured at SPIL include the
following:
• Sampling for VOCs at SPIL under the NMP began in April 2003; because a full
year's worth of data is not available for 2003, a 1-year average is not presented,
although the range of measurements is provided.
10-50
-------
• The only two concentrations of benzene greater than 5 |ig/m3 were measured at SPIL
in 2005.
• The 1-year average benzene concentration has a significant decreasing trend between
2004 and 2009, decreasing from 1.52 |ig/m3 to 0.68 |ig/m3 during this time.
• The 1-year average concentration increased significantly from 2009 to 2010. While
the maximum concentration measured changed little between these two years, five
concentrations measured in 2009 are less than the minimum concentration measured
in 2010. In addition, the number of benzene concentrations greater than 1 |ig/m3
increased from five in 2009 to 19 in 2010.
• Between 2010 and 2014, the 1-year average benzene concentration has an undulating
pattern, varying between 0.74 |ig/m3 (2013) and 0.95 |ig/m3 (2012). The majority of
benzene concentrations measured at SPIL, as indicated by the 5th and 95th
percentiles, fell within roughly the same range between 2010 and 2014, with the
exception of 2013, when the range of benzene concentrations is slightly smaller than
other recent years.
• The decreasing trend in benzene concentrations resumed at SPIL after 2014. The
range of benzene concentrations measured is smallest for 2016, with just over
1 |ig/m3 separating the minimum and maximum benzene concentrations. Further,
several of the statistical parameters are at a minimum for 2016, including the 1-year
average concentration.
Figure 10-33. Yearly Statistical Metrics for 1,3-Butadiene Concentrations
Measured at SPIL
r
T
I
I
X
I
|-J-|
I
—
a
* <
>.
<6.
L2"
£l.
L2"
O
"
&
T
•«
i <
-0
1
¦ V
^ 4
"O
J V
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
10-51
-------
Observations from Figure 10-33 for 1,3-butadiene measurements at SPIL include the
following:
• The maximum concentration of 1,3-butadiene was measured on February 3, 2005
(1.29 |ig/m3) and is the only measurement greater than 1 |ig/m3 measured at SPIL. In
total, eight concentrations greater than 0.5 |ig/m3 have been measured at SPIL, one in
2004, two in 2005, two in 2011, and one each in 2012, 2013, and 2015.
• The detection rate for 1,3-butadiene increased over the first several years of sampling;
beginning with the first full year of sampling, the detection rate increased from
54 percent measured detections in 2004 to a 100 percent detection rate in 2008.
Between 2007 and 2016, no more than one non-detect was measured in any given
year, with the detection rate varying between 98 percent and 100 percent.
• The 1-year average concentration of 1,3-butadiene changed little between 2004 and
2006, then decreased significantly between 2006 and 2009. The significant increase
in the 1-year average concentration from 2009 to 2010 represents a 67 percent
increase and a return to 2006 levels. The slight increase in the 1-year average
concentration for 2011 is mostly attributable to a couple higher concentrations
measured (compared to 2010, as their concentration profiles are fairly similar to each
other). The 1-year average concentration decreases slightly each year after 2011.
• Despite these changes, most of the 1-year average concentrations shown fall between
0.10 |ig/m3 and 0.15 |ig/m3, with only the minimum 1-year average concentration
(0.08 |ig/m3 for 2009) and maximum 1-year average (0.16 |ig/m3 for 2011) falling
outside this range.
10-52
-------
Figure 10-34. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at SPIL
T
I
r
J
¦o
.
o
<
>
i.
o
VI
TT
J
L
>...
r5"
z
i
r5"
7
T
i
T
T
-
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 10-34 for carbon tetrachloride concentrations measured at
SPIL include the following:
• The maximum concentration of carbon tetrachloride (1.20 |ig/m3) was measured three
times, once in 2005 and twice in 2008.
• Six non-detects of carbon tetrachloride have been measured at SPIL. All of these
were measured during the first 2 years of sampling (four in 2003 and two in 2004).
• The 1-year average concentration changed very little between 2004 and 2007, varying
between 0.65 |ig/m3 and 0.70 |ig/m3. The 1-year average then increased significantly
for 2008 (0.84 |ig/m3). The maximum concentration was measured twice in 2008,
along with nine other concentrations greater than 1 |ig/m3; 21 carbon tetrachloride
concentrations measured in 2008 were greater than or equal to the maximum
concentration measured in 2007 (0.89 |ig/m3).
• The 1-year average concentration exhibits a decreasing trend after 2008 that
continued through 2011, when the 1-year average is at a minimum (0.58 |ig/m3). The
increase shown for 2012 brings the 1-year average carbon tetrachloride concentration
back to near 2010 levels.
10-53
-------
• Most of the statistical parameters exhibit a decrease for 2013 and all of them exhibit
additional decreases for 2014. Carbon tetrachloride concentrations measured at SPIL
level out for 2015 and 2016.
Figure 10-35. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
SPIL
Maximum Concentration
for 2008 is 2.71 p.g/m3
(ti
I r*,
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 10-35 for/;-dichlorobenzene concentrations measured at SPIL
include the following:
• The only p-dichlorobenzene concentrations greater than 1 |ig/m3 were measured at
SPIL during a 3-year period between 2007 and 2009.
• The minimum, 5th percentile, and median concentration is zero for the first three
years of sampling due to the presence of non-detects (at least half the measurements).
This is also true for 2010, and 2013 through 2016. The percentage of non-detects has
varied from as little as 16 percent (2007) to as high as 95 percent (2004).
• The 1-year average concentration exhibits an increasing trend between 2004 and
2007; the number of p-dichlorobenzene greater than 0.1 |ig/m3 increased each year
during this time, from three in 2004 to 19 for 2007.
• A decreasing trend in the 1-year average concentration is shown between 2007 and
2010, when the number of non-detects again exceeds 50 percent (and the median
concentration returns to zero).
10-54
-------
• Most of the statistical parameters exhibit increases again in 2011, due to a few higher
concentrations, fewer non-detects, and twice the number of p-dichlorobenzene
concentrations greater than 0.1 |ig/m3 compared to 2010. Additional decreases are
exhibited in 2012 and 2013, after which there is little change through 2016.
Figure 10-36. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations
Measured at SPIL
Maximum Concentration
for 2003 is 0.75 ng/m3
o.oo aim iiSJ
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
I
o
X
1
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 10-36 for 1,2-dichloroethane concentrations measured at SPIL
include the following:
• There were no measured detections of 1,2-dichloroethane in 2004, 2006, 2007, or
2008. For 2003, 2005, and 2009, the percentage of non-detects was 95 percent or
greater. Thus, the minimum, 5th percentile, median, and in some cases, the 1-year
average concentrations are zero through 2009. The median concentration is also zero
for 2010 and 2011, indicating that at least half the measurements are non-detects. The
percentage of non-detects decreased to 80 percent for 2010 and 73 percent for 2011.
After 2011, the percentage of non-detects decreased significantly, with only a few
non-detects measured each year, with the exception of 2015, when non-detects were
not measured.
• The maximum concentration of 1,2-dichloroethane was measured at SPIL in 2003
(0.75 |ig/m3). This is the only measured detection for 2003 as all other measurements
were non-detects. Only one other 1,2-dichloroethane concentration greater than
0.15 |ig/m3 has been measured at SPIL (0.21 |ig/m3, which was measured in 2014).
10-55
-------
• As the number of non-detects decreased and the number of measured detections
increased, the statistical parameters began to increase correspondingly. The median
concentration is greater than zero for the first time for 2012. The sharp decrease in the
number of non-detects from 73 percent to 8 percent from 2011 to 2012 results in a
sharp increase in both the 1-year average and median concentrations shown for 2012.
With even fewer non-detects measured in 2013 and the highest number of
1,2-dichloroethane concentrations greater than 0.1 |ig/m3 (16), both the 1-year
average and median concentrations are at a maximum for 2013.
• Both the 1-year average and median concentrations exhibit decreases between 2013
and 2016. However, the 1-year average concentrations vary by less than 0.01 |ig/m3
during this time, and the median concentrations vary by less than 0.015 |ig/m3.
Figure 10-37. Yearly Statistical Metrics for Formaldehyde Concentrations
Measured at SPIL
Maximum Concentration
for 2006 is 162 p.g/m3
O
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because consistent sampling did not begin until March 2005.
Observations from Figure 10-37 for formaldehyde concentrations measured at SPIL
include the following:
• The maximum formaldehyde concentration (162 |ig/m3) was measured at SPIL on
May 29, 2006 and is more than 10 times the maximum concentration for the other
years shown in Figure 10-37 other than 2005. Of the 29 formaldehyde concentrations
greater than 15 |ig/m3, 12 were measured at SPIL in 2005, 17 were measured in 2006,
and none were measured in the years that followed.
10-56
-------
• The 1-year average concentration for 2006 is 13.76 |ig/m3. After 2006, the 1-year
average concentration decreased each year through 2009, reaching a minimum of
1.85 |ig/m3.
• There is an increasing trend in the 1-year average concentration between 2009 and
2011, after which little change is shown through 2014. The change in the 1-year
average concentration between 2014 and 2015 is the largest in several years, but is
not statistically significant. Between 2011 and 2016, the 1-year average
concentrations varied between 3.07 |ig/m3 (2012) and 3.85 |ig/m3 (2016).
Figure 10-38. Yearly Statistical Metrics for Trichloroethylene Concentrations
Measured at SPIL
Maximum Concentration
! for 2003 is 110 |j.g/m3
Q
o
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2003.
Observations from Figure 10-38 for trichloroethylene concentrations measured at SPIL
include the following:
• The minimum and 5th percentile are both zero for all years of sampling, indicating
that at least 5 percent of the measurements were non-detects for each year since
sampling began at SPIL. The percentage of non-detects has ranged from 13 percent
(2014) to 39 percent (2004).
• The maximum concentration of trichloroethylene (110 |ig/m3) was measured at SPIL
in 2003 and is an order of magnitude greater than the next highest concentration
(17.5 |ig/m3), which was measured in 2012. No other trichloroethylene concentrations
greater than 10 |ig/m3 have been measured at SPIL.
10-57
-------
• The concentrations of trichloroethylene exhibit considerable variability, as indicated
by confidence intervals calculated for the 1-year average concentrations, particularly
for 2012, when the maximum concentration was nearly four times the next highest
concentration measured that year and non-detects made up about one-fifth of the
measurements.
• The 1-year average concentrations have fluctuated between 0.26 |ig/m3 (2013) and
0.79 |ig/m3 (2010), with no distinct trend in the concentrations. The median
concentration, however, has varied relatively little since 2011, ranging from
0.10 |ig/m3 (2013) to 0.15 |ig/m3 (2011, 2014) during this time.
10.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Illinois monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
10.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Illinois sites, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 10-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
10-58
-------
Table 10-4. Risk Approximations for the Illinois Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
Oig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Northbrook, Illinois - NBIL
Acetaldehyde
0.0000022
0.009
57/57
1.42
±0.16
3.13
0.16
59/59
1.47
±0.21
3.23
0.16
Benzene
0.0000078
0.03
54/54
0.52
±0.06
4.03
0.02
59/59
0.43
±0.04
3.38
0.01
1.3 -Butadiene
0.00003
0.002
51/54
0.04
±0.01
1.20
0.02
56/59
0.04
±0.01
1.10
0.02
Carbon Tetrachloride
0.000006
0.1
54/54
0.63
±0.03
3.79
0.01
59/59
0.56
±0.04
3.38
0.01
p-Dichlorobenzene
0.000011
0.8
27/54
0.13
±0.11
1.47
<0.01
23/59
0.03
±0.01
0.35
<0.01
1,2-Dichloroethane
0.000026
2.4
53/54
0.07
± <0.01
1.81
<0.01
50/59
0.07
±0.01
1.79
<0.01
Formaldehyde
0.000013
0.0098
57/57
2.03
±0.25
26.40
0.21
59/59
2.16
±0.39
28.12
0.22
Acenaphthene3
0.000088
55/60
17.48
±5.87
1.54
55/56
18.93
±6.15
1.67
Arsenic (PMio)3
0.0043
0.000015
56/56
0.94
±0.27
4.04
0.06
57/57
0.87
±0.13
3.74
0.06
Fluoranthene3
0.000088
60/60
8.82
±2.90
0.78
56/56
11.38
±3.85
1.00
Fluorene3
0.000088
53/60
16.27
±5.41
1.43
51/56
19.19
±6.42
1.69
Naphthalene3
0.000034
0.003
60/60
89.32
± 22.07
3.04
0.03
56/56
79.55
± 17.47
2.70
0.03
— = a Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 10-4. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
Oig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Schiller Park, Illinois - SPIL
Acetaldehyde
0.0000022
0.009
61/61
2.43
±0.52
5.35
0.27
57/57
2.45
±0.61
5.38
0.27
Benzene
0.0000078
0.03
60/60
0.71
±0.10
5.52
0.02
58/58
0.63
±0.05
4.92
0.02
1.3 -Butadiene
0.00003
0.002
60/60
0.12
±0.02
3.62
0.06
58/58
0.11
±0.01
3.35
0.06
Carbon Tetrachloride
0.000006
0.1
60/60
0.62
±0.02
3.70
0.01
58/58
0.62
±0.03
3.72
0.01
p-Dichlorobenzene
0.000011
0.8
25/60
0.03
±0.01
0.35
<0.01
25/58
0.03
±0.01
0.38
<0.01
1,2-Dichloroethane
0.000026
2.4
60/60
0.08
± <0.01
1.97
<0.01
54/58
0.07
±0.01
1.92
<0.01
Formaldehyde
0.000013
0.0098
61/61
3.85
±0.45
50.00
0.39
57/57
3.54
±0.47
45.98
0.36
Trichloroethylene
0.0000048
0.002
42/60
0.29
±0.16
1.37
0.14
48/58
0.32
±0.13
1.54
0.16
— = a Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 10-4. Risk Approximations for the Illinois Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
Oig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Roxana, Illinois - ROIL
Acetaldehyde
0.0000022
0.009
33/33
NA
NA
NA
NS
NS
NS
NS
Benzene
0.0000078
0.03
32/32
NA
NA
NA
NS
NS
NS
NS
1.3 -Butadiene
0.00003
0.002
30/32
NA
NA
NA
NS
NS
NS
NS
Carbon Tetrachloride
0.000006
0.1
32/32
NA
NA
NA
NS
NS
NS
NS
1,2-Dichloroethane
0.000026
2.4
31/32
NA
NA
NA
NS
NS
NS
NS
Formaldehyde
0.000013
0.0098
33/33
NA
NA
NA
NS
NS
NS
NS
— = a Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Observations for the Illinois sites from Table 10-4 include the following:
• Formaldehyde and acetaldehyde are the pollutants of interest with the highest annual
average concentrations for NBIL and SPIL (and the only ones greater than 1 |ig/m3).
The annual averages for these pollutants are significantly higher for SPIL than NBIL.
• Formaldehyde has the highest cancer risk approximations for both sites. The cancer
risk approximations for SPIL (50.00 in-a-million for 2015 and 45.98 in-a-million for
2016) are greater than those calculated for NBIL (26.40 in a million for 2015 and
28.12 in-a-million for 2016). There were no other pollutants for which a cancer risk
approximation greater than 10 in-a-million was calculated.
• None of the pollutants of interest for NBIL or SPIL have noncancer hazard
approximations greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. The pollutant with the highest
noncancer hazard approximations among the pollutants of interest for the Chicago
sites is formaldehyde (an HQ of 0.39 for 2015 and 0.36 for 2016 for SPIL and an HQ
of 0.21 for 2015 and 0.22 for 2016 for NBIL).
• Because annual average concentrations could not be calculated for ROIL, cancer risk
and noncancer hazard approximations were not calculated for this site.
10.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 10-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 10-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 10-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each Illinois site, as presented in Table 10-4. Cancer risk approximations for 2015 are presented
in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 10-5.
Table 10-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
10-62
-------
Table 10-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Illinois Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Northbrook, Illinois (Cook County) - NBIL
Benzene
1,084.36
Formaldehyde
1.28E-02
Formaldehyde
28.12
Formaldehyde
985.00
Naphthalene
9.29E-03
Formaldehyde
26.40
Ethylbenzene
648.33
Benzene
8.46E-03
Arsenic
4.04
Acetaldehyde
549.09
1,3-Butadiene
5.36E-03
Benzene
4.03
Naphthalene
273.18
Hexavalent Chromium. PM
4.67E-03
Carbon Tetrachloride
3.79
Tetrachloroethylene
207.13
Ethylene oxide
3.81E-03
Arsenic
3.74
1.3 -Butadiene
178.79
POM, Group 2b
1.83E-03
Benzene
3.38
Trichloroethylene
59.27
Ethylbenzene
1.62E-03
Carbon Tetrachloride
3.38
Dichloromethane
37.59
POM, Group 2d
1.32E-03
Acetaldehyde
3.23
POM, Group 2b
20.80
POM, Group 5a
1.26E-03
Acetaldehyde
3.13
Schiller Park, Illinois (Cook County) - SPIL
Benzene
1,084.36
Formaldehyde
1.28E-02
Formaldehyde
50.00
Formaldehyde
985.00
Naphthalene
9.29E-03
Formaldehyde
45.98
Ethylbenzene
648.33
Benzene
8.46E-03
Benzene
5.52
Acetaldehyde
549.09
1,3-Butadiene
5.36E-03
Acetaldehyde
5.38
Naphthalene
273.18
Hexavalent Chromium. PM
4.67E-03
Acetaldehyde
5.35
Tetrachloroethylene
207.13
Ethylene oxide
3.81E-03
Benzene
4.92
1.3 -Butadiene
178.79
POM, Group 2b
1.83E-03
Carbon Tetrachloride
3.72
Trichloroethylene
59.27
Ethylbenzene
1.62E-03
Carbon Tetrachloride
3.70
Dichloromethane
37.59
POM, Group 2d
1.32E-03
1,3-Butadiene
3.62
POM, Group 2b
20.80
POM, Group 5a
1.26E-03
1,3-Butadiene
3.35
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 10-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Illinois Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Roxana, Illinois (Madison County) - ROIL
Formaldehyde
128.06
Coke Oven Emissions, PM
2.96E-02
Benzene
126.13
Formaldehyde
1.66E-03
Ethylbenzene
58.76
Hexavalent Chromium. PM
1.35E-03
Acetaldehyde
50.92
Naphthalene
9.90E-04
Coke Oven Emissions, PM
29.89
Benzene
9.84E-04
Naphthalene
29.11
Arsenic, PM
5.88E-04
1.3 -Butadiene
13.46
Ethylene oxide
4.13E-04
POM, Group 2b
2.19
1,3-Butadiene
4.04E-04
Tetrachloroethylene
1.63
POM, Group 5a
3.39E-04
POM, Group 2d
1.57
POM, Group 2b
1.93E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 10-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Illinois Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Northbrook, Illinois (Cook County) - NBIL
Ethylene glycol
6,317.67
Acrolein
4,321,056.56
Formaldehyde
0.22
Toluene
4,302.61
Cyanide Compounds, gas
122,716.64
Formaldehyde
0.21
Xylenes
2,428.56
Formaldehyde
100,510.67
Acetaldehyde
0.16
Methanol
2,242.96
Naphthalene
91,058.85
Acetaldehyde
0.16
Hexane
1,430.98
1,3-Butadiene
89,397.31
Arsenic
0.06
Benzene
1,084.36
Acetaldehyde
61,009.94
Arsenic
0.06
Formaldehyde
985.00
Cadmium, PM
55,856.40
Naphthalene
0.03
Ethylbenzene
648.33
Maleic anhydride
37,745.29
Naphthalene
0.03
Acetaldehyde
549.09
Benzene
36,145.42
1,3-Butadiene
0.02
Glycol ethers, gas
356.93
T richloroethylene
29,634.05
1,3-Butadiene
0.02
Schiller Park, Illinois (Cook County) - SPIL
Ethylene glycol
6,317.67
Acrolein
4,321,056.56
Formaldehyde
0.39
Toluene
4,302.61
Cyanide Compounds, gas
122,716.64
Formaldehyde
0.36
Xylenes
2,428.56
Formaldehyde
100,510.67
Acetaldehyde
0.27
Methanol
2,242.96
Naphthalene
91,058.85
Acetaldehyde
0.27
Hexane
1,430.98
1,3-Butadiene
89,397.31
T richloroethylene
0.16
Benzene
1,084.36
Acetaldehyde
61,009.94
T richloroethylene
0.14
Formaldehyde
985.00
Cadmium, PM
55,856.40
1,3-Butadiene
0.06
Ethylbenzene
648.33
Maleic anhydride
37,745.29
1,3-Butadiene
0.06
Acetaldehyde
549.09
Benzene
36,145.42
Benzene
0.02
Glycol ethers, gas
356.93
T richloroethylene
29,634.05
Benzene
0.02
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 10-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Illinois Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tl»)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Roxana, Illinois (Madison County) - ROIL
Toluene
355.12
Acrolein
365,050.42
Ethylene glycol
320.64
Chlorine
90,642.90
Xylenes
211.91
Manganese, PM
15,961.75
Hydrochloric acid
149.65
Cyanide Compounds, gas
13,584.64
Methanol
131.56
Formaldehyde
13,067.65
Formaldehyde
128.06
Lead, PM
10,007.39
Benzene
126.13
Naphthalene
9,704.69
Hexane
119.93
Arsenic, PM
9,109.27
Ethylbenzene
58.76
Hydrochloric acid
7,482.26
Acetaldehyde
50.92
Cyanide Compounds, PM
7,297.86
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 10.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 10-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Cook County. These same pollutants are the highest emitted
pollutants with cancer UREs in Madison County, although the order differs. The
quantity of emissions is considerably different between the two counties, with the
emissions for Cook County up to an order of magnitude greater than Madison
County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Cook County are formaldehyde, naphthalene, and benzene. Coke
oven emissions top Madison County's toxicity-weighted emissions, followed by
formaldehyde and hexavalent chromium.
• Six of the highest emitted pollutants in Cook County also have the highest toxicity-
weighted emissions. Six of the highest emitted pollutants in Madison County also
have the highest toxicity-weighted emissions, though the exact pollutants differ
between the counties.
• For NBIL and SPIL, formaldehyde is the pollutant with the highest cancer risk
approximations. This pollutant also has the highest toxicity-weighted emissions and
ranks second for quantity emitted in Cook County. Benzene also appears on all three
lists for both sites. 1,3-Butadiene also appears on all three lists for SPIL;
1,3-butadiene is also a pollutant of interest for NBIL, although its cancer risk
approximations are lower than those shown in Table 10-5. Acetaldehyde is among the
pollutants with the highest cancer risk approximations for both NBIL and SPIL; while
acetaldehyde is among the highest emitted pollutants in Cook County, it is not among
those with the highest toxicity-weighted emissions.
• Carbon tetrachloride is among the pollutants with the highest cancer risk
approximations for both NBIL and SPIL, yet does not appear on either of Cook
County's emissions-based lists. Carbon tetrachloride ranks 23rd in Cook County for
the quantity emitted and 33rd for its toxicity-weighted emissions.
10-67
-------
• Similarly, arsenic is among the pollutants with the highest cancer risk approximations
for NBIL yet does not appear on either of Cook County's emissions-based lists.
Arsenic ranks 24th for its emissions in Cook County and ranks 14th for its toxicity-
weighted emissions.
• Naphthalene has the second highest toxicity-weighted emissions for Cook County and
ranks fifth for quantity emitted. While naphthalene is also a pollutant of interest for
NBIL, its cancer risk approximations are lower than those shown in Table 10-5.
POM, Group 2b ranks 10th for quantity emitted and seventh for toxicity-weighted
emissions in Cook County. POM, Group 2b includes acenaphthene, fluorene, and
fluoranthene, all three of which are pollutants of interest for NBIL.
• Trichloroethylene, which is a pollutant of interest only for SPIL, is the eighth highest
emitted pollutant in Cook County, but does not appear among the pollutants with the
highest toxicity-weighted emissions (this pollutant ranks 17th).
Observations from Table 10-6 include the following:
• Ethylene glycol, toluene, and xylenes are the highest emitted pollutants with
noncancer RfCs in Cook County. These same pollutants are the highest emitted
pollutants with noncancer RfCs in Madison County, although the order differs. The
quantity emitted is significantly higher in Cook County.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for both counties is acrolein. Although acrolein was sampled for at
all three sites, this pollutant was excluded from the pollutants of interest designation,
and thus subsequent risk-based screening evaluations, due to questions about the
consistency and reliability of the measurements, as discussed in Section 3.2.
• Only three of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Cook County (formaldehyde, benzene, and acetaldehyde). The highest
emitted pollutants and the pollutants with the highest toxicity-weighted emissions for
Madison County have only two pollutants in common (formaldehyde and
hydrochloric acid). This speaks to the relative toxicity of a pollutant; a pollutant does
not have to be emitted in high quantities to be hazardous to human health.
• Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
for the Chicago sites (albeit less than an HQ of 1.0). These two pollutants appear on
both emissions-based lists for Cook County. 1,3-Butadiene also has some of the
highest noncancer hazard approximations for these two sites; 1,3-butadiene is among
those with the highest toxicity-weighted emissions for Cook County, but is not among
the highest emitted.
• Naphthalene and arsenic are also pollutants of interest for NBIL and appear among
those with highest noncancer hazard approximations for this site. Naphthalene
appears among those with the highest toxicity-weighted emissions in Cook County
but is not among the highest emitted. Arsenic appears on neither emissions-based list
for Cook County.
10-68
-------
• Benzene and trichloroethylene are pollutants of interest for SPIL and appear among
those with highest noncancer hazard approximations for this site. Benzene appears on
both emissions-based lists for Cook County. Trichloroethylene is among those with
the highest toxicity-weighted emissions for Cook County but is not among the highest
emitted (of the pollutants with noncancer RfCs).
10.5 Summary of the 2015-2016 Monitoring Data for NBIL, SPIL, and ROIL
Results from several of the data analyses described in this section include the following:
~~~ Twenty-three pollutants (two carbonyl compounds, 12 VOCs,five PAHs, andfour
speciated metals) failed screens for NBIL; 14 pollutants (three carbonyl compounds
and 11 VOCs) failed screens for SPIL; and 10 pollutants (three carbonyl compounds
and seven VOCs) failed screens for ROIL.
~~~ Sampling at ROIL was discontinued at the end of July 2015, ending a four-year
monitoring effort at this location.
~~~ Formaldehyde had the highest annual average concentrations among the pollutants
of interest for both NBIL and SPIL, although the annual averages for SPIL were
higher than those calculated for NBIL.
~~~ The maximum concentrations of several pollutants across the program were
measured at NBIL (p-dichlorobenzene, fluoranthene, and naphthalene).
~~~ After several years of increasing, concentrations of acetaldehyde decreased
significantly in 2015 and 2016 at NBIL while the opposite is true of formaldehyde
concentrations measured at this site. Concentrations of fluoranthene have an
increasing trend at NBIL while concentrations of naphthalene have a decreasing
trend. Concentrations of benzene measured at NBIL are at a minimum in 2016; this is
also true at SPIL.
~~~ Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for both Chicago sites. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
10-69
-------
11.0 Sites in Indiana
This section summarizes those data from samples collected at the UATMP sites in
Indiana and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial
and temporal characteristics of the ambient monitoring
concentrations and reviews them through the context of
risk. Readers are encouraged to refer to Sections 1 through
4 for detailed discussions and definitions regarding the
various data analyses presented below.
11.1 Site Characterization
This section characterizes the Indiana monitoring sites by providing a description of the
nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
One Indiana monitoring site (INDEM) is located in the Chicago-Naperville-Elgin, IL-IN-
WI CBSA, and another site (WPIN) is located in the Indianapolis-Carmel-Anderson, IN CBSA.
Figures 11-1 and 11-3 present composite satellite images retrieved from ArcGIS Explorer
showing the monitoring sites and their immediate surroundings. Figures 11-2 and 11-4 identify
nearby point source emissions locations by source category near INDEM and WPIN,
respectively, as reported in the 2014 NEI for point sources, version 1. Note that only sources
within 10 miles of the sites are included in the facility counts provided in Figures 11-2 and 11-4.
A 10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring sites. Further, this boundary provides both the proximity of emissions sources to the
monitoring sites as well as the quantity of such sources within a given distance of the sites.
Sources outside the 10-mile boundary are still visible on each map for reference, but have been
grayed out to emphasize emissions sources within the boundary. Table 11-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. r
11-1
-------
Figure 11-1. Gary, Indiana (INDEM) Monitoring Site
-------
Figure 11-2. NEI Point Sources Located Within 10 Miles of IN DEM
Lake
Michigan
Porter I
County |
Lake
County
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (14)
i Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminal/Bulk Plant (6)
C Chemical Manufacturing Facility (10)
H Coke Battery (1)
I Compressor Station (4)
f Electricity Generation Facility (5)
E Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (2)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
<•> Metals Processing/Fabrication Facility (11)
X Mine/Quarry/Mineral Processing Facility (10)
? Miscellaneous Commercial/Industrial Facility (13)
fj Paint and Coating Manufacturing Facility (1)
^ Petroleum Products Manufacturing Facility (5)
2 Petroleum Refinery (1)
R Plastic, Resin, or Rubber Products Plant (1)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (6)
Q Railroad Engines/Parts Manufacturing Facility (1)
V Steel Mill (10)
Legend
87"20"0"W 87" 15'0"W 87"10'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
INDEM UATMP site
Q 10 mile radius
County boundary
0>
ILLINOIS
R|
11-3
-------
Figure 11-3. Indianapolis, Indiana (WPIN) Monitoring Site
¦HHHjl
£ 33rd,St
¦UneyS
-Jennings-St.
George Washington Park
>E«32ndiSt<
E 31st St •
E 30th St
E 30th St
E 33rd,St
,E 32nd St
.Er31st.St,
E 30th St
28th St
.*E'28th-St
-------
Figure 11-4. NEI Point Sources Located Within 10 Miles of WPIN
Hamilton
County
Marion
County
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (1)
*t" Airport/Airline/Airport Support Operations (25)
% Asphalt Production/Hot Mix Asphalt Plant (2)
Automobile/Truck Manufacturing Facility (2)
B Bulk Terminal/Bulk Plant (3)
C Chemical Manufacturing Facility (4)
i Compressor Station (2)
e Electrical Equipment Manufacturing Facility (1)
# Electricity Generation Facility (3)
Electroplating, Plating, Polishing, Anodizing, and
Coloring Facility (2)
F Food Processing/Agriculture Facility (2)
I Foundry, Iron and Steel (1)
0 Institution (school, hospital, prison, etc.) (1)
m Landfill (1)
Metal Coating, Engraving, and Allied Services to
Manufacturers (1)
<«> Metals Processing/Fabrication Facility (4)
? Miscellaneous Commercial/Industrial Facility (3)
M Municipal Waste Combustor (1)
<_> Pharmaceutical Manufacturing Facility (2)
R Plastic, Resin, or Rubber Products Plant (2)
X Rail Yard/Rail Line Operations (1)
1 Wastewater Treatment Facility (1)
Miles
——1^—;' i1
86"25'0"W 86C20'0"W 86C15'0"W
Legend
WPIN UATMP site Q 10 mile radius | | County boundary
Shelby
County
85°55'0"W
Note: Due to facility density arid collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Hendricks
County
Hancock
County
11-5
-------
Table 11-1. Geographical Information for the Indiana Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
INDEM
18-089-0022
Gary
Lake
Chicago- Naperville-
Elgin, IL-IN-WI
41.606680,
-87.304729
Industrial
Urban/City
Center
41,860
1-90 N of 1-65 Interchange
WPIN
18-097-0078
Indianapolis
Marion
Indianapolis-Carmel-
Anderson, IN
39.811097,
-86.114469
Residential
Suburban
24,917
Keystone Ave, N of 38th St
1AADT reflects 2016 data (IN DOT, 2016)
-------
INDEM is located in Gary, Indiana, approximately 11 miles east of the Indiana-Illinois
border, 25 miles southeast of Chicago, and on the southernmost bank of Lake Michigan. The site
is located just north of 1-65 and 1-90, the edge of which can be seen in the bottom left portion of
Figure 11-1. Although INDEM resides on the Indiana Dunes National Lakeshore, about 1 mile
south of the Lake Michigan shoreline, the surrounding area is highly industrialized, as shown in
Figure 11-1, and several rail lines transverse the area. Figure 11-2 shows that the majority of
point sources within 10 miles of INDEM are located to the west of the site. There is also a
second cluster of facilities located to the east of INDEM in Porter County. The emissions source
categories with the highest number of sources within 10 miles of INDEM include aircraft
operations, which includes airports and related operations as well as small runways and heliports,
such as those associated with hospitals or television stations; metals processing and fabrication;
steel mills; chemical manufacturing; and mine/quarry/mineral processing. The sources closest to
INDEM include two steel mills; a heliport at a police station; several facilities that fall into the
miscellaneous commercial/industrial category; several metals processing/fabrication facilities; a
mine/quarry/mineral processing facility; and a petroleum products manufacturing facility.
WPIN is located in the parking lot of a police station across from George Washington
Park, near East 30th Street in northeast Indianapolis. Figure 11-3 shows that the area surrounding
WPIN is suburban and residential, with little industry in close proximity. A church and a
charitable organization are located across the street from Washington Park, as is Oscar
Charleston Park. Figure 11-4 shows that the majority of point sources are located to the south
and southwest of WPIN, towards the center of Marion County. The source category with the
highest number of sources near WPIN is the airport operations source category. The sources
closest to WPIN are a metals processing/fabrication facility and a heliport, each of which is
located within 2 miles of WPIN.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 11-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site.
INDEM experiences a higher traffic volume than WPIN, although the traffic volumes near these
sites rank in the middle of the range compared to traffic volumes near other NMP sites. These
traffic volumes were obtained for 1-90 near 1-65 for INDEM and North Keystone Avenue north
of 38th Street for WPIN.
11-7
-------
11.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 11-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 11-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. Carbonyl compounds were sampled for at
both INDEM and WPIN.
Table 11-2. 2015-2016 Risk-Based Screening Results for the Indiana Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Gary, Indiana - INDEM
Acetaldehyde
0.45
119
119
100.00
50.00
50.00
Formaldehyde
0.077
119
119
100.00
50.00
100.00
Total
238
238
100.00
Indianapolis, Indiana - WPIN
Formaldehyde
0.077
116
116
100.00
50.00
50.00
Acetaldehyde
0.45
115
116
99.14
49.57
99.57
Propionaldehyde
0.8
1
116
0.86
0.43
100.00
Total
232
348
66.67
Observations from Table 11-2 include the following:
• Acetaldehyde, formaldehyde, and propionaldehyde are the carbonyl compounds with
risk screening values.
• Acetaldehyde and formaldehyde failed 100 percent of screens for INDEM and
contributed equally to the number of failed screens; thus, both pollutants were
identified as pollutants of interest for this site.
• Acetaldehyde, formaldehyde, and propionaldehyde each failed screens for WPIN.
Formaldehyde failed 100 percent of screens; acetaldehyde failed one less screen than
formaldehyde; and propionaldehyde failed a single screen.
• Together, acetaldehyde and formaldehyde account for over 99 percent of failed
screens for WPIN, and thus, are the pollutants of interest identified for this site.
11-8
-------
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
11.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Indiana monitoring sites. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at the Indiana sites are
provided in Appendix M.
11.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Indiana monitoring site, as described in Section 3.1. The quarterly
average concentration of a particular pollutant is simply the average concentration of the
preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated for a given year and where method completeness was greater than or equal to
85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
pollutants of interest for the Indiana monitoring sites are presented in Table 11-3, where
applicable. Note that if a pollutant was not detected in a given calendar quarter, the quarterly
11-9
-------
average simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
Observations for INDEM from Table 11-3 include the following:
• Acetaldehyde and formaldehyde were detected in all of the valid carbonyl compound
samples collected at this site.
• Concentrations of acetaldehyde measured at INDEM varied from 0.509 |ig/m3 to
2.38 |ig/m3, with only a few measurements greater than 2 |ig/m3. Quarterly average
concentrations of acetaldehyde range from 0.97 ± 0.13 |ig/m3 (first quarter 2016) to
1.68 ± 0.23 |ig/m3 (third quarter 2015). The annual averages of acetaldehyde for
INDEM varied little over the two years of sampling.
• Concentrations of formaldehyde measured at INDEM were higher, and more variable,
than the acetaldehyde concentrations. Concentrations of formaldehyde measured at
INDEM varied from 1.04 |ig/m3 to 11.0 |ig/m3. The measurements were more
variable in 2016 than in 2015, based on the quarterly and annual average
concentrations shown in Table 11-3, and their associated confidence intervals.
Thirteen of the 14 highest formaldehyde concentrations were measured at this site in
2016. Both the minimum and maximum formaldehyde concentrations measured at
INDEM were measured during the third quarter of 2016, as were four of the five
formaldehyde concentrations greater than 7 |ig/m3 (with the exception measured at
the end of June 2016). This explains the variability shown in the third quarter average
for 2016.
Observations for WPIN from Table 11-3 include the following:
• Acetaldehyde and formaldehyde were detected in all of the valid carbonyl compound
samples collected at this site.
• Concentrations of acetaldehyde measured at WPIN varied from 0.283 |ig/m3 to
5.60 |ig/m3, with several of the highest concentrations measured during the second
quarter of either year. The confidence intervals for both second quarter averages are
the highest shown for each year.
• Concentrations of formaldehyde measured at WPIN were higher, and more variable,
than the acetaldehyde concentrations measured at this site. Concentrations of
formaldehyde measured at WPIN range from 0.806 |ig/m3 to 11.1 |ig/m3, which is
similar to the maximum concentration measured at INDEM. For 2015, the first
quarter average concentration of formaldehyde is significantly less than the other
quarterly averages shown. A review of the data shows that nine of the 10
formaldehyde concentration less than 2 |ig/m3 measured in 2015 were measured in
January or February, including all three less than 1 |ig/m3 measured at this site. By
comparison, formaldehyde concentration less than 2 |ig/m3 measured in 2016 were
split between the first and fourth quarters of the year (with three each).
11-10
-------
Table 11-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Indiana Monitoring Sites
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
Pollutant
# >MDL/
# Samples
Avg
frig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Average
frig/m3)
# >MDL/
# Samples
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Average
frig/m3)
Gary, Indiana - INDEM
1.24
1.33
1.68
1.17
1.33
0.97
1.41
1.38
1.31
1.26
Acetaldehyde
60/60/60
±0.18
±0.21
±0.23
±0.23
±0.12
59/59/59
±0.13
±0.24
±0.24
±0.32
±0.12
2.14
2.93
4.08
2.74
2.95
2.30
3.94
5.62
2.98
3.67
Formaldehyde
60/60/60
±0.33
±0.49
±0.68
±0.39
±0.28
59/59/59
±0.32
± 1.00
± 1.49
±0.57
±0.54
Indianapolis, Indiana - WPIN
1.53
1.81
1.88
1.59
1.69
1.24
2.14
1.60
1.66
1.64
Acetaldehyde
59/59/59
±0.18
±0.53
±0.20
±0.29
±0.15
57/57/57
±0.18
±0.75
±0.40
±0.43
±0.23
2.04
3.98
4.37
3.29
3.36
2.76
5.16
3.65
3.01
3.59
Formaldehyde
59/59/59
±0.66
±0.96
±0.78
±0.48
±0.41
57/57/57
±0.37
± 1.50
±0.52
±0.66
±0.45
-------
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the Indiana
sites from those tables include the following:
• Neither INDEM nor WPIN appears in Table 4-11 for their annual average
concentrations of acetaldehyde, with annual averages for WPIN ranking greater than
20th and annual averages for INDEM ranking greater than 30th.
• Neither INDEM nor WPIN appears in Table 4-11 for their annual average
concentrations of formaldehyde; INDEM and WPIN rank 11th and 12th, respectively,
for their 2016 annual averages, with their annual averages for 2015 ranking lower.
11.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 11-3 for INDEM and WPIN. Figures 11-5 and 11-6 overlay the sites' minimum, annual
average, and maximum concentrations for each year onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
Figure 11-5. Program vs. Site-Specific Average Acetaldehyde Concentrations
INDEM I I
^
WPIN
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
e
o
11-12
-------
Figure 11-5 presents the box plots for acetaldehyde for both sites and shows the
following:
• The range of acetaldehyde concentrations measured at INDEM in 2015 is fairly
similar to the range of concentrations measured in 2016. The range of acetaldehyde
concentrations measured at WPIN in 2015 is smaller than the range of concentrations
measured in 2016. The range of acetaldehyde concentrations measured at INDEM
each year is smaller than the range measured at WPIN.
• The annual average concentrations of acetaldehyde calculated for WPIN are similar
to each other and to the program-level average concentration of 1.67 |ig/m3. The
annual average concentrations of acetaldehyde calculated for INDEM are similar to
each other and both are less than the program-level median concentration of
1.43 |ig/m3.
Figure 11-6. Program vs. Site-Specific Average Formaldehyde Concentrations
I
¦
I
mm
I
III ii i i i i
0 3 6 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
9
o
Figure 11-6 presents the box plots for formaldehyde for both sites and shows the
following:
• For both INDEM and WPIN, the range of formaldehyde concentrations measured in
2016 is larger than the range of formaldehyde concentrations measured in 2015. The
range of concentrations measured each year is similar across the two sites.
• The annual average concentration for 2015 for INDEM is less than the annual
average for 2016; INDEM's annual average concentration for 2015 is just less than
the program-level average concentration of 3.05 |ig/m3 while the annual average for
2016 is just less than the program-level third quartile (3.78 |ig/m3).
• The annual average concentrations for WPIN vary less, with both falling between the
program-level average concentration and the program-level third quartile.
11-13
-------
11.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
INDEM and WPIN have sampled carbonyl compounds under the NMP since 2004 and 2007,
respectively. Thus, Figures 11-7 through 11-10 present the 1-year statistical metrics for each of
the pollutants of interest first for INDEM, then for WPIN. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented. Note that sampling under the NMP was discontinued at INDEM at the end of 2016.
Figure 11-7. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
INDEM
16.0
1E
.! 80
n
C
m
c
O
u
<
>
<
*
o
—
o.
\ 1 X I rh ^ ^ rh A
^ I
*
^LoJLgJLgJ^LjJLeJLoJ
2004 20051 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented due to a gap in sampling between September and November 2005.
Observations from Figure 11-7 for acetaldehyde concentrations measured at INDEM
include the following:
• Although carbonyl compound sampling under the NMP began in 2003, samples were
only collected for 3 months. Carbonyl compound sampling began in earnest at
INDEM at the beginning of 2004; thus, Figure 11-7 begins with 2004. A 1-year
average concentration is not presented for 2005 due to a break in sampling between
September and November 2005, although the range of measurements is provided.
11-14
-------
The maximum acetaldehyde concentration shown (13.8 |ig/m3) was measured at
INDEM on June 14, 2004. Four additional concentrations greater than 10 |ig/m3 have
been measured at INDEM (one in 2006 and three in 2008).
Although the maximum and 95th percentile increased from 2007 to 2008, the 1-year
average, median, 5th percentile, and minimum concentrations of acetaldehyde exhibit
decreases from 2007 to 2008. Although three concentrations greater than 10 |ig/m3
were measured in 2008 (compared to zero in 2007), the number of measurements at
the lower end of the concentration range increased significantly. The number of
acetaldehyde concentrations less than 2 |ig/m3 increased seven-fold (from three in
2007 to 21 in 2008).
With the exception of the minimum and 5th percentile, the statistical parameters
decreased significantly from 2008 to 2009. The 1-year average and median
concentrations decreased by more than half and the 95th percentile decreased by more
than 80 percent during this time. The carbonyl compound collection system was
replaced in 2009, which seems to have had a significant effect on the concentrations
measured, particularly with respect to formaldehyde, which is discussed in more
detail below.
Acetaldehyde concentrations greater than 4 |ig/m3 were not measured in the years
after 2008. After 2008, the year-to-year changes in the statistical parameters for
acetaldehyde are smaller in magnitude.
Most of the statistical parameters exhibit a slight decreasing trend between 2010 and
2013, with many of them at a minimum for 2013. The median concentration for 2013
is less than 1.00 |ig/m3 and the 1-year average concentration is just slightly greater
than 1.00 |ig/m3.
An increasing trend in acetaldehyde concentrations is shown between 2013 and 2015,
with the both 1-year average and median concentrations at their highest since 2010
and 2008, respectively.
With the exception of the maximum concentration, each of the statistical parameters
exhibits a slight decrease for 2016.
11-15
-------
Figure 11-8. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
INDEM
5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to a break in sampling between September and November 2005.
Observations from Figure 11-8 for formaldehyde concentrations measured at INDEM
include the following:
• Five formaldehyde concentrations greater than 400 |ig/m3 were measured in the
summer of 2008 (ranging from 414 |ig/m3 to 500 |ig/m3). While these are extremely
high values of formaldehyde, concentrations of formaldehyde have been historically
high at this site, as shown by the statistics in Figure 11-8. Thirty-eight concentrations
of formaldehyde greater than 100 |ig/m3 have been measured at INDEM.
• Prior to 2009, the maximum concentration for each year is greater than 150 |ig/m3.
The median concentrations for 2004, 2006, and 2007 are greater than 30 |ig/m3,
indicating that at least half of the concentrations were greater than 30 |ig/m3 for these
years; the median concentration for 2005 and 2008 are both greater than 10 |ig/m3.
• Although the 1-year average concentration doubled from 2007 to 2008, the median
concentration decreased by more than half. This means that although the magnitude
of those higher measurements is driving the 1-year average concentration upward,
there were also a larger number of concentrations at the lower end of the
concentration range. Twenty-four formaldehyde concentrations measured in 2008
were less than the minimum concentration measured in 2007; those 24 measurements
represent 41 percent of the concentrations measured in 2008. The last "high"
formaldehyde concentration was measured on August 4, 2008, after which
formaldehyde concentrations greater than 4 |ig/m3 were not measured that year.
11-16
-------
• All the statistical metrics decreased significantly for 2009. Between 2009 and 2013,
less than 0.5 |ig/m3 separates the 1-year average concentrations, ranging from
2.13 |ig/m3 (2013) to 2.58 |ig/m3 (2014). The number of formaldehyde measurements
greater than 4 |ig/m3 ranged from two to seven for each year during this time,
compared to accounting for more than half of the measurements in each of the
previous years.
• INDEM's formaldehyde concentrations have historically been higher than any other
NMP site sampling carbonyl compounds. During the summer PAMS season, which
begins on June 1, a state-owned multi-channel collection system was used at INDEM
to collect multiple samples per day. At the end of each PAMS season, sample
collection goes back to a state-owned single-channel collection system. The multi-
channel collection system used at INDEM during the PAMS season was replaced in
2009 and this site's formaldehyde concentrations decreased substantially (as did their
acetaldehyde concentrations, but the difference is less dramatic). Given that the
elevated concentrations of formaldehyde were typically measured during the summer,
this change could account for the differences in the concentrations measured before
and after 2009. Thus, the elevated concentrations from previous years were likely
related to the multi-channel collection equipment and may not reflect the actual levels
in ambient air. However, concentrations in the earlier years of sampling must have
still been higher based on the median concentrations shown before and after 2009, as
discussed in the previous bullets.
• Both the 1-year average and median concentrations of formaldehyde have an
increasing trend between 2013 and 2016, with the 1-year average concentration
increasing by more than 1.5 |ig/m3 during this period. For 2016, both of these central
tendency parameters are at their highest since the replacement of the collection
system in 2009. 2016 is the first year that a formaldehyde concentration greater than
10 |ig/m3 has been measured since 2009; the number of formaldehyde measurements
greater than 4 |ig/m3 increased from three in 2013 to nine for 2014 and 2015, to 16
for 2016.
11-17
-------
Figure 11-9. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
WPIN
.o~
T
"1"
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile
Observations from Figure 11-9 for acetaldehyde concentrations measured at WPIN
include the following:
• Although carbonyl compound sampling under the NMP began in 2006, samples were
collected intermittently. Carbonyl compound sampling began in earnest at WPIN at
the beginning of 2007; thus, Figure 11-9 begins with 2007.
• The three highest acetaldehyde concentrations were measured at WPIN in 2010 and
ranged from 5.96 |ig/m3 to 6.72 |ig/m3. Four additional concentrations greater than
5 |ig/m3 have been measured at WPIN (two in 2007, one in 2012, and one in 2016).
• The 1-year average concentration has a decreasing trend through 2009, after which a
significant increase is shown for 2010. All of the statistical parameters exhibit an
increase for 2010, particularly the maximum concentration (which doubled) and the
95th percentile (which increased by nearly 60 percent). The number of acetaldehyde
concentrations greater than 3 |ig/m3 increased five-fold, from three measured in 2009
to 15 measured in 2010. This increase returns the statistical parameters for 2010 to
near 2007 levels.
• Acetaldehyde concentrations measured at WPIN have a decreasing trend after 2010
and, despite a larger range of concentrations measured in 2016, both the 1-year
average and median concentrations are at a minimum for 2016. The rate of decrease
slowed considerably in recent years.
11-18
-------
Figure 11-10. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
WPIN
2011 2012
Year
O 5th Percentile
Maximum o 95th Percentile
Observations from Figure 11-10 for formaldehyde concentrations measured at WPIN
include the following:
• The maximum concentration of formaldehyde measured at WPIN (11.1 |ig/m3) has
been measured twice, once in 2011 and once in 2016. Two additional formaldehyde
concentrations greater than 10 |ig/m3 were measured in 2012.
• The 1-year average concentration has a decreasing trend through 2009, similar to
acetaldehyde, after which an increasing trend is shown through 2012. Although the
1-year average concentration did not change significantly between 2011 and 2012, the
median concentration for 2012 decreased considerably. While the range of
concentrations did not change much between the two years, the biggest change
between the two datasets is in the number of concentrations in the middle of the
concentration range; the number of formaldehyde measurements between 3 |ig/m3
and 4 |ig/m3 more than doubled, from seven in 2011 to 15 in 2012.
• A decreasing trend in formaldehyde concentrations is shown between 2012 and 2014,
with all of the statistical parameters exhibiting decreases during this period. Further,
each statistical parameter is at a minimum for 2014, with the 1-year average
concentration less than 3 |ig/m3 for the first time.
This decreasing trend is followed by an increasing trend for 2015 and 2016.
11-19
-------
11.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Indiana monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
11.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Indiana sites, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 11-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the Indiana sites from Table 11-4 include the following:
• For both sites, the annual average concentrations of formaldehyde are greater than the
annual average concentrations of acetaldehyde. The annual average concentrations of
acetaldehyde for WPIN are greater than the annual averages of acetaldehyde for
INDEM. For formaldehyde, this is true for 2015 but not 2016.
• The cancer risk approximations for formaldehyde are an order of magnitude higher
than the cancer risk approximations for acetaldehyde for both sites. The cancer risk
approximations for formaldehyde for INDEM are 38.34 in-a-million for 2015 and
47.66 in-a-million for 2016, with the cancer risk approximations for WPIN falling in-
between. Cancer risk approximations for acetaldehyde range from 2.77 in-a-million
(INDEM, 2016) to 3.72 in-a-million (WPIN, 2015).
• Neither pollutant of interest for INDEM or WPIN has a noncancer hazard
approximation greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. The highest noncancer hazard
approximation was 0.37, which was calculated for both sites, based on the 2016
annual average concentrations of formaldehyde.
11-20
-------
Table 11-4. Risk Approximations for the Indiana Monitoring Sites
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Gary, Indiana - INDEM
Acetaldehyde
0.0000022
0.009
60/60
1.33
±0.12
2.94
0.15
59/59
1.26
±0.12
2.77
0.14
Formaldehyde
0.000013
0.0098
60/60
2.95
±0.28
38.34
0.30
59/59
3.67
±0.54
47.66
0.37
Indianapolis, Indiana - WPIN
Acetaldehyde
0.0000022
0.009
59/59
1.69
±0.15
3.72
0.19
57/57
1.64
±0.23
3.60
0.18
Formaldehyde
0.000013
0.0098
59/59
3.36
±0.41
43.65
0.34
57/57
3.59
±0.45
46.61
0.37
-------
11.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 11-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 11-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 11-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 11-4. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 11-5. Table 11-6 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 11.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
11-22
-------
Table 11-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Indiana Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Gary, Indiana (Lake County) - INDEM
Benzene
194.05
Formaldehyde
2.10E-03
Formaldehyde
47.66
Formaldehyde
161.83
Benzene
1.51E-03
Formaldehyde
38.34
Ethylbenzene
97.22
Naphthalene
1.28E-03
Acetaldehyde
2.94
Acetaldehyde
96.21
Coke Oven Emissions, PM
1.19E-03
Acetaldehyde
2.77
Naphthalene
37.50
1,3-Butadiene
8.90E-04
1.3 -Butadiene
29.66
POM, Group lb
7.65E-04
Bis(2-ethylhexyl) phthalate, gas
22.83
POM, Group 2b
3.42E-04
POM, Group lb
8.70
Cadmium, PM
3.38E-04
POM, Group 2b
3.88
POM, Group 2d
3.12E-04
POM, Group 2d
3.55
Arsenic, PM
3.08E-04
Indianapolis, Indiana (Marion County) - WPIN
Benzene
349.62
Formaldehyde
4.01E-03
Formaldehyde
46.61
Formaldehyde
308.77
Benzene
2.73E-03
Formaldehyde
43.65
Ethylbenzene
202.77
Naphthalene
1.80E-03
Acetaldehyde
3.72
Acetaldehyde
197.48
1,3-Butadiene
1.54E-03
Acetaldehyde
3.60
Naphthalene
52.95
Arsenic, PM
9.98E-04
1,3-Butadiene
51.37
POM, Group 2b
5.72E-04
Bis(2-ethylhexyl) phthalate, gas
43.23
Ethylbenzene
5.07E-04
Tetrachloroethylene
14.44
POM, Group 2d
4.53E-04
POM, Group 2b
6.50
Acetaldehyde
4.34E-04
POM, Group 2d
5.15
POM, Group 5a
4.01E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 11-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Indiana Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Gary, Indiana (Lake County) - INDEM
Toluene
610.91
Acrolein
537,322.19
Formaldehyde
0.37
Xylenes
360.37
Manganese, PM
58,805.75
Formaldehyde
0.30
Hydrochloric acid
256.95
Lead, PM
38,552.81
Acetaldehyde
0.15
Hexane
246.81
2,4-Toluene diisocyanate
27,925.71
Acetaldehyde
0.14
Methanol
224.95
Cadmium, PM
18,793.80
Benzene
194.05
Formaldehyde
16,512.88
Formaldehyde
161.83
1,3-Butadiene
14,828.51
Ethylbenzene
97.22
Hydrochloric acid
12,847.50
Acetaldehyde
96.21
Naphthalene
12,501.56
Naphthalene
37.50
Acetaldehyde
10,689.83
Indianapolis, Indiana (Marion County) - WPIN
Toluene
1,444.06
Acrolein
1,191,081.87
Formaldehyde
0.37
Xylenes
761.49
2,4-Toluene diisocyanate
52,756.71
Formaldehyde
0.34
Hydrochloric acid
543.66
Formaldehyde
31,506.82
Acetaldehyde
0.19
Methanol
531.90
Hydrochloric acid
27,182.77
Acetaldehyde
0.18
Benzene
349.62
1,3-Butadiene
25,684.69
Formaldehyde
308.77
Acetaldehyde
21,941.93
Hexane
299.13
Lead, PM
18,973.21
Ethylbenzene
202.77
Naphthalene
17,651.26
Acetaldehyde
197.48
Arsenic, PM
15,480.43
Ethylene glycol
78.78
Benzene
11,653.89
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 11-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the three highest emitted pollutants
with cancer UREs in both Marion and Lake Counties, although the quantity emitted is
higher in Marion County. Nine of the 10 pollutants listed are the same between the
two counties; the only difference is the eighth ranked pollutant for each county, POM
Group 1 (for Lake County) and tetrachloroethylene (for Marion County).
• Formaldehyde, benzene, and naphthalene are the pollutants with the highest toxicity-
weighted emissions (of the pollutants with cancer UREs) for both counties.
• Seven of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions; eight of the highest emitted pollutants in Marion County also
have the highest toxicity-weighted emissions.
• Acetaldehyde and formaldehyde are the only pollutants of interest for INDEM.
Acetaldehyde and formaldehyde appear among the highest emitted pollutants for
Lake County, with only formaldehyde appearing among the pollutants with the
highest toxicity-weighted emissions (acetaldehyde ranks 14th). Formaldehyde has the
highest toxicity-weighted emissions in Lake County.
• Acetaldehyde and formaldehyde are also the pollutants of interest for WPIN.
Acetaldehyde and formaldehyde appear among the highest emitted pollutants for
Marion County, and are both among the pollutants with the highest toxicity-weighted
emissions. Formaldehyde also has the highest toxicity-weighted emissions in Marion
County.
Observations from Table 11-6 include the following:
• Toluene, xylenes, and hydrochloric acid are the three highest emitted pollutants with
noncancer RfCs in both Marion and Lake Counties, although the quantity emitted is
higher in Marion County. Nine of the 10 pollutants listed are the same between the
two counties; the only difference is the tenth ranked pollutant for each county,
naphthalene (for Lake County) and ethylene glycol (for Marion County).
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for both counties. Manganese and lead rank second
and third for Lake County, while 2,4-toluene diisocyanate and formaldehyde rank
second and third for Marion County.
• Four of the highest emitted pollutants in Lake County also have the highest toxicity-
weighted emissions. This is also true for Marion County, although the pollutants are
somewhat different.
• Formaldehyde and acetaldehyde appear in all three columns in Table 11-6 for both
sites.
11-25
-------
11.5 Summary of the 2015-2016 Monitoring Data for INDEM and WPIN
Results from several of the data analyses described in this section include the following:
~~~ Carbonyl compounds were sampled for at INDEM and WPIN in 2015 and 2016.
Sampling at INDEM under the NMP was discontinued at the end of 2016, ending a
13-year continuous monitoring effort.
~~~ Acetaldehyde andformaldehyde failed screens for each site and were identified as
pollutants of interest for each site.
~~~ The annual average concentrations of formaldehyde are greater than the annual
average concentrations of acetaldehyde for both sites.
~~~ Concentrations of formaldehyde and acetaldehyde decreased significantly at INDEM
from 2008 to 2009; these changes may be at least partially explained by the
replacement of the collection system. Although considerably less than those measured
before 2009, concentrations of formaldehyde have a slight increasing trend at
INDEM over the last few years; this is also true for acetaldehyde through 2015.
~~~ Acetaldehyde concentrations have been decreasing at WPIN since 2010, although the
rate of decrease slowed considerably in recent years. After a few years of decreasing,
formaldehyde concentrations have an increasing trend at WPIN in 2015 and 2016.
~~~ Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for both sites; none of the pollutants of interest for either site have noncancer
hazard approximations greater than an HQ of 1.0.
11-26
-------
12.0 Sites in Kentucky
This section summarizes those data from samples collected at the NATTS and UATMP
sites in Kentucky and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines the
IXilii jjcncmlct.1 h> sources oilier
spatial and temporal characteristics of the ambient monitoring iluin LR( i. UW's com rue i
A A A <- • i kilionlor\ lor llic \\ll\ are noi
concentrations and reviews them through the context or risk. . . ,, .
° incliklcu m llic (.Uila ;uuil\ ses
Readers are encouraged to refer to Sections 1 through 4 for conuuneil in iliis repori
detailed discussions and definitions regarding the various data
analyses presented below.
12.1 Site Characterization
This section characterizes the Kentucky monitoring sites by providing a description of the
nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
Data from eight monitoring sites in Kentucky are included in this report, one NATTS site
and seven predominantly "source-oriented" sites: three sites are located in northeast Kentucky,
two in Ashland and one near Grayson Lake; one is located south of Evansville, Indiana in the
town of Baskett; three are located in or near the Calvert City area, east of Paducah, Kentucky;
and the final site is located in Lexington, in north-central Kentucky. A composite satellite image
and a facility map are provided for each site in Figures 12-1 through 12-13. The composite
satellite images were retrieved from ArcGIS Explorer and show each monitoring site in its
respective location. The facility maps identify nearby point source emissions locations by source
category, as reported in the 2014 NEI for point sources, version 1. Note that only sources within
10 miles of each site are included in the facility counts provided. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at each monitoring site.
Further, this boundary provides both the proximity of emissions sources to each monitoring site
as well as the quantity of such sources within a given distance of the sites. Sources outside the
10-mile boundaries are still visible on the maps for reference but have been grayed out to
emphasize emissions sources within the boundaries. Table 12-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates for each
site. Each figure and table is discussed in detail in the paragraphs that follow.
12-1
-------
Figure 12-1. Ashland, Kentucky (ASKY) Monitoring Site
-------
Figure 12-2. Ashland, Kentucky (ASKY-M) Monitoring Site
-------
Figure 12-3. NEI Point Sources Located Within 10 Miles of ASKY and ASKY-M
OHIO
Greenup
County
Ohio River
Boyd
County
ASKY-M UATMP site O 10 mile radius | County boundary
Carter
County
Legend
~ ASKY UATMP site
WEST i
VIRGINIA \
I
0 2.5 5 10
1 1 1 1 1 1 1 1 1
Miles
^ i —i
82"45'0"W 82"40'0"W 82°35'0"W 82"30'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Source Category Group (No. of Facilities)
f Airport/Airline/Airport Support Operations (6) x
'i Asphalt Production/Hot Mix Asphalt Plant (1) ?
b Bulk Terminal/Bulk Plant (1) Q
c Chemical Manufacturing Facility (5) <
C\ Coke Battery (1) a
i Compressor Station (3) R
f Electricity Generation Facility (3) V
F Food Processing/Agriculture Facility (2) x
If Gasoline/Diesel Service Station (1) A
¦ Landfill (2) V
© Metals Processing/Fabrication Facility (2) ®
Mine/Quarry/Mineral Processing Facility (7)
Miscellaneous Commercial/Industrial Facility (2)
Paint and Coating Manufacturing Facility (1)
Pesticide Manufacturing Plant (1)
Petroleum Refinery (1)
Plastic, Resin, or Rubber Products Plant (2)
Port and Harbor Operations (3)
Rail Yard/Rail Line Operations (2)
Ship/Boat Manufacturing or Repair Facility (1)
Steel Mill (2)
Testing Laboratory (1)
12-4
-------
Figure 12-4. Grayson, Kentucky (GLKY) Monitoring Site
to
yAppahchian Mountains
PV
-V
#¦
¦o
ce
¦C
I*
f
Si
f
/ /
1 /
Lt\
//9i
sK\
¦5>
\
«
If /
* . "i-
iPBS ' /
*
Grayson Lake
%
X
\c
c
fO"
Source: USGS
Source: NASA, NGA, USGS
© 2008 Microsoft Corp.
-------
Figure 12-5. NEI Point Sources Located Within 10 Miles of GLKY
83°10'0"W
83°10'0"W
Greenup
County
Carter
County
Elliot
County
~ GLKY NATTS site O 10 mile radius County boundary
Source Category Group (No. of Facilities)
1" Airport/Airline/Airport Support Operations (1)
i=i Brick, Structural Clay, or Clay Ceramics Plant (2)
b Bulk Terminal/Bulk Plant (1)
F Food Processing/Agriculture Facility (1)
o Institution (school, hospital, prison, etc.) (1)
x Mine/Quarry/Mineral Processing Facility (2)
l
Rowan
County
Legend
l 1 l •
SS'WW 83°0'0"W 82°55"0"W 82°50'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Boyd
County
\
\
\
\
Lawrence ^
County ^
" " " "\
\
L
12-6
-------
Figure 12-6, Baskett, Kentucky (BAKY) Monitoring Site
Source^ USGS
Source: NASAvNGA, USGS
© 2008 Microsoft Corp.
-------
Figure 12-7. NEI Point Sources Located Within 10 Miles of BAKY
INDIANA
Ohio River
Henderson
County
Daviess
County
Note: Due to facility density and collocation, the total facilities
L6C|6nd displayed may not represent all facilities within the area of interest.
BAKY UATMP site Q 10 mile radius | | County boundary
Source Category Group (No. of Facilities)
T
Airport/Airline/Airport Support Operations (6)
<•>
Metals Processing/Fabrication Facility (4)
*
Asphalt Production/Hot Mix Asphalt Plant (1)
?
Miscellaneous Commercial/Industrial Facility (2)
B
Bulk Terminal/Bulk Plant (1)
Ti
Paint and Coating Manufacturing Facility (1)
C
Chemical Manufacturing Facility (1)
*
Petroleum Products Manufacturing Facility (1)
A
Concrete Batch Plant (1)
R
Plastic, Resin, or Rubber Products Plan (3)
©
Dry Cleaning Facility (2)
m
Pulp and Paper Plant (2)
f
Electricity Generation Facility (4)
X
Rail Yard/Rail Line Operations (1)
E
Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (1)
TT
Telecommunications/Radio Facility (1)
F
Food Processing/Agriculture Facility (3)
«
Testing Laboratory (1)
12-8
-------
Figure 12-8. Calvert City, Kentucky (ATKY) Monitoring Site
>: J '
-------
Figure 12-9. Smithland, Kentucky (BLKY) Monitoring Site
-------
Figure 12-10. Calvert City, Kentucky (TVKY) Monitoring Site
St/urce: USGS
ISourcirrNASA, NGA, USGS
"<£?yg'Q.O8 Microsoft Corp.
-------
Figure 12-11. NET Point Sources Located Within 10 Miles of ATKY, BLKY, and TVKY
Livingston
County
OHIO
Lyon
County
Tennessee
f River
Lake
Barkley
Marshall
County
Kentucky
Lake /
&
TT 0
I
1
# ? I
I
1
1
McCracken |
County ,
88°25'0"W 88r20'0"W 88°15'0"W 88"10'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
Crittenden
County
County
ATKY UATMP site BLKY UATMP site TVKY UATMP site
Q 10 miie radius
Source Category Group (No. of Facilities)
T Airport/Airline/Airport Support Operations (1)
it Asphalt Production/Hot Mix Asphalt Plant (1)
b Bulk Terminal/Bulk Plant (1)
c Chemical Manufacturing Facility (6)
: Compressor Station (1)
* Electricity Generation Facility (2)
+ Industrial Machinery or Equipment Plant (1)
County boundary
© Metals Processing/Fabrication Facility (3)
x Mine/Quarry/Mineral Processing Facility (5)
? Miscellaneous Commercial/Industrial Facility (5)
R Plastic, Resin, or Rubber Products Plant (8)
A Ship/Boat Manufacturing or Repair Facility (4)
V Steel Mill (1)
12-12
-------
Figure 12-12. Lexington, Kentucky (LEKY) Monitoring Site
^ Fred Douglass
&
i3 &&&
Q™ _ b right St ff
%4L I.
^Booker St-
Coolavin Park
SSo.u'rc«: USGS / V
-------
Figure 12-13. NET Point Sources Located Within 10 Miles of LEKY
Scott
County
Fayette
County
Jessamine
County
Miles
i o-p
, Woodford
County
84°45"0"W 84°40"0"W 84°35-0"W e4°30'0"W 54°25,0"W 64°20'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
LEKY UATMP site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
r
*
0
«
«
B
A
(I)
e
*
F
*
Airport/Airline/Airport Support Operations (9)
Asphalt Production/Hot Mix Asphalt Plant (4)
Auto Body Shop/Painters/Automotive Stores (1)
Automobile/Truck Manufacturing Facility (4)
Building/Construction (1)
Bulk Terminal/Bulk Plant (3)
Concrete Batch Plant (5)
Dry Cleaning Facility (1)
Electrical Equipment Manufacturing Facility (3)
Electricity Generation Facility (2)
Electroplating, Plating, Polishing, Anodizing, and Coloring
Facility (1)
Food Processing/Agriculture Facility (2)
Glass Plant (1)
Hotels/Motels/Lodging (4)
Industrial Machinery or Equipment Plant (7)
O Institution (school, hospital, prison, etc.) (14)
ft Landfill (1)
® Metals Processing/Fabrication Facility (1)
A Military Base/National Security Facility (1)
X Mine/Quarry/Mineral Proccesing Facility (4)
? Miscellaneous Commercial/Industrial Facility (4)
[j Paint and Coating Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (2)
Printing/Publishing/Paper Product Manufacturing Facility
P (3)
TT Telecommunications/Radio Facility (1)
M Tobacco Manufacturing Facility (1)
... Woodwork, Furniture, Millwork and Wood Preserving
vv Facility (3)
Bourbon
County
12-14
-------
Table 12-1. Geographical Information for the Kentucky Monitoring Sites
Site Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection Used for Traffic Data
ASKY
21-019-0017
Ashland
Boyd
Huntington- Ashland,
WV-KY-OH
38.459340,
-82.640410
Residential
Suburban
5,934
29th St between Newman St and
Lynwood Ave
ASKY-M
21-019-0002
Ashland
Boyd
Huntington- Ashland,
WV-KY-OH
38.476000,
-82.631370
Industrial
Urban/City
Center
13,241
Greenup (23rd) between 16th St and
17th St
GLKY
21-043-0500
Grayson
Carter
None
38.238870,
-82.988100
Residential
Rural
303
Rd 1496, S of Camp Webb Rd
BAKY
21-101-0014
Baskett
Henderson
Evansville, IN-KY
37.871200,
-87.463750
Commercial
Rural
929
Rte 1078, N of Hwy 60
ATKY
21-157-0016
Calvert
City
Marshall
None
37.041760,
-88.354070
Industrial
Suburban
3,672
Main St (Rte 95), S of Johnson Rilev
Rd
BLKY
21-139-0004
Smithland
Livingston
Paducah, KY-IL
37.071510,
-88.333890
Agricultural
Rural
2,011
Rte 93/453, E of BloodworthRd
TVKY
21-157-0014
Calvert
City
Marshall
None
37.045200,
-88.330870
Industrial
Suburban
1,458
Industrial Pkwy (Rte 1523), E of
Plant Cut-off Rd
LEKY
21-067-0012
Lexington
Fayette
Lexington-Fayette,
KY
38.065030,
-84.497610
Residential
Suburban
18,993
Newton Pike, N of W Loudon Ave
1AADT reflects 2012 data for GLKY; 2014 data for ASKY, LEKY & TVKY; 2015 data for ASKY-M, ATKY, and BAKY; and 2016 data for BLKY (KYTC, 2016)
BOLD ITALICS = EPA-designated NATTS Site
-------
Two Kentucky monitoring sites are located in the town of Ashland. Ashland is located on
the Ohio River, just north of where the borders of Kentucky, West Virginia, and Ohio meet, and
is part of the Huntington-Ashland, WV-KY-OH CBSA. The ASKY site is located behind the
county health department, in a residential area in the center of town, as shown in Figure 12-1.
The ASKY-M site is located on the roof of an oil company complex in the north-central part of
Ashland, which is more industrial. The ASKY-M monitoring site is located less than one-quarter
mile from the Ohio River, as shown in Figure 12-2, and a rail yard, a scrap yard, and other
industries are located between the site and the river.
ASKY and ASKY-M are approximately 1.25 miles apart, as shown in Figure 12-3. Most
of the emissions sources near these sites are located along the Ohio River and its tributary to the
south, the Big Sandy River. These emissions sources reflect a variety of industries including
asphalt production, chemical manufacturing, food processing, metals processing/fabrication,
pesticide manufacturing, petroleum refining, and ship/boat manufacturing, to name a few. A
cluster of emissions sources is located very close to ASKY-M, within a half-mile, such that the
symbol for the site hides the symbols for the facilities. This cluster includes a testing laboratory,
a miscellaneous commercial/industrial facility, a mine/quarry, and a heliport at a hospital. There
are no emissions sources within a half-mile of ASKY. The closest sources to ASKY are the same
ones under the symbol for ASKY-M, although a coke battery and a ship/boat manufacturing or
repair facility are located a little farther to the east of ASKY.
Grayson Lake is located in northeast Kentucky, south of the town of Grayson, and
southwest of the Huntington-Ashland, WV-KY-OH CBSA. The Little Sandy River feeds into
Grayson Lake, which is a U.S. Army Corps of Engineers-managed project, and part of the
Kentucky State Parks system. The lake is narrow and winding, with sandstone cliffs rising to up
to 200 feet above the lake surface (KY, 2018; ACE, 2018). The closest road to the monitoring
site is a service road feeding into Camp Grayson, as shown in Figure 12-4. This site serves as the
Grayson Lake NATTS site. Figure 12-5 shows that few point sources surround GLKY and that
most of them are on the outer periphery of the 10-mile boundary around GLKY. This is not
surprising given the rural nature of the area and that Grayson Lake is located roughly in the
center of the 10-mile boundary in Figure 12-5. Sources within 10 miles of GLKY are involved in
brick/structural clay/clay ceramics manufacturing, food processing, and mining, among others.
12-16
-------
The BAKY monitoring site is located at the Baskett Fire Department in Baskett, a small
rural town in northwest Kentucky. Baskett is northeast of Henderson and south of Evansville,
Indiana. The Ohio River is the border between Kentucky and Indiana and meanders through the
area, with the Green River, a tributary of the Ohio River, just over 1 mile north of the site at the
closest point. The fire department property backs up to a rail line that runs through town. Open
fields surround the town, as shown in Figure 12-6. There are no emissions sources within a few
miles of BAKY, as shown in Figure 12-7. The cluster of emissions sources to the southwest of
BAKY are located in or near Henderson, while the sources to the northwest are located in
Evansville.
Three monitoring sites are located in and around the Calvert City area. Calvert City is
located on the Tennessee River, east of the Paducah metro area, approximately 6 miles southeast
of the Ohio River and the Kentucky/Illinois border. The northern half of the city is highly
industrialized while the southern half is primarily residential, with a rail line that transverses the
area acting as a pseudo-dividing line. The city is home to some 16 industrial plants, including
metal, steel, and chemical plants (Calvert City, 2018).
The ATKY monitoring site is located off Main Street (State Road 95), just south of the
entrance to a chemical manufacturing plant. The majority of the city's industry lies north and
east of ATKY. Just over one mile east-northeast of ATKY is the TVKY monitoring site. This
monitoring site is located at a power substation just south of another chemical manufacturing
plant. BLKY, the third monitoring site in the Calvert City area, is located across the Tennessee
River, north of Calvert City, in Smithland. The site is located on a residential property in an
agricultural area. This site is potentially downwind of the Calvert City industrial area. These sites
roughly form a triangle around the industrial area of Calvert City. Composite satellite images for
these sites are provided in alphabetical order by site in Figures 12-8 through 12-10.
Figure 12-11 is the facility map for the Calvert City sites and provides an indication of
how close these sites are to one another. Most of the emissions sources in Calvert City are
located between ATKY, TVKY, and the Tennessee River. Many of the emissions sources closest
to the Calvert City sites are in the chemical manufacturing or plastic, resin, or rubber product
source categories. Industries located farther away from the sites but within 10 miles include
ship/boat manufacturing or repair; mine, quarry, or mineral processing; a steel mill; metals
processing/fabrication, and an asphalt production/hot mix asphalt plant.
12-17
-------
The LEKY monitoring site is located in the city of Lexington in north-central Kentucky.
The site is located on the property of the county health department in a primarily residential area
of northern Lexington. A YMCA is located adjacent to the health department along W. Loudon
Avenue and a community college is located immediately to the south. The mental health facility
formerly located on the property has been demolished after relocating. Although the area is
residential and suburban, most of the residences are located to the west of Newtown Pike (922).
An electrical equipment and ink manufacturer is located to the northeast of the site, as shown in
Figure 12-12. LEKY is located just over a half-mile south of New Circle Road (4/421), a loop
encircling the city of Lexington. Figure 12-13 shows that most of the emissions sources within
10 miles of LEKY are within a few miles of the site. Emissions sources within 1 mile of LEKY
include the aforementioned electrical equipment manufacturing plant, a food processing plant, an
institution, and a metals processing and fabrication facility.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 12-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site.
Among these sites, traffic volume is highest near LEKY and ASKY-M and lowest near GLKY
and BAKY. Traffic counts for all of the Kentucky sites are in the bottom half of the range
compared to other NMP sites, with the traffic near GLKY the lowest among all NMP sites.
12.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
monitoring site to identify site-specific "pollutants of interest," which allows analysts and
readers to focus on a subset of pollutants through the context of risk. For each site, each
pollutant's preprocessed daily measurement was compared to its associated risk screening value.
If the concentration was greater than the risk screening value, then the concentration "failed the
screen." Pollutants of interest are those for which the individual pollutant's total failed screens
contribute to the top 95 percent of the site's total failed screens.
It is important to note which pollutants were sampled for at each site when reviewing the
results of this analysis. Table 12-2 provides an overview of which pollutant groups were sampled
for at each site. The site-specific results of the risk-based screening process are presented in
Table 12-3 and incorporate measurements from both 2015 and 2016, with the pollutants of
interest for each site shaded in gray in Table 12-3.
12-18
-------
Table 12-2. Overview of Pollutant Groups Sampled for at the Kentucky Monitoring Sites
Site
VOCs
Carbonyl
Compounds
PAHs
PMio
Metals
ASKY
V
__
__
ASKY-M
__
__
V
GLKY
V
BAKY
__
__
V
ATKY
V
__
__
BLKY
V
__
TVKY
V
__
__
LEKY
V
—
—
-- = This pollutant group was not sampled for at this site.
BOLD ITALICS = EPA-designated NATTS Site
Observations from Table 12-2 include the following:
• Carbonyl compounds, VOCs, PAHs, and PMio metals were sampled for at GLKY.
• Additional sites sampling PMio metals include ASKY-M, BAKY, BLKY, and LEKY.
• Additional sites sampling VOCs include ASKY, ATKY, BLKY, TVKY, and LEKY.
• No additional sites sampled carbonyl compounds or PAHs.
Table 12-3. 2015-2016 Risk-Based Screening Results for the Kentucky Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Health Department, Ashland, Kentucky - ASKY
Benzene
0.13
120
120
100.00
25.21
25.21
Carbon Tetrachloride
0.17
119
120
99.17
25.00
50.21
1,2 -Dichloroethane
0.038
111
112
99.11
23.32
73.53
1.3 -Butadiene
0.03
98
111
88.29
20.59
94.12
Hexachloro-1,3 -butadiene
0.045
14
17
82.35
2.94
97.06
p-Dichlorobcnzcnc
0.091
6
59
10.17
1.26
98.32
Ethylbenzene
0.4
5
120
4.17
1.05
99.37
1,2 -Dibromoethane
0.0017
3
3
100.00
0.63
100.00
Total
476
662
71.90
12-19
-------
Table 12-3. 2015-2016 Risk-Based Screening Results for the Kentucky Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMio)
0.00023
106
110
96.36
53.81
53.81
Nickel (PMio)
0.0021
38
110
34.55
19.29
73.10
Manganese (PMio)
0.03
25
110
22.73
12.69
85.79
Cadmium (PMio)
0.00056
15
110
13.64
7.61
93.40
Lead (PMio)
0.015
13
110
11.82
6.60
100.00
Total
197
550
35.82
Grayson, Kentucky - GLKY
Carbon Tetrachloride
0.17
121
121
100.00
15.71
15.71
Formaldehyde
0.077
121
121
100.00
15.71
31.43
Benzene
0.13
120
121
99.17
15.58
47.01
Acetaldehyde
0.45
118
121
97.52
15.32
62.34
1,2 -Dichloroethane
0.038
106
109
97.25
13.77
76.10
Arsenic (PMio)
0.00023
103
116
88.79
13.38
89.48
1.3 -Butadiene
0.03
58
98
59.18
7.53
97.01
Naphthalene
0.029
10
120
8.33
1.30
98.31
Hexachloro-1,3 -butadiene
0.045
7
9
77.78
0.91
99.22
1,2 -Dibromoethane
0.0017
3
3
100.00
0.39
99.61
Cadmium (PMio)
0.00056
1
117
0.85
0.13
99.74
Manganese (PMio)
0.03
1
117
0.85
0.13
99.87
Nickel (PMio)
0.0021
1
117
0.85
0.13
100.00
Total
770
1,290
59.69
Baskett, Kentucky - BAKY
Arsenic (PMio)
0.00023
108
114
94.74
95.58
95.58
Nickel (PMio)
0.0021
2
113
1.77
1.77
97.35
Antimony (PMio)
0.02
1
114
0.88
0.88
98.23
Lead (PMio)
0.015
1
114
0.88
0.88
99.12
Manganese (PMio)
0.03
1
114
0.88
0.88
100.00
Total
113
569
19.86
12-20
-------
Table 12-3. 2015-2016 Risk-Based Screening Results for the Kentucky Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.13
120
120
100.00
22.47
22.47
Carbon Tetrachloride
0.17
120
120
100.00
22.47
44.94
1,2 -Dichloroethane
0.038
119
119
100.00
22.28
67.23
1.3 -Butadiene
0.03
77
108
71.30
14.42
81.65
Vinyl chloride
0.11
56
95
58.95
10.49
92.13
Hexachloro-1,3 -butadiene
0.045
15
17
88.24
2.81
94.94
1,1,2-Trichloroethane
0.0625
12
15
80.00
2.25
97.19
1,2 -Dibromoethane
0.0017
6
6
100.00
1.12
98.31
Ethylbenzene
0.4
3
120
2.50
0.56
98.88
T richloroethylene
0.2
3
32
9.38
0.56
99.44
p-Dichlorobenzene
0.091
1
27
3.70
0.19
99.63
1,1 -Dichloroethane
0.625
1
31
3.23
0.19
99.81
Xylenes
10
1
120
0.83
0.19
100.00
Total
534
930
57.42
Smithland, Kentucky - BLKY
Benzene
0.13
119
119
100.00
19.51
19.51
Carbon Tetrachloride
0.17
119
119
100.00
19.51
39.02
1,2 -Dichloroethane
0.038
117
117
100.00
19.18
58.20
Arsenic (PMio)
0.00023
98
109
89.91
16.07
74.26
1.3 -Butadiene
0.03
66
100
66.00
10.82
85.08
Vinyl chloride
0.11
46
88
52.27
7.54
92.62
Hexachloro-1,3 -butadiene
0.045
18
23
78.26
2.95
95.57
1.1,2-Trichloroethane
0.0625
11
14
78.57
1.80
97.38
1,2 -Dibromoethane
0.0017
10
10
100.00
1.64
99.02
Manganese (PMio)
0.03
2
110
1.82
0.33
99.34
Nickel (PMio)
0.0021
2
109
1.83
0.33
99.67
p-Dichlorobenzene
0.091
1
29
3.45
0.16
99.84
T richloroethylene
0.2
1
24
4.17
0.16
100.00
Total
610
971
62.82
12-21
-------
Table 12-3. 2015-2016 Risk-Based Screening Results for the Kentucky Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
TVA Substation, Calvert City, Kentucky - TVKY
Benzene
0.13
123
123
100.00
22.32
22.32
Carbon Tetrachloride
0.17
123
123
100.00
22.32
44.65
1,2 -Dichloroethane
0.038
122
122
100.00
22.14
66.79
1.3 -Butadiene
0.03
82
108
75.93
14.88
81.67
Vinyl chloride
0.11
58
95
61.05
10.53
92.20
1,1,2-Trichloroethane
0.0625
23
29
79.31
4.17
96.37
Hexachloro-1,3 -butadiene
0.045
12
16
75.00
2.18
98.55
1,2 -Dibromoethane
0.0017
4
4
100.00
0.73
99.27
Chloroprene
0.0021
2
2
100.00
0.36
99.64
1,1 -Dichloroethane
0.625
2
46
4.35
0.36
100.00
Total
551
668
82.49
Lexin
gton, Kentucky - LEKY
Arsenic (PMio)
0.00023
97
103
94.17
22.61
22.61
Benzene
0.13
82
82
100.00
19.11
41.72
Carbon Tetrachloride
0.17
82
82
100.00
19.11
60.84
1,2 -Dichloroethane
0.038
78
80
97.50
18.18
79.02
1.3 -Butadiene
0.03
66
75
88.00
15.38
94.41
p-Dichlorobcnzcnc
0.091
7
35
20.00
1.63
96.04
Hexachloro-1,3 -butadiene
0.045
6
6
100.00
1.40
97.44
Ethylbenzene
0.4
5
82
6.10
1.17
98.60
1,2 -Dibromoethane
0.0017
2
2
100.00
0.47
99.07
Manganese (PMio)
0.03
1
103
0.97
0.23
99.30
Nickel (PMio)
0.0021
1
103
0.97
0.23
99.53
1,1,2-Trichloroethane
0.0625
1
2
50.00
0.23
99.77
T richloroethylene
0.2
1
8
12.50
0.23
100.00
Total
429
763
56.23
Observations for the Ashland sites from Table 12-3 include the following:
• The pollutants failing screens is very different between these two monitoring sites;
this is expected given the different pollutants measured at each site. As shown in
Table 12-2, VOCs were sampled for at ASKY while PMio metals were sampled for at
ASKY-M.
• Concentrations of eight VOCs failed at least one screen for ASKY, with 72 percent of
concentrations for these eight pollutants greater than their associated risk screening
value (or failing screens).
12-22
-------
• Five VOCs contributed to 95 percent of failed screens for ASKY and therefore were
identified as pollutants of interest. These five are benzene, 1,3-butadiene, carbon
tetrachloride, 1,2-dichloroethane, and hexachloro-l,3-butadiene.
• Concentrations of five metals failed at least one screen for ASKY-M, with 36 percent
of concentrations for these five pollutants greater than their associated risk screening
value (or failing screens).
• All five of these metals (arsenic, nickel, manganese, cadmium, and lead) contributed
to 95 percent of failed screens for ASKY-M and therefore were identified as
pollutants of interest. ASKY-M is the only NMP site with lead as a pollutant of
interest. ASKY-M is one of only two NMP sites with manganese as a pollutant of
interest (TOOK is the other). This is also true for cadmium (BOMA is the other site).
Observations for GLKY from Table 12-3 include the following:
• All four pollutant groups shown in Table 12-2 were sampled for at GLKY.
• Concentrations of 13 pollutants failed at least one screen for GLKY, with nearly
60 percent of concentrations for these 13 pollutants greater than their associated risk
screening value (or failing screens).
• Seven pollutants contributed to 95 percent of failed screens for GLKY and therefore
were identified as pollutants of interest. These include two carbonyl compounds, four
VOCs, and one metal.
Observations for BAKY from Table 12-3 include the following:
• BAKY sampled for PMio metals only.
• Concentrations of five PMio metals failed at least one screen for BAKY, although a
single metal (arsenic) accounts for 108 of 113 of the total failed screens. Only one or
two concentrations for each of the other PMio metals failed screens.
• With arsenic contributing to 96 percent of the failed screens for BAKY, this pollutant
was identified as BAKY's sole pollutant of interest.
Observations for the Calvert City sites from Table 12-3 include the following:
• VOCs were sampled for at all three Calvert City sites. PMio metals were also sampled
for at BLKY.
• Concentrations of 13 VOCs failed screens for ATKY; seven of these VOCs
contributed to 95 percent of failed screens for ATKY and thus, were identified as
pollutants of interest for this site.
• Concentrations of 13 pollutants failed screens for BLKY; seven VOCs and one PMio
metal (arsenic) contributed to 95 percent of failed screens for BLKY and thus, were
identified as pollutants of interest for this site.
12-23
-------
• Concentrations of 10 VOCs failed screens for TVKY; six of these VOCs contributed
to 95 percent of failed screens for TVKY and thus, were identified as pollutants of
interest for this site.
• Benzene, carbon tetrachloride, 1,2-dichloroethane, 1,3-butadiene, and vinyl chloride
were identified as pollutants of interest for all thee Calvert City sites. These sites are
the only NMP sites with vinyl chloride as a pollutant of interest.
Observations for LEKY from Table 12-3 include the following:
• VOCs and PMio metals were sampled for at LEKY.
• Concentrations of 13 pollutants failed at least one screen for LEKY, with 56 percent
of concentrations of these 13 pollutants greater than their associated risk screening
value (or failing screens).
• Six pollutants contributed to 95 percent of failed screens for LEKY and therefore
were identified as pollutants of interest. These include five VOCs and one metal.
12.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Kentucky monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at the Kentucky
monitoring sites are provided in Appendices J, M, N, and O.
12-24
-------
12.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages were calculated for 2015 and 2016 for the
pollutants of interest for the Kentucky sites, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the Kentucky
monitoring sites are presented in Table 12-4, where applicable. Note that concentrations of the
PAHs and metals are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
12-25
-------
Table 12-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Kentucky Monitoring Sites
to
to
On
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
Pollutant
# >MDL/
# Samples
Avg
frig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Average
frig/m3)
# >MDL/
# Samples
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Average
frig/m3)
Health Department, Ashland, Kentucky - ASKY
0.80
0.74
0.69
0.88
0.78
0.91
0.49
0.51
0.69
0.65
Benzene
60/60/60
±0.19
±0.33
±0.16
±0.22
±0.11
60/60/60
±0.22
±0.10
±0.11
±0.25
±0.10
0.06
0.07
0.05
0.08
0.06
0.09
0.06
0.04
0.07
0.06
1.3 -Butadiene
56/50/60
±0.02
±0.02
±0.01
±0.02
±0.01
55/25/60
±0.03
±0.02
±0.01
±0.03
±0.01
0.55
0.62
0.67
0.63
0.62
0.63
0.70
0.65
0.58
0.64
Carbon Tetrachloride
60/60/60
±0.09
±0.03
±0.02
±0.05
±0.03
60/60/60
±0.04
±0.05
±0.04
±0.04
±0.02
0.09
0.07
0.06
0.08
0.08
0.09
0.08
0.05
0.06
0.07
1,2 -Dichloroethane
59/49/60
±0.01
±0.02
±0.01
±0.01
±0.01
53/50/60
±0.01
±0.01
±0.02
±0.02
±0.01
0.02
0.02
0.01
0.02
0.02
Hexachloro-1,3 -butadiene
2/0/60
NR
NR
NR
NR
NR
15/0/60
±0.02
±0.02
±0.02
±0.02
±0.01
21st and Greenup, Ashland, Kentucky
- ASKY-M
0.78
1.94
1.44
1.32
1.38
1.00
1.34
1.18
1.13
Arsenic (PMi0)a
55/55/55
±0.26
±0.95
±0.52
±0.47
±0.32
55/55/55
±0.47
±0.45
NA
±0.55
±0.23
0.32
0.44
0.29
0.23
0.32
0.68
0.31
0.20
0.39
Cadmium (PMi0)a
55/55/55
±0.20
±0.23
±0.07
±0.08
±0.08
55/55/55
±0.74
±0.13
NA
±0.09
±0.22
8.34
11.57
8.88
6.34
8.83
6.78
8.81
7.40
6.97
Lead (PMi0)a
55/55/55
±4.54
±5.98
±2.48
±2.26
±2.08
55/55/55
±2.93
±4.08
NA
±4.92
± 1.86
25.75
33.28
23.71
16.67
25.05
20.65
26.86
16.77
19.85
Manganese (PMi0)a
55/55/55
± 13.87
± 15.13
±6.32
±6.30
±5.70
55/55/55
± 10.28
± 12.18
NA
±7.69
±4.75
2.24
2.26
2.21
2.09
2.20
3.26
2.06
1.71
2.12
Nickel (PMi,:,)a
55/53/55
± 1.52
±0.86
±0.83
±0.72
±0.49
55/55/55
± 1.78
±0.87
NA
±0.90
±0.62
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 12-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Kentucky Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
Grayson, Kentucky - GLKY
0.92
1.13
0.77
0.67
0.87
0.69
1.06
0.76
0.93
0.86
Acetaldehyde
60/60/60
±0.11
±0.17
±0.10
±0.12
±0.07
61/61/61
±0.07
±0.27
±0.09
±0.30
±0.10
0.56
0.28
0.31
0.42
0.39
0.46
0.26
0.28
0.54
0.39
Benzene
60/60/60
±0.10
±0.05
±0.05
±0.08
±0.04
61/61/61
±0.05
±0.03
±0.05
±0.12
±0.04
0.03
0.03
0.03
0.04
0.03
0.04
0.02
0.03
0.04
0.03
1.3 -Butadiene
51/28/60
±0.01
±0.01
±0.01
±0.01
±0.01
47/7/61
±0.01
±0.01
±0.01
±0.01
±0.01
0.63
0.60
0.65
0.62
0.62
0.61
0.68
0.64
0.59
0.63
Carbon Tetrachloride
60/60/60
±0.03
±0.03
±0.04
±0.07
±0.02
61/61/61
±0.04
±0.04
±0.06
±0.05
±0.02
0.08
0.07
0.04
0.06
0.06
0.07
0.07
0.03
0.05
0.06
1,2 -Dichloroethane
58/40/60
±0.01
±0.01
±0.01
±0.01
±0.01
51/44/61
± <0.01
±0.01
±0.02
±0.02
±0.01
1.06
2.55
2.40
1.03
1.76
1.15
2.40
2.34
1.65
1.87
Formaldehyde
60/60/60
±0.21
±0.54
±0.35
±0.23
±0.25
61/61/61
±0.17
±0.57
±0.31
±0.53
±0.24
0.40
0.53
0.67
0.57
0.54
0.46
0.60
0.63
0.44
0.53
Arsenic (PMi0)a
57/56/58
±0.13
±0.19
±0.20
±0.22
±0.09
59/59/59
±0.25
±0.18
±0.32
±0.12
±0.11
Baskett, Kentucky - BAKY
0.59
0.95
1.49
0.74
0.96
0.49
1.07
1.34
0.84
0.92
Arsenic (PMio)a
56/54/56
±0.13
±0.26
±0.44
±0.41
±0.19
58/58/58
±0.14
±0.33
±0.26
±0.34
±0.15
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 12-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Kentucky Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
Atmos Energy, Calvert City, Kentucky - ATKY
0.68
0.90
1.22
1.01
0.96
0.55
0.49
0.72
0.53
0.57
Benzene
59/59/59
±0.15
±0.71
±0.71
±0.89
±0.32
61/61/61
±0.11
±0.08
±0.41
±0.09
±0.10
0.09
0.05
0.06
0.04
0.06
0.03
0.05
0.06
0.05
0.05
1.3 -Butadiene
56/39/59
±0.04
±0.02
±0.04
±0.02
±0.02
52/22/61
±0.02
±0.02
±0.03
±0.02
±0.01
0.66
0.65
0.73
0.65
0.67
0.64
0.72
0.73
0.66
0.69
Carbon Tetrachloride
59/59/59
±0.03
±0.03
±0.10
±0.03
±0.03
61/61/61
±0.04
±0.04
±0.09
±0.05
±0.03
0.37
0.31
0.77
0.19
0.41
0.12
1.39
1.89
0.26
0.90
1,2 -Dichloroethane
59/58/59
±0.25
±0.24
±0.67
±0.12
±0.19
60/59/61
±0.05
± 1.72
± 1.42
±0.15
±0.55
0.03
0.03
0.02
0.02
Hexachloro-1,3 -butadiene
3/0/59
NR
NR
NR
NR
NR
14/0/61
0
±0.02
±0.03
±0.02
±0.01
<0.01
0.01
0.02
0.01
0.02
0.30
0.04
0.09
1,1,2-Trichloroethane
6/1/59
0
±0.01
±0.01
±0.03
±0.01
9/9/61
0
±0.03
±0.33
±0.05
±0.08
0.87
0.22
1.00
0.58
0.69
0.28
0.72
1.37
1.00
0.83
Vinyl chloride
45/41/59
±0.62
±0.29
±0.67
±0.87
±0.32
50/30/61
±0.34
±0.46
± 1.34
± 1.78
±0.54
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 12-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Kentucky Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
Smithland, Kentucky - BLKY
0.55
0.59
0.57
0.54
0.56
0.57
0.43
0.45
0.53
0.49
Benzene
59/59/59
±0.09
±0.15
±0.15
±0.14
±0.06
60/60/60
±0.10
±0.13
±0.18
±0.15
±0.07
0.03
0.11
0.03
0.07
0.06
0.05
0.05
0.08
0.06
0.06
1.3 -Butadiene
51/33/59
±0.02
±0.05
±0.01
±0.04
±0.02
49/18/60
±0.03
±0.04
±0.09
±0.03
±0.03
0.60
0.67
0.68
0.73
0.67
0.69
0.79
0.76
0.70
0.73
Carbon Tetrachloride
59/59/59
±0.06
±0.04
±0.04
±0.10
±0.03
60/60/60
±0.09
±0.07
±0.08
±0.08
±0.04
0.16
1.37
0.52
0.79
0.72
3.47
0.94
1.91
1.24
1.89
1,2 -Dichloroethane
59/56/59
±0.10
±0.75
±0.33
±0.77
±0.29
58/57/60
±6.23
±0.78
± 1.57
±0.83
± 1.55
0.01
0.05
0.02
0.02
0.03
Hexachloro-1,3 -butadiene
6/0/59
NR
NR
NR
NR
NR
17/0/60
±0.01
±0.03
±0.03
±0.02
±0.01
0.04
0.14
0.10
0.13
0.10
0.28
0.08
0.15
0.16
0.17
Vinyl chloride
41/37/59
±0.05
±0.07
±0.08
±0.08
±0.04
47/27/60
±0.25
±0.07
±0.10
±0.11
±0.07
0.55
0.55
0.67
0.57
0.40
0.63
0.61
0.56
0.54
Arsenic (PMio)a
50/49/51
±0.16
±0.16
±0.15
NA
±0.09
59/59/59
±0.09
±0.17
±0.12
±0.15
±0.07
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 12-4. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Kentucky Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
TVA Substation,
Calvert City, Kentucky - TVKY
1.20
1.04
1.14
0.78
1.04
0.62
0.53
0.43
0.57
0.54
Benzene
62/62/62
±0.66
±0.73
±0.72
±0.36
±0.30
61/61/61
±0.17
±0.17
±0.24
±0.16
±0.09
0.44
0.35
0.33
0.29
0.35
0.11
0.39
0.05
0.08
0.16
1.3 -Butadiene
56/43/62
±0.38
±0.35
±0.33
±0.49
±0.19
52/21/61
±0.11
±0.49
±0.05
±0.07
±0.12
0.93
0.70
0.94
0.83
0.85
0.70
0.93
0.85
0.72
0.80
Carbon Tetrachloride
62/62/62
±0.27
±0.04
±0.42
±0.24
±0.13
61/61/61
±0.05
±0.32
±0.24
±0.06
±0.10
4.73
5.02
2.72
2.57
3.75
4.63
3.47
1.14
4.63
3.49
1,2 -Dichloroethane
62/61/62
±3.95
±3.76
±2.48
±2.55
± 1.56
60/60/61
±3.56
±2.31
±0.96
±3.33
± 1.36
0.03
0.01
0.02
0.06
0.03
0.03
0.05
<0.01
0.32
0.09
1,1,2-Trichloroethane
13/9/62
±0.02
±0.01
±0.03
±0.08
±0.02
16/10/61
±0.03
±0.05
±0.01
±0.33
±0.08
0.30
0.23
0.30
0.35
0.30
0.21
0.44
0.11
0.53
0.31
Vinyl chloride
42/41/62
±0.20
±0.16
±0.33
±0.30
±0.12
53/32/61
±0.14
±0.30
±0.08
±0.29
±0.11
Lexington, Kentucky - LEKY
0.58
0.49
0.45
0.51
0.52
0.35
Benzene
53/53/53
±0.09
±0.13
±0.10
NA
±0.06
29/29/29
±0.10
±0.06
NA
NS
NA
0.04
0.07
0.06
0.06
0.05
0.05
1.3 -Butadiene
47/44/53
±0.02
±0.03
±0.02
NA
±0.01
28/9/29
±0.03
±0.02
NA
NS
NA
0.61
0.60
0.66
0.63
0.68
0.71
Carbon Tetrachloride
53/53/53
±0.04
±0.04
±0.03
NA
±0.02
29/29/29
±0.04
±0.03
NA
NS
NA
0.02
0.04
0.04
0.03
0.02
0.03
p-Dichlorobenzene
26/2/53
±0.01
±0.03
±0.02
NA
±0.01
9/0/29
±0.02
±0.02
NA
NS
NA
0.08
0.07
0.05
0.07
0.09
0.08
1,2 -Dichloroethane
51/40/53
±0.01
±0.01
±0.01
NA
±0.01
29/29/29
±0.01
±0.01
NA
NS
NA
0.60
0.75
0.80
1.06
0.81
0.78
0.81
Arsenic (PMi0)a
56/56/56
±0.14
±0.18
±0.18
±0.60
±0.17
47/47/47
NA
±0.16
NA
±0.28
NA
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Observations for the Ashland sites from Table 12-4 include the following:
• VOCs were sampled for at ASKY and PMio metals were sampled for at ASKY-M.
Thus, these sites have no pollutants of interest in common.
• For 2015, benzene is the pollutant of interest with the highest annual average
concentration for ASKY (0.78 ±0.11 |ig/m3), followed by carbon tetrachloride
(0.62 ± 0.03 |ig/m3). The annual averages for these two pollutants are an order of
magnitude greater than the annual averages for 1,3-butadiene and 1,2-dichloroethane.
For 2016, this is also true, but the difference between the annual averages of benzene
(0.65 ± 0.10 |ig/m3) and carbon tetrachloride (0.64 ± 0.02 |ig/m3) is negligible.
• Concentrations of benzene measured at ASKY range from 0.157 |ig/m3 to
2.80 |ig/m3. Quarterly average concentrations exhibit considerable variability,
particularly for 2016, when the first quarter average (0.91 ± 0.22 |ig/m3) is almost
twice the second quarter average (0.49 ± 0.10 |ig/m3). In 2016, two-thirds of the nine
benzene concentrations greater than 1 |ig/m3 were measured during the first quarter of
the year, compared to none during the second and third quarters and three during the
fourth. By comparison, 14 benzene concentrations greater than 1 |ig/m3 were
measured in 2015, but at least one was measured during each calendar quarter (four
were measured during the first quarter, one (the maximum) during the second, two
during the third, and seven during the fourth).
• Concentrations of carbon tetrachloride measured at ASKY range from 0.132 |ig/m3 to
0.820 |ig/m3, although few measurements fall outside the 0.50 |ig/m3 to 0.75 |ig/m3
range. The six concentrations greater than 0.75 |ig/m3 were all measured at ASKY in
2016, with four measured during the second quarter of the year, including three back-
to-back sample days in May, explaining the slightly higher quarterly average for this
calendar quarter.
• A few non-detects of 1,3-butadiene and 1,2-dichloroethane were measured at ASKY.
For 1,3-butadiene, the annual average concentration for 2015 is the same as the
annual average for 2016 (0.06 ± 0.01 |ig/m3), although the quarterly average
concentrations exhibit slightly more variability in 2016. The annual averages of
1,2-dichloroethane for each year are also similar to each other.
• Quarterly and annual averages for hexachloro-1,3-butadiene for 2015 are not
presented in Table 12-4 due to the use of a contaminated internal standard at the
laboratory for Method TO-15, which resulted in the invalidation of the results from
early March 2015 through mid-December 2015, as described in Section 2.4. For
2016, the quarterly average concentrations vary little across the calendar quarters.
Note that none of the 17 measured detections of hexachloro-l,3-butadiene measured
at ASKY were greater than the MDL for this pollutant.
• Table 4-10 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. ASKY does not appear on
this list.
12-31
-------
Each of the five metal pollutants of interest for ASKY-M were detected in all of the
valid samples collected at this site in 2015 and 2016.
The pollutant of interest with the highest annual average concentrations for ASKY-M
is manganese (25.05 ± 5.70 ng/m3 for 2015 and 19.85 ± 4.75 ng/m3 for 2016),
followed by lead (8.83 ± 2.08 ng/m3 for 2015 and 6.97 ± 1.86 ng/m3 for 2016) and
nickel (2.20 ± 0.49 ng/m3 for 2015 and 2.12 ± 0.62 ng/m3 for 2016).
Concentrations of manganese measured at ASKY-M range from 1.49 ng/m3 to
117 ng/m3, which is the second highest manganese concentration measured across the
program. Manganese concentrations greater than 50 ng/m3 were measured at four
NMP sites sampling PMio metals, with measurements from ASKY-M accounting for
most of them (10 out of 14). Nine of these 10 manganese concentrations were
measured at ASKY-M during the first and second calendar quarters; five were
measured in 2015, three in March and two in May (including the maximum
concentration) and four were measured in 2016, but these were more spread out
across the months of the first and second quarters.
Concentrations of lead measured at ASKY-M range from 0.640 ng/m3 to 41.4 ng/m3,
which is the second highest lead concentration measured across the program. The
maximum lead concentration was also measured at ASKY-M during May 2015, but
on a different sample day than the highest manganese concentration was measured.
Concentrations of lead measured at ASKY-M account for six of the 10 lead
measurements greater than 20 ng/m3 measured at NMP sites sampling PMio metals;
ASKY-M is the only NMP site at which more than one was measured.
Concentrations of arsenic measured at ASKY-M range from 0.153 ng/m3 to
7.36 ng/m3. The maximum arsenic concentration measured at ASKY-M is the
maximum arsenic concentration measured across the program. This concentration
was measured on May 24, 2015, the same day that the highest manganese
concentration was measured at ASKY-M. This explains, at least in part, the relatively
large confidence interval shown in Table 12-4 for the second quarter average of 2015.
This is also the calendar quarter in which the most arsenic concentrations greater than
2 ng/m3 were measured (5). ASKY-M has the highest number of arsenic
measurements greater than 2 ng/m3 compared to any other NMP site (16, which is
twice the site(s) with next highest, for which BAKY and BTUT are tied).
Concentrations of cadmium measured at ASKY-M range from 0.032 ng/m3 to
5.99 ng/m3. The maximum cadmium concentration measured at ASKY-M is the
maximum cadmium concentration measured across the program. This concentration
is more than three times higher than the next highest concentration measured at
ASKY-M (1.75 ng/m3); this concentration was measured on January 25, 2016 and the
effects of this outlier can be seen in the first quarter average concentration for 2016 as
the associated confidence interval is greater than the average itself. Of the 10
cadmium concentrations greater than 1 ng/m3 measured across the program, half were
measured at ASKY-M.
12-32
-------
• Quarterly average concentrations for the third quarter of 2016 could not be calculated
because there were too many invalid samples during this calendar quarter.
• Table 4-13 presents the NMP sites with the 10 highest annual average concentrations
for the program-level metal pollutant of interest (arsenic). This table shows that the
highest annual average concentrations for arsenic were calculated for ASKY-M
(2015, followed by 2016). Similar observations were made in the 2013 and 2014
NMP reports. ASKY-M has the only annual average concentrations of arsenic greater
than 1 ng/m3 (1.38 ± 0.32 ng/m3 for 2015 and 1.13 ± 0.23 ng/m3 for 2016).
Observations for GLKY from Table 12-4 include the following:
• GLKY sampled VOCs, carbonyl compounds, metals (PMio), and PAHs.
• The only pollutant of interest for GLKY with an annual average concentration greater
than 1 |ig/m3 is formaldehyde (1.76 ± 0.25 |ig/m3 for 2015 and 1.87 ± 0.24 |ig/m3 for
2016). However, these are among some of the lowest annual averages of
formaldehyde calculated among NMP sites sampling carbonyl compounds. The
annual average concentrations of formaldehyde also exhibit the largest year-to-year
difference; the annual average concentrations for each of the other pollutants of
interest for GLKY vary by 0.01 |ig/m3 or less between the two years of sampling.
• Concentrations of formaldehyde measured at GLKY range from 0.446 |ig/m3 to
5.43 |ig/m3. The highest formaldehyde concentration for each year was measured on
the same day, June 11th. Concentrations of formaldehyde were higher during the
warmer months of 2015, based on the quarterly averages. This is mostly true for
2016, although a few higher concentrations were measured in October and November
2016. The 31 formaldehyde concentrations greater than 2.50 |ig/m3 measured at
GLKY were split nearly evenly across the two years of sampling. For 2015, all 16 of
these concentrations were measured during the second or third quarters of the year.
For 2016, 12 of these concentrations were measured during the second or third
quarters of the year, with the other three measured on back-to-back sample days
during the fourth quarter.
• Concentrations of acetaldehyde measured at GLKY range from 0.271 |ig/m3 to
2.55 |ig/m3, Acetaldehyde concentrations greater than 1 |ig/m3 were not measured at
GLKY for a nine-month stretch between August 2015 and March 2016, which is
reflected in the quarterly average concentrations shown for this pollutant.
• Among the VOC pollutants of interest for GLKY, carbon tetrachloride has the highest
annual average concentrations. The quarterly average concentrations of carbon
tetrachloride exhibit little variability across the two years of sampling, varying by less
than 0.1 |ig/m3. Concentrations of carbon tetrachloride measured at GLKY range
from 0.182 |ig/m3 to 0.820 |ig/m3, with only a few measurements falling outside the
0.50 |ig/m3 to 0.75 |ig/m3 range.
12-33
-------
• Based on the quarterly average concentrations shown, benzene concentrations appear
higher during the colder months of the year. Concentrations of benzene measured at
GLKY range from 0.128 |ig/m3 to 1.14 |ig/m3, with no concentrations greater than
0.5 |ig/m3 measured outside the first or fourth quarters of either year.
• All of GLKY's 1,3-butadiene concentrations are less than 0.1 |ig/m3; GLKY is one of
three NMP sites sampling 1,3-butadiene with Method TO-15 for which this is true.
There is little variability in the quarterly average concentrations shown for this
pollutant.
• All but one of GLKY's 1,2-dichloroethane concentrations are also less than
0.1 |ig/m3. Twelve non-detects of 1,2-dichloroethane were measured at GLKY, two in
2015 and 10 in 2016. The two non-detects in 2015 were measured in August and
September; seven of the 10 non-detects in 2016 were also measured in August and
September (with one measured during each month during the fourth quarter). These
non-detects are reflecting the quarterly average concentrations shown in Table 12-4.
• Arsenic is the only other pollutant of interest for GLKY that is not a VOC or carbonyl
compound. Concentrations of arsenic measured at GLKY range from 0.099 ng/m3 to
2.35 ng/m3, plus one non-detect. For both years, the third quarter has the highest
quarterly average concentration (although the differences are not statistically
significant). The maximum arsenic concentration was measured during the third
quarter of each year. In addition, the third quarter is the only calendar quarter for each
year with a median arsenic concentration greater than 0.5 ng/m3.
• GLKY is not listed in Tables 4-10 through 4-13, which present the NMP sites with
the highest annual average concentrations for each of the program-level pollutants of
interest. The annual average concentrations for GLKY's pollutants of interest are
among some of the lowest across the program. Only one NMP site has an annual
average benzene concentration less than GLKY's.
Observations for BAKY from Table 12-4 include the following:
• Speciated metals were sampled for at BAKY; only arsenic was identified as a
pollutant of interest for BAKY.
• Arsenic was measured in all 114 valid metals samples collected at BAKY.
• Arsenic concentrations measured at BAKY range from 0.008 ng/m3 to 3.15 ng/m3,
which was measured on July 5, 2015. The COC for this sample indicated fireworks
during sampling. The four highest arsenic concentrations were measured at BAKY in
2015, with three of the four measured during the third quarter. The third quarter
average concentration is the highest quarterly average for each year, although there is
considerable variability in the arsenic concentrations measured each quarter. Half of
the 18 arsenic concentrations greater than 1 ng/m3 measured at BAKY in 2015 were
measured during the third quarter; nearly half (10) of the 22 arsenic concentrations
greater than 1 ng/m3 in 2016 were measured during the third quarter.
12-34
-------
• Among NMP sites sampling PMio metals, BAKY has the third (2015) and fifth
(2016) highest annual average concentrations of arsenic, behind only ASKY-M
(which ranks first and second) and NBIL (which ranks fourth), as shown in
Table 4-13. Similar observations were made in the 2014 and 2013 NMP reports.
Observations for the Calvert City monitoring sites from Table 12-4 include the following:
• VOC samples were collected at all three Calvert City sites (ATKY, BLKY, and
TVKY); PMio metals were also sampled for at BLKY.
• Benzene, carbon tetrachloride, 1,2-dichloroethane, 1,3-butadiene, and vinyl chloride
were identified as pollutants of interest for all three Calvert City sites. These sites are
the only NMP sites with vinyl chloride as a pollutant of interest.
• Some of the highest concentrations of VOCs were measured at the Calvert City sites
and these data are reviewed in the bullets that follow.
• When detected, vinyl chloride is generally measured at relatively low levels under the
NMP. Across the program, this pollutant was detected in 32 percent of the total
samples collected, and of these, two-thirds were less than the MDL. Together, the
Calvert City sites account for 278 of the 914 measured detections of this pollutant.
The Calvert City sites account for all 146 concentrations of vinyl chloride greater
than 0.15 |ig/m3 measured across the program, including the 39 measurements greater
than 1 |ig/m3. The seven highest concentrations of vinyl chloride across the program
were measured at ATKY, with these measurements ranging from 3.42 |ig/m3 to
13.1 |ig/m3. Twenty-five vinyl chloride concentrations of at least 1 |ig/m3 were
measured at ATKY, with 12 measured at TVKY and two measured at BLKY.
• As shown in Table 12-4, annual average concentrations of vinyl chloride for these
three sites range from 0.10 ± 0.04 |ig/m3 for BLKY in 2015 to 0.83 ± 0.54 |ig/m3 for
ATKY in 2016. The annual average and quarterly average concentrations of vinyl
chloride for these sites have relatively large confidence intervals, including several
that are larger than the average itself (such as ATKY's fourth quarter average for
2015, 0.58 ± 0.87 |ig/m3). This is indicative of the large amount of variability
associated with these measurements.
• Another pollutant for which the highest concentrations program-wide were measured
at the Calvert City sites is 1,2-dichloroethane. The 174 highest concentrations of
1,2-dichloroethane across the program were measured at the Calvert City sites. This
includes all 100 measurements greater than 1 |ig/m3 and 22 greater than 10 |ig/m3.
• Annual average concentrations of 1,2-dichloroethane for these sites range from
0.41 ± 0.19 |ig/m3 for ATKY (2015) to 3.75 ± 1.56 |ig/m3 for TVKY (2015). Each of
the Calvert City sites has at least two quarterly average concentrations of
1,2-dichloroethane greater than 1 |ig/m3; in the case of TVKY, all of the quarterly
averages are greater than 1 |ig/m3 and include one greater than 5 |ig/m3. Nearly all the
quarterly and annual averages have a relatively large confidence interval associated
with them, some greater than the average itself, indicating the relatively large amount
of variability associated with these measurements.
12-35
-------
The highest measurements of carbon tetrachloride across the program were also
measured at the Calvert City sites. All 22 carbon tetrachloride concentrations greater
than or equal to 1 |ig/m3 measured across the program were measured at the Calvert
City sites (12 measured at TVKY, eight at BLKY, and two at ATKY). All five carbon
tetrachloride concentrations greater than 2 |ig/m3 were measured at TVKY.
Annual average concentrations of carbon tetrachloride for the Calvert City sites range
from 0.67 ± 0.03 |ig/m3 for ATKY and BLKY (2015) to 0.85 ± 0.13 |ig/m3 for TVKY
(2015). Quarterly average concentrations for TVKY exhibit the most variability,
ranging from 0.70 ± 0.04 |ig/m3 for the second quarter of 2015 to 0.94 ± 0.42 |ig/m3
for the third quarter of 2015. Most of the quarterly average concentrations calculated
for NMP sites sampling carbon tetrachloride fall between 0.55 |ig/m3 and 0.75 |ig/m3;
five of TVKY's quarterly averages and two of BLKY's quarterly averages are outside
this range.
Nine of the 10 1,3-butadiene concentrations greater than 1 |ig/m3 measured across the
program were measured at TVKY. Concentrations of 1,3-butadiene greater than
0.5 |ig/m3 were measured at only five NMP sites in 2015 and 2016; concentrations
measured at TVKY account for 15 of these 41 concentrations (tying with PXSS for
the most). A single measurement greater than 0.5 |ig/m3 was measured at BLKY.
Annual average concentrations of 1,3-butadiene for the Calvert City sites range from
0.05 ± 0.01 |ig/m3 for ATKY (2016) to 0.35 ± 0.19 |ig/m3 for TVKY (2015).
Quarterly average concentrations of 1,3-butadiene for TVKY exhibit considerable
variability, ranging from 0.05 ± 0.05 |ig/m3 for the third quarter of 2016 to
0.44 ± 0.38 |ig/m3 for the first quarter of 2015. Many of the quarterly average
concentrations for TVKY have relatively large confidence intervals associated with
them, several of which are equivalent to or greater than the average itself.
Benzene is the only other VOC that is a pollutant of interest across the Calvert City
sites. The maximum benzene concentration measured across the program
(6.85 |ig/m3) was measured at ATKY on October 9, 2015. Concentrations measured
at ATKY and TVKY account for seven of the 10 benzene measurements greater than
4 |ig/m3 measured across the program. These seven concentrations were all measured
in 2015. Most of the benzene concentrations greater than 1 |ig/m3 measured at these
two sites were measured in 2015 (30 out of 38). The quarterly and annual average
concentrations of benzene for ATKY and TVKY reflect this; for both sites, the annual
average concentration for 2015 is roughly twice the annual average for 2016 (for
example, TVKY's annual average benzene concentration for 2015 is
1.04 ± 0.30 |ig/m3, compared to TVKY's annual average of 0.54 ± 0.09 |ig/m3 for
2016). The difference in the magnitude of benzene concentrations between the two
years of sampling is much less at BLKY.
Hexachloro-1,3-butadiene is a pollutant of interest for ATKY and BLKY. Quarterly
and annual averages for hexachloro-l,3-butadiene for 2015 are not presented in
Table 12-4 due to the use of a contaminated internal standard at the laboratory for
Method TO-15, which resulted in the invalidation of measurements from early
March 2015 through mid-December 2015, as described in Section 2.4. The maximum
concentration of hexachloro-1,3-butadiene measured at BLKY (0.15 |ig/m3) ties with
12-36
-------
the same concentration measured at five other sites (including TVKY) as the second
highest concentration of this pollutant across the program. However, even this
measurement is less than the MDL (only one concentration measured over the two-
year sampling period is greater than the MDL for hexachloro-l,3-butadiene).
• 1,1,2-Trichloroethane is a pollutant of interest for ATKY and TVKY, the only NMP
sites for which this is true. This pollutant was detected in 88 samples collected across
the program in 2015 and 2016, with measurements from the Calvert City sites
accounting for 58 of them. This pollutant was detected 29 times at TVKY, 15 times at
ATKY, and 14 times at BLKY. TVKY, ATKY, and BLKY account for all 32
concentrations of this pollutant greater than 0.1 |ig/m3, including four greater than
1 |ig/m3 and one greater than 2 |ig/m3 (all of which were measured in 2016). This
pollutant was still detected infrequently at these sites, though, and in some cases, was
not detected at every site during some calendar quarters. Combining many zeros
substituted for non-detects with a few relatively high measurements, results in
quarterly and annual averages exhibiting large confidence intervals, such as ATKY
and TVKY's 2016 annual averages, both 0.09 ± 0.08 |ig/m3.
• Table 4-10 presents the NMP sites with the 10 highest annual average concentrations
for each of the program-level VOC pollutants of interest. The Calvert City sites
appear in Table 4-10 17 times. Both years' annual averages for the three Calvert City
sites appear in Table 4-10 for carbon tetrachloride and 1,2-dichloroethane, with the
averages for TVKY ranking highest for each pollutant. TVKY also has the highest
(2015) and sixth highest (2016) annual average concentrations of 1,3-butadiene
across the program. TVKY also has the tenth highest annual average benzene
concentration (2015) among sites sampling this pollutant. BLKY and ATKY's 2016
annual average concentrations of hexachloro-1,3-butadiene rank fifth and eighth
highest, respectively, among NMP sites sampling this pollutant.
• Metals (PMio) were also sampled for at BLKY; arsenic is the only non-VOC
pollutant of interest for BLKY. Concentrations of arsenic measured at BLKY range
from 0.018 ng/m3 to 1.62 ng/m3, plus one non-detect. BLKY's annual average arsenic
concentration for 2015 (0.57 ± 0.09 |ig/m3) is similar to this site's 2016 annual
average (0.54 ± 0.07 |ig/m3).
Observations for LEKY from Table 12-4 include the following:
• VOC and speciated metals samples were collected at LEKY in 2015 and 2016,
although VOC sampling was discontinued at the end of July 2016.
• Based on the available quarterly and annual average concentrations available for
LEKY, the pollutant of interest with the highest concentrations is carbon
tetrachloride. Concentrations of carbon tetrachloride measured at LEKY range from
0.473 |ig/m3 to 0.832 |ig/m3, with only a few measurements outside the 0.50 |ig/m3 to
0.75 |ig/m3 range. Five of the six carbon tetrachloride concentrations greater than
0.75 |ig/m3 were measured in 2016, explaining, at least in part, the higher quarterly
average concentrations shown for 2016 compared to those for 2015. At the other end
of the concentration scale, a greater number of concentrations less than 0.6 |ig/m3
were measured in 2015 (17) compared to 2016 (one), further explaining the lower
12-37
-------
quarterly average concentrations for 2015 compared to 2016. However, a full year's
worth of data for 2016 may result in different findings.
• Although LEKY's quarterly and annual average concentrations of carbon
tetrachloride are higher in magnitude, quarterly and annual average concentrations of
benzene exhibit more variability, based on both the range of averages and the
confidence intervals shown. Benzene concentrations measured at LEKY range from
0.182 |ig/m3 to 1.10 |ig/m3, FewNMP sites have a quarterly average concentration of
benzene less than LEKY's second quarter average for 2016 (0.35 ± 0.06 |ig/m3). A
review of the data shows that this calendar quarter is the only one without a benzene
concentration greater than 0.6 |ig/m3 (all others have at least two, with the number
ranging from two to six). In addition, the only two benzene concentrations less than
0.2 |ig/m3 were also measured during this calendar quarter.
• Concentrations of arsenic measured at LEKY range from 0.150 ng/m3 to 4.97 ng/m3,
which is the fifth highest arsenic concentration measured across the program. The
fourth quarter average concentration for 2015 is the only quarterly average greater
than 1 ng/m3 (1.06 ± 0.60 ng/m3) and has a relatively large confidence interval
associated with it. This is attributable to the maximum arsenic concentration, which
was measured at LEKY during the fourth quarter of 2015, and is more than three
times greater than the next highest concentration measured during this calendar
quarter. Among NMP sites sampling PMio metals, LEKY has the ninth highest annual
average concentration of arsenic (0.81 ± 0.13 ng/m3, 2015), as shown in Table 4-13.
An annual average for 2016 is not provided because the completeness criteria
specified in Section 2.4 was not met. A number of metals samples were invalid
throughout 2016, including a series of samples collected at LEKY in March and April
2016 with QA-related issues.
12.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where annual averages are available. Thus, box plots were created for each
of the pollutants listed in Table 12-4 for the Kentucky sites. Figures 12-14 through 12-28 overlay
the sites' minimum, annual average, and maximum concentrations for each year onto the
program-level minimum, first quartile, median, average, third quartile, and maximum
concentrations, as described in Section 3.4.2.1. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
12-38
-------
Figure 12-14. Program vs. Site-Specific Average Acetaldehyde Concentrations
0 2 4 6 8 10 12 14 16 18
Concentration (]ug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 12-14 presents the box plot for acetaldehyde for GLKY and shows the following:
• The range of acetaldehyde concentrations measured at GLKY is relatively small,
particularly for 2015, when all of the measurements are less than 2 |ig/m3 (and the
program-level third quartile).
• Both annual average concentrations for GLKY are less than the program-level first
quartile (0.96 |ig/m3).
12-39
-------
Figure 12-15. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
—o
j
¦—h
-a
E
1
\
1
r
!"t
l
2
3 4 5
Concentration (ng/m3)
6
7
Progra m:
1st Qua rtile
2ndQuartile 3rdQuartile
4th Qua rtile
Ave ra ge
¦
~
~
~
i
Site:
2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 12-15 presents the box plots for arsenic for the five Kentucky sites sampling PMio
metals and shows the following:
• Among the Kentucky sites, the range of arsenic concentrations measured is smallest
for BLKY and largest for ASKY-M. Several of the highest arsenic concentrations
measured across the program were measured at Kentucky sites, including the
program-level maximum arsenic concentration (7.36 ng/m3), which was measured at
ASKY-M in 2015.
• The annual average concentrations of arsenic for ASKY-M, BAKY, and LEKY
(2015) are greater than the program-level average concentration while the annual
average concentrations for BLKY and GLKY are less than the program-level average
12-40
-------
concentration. The annual average concentrations for ASYK-M and BAKY are also
greater than the program-level third quartile.
• Two non-detects of arsenic were measured, one each at BLKY and GLKY. While it
appears that a non-detect was also measured at BAKY, the minimum arsenic
concentration measured at BAKY is 0.008 ng/m3, the second lowest measured
detection of arsenic measured across the program. Two of the three arsenic
concentrations less than 0.01 ng/m3 measured across the program were measured at
BAKY.
Figure 12-16. Program vs. Site-Specific Average Benzene Concentrations
ASKY
m
ATKY
BLKY
i?
GLKY
Ht
TVKY
Concentration ([ig/m3]
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
12-41
-------
Figure 12-16 presents the box plots for benzene for the six Kentucky sites sampling
VOCs and shows the following:
• The box plots show that the range of benzene concentrations measured is smallest for
GLKY and LEKY and largest for TVKY and ATKY, particularly for 2015. The range
of benzene concentrations measured at TVKY and ATKY in 2016 is considerably
smaller than the range measured in 2015. This is also true for ASKY, and to a lesser
extent, LEKY (although 2016 does not include a full year's worth of sampling).
• The annual average concentrations of benzene across the Kentucky sites range from
0.39 ± 0.04 |ig/m3 (GLKY for both years) to 1.04 ± 0.30 |ig/m3 (TVKY, 2015). The
2015 annual averages for ASKY, ATKY, and TVKY are greater than the program-
level average concentration (0.72 |ig/m3), while their 2016 annual averages are less
than the program-level average. The annual average concentrations of benzene for
GLKY are less than the annual averages calculated for the other Kentucky sites and
less than the program-level first quartile (0.42 |ig/m3).
12-42
-------
Figure 12-17. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
ASKY 1 Program Max Concentration = 3.90 jxg/m3
| Program Max Concentration = 3.90 (.ig/m3
¦
I0j
r
| Program MaxConcentration = 3.90 (.ig/m3
1
GLKY
.
Program Max Concentration = 3.90 (.ig/m3
i j ¦+ 1
! Program Max Concentration = 3.90 Lig/m3 i
¦ 1 Q — " " "
I 1 , ^
! TVKY 2015 Max = 3.90 (ig/m3; 2016 Max = 3.57 (ig/m3 jl
j
i
Program Max Concentration = 3.90 (.ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 12-17 presents the box plots for 1,3-butadiene for the six Kentucky sites sampling
VOCs and shows the following:
• The program-level maximum concentration (3.90 |ig/m3) is not shown directly on the
box plots because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of the box plots has
been reduced to 1.75 |ig/m3. Also, since the maximum 1,3-butadiene concentration
12-43
-------
measured each year at TVKY is greater than the scale of the box plots, the site-
specific maximum concentrations are labeled for this site.
The maximum 1,3-butadiene concentration measured across the program was
measured at TVKY in 2015, although the maximum concentration measured at this
site in 2016 is similar in magnitude. 1,3-Butadiene concentrations greater than
0.75 |ig/m3 were not measured at the other Kentucky sites; BLKY is the only other
Kentucky site at which 1,3-butadiene concentrations greater than 0.3 |ig/m3 were
measured.
TVKY is the only Kentucky site for which an annual average concentration of
1,3-butadiene greater than the program-level average concentration (0.09 |ig/m3) was
calculated. Although TVKY's annual average for 2015 is twice the annual average
for 2016, both are still greater than the program-level average concentration.
The annual average concentrations for the remaining sites vary little, ranging from
0.03 ± 0.14 |ig/m3 for GLKY (both years) to 0.06 ± 0.03 |ig/m3 for BLKY (2016).
The program-level average concentration is similar to the third quartile, indicating
that the 1,3-butadiene concentrations on the upper end of the concentration range are
driving the program-level average upward.
Non-detects of 1,3-butadiene were measured at each of the Kentucky sites. Among
the five sites sampling VOC for two full years, the number of non-detects was highest
for GLKY (23) and lowest for ASKY (9).
12-44
-------
Figure 12-18. Program vs. Site-Specific Average Cadmium (PMio) Concentrations
;
Program Max Concentration =
5.99 ng/m3
1
1
1 1
1
¦ 1
1
|
¦ ASKY-M 2016 Max:
= 5.99 ng/m3
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 12-18 presents the box plot for cadmium for ASKY-M and shows the following:
• Similar to 1,3-butadiene, the program-level maximum cadmium concentration
(5.99 ng/m3) is not shown directly on the box plot as the scale of the box plot has
been reduced to 2 ng/m3 in order to allow for the observation of data points at the
lower end of the concentration range.
• ASKY-M is one of only two NMP sites sampling PMio metals for which cadmium is
a pollutant of interest.
• The maximum concentration measured at ASKY-M in 2016 is the maximum
concentration measured across the program. Although the maximum cadmium
concentration measured at this site in 2015 is one-third the magnitude, it is still one of
the higher measurements among NMP sites sampling metals (seventh highest).
• ASKY-M's annual average concentration of cadmium for 2016 is more than three
times greater than the program-level average concentration (0.12 ng/m3) while
ASKY-M's annual average concentration of cadmium for 2015 is slightly less.
• The program-level average cadmium concentration is similar to the third quartile,
indicating that cadmium concentrations on the upper end of the concentration range
are driving the program-level average upward.
12-45
-------
Figure 12-19. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
ATKY
BLKY
GLKY
TVKY
LEKY
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 12-19 presents the box plots for carbon tetrachloride for the six Kentucky sites
sampling VOCs and shows the following:
• Approximately 0.11 |ig/m3 separates the first and third quartiles, indicating that
roughly 50 percent of the concentrations of carbon tetrachloride measured in 2015
and 2016 fall into a fairly tight range of measurements. The program-level median
and average concentrations are similar to each other (both approximately 0.64 |ig/m3)
and plotted on top of each other.
12-46
-------
• The range of carbon tetrachloride measurements was largest for TVKY and smallest
for LEKY. The six highest carbon tetrachloride concentrations measured across the
program were measured at TVKY, although concentrations greater than 1 |ig/m3 were
also measured at BLKY and ATKY.
• The annual average concentrations of carbon tetrachloride for the three Calvert City
sites are greater than the program-level average (0.64 |ig/m3) and each of these sites
has at least one annual average greater than or similar to the program-level third
quartile (0.69 |ig/m3). For the remaining Kentucky sites sampling carbon
tetrachloride, the annual average concentrations vary by less than 0.025 |ig/m3.
• Across the program, most annual average concentrations of carbon tetrachloride do
not vary by more than 0.1 |ig/m3, generally falling between 0.6 |ig/m3 and 0.7 |ig/m3.
BLKY and TVKY are the only sites with annual averages greater than 0.7 |ig/m3 and
only TVKY has an annual average greater than 0.8 |ig/m3.
Figure 12-20. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
w—
1
1
Program Max Concentration
= 2.78 |ig/m3
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 12-20 presents the box plot for p-dichlorobenzene for LEKY and shows the
following:
• Similar to other pollutants, the program-level maximum p-dichlorobenzene
concentration (2.78 |ig/m3) is not shown directly on the box plot as the scale of the
box plot has been reduced to allow for the observation of data points at the lower end
of the concentration range. Note that the program-level first and second quartiles are
zero for this pollutant, indicating that at least 50 percent of the measurements across
the program are non-detects and thus, are not visible on the box plot.
• LEKY is the only site for which p-dichlorobenzene was identified as a pollutant of
interest.
• The range ofp-dichlorobenzene concentrations measured at LEKY is fairly small,
particularly for 2016, though sampling was discontinued at this site at the end of July.
All four /;-dichlorobenzene concentrations greater than 0.1 |ig/m3 were measured at
LEKY in 2015.
12-47
-------
• The annual average concentration of/?-dichlorobenzene for LEKY for 2015 is less
than the program-level average of 0.05 |ig/m3.
• Forty-seven non-detects of/?-dichlorobenzene were measured at LEKY, 27 in 2015
and 20 in 2016.
Figure 12-21. Program vs. Site-Specific Average 1,2-DichIoroethane Concentrations
Program Max Concentration = 45.8 ng/m3 \
ASKY
ATKY
r — — ~ 1
Program Max Concentration = 45.8 ng/m3
¦
w
r
ATKY 2015 Max= 4.99 |ig/m3; 2016 Max = 11.8 ng/m3 ;
Program Max Concentration =45.8 ug/m3
¦
n
¦
1 —
BLKY 2015 Max = 5.80 |ig/m3; 2016 Max = 45.8 ng/m3; 2016 Avg = 1.89 |ag/m3
- - - .....
P-
Pro
gram Max Concentration =45.8 ng/m3
GLKY
, i
i Program Max Concentration = 45.8 }ig/m3 [ !
¦ l
"TVKY 2015 Max = 24.1 jig/m3 & Avg = 3.75 jig/m3; 2016 Max = 22.8 [ig/rrr & Avg = 3.49 |ig/m3 J
w
Program Max Concentration = 45.8 ng/m3
m
LEKY
0.00
0.25
0.50
0.75
Concentration (fig/m3'
1.00
1.25
1.50
Program: IstQuartile 2ndQuartile 3rdQuartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Aveage Concentration Range, 2015&2016
• o
12-48
-------
Figure 12-21 presents the box plots for 1,2-dichloroethane for the six Kentucky sites
sampling VOCs and shows the following:
• The program-level maximum concentration (45.8 |ig/m3) is not shown directly on the
box plots for 1,2-dichloroethane as the scale of the box plots has been reduced to
1.5 |ig/m3 in order to allow for the observation of data points at the lower end of the
concentration range. Also, since the maximum 1,2-dichloroethane concentration
measured at several sites is greater than the scale of the box plots, the site-specific
maximum concentrations are labeled for these sites. Note that the program-level
average concentration is nearly three times greater than the third quartile, indicating
that the 1,2-dichloroethane concentrations on the upper end of the concentration range
are driving the program-level average upward.
• The maximum 1,2-dichloroethane concentrations measured at ASKY, GLKY, and
LEKY fall between 0.10 |ig/m3 and 0.15 |ig/m3, similar to many other NMP sites
sampling this pollutant. By comparison, the maximum concentration measured at
each of the Calvert City sites exceeds, and is often an order of magnitude higher than,
the scale of the box plots. The magnitude of some of these measurements is such that
even the annual average concentrations exceed the scale of the box plots. These
include both of TVKY's annual average concentrations and BLKY's 2016 annual
average.
• The annual average concentrations for the Calvert City sites are the highest annual
average concentrations of 1,2-dichloroethane among NMP sites sampling VOCs. No
other NMP site has an annual average concentration of this pollutant greater than
0.10 |ig/m3; ATKY has the lowest annual average concentration of 1,2-
dichloroethane among these three sites (0.41 ±0.19 |ig/m3), although it is still four
times greater than the annual average for the NMP monitoring site with the next
highest annual average (TMOK), as shown in Table 4-10.
Figure 12-22. Program vs. Site-Specific Average Formaldehyde Concentrations
em I |
0 3 6 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 12-22 presents the box plot for formaldehyde for GLKY and shows the following:
• The maximum formaldehyde concentration measured at GLKY is about one-fifth the
maximum formaldehyde concentration measured across the program.
12-49
-------
• GLKY's annual averages are both greater than the program-level first quartile
(1.54 |ig/m3) but less than the program-level median concentration (2.47 |ig/m3).
Eight NMP sites, including GLKY, have at least one annual average concentration of
formaldehyde less than 2 |ig/m3.
Figure 12-23. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
ASKY
r
i
i
1
Program Max Concentration = 1.02 (.ig/m3
1
ATKY
J
| Program Max Concentration = 1.02 (.ig/m3
1 i
1 ,
! -
Program MaxConcentration = 1.02 (.ig/m3
•
0.00 0.05 0.10 0.15 0.20 0.25
Concentration (jig/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 12-23 presents the box plots for hexachloro-1,3-butadiene for the three Kentucky
sites for which this pollutant was identified as a pollutant of interest and shows the following:
• The program-level maximum concentration of hexachloro-1,3-butadiene (1.02 |ig/m3)
is not shown directly on the box plots as the scale has been reduced to allow for the
observations data points at the lower end of the concentration range. The program-
level first, second, and third quartiles are all zero for this pollutant, indicating that at
least 75 percent of the measurements across the program are non-detects and thus, are
not visible on the box plots.
• Annual average concentrations for hexachl oro-1,3-butadiene were not calculated for
2015 due to the use of a contaminated internal standard at the laboratory for Method
TO-15, which resulted in the invalidation of the results from early March 2015
through mid-December 2015, as described above and in Section 2.4.
• Hexachl oro-1,3-butadiene concentrations greater than 0.15 |ig/m3 were not measured
at these sites (or at any other NMP site other than the site where the maximum
concentration was measured).
12-50
-------
• The number of measured detections of this pollutant for these sites ranged from 17
(ASKY and ATKY) to 23 (BLKY), and thus, hexachloro-1,3-butadiene was detected
in fewer than one-third of the valid samples collected at these sites.
• Less than 0.01 |ig/m3 separates the annual average concentrations of
hexachloro-1,3-butadiene for these Kentucky shown; however, the annual averages of
this pollutant for all NMP sites sampling VOCs vary by less than 0.05 |ig/m3,
including two sites at which hexachloro-1,3-butadiene was not detected).
Figure 12-24. Program vs. Site-Specific Average Lead (PMio) Concentrations
[ Program Max Concentration = 107 ng/m3
¦—
¦ 1 -
0 5 10 15 20 25 30 35 40 45 50
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
¦
~ ~ ~ 1
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 12-24 presents the box plot for lead for ASKY-M and shows the following:
• The program-level maximum lead concentration (107 ng/m3) is not shown directly on
the box plot as the scale of the box plot has been reduced to allow for the observation
of data points at the lower end of the concentration range.
• ASKY-M is the only NMP site for which lead is a pollutant of interest.
• The maximum lead concentration measured at ASKY-M in 2015 (41.4 ng/m3) is the
second highest concentration of lead measured across the program; the maximum
concentration measured at ASKY-M in 2016 (36.4 ng/m3) is the fourth highest lead
concentration measured across the program.
• Both of ASKY-M's annual average concentrations of lead are more than two times
greater than the program-level average concentration (3.07 ng/m3), with ASKY-M's
annual average concentration for 2016 nearly three times higher than the program-
level average.
12-51
-------
Figure 12-25. Program vs. Site-Specific Average Manganese (PMio) Concentrations
1 A
E 1
1 r
>
G
¦
, ,
i Program Max Concentration = 202 ng/m3
0 20 40 60 80 100 120
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 12-25 presents the box plot for manganese for ASKY-M and shows the following:
• The program-level maximum manganese concentration (202 ng/m3) is not shown
directly on the box plot as the scale of this box plot has also been reduced.
• ASKY-M is one of only two NMP sites for which manganese is a pollutant of
interest.
• The maximum manganese concentration measured at ASKY-M in 2015 (117 ng/m3)
is the second highest concentration of manganese measured across the program; two
additional manganese concentrations greater than 80 ng/m3 were measured at
ASKY-M, one in 2015 and one in 2016. Manganese concentrations measured at
ASKY-M account for five of the nine highest concentrations measured across the
program.
• Both of ASKY-M's annual average concentrations of manganese are more than two
times greater than the program-level average concentration (8.51 ng/m3), with
ASKY-M's annual average concentration for 2015 nearly three times higher than the
program-level average.
12-52
-------
Figure 12-26. Program vs. Site-Specific Average Nickel (PMio) Concentrations
1
¦ —
EI •
! Program Max Concentration =
69.5 ng/m3 1
E i ^
0 3 6 9 12 15 18 21 24
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 12-26 presents the box plot for nickel for ASKY-M and shows the following:
• The program-level maximum nickel concentration (69.5 ng/m3) is not shown directly
on the box plot as the scale of this box plot has also been reduced.
• Although the maximum nickel concentration measured each year at ASKY-M is
considerably less than the maximum nickel concentration measured across the
program, they are still among some of these highest nickel measurements.
• The annual average concentrations of nickel for ASKY-M are fairly similar to each
other, with both approximately two times greater than the program-level average
concentration (1.09 ng/m3). Only one other NMP site has an annual average
concentration of nickel greater than ASKY-M (among NMP sites sampling PMio
metals).
12-53
-------
Figure 12-27. Program vs. Site-Specific Average 1,1,2-Trichloroethane Concentrations
ATKY
Program Max Concentration =2.07 Lig/m3
ib i
T
i ^
, ,
i ATKY 2016 Max = 1.99 ng/m3 ;
1 (ft
; Program Max Concentration = 2.07 (xg/m3 !
j
1 ®
Q
1
r 1
I TVKY 2016 Max = 2.07 |ig/m3 i
0 0.15 0.3 0.45 0.6 0.75
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 12-27 presents the box plot for 1,1,2-trichloroethane for ATKY and TVKY and
shows the following:
• The program-level maximum concentration of 1,1,2-trichloroethane (2.07 |ig/m3)
is not shown directly on the box plots as the scale has been reduced to allow for the
observation of data points at the lower end of the concentration range. The program-
level first, second, and third quartiles are all zero for this pollutant, indicating that at
least 75 percent of the measurements across the program are non-detects and thus, are
not visible on the box plots.
• ATKY and TVKY are the only NMP sites with 1,1,2-trichloroethane as a pollutant of
interest.
• The two highest 1,1,2-trichloroethane concentrations measured across the program
were measured at TVKY and ATKY, both in 2016. While the maximum
concentrations of this pollutant measured at these sites in 2015 were considerably
less, they are still among some of the highest measured across the program. These
two sites account for 27 of the 32 1,1,2-trichloroethane concentrations greater than
0.1 |ig/m3 measured across the program (with BLKY accounting for the other five).
• For both sites, the annual average concentration for 2016 is greater than the annual
average for 2015; the highest 1,1,2-trichloroethane concentrations were measured in
2016. Two concentrations greater than 1 |ig/m3 were measured at each of these sites
in 2016; further, of the 27 1,1,2-trichloroethane concentrations greater than 0.1 |ig/m3
measured at these sites, 20 were measured in 2016 (nine at ATKY and 11 at TVKY).
12-54
-------
Figure 12-28. Program vs. Site-Specific Average Vinyl Chloride Concentrations
H•—
! '
t»
4 6 S
Concentration (jug/m3]
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 12-28 presents the box plots for vinyl chloride for the three Kentucky sites for
which this pollutant was identified as a pollutant of interest and shows the following:
• The program-level first, second, and third quartiles are all zero for this pollutant,
indicating that at least 75 percent of the measurements across the program are non-
detects and thus, are not visible on the box plots.
• The maximum vinyl chloride concentration measured at ATKY in 2016 (13.1 |ig/m3)
is the maximum concentration measured across the program; the eight highest vinyl
chloride concentrations measured across the program were measured at ATKY.
Although the concentrations measured at BLKY and TVKY are lower in magnitude,
its worth nothing that the 146 highest vinyl chloride concentrations measured across
the program (those greater than 0.15 |ig/m3) were measured at these three sites.
• The annual average vinyl chloride concentrations for these sites range from
0.10 ± 0.04 |ig/m3 for BLKY (2015) to 0.83 ± 0.54 |ig/m3 for ATKY (2016), all of
which are greater than the program-level average concentration of 0.05 |ig/m3. The
annual averages for ATKY are considerably higher than the annual averages for the
remaining two sites.
12-55
-------
12.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
GLKY is the longest running NMP site in Kentucky; VOCs and PAHs have been sampled under
the NMP since 2010, and sampling for carbonyl compounds and PMio metals began in 2011. The
remaining Kentucky sites began sampling under the NMP in 2012. Thus, Figures 12-29 through
12-71 present the 1-year statistical metrics for each of the pollutants of interest for the Kentucky
sites in the same order presented in the previous sections. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
Figure 12-29. Yearly Statistical Metrics for Benzene Concentrations Measured at ASKY
Maximum
Concentration for !
2013 is 43.5 (J.g/m3 !
I
—i r^~~l
•——ft—
2014
Year
O 5th Percentile — Minimum — Median
Maximum o 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
12-56
-------
Observations from Figure 12-29 for benzene concentrations measured at ASKY include
the following:
• Although sampling for VOCs at ASKY under the NMP began in March 2012, there
was a 3-month break in sampling, after which sampling resumed in mid-July.
Because a full year's worth of data is not available, a 1-year average concentration for
2012 is not presented, although the range of measurements is provided.
• The maximum benzene concentration measured at ASKY was measured on
November 6, 2013 (43.5 |ig/m3). The second highest benzene concentration, which
was measured in 2014, is nearly one-third the magnitude (12.4 |ig/m3). Only one
other benzene concentration greater than 3 |ig/m3, which was also measured in 2013,
has been measured at this site. The 1-year average concentration for 2013, though, is
being driven upward by the outlier concentration measured; if this concentration was
excluded from the calculation, the 1-year average concentration for 2013 would
decrease by almost half.
• With the exception of a slight increase for 2015, the median benzene concentration
has decreased each year through 2016, decreasing from 0.85 |ig/m3 for 2012 to
0.54 |ig/m3 from 2016. The 1-year average concentration has also decreased, from
1.52 |ig/m3 for 2013 to 0.65 |ig/m3 for 2016, representing a 57 percent decrease.
• With the exception of the 95th percentile, each of the statistical parameters is at a
minimum for 2016.
Figure 12-30. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
ASKY
2014
Year
O 5th Percentile — Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
12-57
-------
Observations from Figure 12-30 for 1,3-butadiene concentrations measured at ASKY
include the following:
• The only two 1,3-butadiene concentrations greater than 0.30 |ig/m3 were measured at
ASKY in 2012; 2012 has three times the number of 1,3-butadiene concentrations
greater than 0.15 |ig/m3 (nine), compared to other years of sampling (three or less
each year). 2012 is the only year in which the 5th percentile is greater than zero. Only
one non-detect of 1,3-butadiene was measured in 2012; between four (2015) and
fourteen (2013) were measured in the years that follow.
• All of the statistical parameters exhibit a decrease from 2012 to 2013 (except the
minimum concentration, which did not change). The median concentration decreased
by nearly 30 percent during this time; the number of 1,3-butadiene concentrations less
than 0.1 |ig/m3 tripled between 2012 and 2013, increasing from 18 to 53.
• Between 2013 and 2016, the median concentration varied little, falling between
0.5 |ig/m3 and 0.6 |ig/m3 each year.
• Although the 1-year average concentration of 1,3-butadiene appears to increase
between 2013 and 2016, the difference is not statistically significant; the change in
the 1-year average concentration over these four years is 0.008 |ig/m3.
Figure 12-31. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at ASKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
12-58
-------
Observations from Figure 12-31 for carbon tetrachloride concentrations measured at
ASKY include the following:
• Nearly 90 percent of the carbon tetrachloride concentrations measured at ASKY fall
between 0.50 |ig/m3 and 0.75 |ig/m3. The two carbon tetrachloride concentrations
greater than 0.90 |ig/m3 were measured at ASKY in 2013; the three carbon
tetrachloride concentrations less than 0.40 |ig/m3 were measured at ASKY in 2015.
• The median concentration decreased from 0.71 |ig/m3 for 2012 to 0.63 |ig/m3 for
2013. Although similar in magnitude, carbon tetrachloride concentrations greater than
0.7 |ig/m3 accounted for 55 percent of the measurements in 2012; by comparison,
these concentrations accounted for only 26 percent of the measurements in 2013. The
median concentration varied between 0.63 |ig/m3 and 0.64 |ig/m3 for each of the
years of sampling between 2013 and 2016.
• The 1-year average concentration varies by less than 0.04 |ig/m3 over the period
shown. If 2015 is excluded, less than 0.02 |ig/m3 separates these parameters. The four
lowest carbon tetrachloride concentrations were measured at ASKY in 2015,
including the only measurement less than 0.25 |ig/m3. In addition, 2015 is the only
year in which a measurement greater than 0.75 |ig/m3 was not measured.
Figure 12-32. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at ASKY
0.16
0.14
0.12
0.10
E
1
1 008
c
dj
c
O
u
0.06
0.04
0.02
0.00
20121 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
12-59
-------
Observations from Figure 12-32 for 1,2-dichloroethane concentrations measured at
ASKY include the following:
• The maximum 1,2-dichloroethane concentration measured each year has varied little,
ranging from 0.105 |ig/m3 (2014) to 0.138 |ig/m3 (2016).
• The minimum concentration measured each year is zero, indicating the substitution of
zero(s) for at least one non-detect; with the exception of 2015, the 5th percentile is
also zero. The number of non-detects measured has ranged from one (2015) to eight
(2013).
• The median 1,2-dichloroethane concentration has changed little over the years of
sampling, falling between 0.07 |ig/m3 and 0.08 |ig/m3. This is also true for the 1-year
average concentration.
Figure 12-33. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at ASKY
0.12
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented due to a laboratory contamination issue affecting numerous samples.
Observations from Figure 12-33 for hexachloro-l,3-butadiene concentrations measured at
ASKY include the following:
• The use of a contaminated internal standard at the laboratory for Method TO-15
resulted in the invalidation of the hexachloro-l,3-butadiene measurements from early
March 2015 through mid-December 2015, as described in Section 2.4.
12-60
-------
• The minimum, 5th percentile, and median concentration for each year is zero, where
available, indicating that at least 50 percent of the measurements were non-detects.
The percentage of non-detects measured has ranged from 75 percent (2016) to
92 percent (2014).
• Three of the four highest hexachloro-1,3 -butadiene concentrations were measured at
ASKY in 2016. The maximum concentration measured is 0.118 |ig/m3; the next
highest concentration (0.107 |ig/m3) has been measured three times, once in 2013 and
twice in 2016. No other concentrations greater than 0.1 |ig/m3 have been measured at
this site. The majority of measured detections fall between 0.05 |ig/m3 and 0.1 |ig/m3
(though accounting for no more than 25 percent of measurements for any given year).
Figure 12-34. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
ASKY-M
O 5th Percentile
— Minimum
O 95th Percentile
Observations from Figure 12-34 for arsenic (PMio) concentrations measured at ASKY-M
include the following:
• The maximum arsenic concentration (10.1 ng/m3) was measured at ASKY-M in
2014. Arsenic concentrations greater that 4 ng/m3 were measured during each year of
sampling except 2016.
• Both the 1-year average and median concentrations are at a maximum for 2012, after
which each exhibits a decrease through 2014. The number of arsenic concentrations
greater than 2 ng/m3 is at its highest for 2012 (26), after which the number decreased
by half for 2013 (13), then in half again for 2014 (7).
12-61
-------
• Most of the statistical parameters exhibit a slight increase for 2015, which is followed
by a slight decrease for 2016, when the smallest range of arsenic concentrations was
measured.
Figure 12-35. Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
ASKY-M
T
o
>
2012 20
13 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
Observations from Figure 12-35 for cadmium (PMio) concentrations measured at
ASKY-M include the following:
• A significant decreasing trend in cadmium concentrations measured at ASKY-M is
shown between 2012 and 2014. The number of cadmium concentrations greater than
1 ng/m3 measured at ASKY-M decreased from 12 in 2012 to three in 2013 to none in
2014. Both the 1-year average and median concentrations decreased by more than
half during this time and most of the statistical parameters are at a minimum for 2014.
• Cadmium concentrations of increasing magnitude were measured in 2015 and 2016,
when the maximum cadmium concentration was measured (5.99 ng/m3). The 1-year
average concentration increases for 2015 and again for 2016, although the increase
for 2016 is mostly due to the magnitude of the maximum concentration measured (the
next highest concentration is one-fifth as high).
12-62
-------
Figure 12-36. Yearly Statistical Metrics for Lead (PMio) Concentrations Measured at
ASKY-M
O 5th Percentile
O 95th Percentile
Observations from Figure 12-36 for lead (PMio) concentrations measured at ASKY-M
include the following:
• The five highest lead concentrations were all measured at ASKY-M in 2012,
including the maximum concentration (100.1 ng/m3). The second, third, fourth, and
fifth highest lead concentrations were half as high, falling between 40 ng/m3 and
50 ng/m3. Concentrations of lead greater than 40 ng/m3 were also measured in 2013
and 2015.
• Similar to cadmium, concentrations of lead decreased significantly between 2012 and
2014, with the 1-year average concentration decreasing from 14.76 ng/m3 to
6.44 ng/m3 during this time. The median concentration exhibits a similar pattern.
• Most of the statistical parameters exhibit increases from 2014 to 2015, which is
followed by slight decreases for 2016.
12-63
-------
Figure 12-37. Yearly Statistical Metrics for Manganese (PMio) Concentrations Measured at
ASKY-M
O 5th Percentile
O 95th Percentile
Observations from Figure 12-37 for manganese (PMio) concentrations measured at
ASKY-M include the following:
• The trends graph for manganese resembles the trends graph for several of the other
metals pollutants of interest for ASKY-M.
• The maximum manganese concentration (236 ng/m3) was measured at ASKY-M in
2012. Two additional manganese concentrations greater than 100 ng/m3 have been
also measured at ASKY-M, one in 2015 (117 ng/m3) and the other in 2012
(115 ng/m3). Manganese concentrations greater than 50 ng/m3 have been measured at
ASKY-M during each year of sampling (a total of 29), with the most measured in
2012 (12).
• Similar to lead, concentrations of manganese decreased significantly between 2012
and 2014, with the 1-year average concentration decreasing from 33.95 ng/m3 to
18.21 ng/m3 during this time. The median concentration exhibits a similar pattern.
• Most of the statistical parameters exhibit increases from 2014 to 2015, which is
followed by slight decreases for 2016.
12-64
-------
Figure 12-38. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at
ASKY-M
2014
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 12-38 for nickel (PMio) concentrations measured at ASKY-M
include the following:
• The maximum nickel concentration was measured at ASKY-M in 2013 (21.2 ng/m3).
Nickel concentrations greater than 10 ng/m3 have been measured at ASKY-M during
each year of sampling except 2014.
• Although the 1-year average concentration of nickel has decreased by nearly 1 ng/m3
over the course of sampling, most of this decrease occurred between 2012 and 2013.
The median has a similar pattern until 2015, when a slight increase is shown; this is
followed by additional decreases for 2016. Both central tendency parameters are at a
minimum for 2016. However, confidence intervals calculated on the data indicate that
the change is not statistically significant.
12-65
-------
Figure 12-39. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
GLKY
3.0
E
3
c
¦° 15
C
m
c
O
u
1.0
o
1
t
20
12
2013
20
Y
14
ear
20
15
20
16
O
5th Pe
rcenti
e — M
nimum
— Median
-
Maximum
O
95th
Percentile
Aver a
e
Observations from Figure 12-39 for acetaldehyde concentrations measured at GLKY
include the following:
• GLKY began sampling carbonyl compounds under the NMP in August 2011.
However, data from 2011 is excluded because this represents less than 6 months of
sampling, as described in Section 3.4.2.2, and thus, Figure 12-39 begins with 2012.
• The range of acetaldehyde concentrations measured at GLKY compressed from 2012
to 2013, although the majority of concentrations fell within a relatively similar range
both years, as indicated by the 5th and 95th percentiles. The decrease shown in the
1-year average and median concentrations for 2013 results not just from the lower
maximum concentration; the number of acetaldehyde concentrations greater than
1 |ig/m3 fell to eight in 2013, from 13 in 2012, which included four measurements
greater than the maximum concentration measured in 2013.
• All of the statistical parameters exhibited increases for 2014, when the number of
acetaldehyde concentrations greater than 1 |ig/m3 doubled (16) and concentrations
greater than 0.75 |ig/m3 account for more than half of measurements. The fifth
percentile is also greater than 0.5 |ig/m3, indicating that fewer "low" concentrations
were measured.
• The 1-year average concentration holds steady between 2014 and 2016, despite
increasingly higher acetaldehyde concentrations being measured. The median
concentration is at a maximum for 2015, then decreases somewhat for 2016.
12-66
-------
Figure 12-40. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
GLKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2011.
Observations from Figure 12-40 for arsenic (PMio) concentrations measured at GLKY
include the following:
• Sampling for PMio metals at GLKY under the NMP began in May 2011; because a
full year's worth of data is not available, a 1-year average concentration for 2011 is
not presented, although the range of measurements is provided.
• The maximum arsenic concentration measured at GLKY was measured in 2011
(3.56 ng/m3); no other arsenic concentrations greater than 3 ng/m3 have been
measured at GLKY. Three arsenic concentrations greater than 2 ng/m3 have been
measured, one in 2013 and two in 2016.
• The median arsenic concentration exhibits little change from 2011 to 2012, despite
the changes shown in the concentration profiles for these years (e.g., the decrease in
the maximum concentration measured, the increase in the 5th and 95th percentiles).
• All of the statistical parameters exhibit a decreasing trend between 2012 and 2014,
when all of the parameters are at a minimum, and when the first non-detects (four) of
arsenic were measured.
• With the exception of the minimum concentration, each of the statistical parameters
increases for 2015, each returning to near 2013 levels. Most of the statistical
parameters exhibit further increases for 2016, although the 1-year average changes
12-67
-------
little and the median exhibits a slight decrease. Non-detects were not measured at
GLKY in 2016.
Figure 12-41. Yearly Statistical Metrics for Benzene Concentrations Measured at GLKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-41 for benzene concentrations measured at GLKY include
the following:
• GLKY began sampling VOCs under the NMP in June 2010. Because a full year's
worth of data is not available, a 1-year average concentration for 2010 is not
presented, although the range of measurements is provided.
• The maximum benzene concentration was measured at GLKY in August 2013
(2.75 |ig/m3); this measurement is one of only three benzene concentrations greater
than 2 |ig/m3 measured at this site.
• The median benzene concentration exhibits a decreasing trend through 2015,
decreasing from 0.59 |ig/m3 for 2010 to 0.34 |ig/m3 for 2015 (with little change for
2016). The 1-year average concentration exhibits a similar pattern, decreasing from
0.58 |ig/m3 in 2011 to 0.39 |ig/m3 in 2015 and 2016, although there is a slight
increase shown from 2012 to 2013. If the maximum concentration measured in 2013
was excluded from the dataset, the 1-year average concentration would exhibit a
pattern similar to the 1-year median concentration.
12-68
-------
Figure 12-42. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
GLKY
.••o
2013
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-42 for 1,3-butadiene concentrations measured at GLKY
include the following:
• The eight highest 1,3-butadiene concentrations were measured at GLKY in 2012 and
account for all of the measurements greater than 0.15 |ig/m3.
• In addition to having the concentrations of highest magnitude, 2012 is also the year
with the fewest non-detects; two non-detects were measured in 2012, compared to
between nine (2014, 2015) and 18 (2011) for the remaining years.
• The 1-year average concentration nearly doubled from 2011 to 2012, as a result of the
higher concentrations and reduced number of non-detects.
• The 1-year average concentration decreased significantly from 2012 to 2013, with
additional decreases shown each year through 2016, when the 1-year average is at a
minimum (0.031 |ig/m3). During this time, the majority of concentrations, as
indicated by the difference in the 5th and 95th percentiles, fell within a similar range.
• The median concentration does not exhibit quite the same pattern as the 1-year
average. The median concentration increases each year between 2010 and 2013,
reaching a maximum one year later than the 1-year average, although the difference
between the median concentrations for these two years is rather small. The median
concentration decreases after 2013, falling to less than 0.03 |ig/m3 for 2015 and 2016.
12-69
-------
Figure 12-43. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at GLKY
O 5th Percentile
Maximum O 95th Percentile Averc
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-43 for carbon tetrachloride concentrations measured at
GLKY include the following:
• Only one carbon tetrachloride concentration greater than 1 |ig/m3 has been measured
at GLKY (March 2012).
• All of the statistical parameters exhibit an increase from 2011 to 2012, although the
majority of concentrations, as indicated by the 5th and 95th percentiles, fell into a
similar range. The number of carbon tetrachloride measurements between 0.7 |ig/m3
and 0.8 |ig/m3 more than doubled, from 11 measured in 2011 to 25 in 2012.
• Decreases in the 1-year average concentration are shown from 2012 to 2013 as the
number of carbon tetrachloride measurements between 0.7 |ig/m3 and 0.8 |ig/m3
accounts for fewer measurements (falling to 14 from 25). Most of the statistical
parameters exhibit decreases from 2013 to 2014, with the exception of the minimum
concentration, which did not change. Several parameters are at a minimum for 2014,
which is the first year without a measurement greater than 0.8 |ig/m3.
• Slight increases in the 1-year average concentration are shown for 2015 and 2016,
though the changes are not statistically significant. In fact, the 1-year average
concentrations vary by less than 0.1 |ig/m3 over the period of sampling. This is also
true for the median concentration.
12-70
-------
Figure 12-44. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at GLKY
«
2013
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2010.
Observations from Figure 12-44 for 1,2-dichloroethane concentrations measured at
GLKY include the following:
• There was one measured detection of 1,2-dichloroethane in 2010. The number of
measured detections increased to 19 for 2011, and 54 in 2012, accounting for
90 percent of the measurements. Between 2012 and 2016, the number of measured
detections has ranged from 48 (2014) to 58 (2015).
• As the number of non-detects decreased and measured detections increased, the
1-year average and median concentrations increased correspondingly, each reaching a
maximum for 2013. A significant decrease in the 1-year average concentration is
shown after 2013, and by 2016 1-year average is less than 0.06 |ig/m3 for the first
time since 2011.
• The median concentration is greater than the 1-year average concentration for each
year between 2012 and 2016. This is because there were still several non-detects (or
zeros) factoring into the 1-year average concentration for each year: 2012 (6), 2013
(5), 2014 (8), 2015 (2), and 2016 (10), which drive the 1-year average concentrations
down in the same manner that a maximum or outlier concentration can drive the
average up. The increase in non-detects is the primary reason for the decreases in
several of the parameters shown for 2016.
12-71
-------
Figure 12-45. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
GLKY
O 5th Percentile
O 95th Percentile
Observations from Figure 12-45 for formaldehyde concentrations measured at GLKY
include the following:
• The maximum formaldehyde concentration (5.96 |ig/m3) was measured at GLKY in
2012. Three additional formaldehyde concentrations greater than 5 |ig/m3 have been
measured, another in 2012, plus one each in 2014 and 2016.
• Each of the statistical parameters exhibits a decrease from 2012 to 2013, particularly
those representing the upper end of the concentration range.
• A steady increasing trend in formaldehyde concentrations measured at GLKY is
shown each year after 2013, with both the 1-year average and median concentration at
a maximum for 2016.
12-72
-------
Figure 12-46. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
BAKY
2014
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 12-46 for arsenic (PMio) concentrations measured at BAKY
include the following:
• BAKY began sampling PMio metals in March 2012.
• The maximum arsenic concentration measured at BAKY was measured in 2013
(6.37 ng/m3). Additional arsenic concentrations greater than 4 ng/m3 have not been
measured at BAKY.
• Although concentrations at the upper end of the concentration range have varied
across the years, the majority of arsenic concentrations measured have changed little
over the years of sampling. The central tendency parameters have varied by less than
0.2 ng/m3 over the years: the 1-year average concentration has ranged from
0.82 ng/m3 (2013) to 0.96 ng/m3 (2015). The median concentration has ranged from
0.66 ng/m3 (2013) to 0.83 ng/m3 (2015).
12-73
-------
Figure 12-47. Yearly Statistical Metrics for Benzene Concentrations Measured at ATKY
2014
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-47 for benzene concentrations measured at ATKY include
the following:
• Although sampling for VOCs at ATKY under the NMP began in March 2012, there
was a 3-month break in sampling, then sampling resumed in mid-July. Because a full
year's worth of data is not available, a 1-year average concentration for 2012 is not
presented, although the range of measurements is provided.
• The three highest benzene concentrations were measured in 2015, ranging from
5.47 |ig/m3 to 6.85 |ig/m3. Benzene concentrations greater than 1.25 |ig/m3 were not
measured prior to 2014, with 11 of these 16 concentrations measured in 2015 (with
four measured in 2014 and one measured in 2016).
• The entire range of benzene concentrations measured at ATKY in 2012 spans just
over 1 |ig/m3. The range is even smaller for 2013. The range of measurements nearly
doubles from 2013 to 2014, then increases four-fold for 2015. The 1-year average
concentration nearly doubles from 2013 to 2015; only two concentrations measured
in 2013 and seven measured in 2014 are greater than the 1-year average concentration
for 2015.
• With the exception of the maximum concentration (3.33 |ig/m3), the concentration
profile for 2016 more closely resembles the concentration profile for 2013. If the
maximum concentration was excluded, this would be even more true.
12-74
-------
Figure 12-48. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
ATKY
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-48 for 1,3-butadiene concentrations measured at ATKY
include the following:
• Three 1,3-butadiene concentrations greater than 0.6 |ig/m3 have been measured at
ATKY, one measured each year between 2012 and 2014. The maximum
1,3-butadiene concentration decreased by more than half for 2015 and is at a
minimum for 2016, when 1,3-butadiene concentrations greater than 0.2 |ig/m3 were
not measured.
• The range of 1,3-butadiene concentrations measured during the first three years of
sampling did not vary much. The median concentration did not change from 2012 to
2013, despite twice the number of samples collected and a four-fold increase in the
number of non-detects measured (from four to 18).
• The central tendency parameters both exhibit increases for 2014. This is due, at least
in part, to fewer non-detects, which decreased by half from 2013 (18) to 2014 (9).
• With the exception of the minimum concentration (which did not change) and the 5th
percentile (which increased slightly), the statistical metrics exhibit decreases for
2015: additional decreases are shown for 2016.
12-75
-------
Figure 12-49. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at ATKY
5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-49 for carbon tetrachloride concentrations measured at
ATKY include the following:
• Concentrations of carbon tetrachloride measured at ATKY span less than 1 |ig/m3
across the years of sampling, ranging from 0.43 |ig/m3 (2013) to 1.37 |ig/m3 (2015).
The majority of concentrations measured fall into an even tighter range, as indicated
by the 5th and 95th percentiles.
• Less than 0.015 |ig/m3 separates the available 1-year average concentrations. The
median concentrations vary a little more. The median concentration for 2012 is
0.73 |ig/m3, then decreases to 0.67 |ig/m3 for 2013, which represents the largest year-
to-year change in this parameter. The median concentration varies between
0.66 |ig/m3 and 0.69 |ig/m3 for the years between 2013 and 2016. The percentage of
carbon tetrachloride concentrations greater than 0.8 |ig/m3 is at a maximum for 2012
(27 percent), accounting for less than 10 percent of measurements in the years that
follow.
12-76
-------
Figure 12-50. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at ATKY
' o }
O 5th Percentile
Maximum o 95th Percentile Averc
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-50 for 1,2-dichloroethane concentrations measured at
ATKY include the following:
• Concentrations of 1,2-dichloroethane measured at ATKY are highly variable. With
the exception of 2015, at least one non-detect has been measured each year. Measured
detections have ranged from 0.049 |ig/m3 to 11.8 |ig/m3. Concentrations greater than
1 |ig/m3 have been measured each year, varying in number from as low as three
(2013) to as many as 14 (2014). Yet, the median concentration, or the mid-point of
the dataset, for each year is 0.15 |ig/m3 or less.
• The 1-year average concentrations have an undulating pattern, with a year with a
"higher" average concentration following a year with a "lower" average. The 1-year
average concentrations have ranged from 0.30 |ig/m3 to 0.90 |ig/m3, reaching a
maximum for 2016. The four highest 1,2-dichloroethane concentrations measured at
ATKY were all measured in 2016 and range from 5.96 |ig/m3 to 11.8 |ig/m3.
12-77
-------
Figure 12-51. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at ATKY
0.14
0.12
0.10
E
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile •••~•••Average
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented due to a laboratory contamination issue affecting numerous samples.
Observations from Figure 12-51 for hexachloro-l,3-butadiene concentrations measured at
ATKY include the following:
• The use of a contaminated internal standard at the laboratory for Method TO-15
resulted in the invalidation of the hexachloro-l,3-butadiene measurements from early
March 2015 through mid-December 2015, as described in Section 2.4.
• The minimum, 5th percentile, and median concentrations of hexachloro-l,3-butadiene
for each year are zero, indicating that at least half of the measurements were non-
detects each year. The percentage of non-detects has ranged from 77 percent (2016)
to 85 percent (2013), excluding 2015, which was affected by a contamination issue.
• Hexachloro-1,3-butadiene concentrations greater than 0.1 |ig/m3 have been measured
each year of sampling at ATKY (except 2015, for which there were only 10 valid
samples), with the number increasing over time, from one each year in 2012 and
2013, to four in 2014, to six in 2016.
12-78
-------
Figure 12-52. Yearly Statistical Metrics for 1,1,2-Trichloroethane Concentrations
Measured at ATKY
i Maximum Concentration |
j for 2016 is 1.99 ^g/m3
T I
_ J- 1 c
o
O 5th Percentile
Maximum o 95th Percentile Averc
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-52 for 1,1,2-trichloroethane concentrations measured at
ATKY include the following:
• Similar to hexachloro-1,3-butadiene, the minimum, 5th percentile, and median
concentrations of 1,1,2-trichloroethane for ATKY are all zero, indicating that at least
half of the measurements were non-detects each year. The percentage of non-detects
has ranged from 85 percent (2016) to 97 percent (2013 and 2014), explaining why
even the 95th percentile is zero for these two years.
• Not only is the number of non-detects at a minimum for 2016, but the concentrations
measured were also higher than in previous years. The five highest
1,1,2-trichloroethane concentrations measured at ATKY were measured in 2016 and
range from 0.317 |ig/m3 to 1.99 |ig/m3; this is the only year in which a
1,1,2-trichloroethane concentration greater than 0.3 |ig/m3 has been measured.
12-79
-------
Figure 12-53. Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at
ATKY
14.0
12.0
10.0
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-53 for vinyl chloride concentrations measured at ATKY
include the following:
• The minimum concentration and 5th percentile for each year is zero, indicating the
presence of non-detects. The percentage of non-detects has ranged from 18 percent
(2016) to 42 percent (2012).
• Measured detections have ranged from 0.00256 |ig/m3 to 13.1 |ig/m3, both of which
were measured in 2016. Concentrations greater than 1 |ig/m3 account for about
20 percent of the measurements each year, yet, the median concentration, or the mid-
point of the dataset, is less than 0.1 |ig/m3 for all years except 2015 (for which the
median is 0.15 |ig/m3).
• The 1-year average concentration has varied by less than 0.25 |ig/m3 across the years,
ranging from 0.60 |ig/m3 (2013) to 0.83 |ig/m3 (2016).
12-80
-------
Figure 12-54. Yearly Statistical Metrics for Benzene Concentrations Measured at BLKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-54 for benzene concentrations measured at BLKY include
the following:
• Although sampling for VOCs at BLKY under the NMP began in March 2012, there
was a 3-month break in sampling, then sampling resumed in mid-July. Because a full
year's worth of data is not available, a 1-year average concentration for 2012 is not
presented, although the range of measurements is provided.
• All six benzene concentrations greater than 2 |ig/m3 were measured at BLKY in
either 2013 or 2014. Fifteen of the 24 benzene concentrations greater than 1 |ig/m3
were measured during these two years.
• The 1-year average concentration has decreased each year since 2013, decreasing
from 0.67 |ig/m3 to 0.49 |ig/m3. However, confidence intervals calculated for these
averages indicate that the changes are not statistically significant, mostly due to the
relatively large amount of variability in the measurements from 2013 and 2014.
12-81
-------
Figure 12-55. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
BLKY
12.0
10.0
2.0
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-55 for 1,3-butadiene concentrations measured at BLKY
include the following:
• The minimum concentration and 5th percentile for each year is zero, indicating the
presence of non-detects. The percentage of non-detects has ranged from 14 percent
(2015) to 36 percent (2013).
• All 11 1,3-butadiene concentrations greater than 1 |ig/m3 were measured at BLKY
prior to 2015 (with two measured in 2012, eight in 2013, and one in 2014).
• The 1-year average concentration decreased by an order of magnitude between 2013
and 2016, decreasing from 0.63 |ig/m3 for 2013 to 0.06 |ig/m3 for 2015 and 2016.
• The median concentration, however, has changed little, varying between 0.04 |ig/m3
and 0.05 |ig/m3 across the years of sampling.
12-82
-------
Figure 12-56. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at BLKY
Maximum Concentration
for 2013 is 23.7 |J.g/m3
2014
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-56 for carbon tetrachloride concentrations measured at
BLKY include the following:
• The maximum carbon tetrachloride concentration was measured at BLKY in 2013
(23.7 |ig/m3) and is an order of magnitude greater than the next highest concentration
(2.30 |ig/m3), which was measured in 2014.
• The magnitude of the outlier measured in 2013 is such that the 1-year average
concentration calculated for 2013 (1.11 |ig/m3) is greater than the 95th percentile for
that year. The median concentration, which is less affected by outliers, for 2013 is
0.70 |ig/m3.
• The median concentration decreased slightly each year through 2015. This is also true
for the 1-year average concentration, although the largest change is exhibited between
2013 and 2014. If the maximum concentration was excluded from the dataset, the
1-year average would exhibit virtually no change between 2013 and 2014.
• Both central tendency parameters exhibit increases for 2016. Although the range of
concentrations measured exhibits little change from 2015 to 2016, the range within
which the majority of concentrations fall, as indicated by the 5th and 95th percentiles,
doubles from 2015 to 2016.
12-83
-------
Figure 12-57. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at BLKY
Maximum Concentrati on
for 2016 is 45.8 |J.g/m3
2014
Year
O 5th Percentile
Maximum O 95th Percentile Avera
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-57 for 1,2-dichloroethane concentrations measured at
BLKY include the following:
• Concentrations of 1,2-dichloroethane measured at BLKY are highly variable. Two to
three non-detects have been measured each year, with the exception of 2015.
Measured detections span three orders of magnitude, ranging from 0.041 |ig/m3 to
45.8 |ig/m3. Concentrations greater than 5 |ig/m3 have been measured each year,
varying in number from as few as one (2012 and 2015) to as many as six (2013). At
the other end of the concentration range, multiple concentrations less than 0.05 |ig/m3
have also been measured each year, varying in number from two (2015) to five
(2012).
• Although difficult to discern in Figure 12-57, the median concentration has varied by
less than 0.25 |ig/m3 over the years of sampling, ranging from 0.13 |ig/m3 (2012) to
0.33 |ig/m3 (2016).
• The 1-year average concentration decreases somewhat between 2013 and 2015,
before increasing considerably for 2016. Even if the maximum concentration
measured in 2016 was excluded from the dataset, the 1-year average concentration for
2016 would still exhibit an increase.
12-84
-------
Figure 12-58. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at BLKY
o-
-o
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented due to a laboratory contamination issue affecting numerous samples.
Observations from Figure 12-58 for hexachloro-l,3-butadiene concentrations measured at
BLKY include the following:
• The use of a contaminated internal standard at the laboratory for Method TO-15
resulted in the invalidation of the hexachloro-l,3-butadiene measurements from early
March 2015 through mid-December 2015, as described in Section 2.4.
• The minimum, 5th percentile, and median concentrations of hexachloro-l,3-butadiene
for BLKY are zero for all years of sampling except 2015, indicating that at least half
of the measurements were non-detects each year. Excluding 2015, the percentage of
non-detects has ranged from 72 percent (2016) to 93 percent (2012).
• The maximum hexachloro-1,3-butadiene concentration (0.29 |ig/m3) was measured in
2013, along with the two other concentrations greater than 0.15 |ig/m3.
Concentrations greater than 0.1 |ig/m3 have been measured each year of sampling at
BLKY except 2015, with the number varying from one (2012) to seven (2016).
12-85
-------
Figure 12-59. Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at
BLKY
2.0
1.8
1.6
1.4
u 0.8
0.6
0.4
0.2
0.0
2012 1 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-59 for vinyl chloride concentrations measured at BLKY
include the following:
• The minimum concentration and 5th percentile for each year is zero, indicating the
presence of non-detects. The percentage of non-detects was greater than 40 percent
for each of the first three years, decreased to 31 percent for 2015 and to 22 percent for
2016.
• Measured detections have ranged from 0.00256 |ig/m3 to 4.56 |ig/m3; the maximum
vinyl chloride concentration measured at BLKY is three times greater than the next
highest concentration measured at this site (1.68 |ig/m3). In total, six concentrations
greater than 1 |ig/m3 have been measured at BLKY since 2012.
• The median concentration is approximately 0.04 |ig/m3 for all years except 2014, for
which the median is 0.06 |ig/m3.
• The 1-year average concentration decreases slightly between 2013 and 2015, before
returning to near-2013 levels for 2016. The 1-year average concentration has varied
by less than 0.25 |ig/m3 across the years, ranging from 0.60 |ig/m3 to 0.83 |ig/m3,
reaching a maximum for 2016.
Maximum Concentrati on
for 2013 is 4.56 ng/m3
12-86
-------
Figure 12-60. Yearly Statistical Metrics for Benzene Concentrations Measured at TVKY
5th Percentile
95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-60 for benzene concentrations measured at TVKY include
the following:
• Although sampling for VOCs at TVKY under the NMP began in March 2012, there
was a 3-month break in sampling, then sampling resumed in mid-July. Because a full
year's worth of data is not available, a 1-year average concentration for 2012 is not
presented, although the range of measurements is provided.
• The maximum concentration of benzene was measured at TVKY in 2014
(9.92 |ig/m3). In total, six benzene concentrations greater than 5 |ig/m3 have been
measured at TVKY.
• Despite fluctuations in the concentrations at the upper end of the concentration range,
the central tendency parameters exhibit little change across the first four years of
sampling. During this time, the 1-year average concentration varied between 1 |ig/m3
and 1.05 |ig/m3 and the median concentration varied between 0.50 |ig/m3 and
0.60 |ig/m3.
• A significant decrease in the magnitude of benzene concentrations is shown for 2016.
The maximum concentration measured in 2016 (2.03 |ig/m3) is considerably less than
the maximum concentration measured for each of the other years, and is also less than
the 95th percentile for each of the previous years. The 1-year average concentration
decreased by half from 2015 to 2016.
12-87
-------
Figure 12-61. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
TVKY
i Maximum Concentration |
| for 2013 is 21.5 (J.g/m3
^
J
2014
Year
O 5th Percentile
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-61 for 1,3-butadiene concentrations measured at TVKY
include the following:
• The maximum concentration of 1,3-butadiene was measured at TVKY in October
2013 (21.5 |ig/m3), although a concentration of similar magnitude was also measured
a few months earlier (20.8 |ig/m3). These are the only two 1,3-butadiene
concentrations greater than 6 |ig/m3 measured at this site.
• The median concentration is less than 0.1 |ig/m3 for all years of sampling, indicating
that at least half of the measurements collected each year are less than this (including
non-detects).
• The 1-year average concentration has decreased each year since 2013, decreasing
from a maximum of 1.03 |ig/m3 for 2013 to a minimum of 0.16 |ig/m3 for 2016.
However, the confidence intervals associated with these averages are considerably
large, indicating such a large amount of variability in the measurements for all years
of sampling.
12-88
-------
Figure 12-62. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at TVKY
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-62 for carbon tetrachloride concentrations measured at
TVKY include the following:
• The two highest carbon tetrachloride concentrations measured at TVKY were both
measured in 2012 and are the only measurements greater than 4 |ig/m3. Between two
and four carbon tetrachloride concentrations greater than 2 |ig/m3 have been
measured at this site each year of sampling, for a total of 14.
• Despite fluctuations in the concentrations at the upper end of the concentration range,
the central tendency parameters exhibit little change across the years of sampling
(less than 0.1 |ig/m3 separates them). During this time, the 1-year average
concentration has varied between 0.79 |ig/m3 (2013) and 0.87 |ig/m3 (2014) and the
median concentration has varied between 0.69 |ig/m3 (2015) and 0.77 |ig/m3 (2012).
12-89
-------
Figure 12-63. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at TVKY
Maxi mum Concentration
for 2013 is lll^g/m3
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-63 for 1,2-dichloroethane concentrations measured at
TVKY include the following:
• Concentrations of 1,2-dichloroethane are the most variable among TVKY's pollutants
of interest. Measured detections of 1,2-dichloroethane span four orders of magnitude,
ranging from 0.0385 |ig/m3 to 111 |ig/m3.
• The maximum 1,2-dichloroethane concentration measured at TVKY in 2013 is four
times higher than the next highest 1,2-dichloroethane concentration measured at this
site (27.4 |ig/m3). While the maximum concentration measured is an obvious outlier,
higher 1,2-dichloroethane concentrations are not an anomaly at this site;
1,2-dichloroethane concentrations greater than 10 |ig/m3 have been measured during
each year of sampling, from as few as two (2012) to as many as 10 (2015).
• Despite the considerable fluctuations in the concentrations at the upper end of the
concentration range, the central tendency parameters exhibit relatively little change
across the years of sampling. During this time, the 1-year average concentration has
varied by just over 0.25 |ig/m3, ranging between 3.49 |ig/m3 (2016) and 3.75 |ig/m3
(2015).
• The median concentration calculated for each year is an order of magnitude less than
the 1-year average concentration for each year, varying between 0.22 |ig/m3 (2015)
and 0.58 |ig/m3 (2012).
12-90
-------
Figure 12-64. Yearly Statistical Metrics for 1,1,2-Trichloroethane Concentrations
Measured at TVKY
i n!
a
20121 2013 2014 2015 2016
Year
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-64 for 1,1,2-trichloroethane concentrations measured at
TVKY include the following:
• Three 1,1,2-trichloroethane concentrations greater than 1 |ig/m3 have been measured
at TVKY; one in 2013 (2.15 |ig/m3) and two in 2016 (1.90 |ig/m3 and 1.06 |ig/m3).
All other 1,1,2-trichloroethane concentrations measured at TVKY are less than
0.65 |ig/m3 and most are less than 0.45 |ig/m3.
• The minimum, 5th percentile, and median concentrations of 1,1,2-trichloroethane for
TVKY are zero for all years of sampling, indicating that at least half of the
measurements were non-detects each year. The percentage of non-detects has ranged
from 72 percent (2016) to 87 percent (2013).
• The 1-year average concentration decreased slightly from 2013 to 2014, changed little
for 2015, then increased by a factor of three for 2016. With the percentage of non-
detects at a minimum, the 95th percentile at a maximum, and the two of the three
highest concentrations measured, this increase is not surprising. The number of
1,1,2-trichloroethane concentrations greater than 0.1 |ig/m3 is also at a maximum (11)
for 2016.
12-91
-------
Figure 12-65. Yearly Statistical Metrics for Vinyl Chloride Concentrations Measured at
TVKY
<>.
2013
Maximum Concentration
for 2014 is 22.2 |Jg/m3
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
Observations from Figure 12-65 for vinyl chloride concentrations measured at TVKY
include the following:
• The minimum concentration and 5th percentile for each year is zero, indicating the
presence of non-detects. The percentage of non-detects increased from 31 percent to
42 percent between 2012 and 2013, then decreased each year after, reaching a
minimum of 13 percent in 2016.
• Measured detections have ranged from 0.00512 |ig/m3 to 22.2 |ig/m3; the maximum
vinyl chloride concentration measured at TVKY is nearly four times greater than the
next highest concentration measured at this site (6.49 |ig/m3). Vinyl chloride
concentrations greater than 2 |ig/m3 have been measured each year of sampling
except 2016.
• The median concentration decreased by more than 70 percent from 2012 to 2013,
from 0.14 |ig/m3 to 0.04 |ig/m3, although this is difficult to see in Figure 12-65. This
is followed by an increase each year through 2015, such that the median
concentration approaches 0.10 |ig/m3 again for 2015 (and is unchanged for 2016).
• Although the 1-year average concentration doubles between 2013 and 2014, if the
maximum concentration was excluded from the calculation, the 1-year average would
change little among the years of sampling, with each approximately 0.30 |ig/m3; thus,
the increase shown for 2014 is attributable to the outlier measured that year.
12-92
-------
Figure 12-66. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
LEKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because the completeness criteria were not met for 2016.
Observations from Figure 12-66 for arsenic (PMio) concentrations measured at LEKY
include the following:
• The two highest arsenic concentrations measured LEKY were measured one month
apart (4.97 ng/m3, December 2015 and 3.08 ng/m3, January 2016).
• The 95th percentile for 2012 is at a maximum across the years of sampling; this is
also true for the 1-year average (0.92 ng/m3) and median (0.76 ng/m3) concentrations.
The number of arsenic concentrations greater than 1 ng/m3 is at a maximum for 2012,
accounting for more than one-third of the measurements, which decreases by almost
half for the following year.
• The sole non-detect of arsenic was measured in 2013. This, combined nearly twice
the number of arsenic concentrations less than 0.25 ng/m3, along with fewer arsenic
concentrations at the upper end of the concentration range, helps explain the decrease
in the 1-year average concentration shown for 2013. Little change is shown in the
concentration profile for 2014.
• All of the statistical parameters exhibit an increase from 2014 to 2015. While some of
this is related to the maximum concentration measured that year, it is not the only
reason. The minimum concentration increased two-fold from 2014 to 2015; the 5th
percentile exhibits a similar increase. The number of arsenic concentrations less than
0.3 ng/m3 decreased from 12 in 2014 to two in 2015.
12-93
-------
• A 1-year average concentration for 2016 is not provided because the completeness
criteria were not met; a number of metals samples collected at LEKY in March and
April 2016 had QA-related issues according to the state of Kentucky, which
combined with additional invalid samples throughout the year, resulted in a completes
of 77 percent. The median concentration for 2016 exhibits an increase, returning to
near 2012-levels.
Figure 12-67. Yearly Statistical Metrics for Benzene Concentrations Measured at LEKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented as collection system issues were experienced in 2013.
3 A 1-year average is not presented because sampling was discontinued at the end of July 2016.
Observations from Figure 12-67 for benzene concentrations measured at LEKY include
the following:
• Although sampling for VOCs at LEKY under the NMP began in March 2012, there
was a 3-month break in sampling, then sampling resumed in mid-July. Because a full
year's worth of data is not available, a 1-year average concentration for 2012 is not
presented, although the range of measurements is provided. Issues with the collection
system experienced during the first half of 2013 resulted in the invalidation of most
samples collected between February and April; thus, a 1-year average concentration
for 2013 is not presented. Finally, a 1-year average concentration is not presented for
2016 because VOC sampling was discontinued at the end of July.
• Seven of the 13 benzene concentrations greater than 1 |ig/m3 measured at LEKY were
measured in 2013, including the maximum concentration (1.63 |ig/m3).
12-94
-------
• Concentrations of benzene appear to have a decreasing trend at LEKY after 2013.
The median benzene concentration increased slightly from 2012 to 2013, then
decreased each year after, reaching a minimum of 0.39 |ig/m3 for 2016. 2016 is the
first year a benzene concentration greater than 1 |ig/m3 is not measured, although the
concentration profile for 2016 represents only seven months of sampling.
Figure 12-68. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
LEKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented as collection system issues were experienced in 2013.
3 A 1-year average is not presented because sampling was discontinued at the end of July 2016.
Observations from Figure 12-68 for 1,3-butadiene concentrations measured at LEKY
include the following:
• The six highest 1,3-butadiene concentrations were all measured at LEKY in 2012.
Concentrations greater than 0.20 |ig/m3 have not been measured in the years since.
• Each of the statistical parameters shown (with the exception of the minimum
concentration, which did not change) exhibits a decrease between 2012 and 2013.
• Although the range of concentrations measured is fairly consistent, concentrations of
1,3-butadiene also appear to have a decreasing trend at this site. The median
concentration has decreased by nearly half between 2012 and 2016, decreasing from
0.08 |ig/m3 to 0.04 |ig/m3 during the period of sampling.
12-95
-------
Figure 12-69. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at LEKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented as collection system issues were experienced in 2013.
3 A 1-year average is not presented because sampling was discontinued at the end of July 2016.
Observations from Figure 12-69 for carbon tetrachloride concentrations measured at
LEKY include the following:
• The entire range of carbon tetrachloride concentrations measured at LEKY spans less
than 0.75 |ig/m3. If the minimum and maximum concentrations measured in 2013 are
excluded, the entire range would vary by less than 0.45 |ig/m3.
• Seven carbon tetrachloride concentrations greater than 0.8 |ig/m3 were measured at
LEKY, all of which were measured in either 2013 (three) or 2016 (four).
• Despite the difference in the range of concentrations measured, the median
concentration for 2012 is the same as the median concentration for 2013
(0.65 |ig/m3). Between 2012 and 2015, the median concentration varies by less than
0.04 |ig/m3, ranging from 0.61 |ig/m3 (2014) to 0.65 |ig/m3 (2012, 2013).
• All of the statistical parameters exhibit an increase for 2016, with several at a
maximum over the period of sampling. Four carbon tetrachloride concentrations
measured in 2016 are greater than the maximum concentration measured in 2015
(also true for 2014). At the other end of the concentration range, only one carbon
tetrachloride concentrations less than 0.6 |ig/m3 was measured in 2016, compared to
17 for 2015 (and 22 for 2014).
12-96
-------
Figure 12-70. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
LEKY
o-
2014
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented as collection system issues were experienced in 2013.
3 A 1-year average is not presented because sampling was discontinued at the end of July 2016.
Observations from Figure 12-70 for /;-dichlorobenzene concentrations measured at
LEKY include the following:
• With the exception of 2013, non-detects account for more than half of the
/;-dichlorobenzene measurements collected at this site each year, as the minimum, 5th
percentile, and median concentration (for most years) are zero. For 2013, non-detects
account for less than 40 percent of the measurements.
• Although it appears that there is considerable variation in the upper end of the
concentrations measured, the relatively small scale of Figure 12-70 should be noted.
Less than 0.1 |ig/m3 separates the maximum concentrations measured across the years
of sampling and less than 0.04 |ig/m3 separates the 95th percentiles shown.
12-97
-------
Figure 12-71. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at LEKY
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because consistent sampling under the NMP began in July 2012.
2 A 1-year average is not presented as collection system issues were experienced in 2013.
3 A 1-year average is not presented because sampling was discontinued at the end of July 2016.
Observations from Figure 12-71 for 1,2-dichloroethane concentrations measured at
LEKY include the following:
• The number of non-detects is at a maximum for 2012 (five). Two or fewer non-
detects were measured between 2013 and 2015, with none measured in 2016.
• The maximum 1,2-dichloroethane concentration was measured in 2014
(0.211 |ig/m3). The maximum concentration for the remaining years falls within a
relatively tight range, between 0.09 |ig/m3 and 0.13 |ig/m3.
• The median concentration increases slightly each year between 2012 and 2014,
decreases for 2015 (as does the 1-year average concentration), then increases again
for 2016. The changes in the median concentration, however, vary by less than
0.02 |ig/m3 over the years of sampling.
12-98
-------
12.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the Kentucky monitoring sites. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
12.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Kentucky monitoring sites, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 12-5, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
12-99
-------
Table 12-5. Risk Approximations for the Kentucky Monitoring Sites
Pollutant
Cancer URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Health Department, Ashland, Kentucky - ASKY
Benzene
0.0000078
0.03
60/60
0.78
±0.11
6.08
0.03
60/60
0.65
±0.10
5.10
0.02
1.3 -Butadiene
0.00003
0.002
56/60
0.06
±0.01
1.92
0.03
55/60
0.06
±0.01
1.94
0.03
Carbon Tetrachloride
0.000006
0.1
60/60
0.62
±0.03
3.71
0.01
60/60
0.64
±0.02
3.85
0.01
1,2 -Dichloroethane
0.000026
2.4
59/60
0.08
±0.01
1.96
<0.01
53/60
0.07
±0.01
1.78
<0.01
Hexachloro-1,3 -butadiene
0.000022
0.09
2/60
NR
NR
NR
15/60
0.02
±0.01
0.41
<0.01
21st and Greenup, Ashland, Kentucky - ASKY-M
Arsenic (PMi0)a
0.0043
0.000015
55/55
1.38
±0.32
5.93
0.09
55/55
1.13
±0.23
4.86
0.08
Cadmium (PMi0)a
0.0018
0.00001
55/55
0.32
±0.08
0.58
0.03
55/55
0.39
±0.22
0.71
0.04
Lead (PMi0)a
0.00015
55/55
8.83
±2.08
0.06
55/55
6.97
± 1.86
0.05
Manganese (PMi0)a
0.0003
55/55
25.05
±5.70
0.08
55/55
19.85
±4.75
0.07
Nickel (PMi,:,)a
0.00048
0.00009
55/55
2.20
±0.49
1.06
0.02
55/55
2.12
±0.62
1.02
0.02
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Table 12-5. Risk Approximations for the Kentucky Monitoring Sites (Continued)
2015
2016
# of
Risk Approximations
# of
Risk Approximations
Measured
Measured
Noncancer
Detections
Annual
Detections
Annual
Cancer
Cancer URE
RfC
vs. # of
Average
Cancer
Noncancer
vs. # of
Average
(in-a-
Noncancer
Pollutant
(jig/m3)1
(mg/m3)
Samples
frig/m3)
(in-a-million)
(HQ)
Samples
frig/m3)
million)
(HQ)
Grayson, Kentucky - GLKY
0.87
0.86
Acetaldehyde
0.0000022
0.009
60/60
±0.07
1.92
0.10
61/61
±0.10
1.89
0.10
0.39
0.39
Benzene
0.0000078
0.03
60/60
±0.04
3.05
0.01
61/61
±0.04
3.01
0.01
0.03
0.03
1.3 -Butadiene
0.00003
0.002
51/60
±0.01
0.98
0.02
47/61
±0.01
0.93
0.02
0.62
0.63
Carbon Tetrachloride
0.000006
0.1
60/60
±0.02
3.74
0.01
61/61
±0.02
3.79
0.01
0.06
0.06
1,2 -Dichloroethane
0.000026
2.4
58/60
±0.01
1.61
<0.01
51/61
±0.01
1.47
<0.01
1.76
1.87
Formaldehyde
0.000013
0.0098
60/60
±0.25
22.86
0.18
61/61
±0.24
24.35
0.19
0.54
0.53
Arsenic (PMi0)a
0.0043
0.000015
57/58
±0.09
2.33
0.04
59/59
±0.11
2.27
0.04
Baskett, Kentucky - BAKY
0.96
0.92
Arsenic (PMi0)a
0.0043
0.000015
56/56
±0.19
4.15
0.06
58/58
±0.15
3.96
0.06
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Table 12-5. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Atmos Energy, Calvert City, Kentucky - ATKY
Benzene
0.0000078
0.03
59/59
0.96
±0.32
7.48
0.03
61/61
0.57
±0.10
4.45
0.02
1.3 -Butadiene
0.00003
0.002
56/59
0.06
±0.02
1.84
0.03
52/61
0.05
±0.01
1.47
0.02
Carbon Tetrachloride
0.000006
0.1
59/59
0.67
±0.03
4.04
0.01
61/61
0.69
±0.03
4.11
0.01
1,2 -Dichloroethane
0.000026
2.4
59/59
0.41
±0.19
10.65
<0.01
60/61
0.90
±0.55
23.46
<0.01
Hexachloro-1,3 -butadiene
0.000022
0.09
3/59
NR
NR
NR
14/61
0.02
±0.01
0.46
<0.01
1,1,2-Trichloroethane
0.000016
0.4
6/59
0.01
±0.01
0.14
<0.01
9/61
0.09
±0.08
1.43
<0.01
Vinvl chloride
0.0000088
0.1
45/59
0.69
±0.32
6.07
0.01
50/61
0.83
±0.54
7.33
0.01
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Table 12-5. Risk Approximations for the Kentucky Monitoring Sites (Continued)
Pollutant
Cancer URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Smithland, Kentucky - BLKY
Benzene
0.0000078
0.03
59/59
0.56
±0.06
4.39
0.02
60/60
0.49
±0.07
3.85
0.02
1.3 -Butadiene
0.00003
0.002
51/59
0.06
±0.02
1.84
0.03
49/60
0.06
±0.03
1.87
0.03
Carbon Tetrachloride
0.000006
0.1
59/59
0.67
±0.03
4.02
0.01
60/60
0.73
±0.04
4.40
0.01
1,2 -Dichloroethane
0.000026
2.4
59/59
0.72
±0.29
18.64
<0.01
58/60
1.89
± 1.55
49.10
<0.01
Hexachloro-1,3 -butadiene
0.000022
0.09
6/59
NR
NR
NR
17/60
0.03
±0.01
0.57
<0.01
Vinyl chloride
0.0000088
0.1
41/59
0.10
±0.04
0.92
<0.01
47/60
0.17
±0.07
1.48
<0.01
Arsenic (PMi0)a
0.0043
0.000015
50/51
0.57
±0.09
2.43
0.04
59/59
0.54
±0.07
2.34
0.04
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Table 12-5. Risk Approximations for the Kentucky Monitoring Sites (Continued)
2015
2016
# of
Risk Approximations
# of
Risk Approximations
Measured
Measured
Noncancer
Detections
Annual
Detections
Annual
Cancer
Cancer URE
RfC
vs. # of
Average
Cancer
Noncancer
vs. # of
Average
(in-a-
Noncancer
Pollutant
Gig/m3)"1
(mg/m3)
Samples
frig/m3)
(in-a-million)
(HQ)
Samples
frig/m3)
million)
(HQ)
TVA Substation, Calvert City, Kentucky - TVKY
1.04
0.54
Benzene
0.0000078
0.03
62/62
±0.30
8.10
0.03
61/61
±0.09
4.21
0.02
0.35
0.16
1.3 -Butadiene
0.00003
0.002
56/62
±0.19
10.55
0.18
52/61
±0.12
4.75
0.08
0.85
0.80
Carbon Tetrachloride
0.000006
0.1
62/62
±0.13
5.10
0.01
61/61
±0.10
4.78
0.01
3.75
3.49
1,2 -Dichloroethane
0.000026
2.4
62/62
± 1.56
97.62
<0.01
60/61
± 1.36
90.70
<0.01
0.03
0.09
1,1,2-Trichloroethane
0.000016
0.4
13/62
±0.02
0.47
<0.01
16/61
±0.08
1.52
<0.01
0.30
0.31
Vinvl chloride
0.0000088
0.1
42/62
±0.12
2.60
<0.01
53/61
±0.11
2.77
<0.01
Lexington, Kentucky - LEKY
0.51
Benzene
0.0000078
0.03
53/53
±0.06
3.96
0.02
29/29
NA
NA
NA
0.06
1.3 -Butadiene
0.00003
0.002
47/53
±0.01
1.71
0.03
28/29
NA
NA
NA
0.63
Carbon Tetrachloride
0.000006
0.1
53/53
±0.02
3.77
0.01
29/29
NA
NA
NA
0.03
p-Dichlorobcnzcnc
0.000011
0.8
26/53
±0.01
0.36
<0.01
9/29
NA
NA
NA
0.07
1,2 -Dichloroethane
0.000026
2.4
51/53
±0.01
1.73
<0.01
29/29
NA
NA
NA
0.81
Arsenic (PMi0)a
0.0043
0.000015
56/56
±0.17
3.50
0.05
47/47
NA
NA
NA
a Average concentrations provided below the blue line for this site and/or pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Observations for the Kentucky monitoring sites from Table 12-5 include the following:
• The pollutants with the highest annual average concentrations for both years for
ASKY are benzene and carbon tetrachloride. These two pollutants also have the
highest cancer risk approximations for ASKY. All of the noncancer hazard
approximations for the pollutants of interest for ASKY are considerably less than an
HQ of 1.0 (0.03 or less), indicating that no adverse noncancer health effects are
expected from these individual pollutants.
• The pollutant of interest with the highest annual average concentrations for ASKY-M
is manganese. Arsenic has the highest cancer risk approximations among ASKY-M's
pollutants of interest. (Note that lead and manganese do not have cancer UREs.) All
of the noncancer hazard approximations for the pollutants of interest for ASKY-M are
considerably less than an HQ of 1.0 (0.09 or less), indicating that no adverse
noncancer health effects are expected from these individual pollutants.
• Formaldehyde is the only pollutant of interest for GLKY with annual average
concentrations greater than 1 |ig/m3. This pollutant also has the highest cancer risk
approximations for GLKY. All of the noncancer hazard approximations for the
pollutants of interest for GLKY are considerably less than an HQ of 1.0 (0.19 or less),
indicating that no adverse noncancer health effects are expected from these individual
pollutants.
• Arsenic is the only pollutant of interest for BAKY. BAKY's cancer risk
approximation for arsenic for 2015 is just greater than 4 in-a-million while the cancer
risk approximation for 2016 is just less than 4 in-a-million. The noncancer hazard
approximations for arsenic for BAKY are considerably less than an HQ of 1.0 (both
are 0.06), indicating that no adverse noncancer health effects are expected from this
individual pollutant.
• 1,2-Dichloroethane has the highest cancer risk approximations among the pollutants
of interest for the Calvert City sites, each one greater than 10 in-a-million and ranging
from 10.65 in-a-million (ATKY, 2015) to 97.62 in-a-million (TVKY, 2015). Both of
TVKY's cancer risk approximations are greater than 90 in-a-million. These cancer
risk approximations for TVKY are the highest ones calculated in the 2015-2016 NMP
report. TVKY's cancer risk approximation for 1,3-butadiene for 2015 is also greater
than 10 in-a-million, which is the highest approximation calculated for this pollutant
across the program.
• All of the noncancer hazard approximations for the pollutants of interest for the
Calvert City sites are considerably less than an HQ of 1.0 (0.18 or less), indicating
that no adverse noncancer health effects are expected from these individual
pollutants.
• Carbon tetrachloride and benzene are the pollutants of interest for LEKY with the
highest annual average concentrations (2015 only). These pollutants also have the
highest cancer risk approximations for LEKY, both of which are just less than
4-in-a-million. All of the noncancer hazard approximations for the pollutants of
interest for LEKY are considerably less than an HQ of 1.0 (0.05 or less), indicating
12-105
-------
that no adverse noncancer health effects are expected from these individual
pollutants.
As an extension of this analysis, pollution roses were created for each of the site-specific
pollutants of interest that have a cancer risk approximation greater than 75 in-a-million and/or a
noncancer hazard approximation greater than 1.0, where applicable. Thus, a pollution rose was
created for TVKY's 1,2-dichloroethane measurements. A pollution rose is a plot of the ambient
concentration versus the wind direction; the magnitude of the concentration is indicated using
different colored dots and are shown in relation to the average wind direction oriented about a
16-point compass. Thus, high concentrations may be shown in relation to the direction of
potential emissions sources. Hourly wind observations collected at the NWS station at the
Barkley Regional Airport and obtained from NOAA are used in this analysis and were averaged
(using vector averaging techniques) to compute daily wind direction averages for comparison to
the 24-hour concentration data. This analysis is intended to help identify the geographical area
where the emissions sources of these pollutants may have originated. Additional information
regarding this analysis is also presented in Section 3.4.2.3. Figure 12-72 presents the pollution
rose for all 122 1,2-dichloroethane concentrations measured at TVKY over the two-year
sampling period.
Observations for Figure 12-72 include the following:
• Concentrations greater than 10 |ig/m3 are plotted on the pollution rose in blue, with
pink representing concentrations between 1 |ig/m3 and 10 |ig/m3 and yellow
representing concentrations less than 1 |ig/m3.
• The pollution rose shows that the concentrations greater than 10 |ig/m3 tended to be
measured at TVKY primarily on days with an average wind direction between 0°
(north) and 45° (northeast) or between 135° (southeast) and 225° (southwest), with a
few exceptions. Concentrations between 1 |ig/m3 and 10 |ig/m3 were most often
measured on sample days with an average wind direction between 315° (northwest)
and 45° (northeast) or between 135° (southeast) and 225° (southwest), although there
are exceptions to this as well. Few 1,2-dichloroethane concentrations greater than 1
|ig/m3 were measured on days with an average wind direction from the east or west.
• This pollution rose shows that 1,2-dichloroethane concentrations less than 1 |ig/m3
were measured at TVKY on sample days with a variety of average wind directions.
However, these lower concentrations were infrequently measured on days with an
average wind direction between 45° (northeast) and 135° (southeast).
12-106
-------
Figure 12-72. Pollution Roses for 1,2-Dichloroethane Concentrations Measured at TVKY
360/0
45
315,
25 jig/rrv
270
CD 0
225
135
180
O < 1 |ig/m3 O 1-10 |ig/m3 O > 10 |ig/m3
12-107
-------
12.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, Tables 12-6 and 12-7 present an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 12-6 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 12-6 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 12-6 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 12-5. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 12-6. Table 12-7 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 12.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
12-108
-------
Table 12-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Kentucky Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Benzene
29.01
Nickel, PM
8.18E-04
Benzene
6.08
Formaldehyde
16.75
Hexavalent Chromium. PM
7.61E-04
Benzene
5.10
Acetaldehyde
10.14
2,4-Dinitrotoluene
3.91E-04
Carbon Tetrachloride
3.85
Ethylbenzene
9.11
Benzene
2.26E-04
Carbon Tetrachloride
3.71
2,4-Dinitrotoluene
4.39
Formaldehyde
2.18E-04
1,2-Dichloroethane
1.96
Naphthalene
2.52
Naphthalene
8.58E-05
1,3-Butadiene
1.94
1.3 -Butadiene
2.32
Nitrobenzene
7.27E-05
1,3-Butadiene
1.92
Bis(2-ethylhexyl) phthalate, gas
2.27
1,3-Butadiene
6.97E-05
1,2-Dichloroethane
1.78
Nitrobenzene
1.82
Cadmium, PM
4.78E-05
Hexachloro-1,3 -butadiene
0.41
Nickel, PM
1.70
POM, Group 2b
3.14E-05
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Benzene
29.01
Nickel, PM
8.18E-04
Arsenic
5.93
Formaldehyde
16.75
Hexavalent Chromium. PM
7.61E-04
Arsenic
4.86
Acetaldehyde
10.14
2,4-Dinitrotoluene
3.91E-04
Nickel
1.06
Ethylbenzene
9.11
Benzene
2.26E-04
Nickel
1.02
2,4-Dinitrotoluene
4.39
Formaldehyde
2.18E-04
Cadmium
0.71
Naphthalene
2.52
Naphthalene
8.58E-05
Cadmium
0.58
1,3-Butadiene
2.32
Nitrobenzene
7.27E-05
Bis(2-ethylhexyl) phthalate, gas
2.27
1,3-Butadiene
6.97E-05
Nitrobenzene
1.82
Cadmium, PM
4.78E-05
Nickel, PM
1.70
POM, Group 2b
3.14E-05
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Grayson, Kentucky (Carter County) - GLKY
Benzene
14.81
Formaldehyde
1.78E-04
Formaldehyde
24.35
Formaldehyde
13.70
Benzene
1.16E-04
Formaldehyde
22.86
Acetaldehyde
8.80
Naphthalene
7.26E-05
Carbon Tetrachloride
3.79
Ethylbenzene
5.87
1,3-Butadiene
5.24E-05
Carbon Tetrachloride
3.74
Naphthalene
2.14
POM, Group 2b
3.25E-05
Benzene
3.05
1.3 -Butadiene
1.75
POM, Group 2d
2.54E-05
Benzene
3.01
Bis(2-ethylhexyl) phthalate, gas
1.27
POM, Group 5a
2.36E-05
Arsenic
2.33
POM, Group 2b
0.37
Acetaldehyde
1.94E-05
Arsenic
2.27
POM, Group 2d
0.29
Ethylbenzene
1.47E-05
Acetaldehyde
1.92
2,4-Toluene diisocyanate
0.11
POM, Group 6
6.72E-06
Acetaldehyde
1.89
Baskett, Kentucky (Henderson County) - BAKY
Benzene
28.00
Naphthalene
6.53E-04
Arsenic
4.15
Formaldehyde
27.39
Formaldehyde
3.56E-04
Arsenic
3.96
Naphthalene
19.21
Nickel, PM
3.12E-04
Acetaldehyde
15.32
Benzene
2.18E-04
Ethylbenzene
15.19
1,3-Butadiene
1.01E-04
Tetrachloroethylene
4.71
POM, Group 2b
4.55E-05
1,3-Butadiene
3.36
Cadmium, PM
4.47E-05
Bis(2-ethylhexyl) phthalate, gas
2.15
Ethylbenzene
3.80E-05
Dichloromethane
0.85
POM, Group 2d
3.62E-05
Nickel, PM
0.65
Acetaldehyde
3.37E-05
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Benzene
62.95
Benzene
4.91E-04
1,2-Dichloroethane
23.46
Ethylbenzene
37.64
Formaldehyde
4.06E-04
1,2-Dichloroethane
10.65
Formaldehyde
31.25
Arsenic, PM
3.16E-04
Benzene
7.48
Acetaldehyde
27.52
Vinyl chloride
2.41E-04
Vinyl chloride
7.33
Vinyl chloride
27.36
1,2-Dichloroethane
2.34E-04
Vinyl chloride
6.07
1,2-Dichloroethane
9.01
1,3-Butadiene
2.34E-04
Benzene
4.45
1.3 -Butadiene
7.80
Hexachlorobenzene
2.18E-04
Carbon Tetrachloride
4.11
Naphthalene
5.94
Naphthalene
2.02E-04
Carbon Tetrachloride
4.04
Bis(2-ethylhexyl) phthalate, gas
1.45
Hexavalent Chromium. PM
9.68E-05
1,3-Butadiene
1.84
Carbon tetrachloride
1.43
Ethylbenzene
9.41E-05
1,3-Butadiene
1.47
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Benzene
62.95
Benzene
4.91E-04
1,2-Dichloroethane
97.62
Ethylbenzene
37.64
Formaldehyde
4.06E-04
1,2-Dichloroethane
90.70
Formaldehyde
31.25
Arsenic, PM
3.16E-04
1,3-Butadiene
10.55
Acetaldehyde
27.52
Vinyl chloride
2.41E-04
Benzene
8.10
Vinyl chloride
27.36
1,2-Dichloroethane
2.34E-04
Carbon Tetrachloride
5.10
1,2-Dichloroethane
9.01
1,3-Butadiene
2.34E-04
Carbon Tetrachloride
4.78
1.3 -Butadiene
7.80
Hexachlorobenzene
2.18E-04
1,3-Butadiene
4.75
Naphthalene
5.94
Naphthalene
2.02E-04
Benzene
4.21
Bis(2-ethylhexyl) phthalate, gas
1.45
Hexavalent Chromium. PM
9.68E-05
Vinyl chloride
2.77
Carbon tetrachloride
1.43
Ethylbenzene
9.41E-05
Vinyl chloride
2.60
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Smithland, Kentucky (Livingston County) - BLKY
Benzene
11.51
Formaldehyde
1.19E-04
1,2-Dichloroethane
49.10
Formaldehyde
9.13
Benzene
8.98E-05
1,2-Dichloroethane
18.64
Ethylbenzene
7.04
1,3-Butadiene
4.81E-05
Carbon Tetrachloride
4.40
Acetaldehyde
5.22
Naphthalene
4.07E-05
Benzene
4.39
1.3 -Butadiene
1.60
POM, Group 2b
2.09E-05
Carbon Tetrachloride
4.02
Naphthalene
1.20
Ethylbenzene
1.76E-05
Benzene
3.85
Bis(2-ethylhexyl) phthalate, gas
0.44
POM, Group 2d
1.65E-05
Arsenic
2.43
POM, Group 2b
0.24
POM, Group 5a
1.36E-05
Arsenic
2.34
POM, Group 2d
0.19
Acetaldehyde
1.15E-05
1,3-Butadiene
1.87
2,4-Toluene diisocyanate
0.04
Nickel, PM
9.57E-06
1,3-Butadiene
1.84
Lexington, Kentucky (Fayette County) - LEKY
Benzene
106.03
Formaldehyde
1.15E-03
Benzene
3.96
Formaldehyde
88.65
Benzene
8.27E-04
Carbon Tetrachloride
3.77
Ethylbenzene
69.29
Naphthalene
5.92E-04
Arsenic
3.50
Acetaldehyde
58.48
1,3-Butadiene
4.63E-04
1,2-Dichloroethane
1.73
Naphthalene
17.42
Ethylene oxide
2.98E-04
1,3-Butadiene
1.71
1,3-Butadiene
15.43
POM, Group 2b
1.82E-04
p-Dichlorobenzene
0.36
Bis(2-ethylhexyl) phthalate, gas
14.33
Ethylbenzene
1.73E-04
POM, Group 2b
2.07
POM, Group 2d
1.47E-04
POM, Group 2d
1.67
Acetaldehyde
1.29E-04
2,4-Toluene diisocyanate
1.23
POM, Group 5a
1.17E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Kentucky Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Health Department, Ashland, Kentucky (Boyd County) - ASKY
Toluene
79.35
Acrolein
56,574.58
1,3-Butadiene
0.03
Methanol
34.79
Chlorine
24,334.81
1,3-Butadiene
0.03
Xylenes
33.38
Manganese, PM
21,583.40
Benzene
0.03
Benzene
29.01
Nickel, PM
18,936.65
Benzene
0.02
Hexane
17.21
Lead, PM
4,636.36
Carbon Tetrachloride
0.01
Formaldehyde
16.75
2,4-Toluene diisocyanate
2,782.50
Carbon Tetrachloride
0.01
Hydrochloric acid
12.92
Cyanide Compounds, gas
2,677.09
Hexachloro-1,3 -butadiene
<0.01
Acetaldehyde
10.14
Cadmium, PM
2,652.82
1,2-Dichloroethane
<0.01
Ethylbenzene
9.11
Formaldehyde
1,709.26
1,2-Dichloroethane
<0.01
Manganese, PM
6.48
1,3-Butadiene
1,161.65
21st and Greenup, Ashland, Kentucky (Boyd County) - ASKY-M
Toluene
79.35
Acrolein
56,574.58
Arsenic
0.09
Methanol
34.79
Chlorine
24,334.81
Manganese
0.08
Xylenes
33.38
Manganese, PM
21,583.40
Arsenic
0.08
Benzene
29.01
Nickel, PM
18,936.65
Manganese
0.07
Hexane
17.21
Lead, PM
4,636.36
Lead
0.06
Formaldehyde
16.75
2,4-Toluene diisocyanate
2,782.50
Lead
0.05
Hydrochloric acid
12.92
Cyanide Compounds, gas
2,677.09
Cadmium
0.04
Acetaldehyde
10.14
Cadmium, PM
2,652.82
Cadmium
0.03
Ethylbenzene
9.11
Formaldehyde
1,709.26
Nickel
0.02
Manganese, PM
6.48
1,3-Butadiene
1,161.65
Nickel
0.02
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Grayson, Kentucky (Carter County) - GLKY
Toluene
37.11
Acrolein
44,756.45
Formaldehyde
0.19
Xylenes
21.60
2,4-Toluene diisocyanate
1,550.91
Formaldehyde
0.18
Benzene
14.81
Formaldehyde
1,397.83
Acetaldehyde
0.10
Formaldehyde
13.70
Cyanide Compounds, gas
1,234.28
Acetaldehyde
0.10
Methanol
11.57
Acetaldehyde
978.25
Arsenic
0.04
Acetaldehyde
8.80
1.3 -Butadiene
873.28
Arsenic
0.04
Hexane
8.17
Naphthalene
712.11
1,3-Butadiene
0.02
Ethylbenzene
5.87
Benzene
493.75
1,3-Butadiene
0.02
Naphthalene
2.14
Xylenes
216.03
Benzene
0.01
Styrene
2.04
Bis(2-ethylhexyl) phthalate, gas
126.76
Benzene
0.01
Baskett, Kentucky (Henderson County) - BAKY
Carbonyl sulfide
269.44
Acrolein
110,810.57
Arsenic
0.06
Hydrofluoric acid
131.41
Hydrofluoric acid
9,386.18
Arsenic
0.06
Toluene
82.09
Manganese, PM
8,045.76
Xylenes
54.05
Nickel, PM
7,213.21
Benzene
28.00
Naphthalene
6,403.35
Formaldehyde
27.39
Chlorine
3,106.53
Methanol
23.45
4,4-Methylenediphenyl diisocyanate, gas
3,053.82
Hexane
19.36
Formaldehyde
2,794.74
Naphthalene
19.21
2,4-Toluene diisocyanate
2,636.22
Acetaldehyde
15.32
Cadmium, PM
2,481.69
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Atmos Energy, Calvert City, Kentucky (Marshall County) - ATKY
Methanol
548.56
Chlorine
921,670.25
Benzene
0.03
Toluene
155.41
Acrolein
91,798.74
1,3-Butadiene
0.03
Xylenes
140.83
Hydrochloric acid
6,166.28
1,3-Butadiene
0.02
Chlorine
138.25
Arsenic, PM
4,904.75
Benzene
0.02
Hydrochloric acid
123.33
1.3 -Butadiene
3,898.17
Vinyl chloride
0.01
Benzene
62.95
Formaldehyde
3,188.52
Vinyl chloride
0.01
Vinyl acetate
57.82
Acetaldehyde
3,057.90
Carbon Tetrachloride
0.01
Ethylbenzene
37.64
Manganese, PM
2,784.81
Carbon Tetrachloride
0.01
Formaldehyde
31.25
Acrylic acid
2,725.48
1,2-Dichloroethane
<0.01
Hexane
27.94
Benzene
2,098.49
Hexachloro-1,3 -butadiene
<0.01
TVA Substation, Calvert City, Kentucky (Marshall County) - TVKY
Methanol
548.56
Chlorine
921,670.25
1,3-Butadiene
0.18
Toluene
155.41
Acrolein
91,798.74
1,3-Butadiene
0.08
Xylenes
140.83
Hydrochloric acid
6,166.28
Benzene
0.03
Chlorine
138.25
Arsenic, PM
4,904.75
Benzene
0.02
Hydrochloric acid
123.33
1,3-Butadiene
3,898.17
Carbon Tetrachloride
0.01
Benzene
62.95
Formaldehyde
3,188.52
Carbon Tetrachloride
0.01
Vinyl acetate
57.82
Acetaldehyde
3,057.90
Vinyl chloride
<0.01
Ethylbenzene
37.64
Manganese, PM
2,784.81
Vinyl chloride
<0.01
Formaldehyde
31.25
Acrylic acid
2,725.48
1,2-Dichloroethane
<0.01
Hexane
27.94
Benzene
2,098.49
1,2-Dichloroethane
<0.01
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 12-7. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Kentucky Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Smithland, Kentucky (Livingston County) - BLKY
Toluene
37.15
Acrolein
31,963.47
Arsenic
0.04
Xylenes
24.92
Formaldehyde
931.75
Arsenic
0.04
Benzene
11.51
1,3-Butadiene
802.24
1,3-Butadiene
0.03
Formaldehyde
9.13
Acetaldehyde
579.95
1,3-Butadiene
0.03
Hexane
7.82
2,4-Toluene diisocyanate
532.37
Benzene
0.02
Ethylbenzene
7.04
Cyanide Compounds, gas
509.85
Benzene
0.02
Acetaldehyde
5.22
Naphthalene
399.44
Carbon Tetrachloride
0.01
Methanol
3.91
Benzene
383.60
Carbon Tetrachloride
0.01
1.3 -Butadiene
1.60
Manganese, PM
303.10
Vinyl chloride
<0.01
Styrene
1.35
Xylenes
249.19
Vinyl chloride
<0.01
Lexington, Kentucky (Fayette County) - LEKY
Toluene
434.79
Acrolein
305,893.45
Arsenic
0.05
Xylenes
248.80
2,4-Toluene diisocyanate
17,526.57
1,3-Butadiene
0.03
Methanol
152.86
Formaldehyde
9,045.61
Benzene
0.02
Benzene
106.03
1,3-Butadiene
7,715.19
Carbon Tetrachloride
0.01
Formaldehyde
88.65
Acetaldehyde
6,497.82
p-Dichlorobcnzcnc
<0.01
Hexane
82.79
Manganese, PM
5,988.12
1,2-Dichloroethane
<0.01
Ethylbenzene
69.29
Naphthalene
5,806.27
Acetaldehyde
58.48
Benzene
3,534.28
Ethylene glycol
19.99
Hexamethylene-1,6-diisocyanate, gas
3,371.90
Naphthalene
17.42
Xylenes
2,488.00
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 12-6 include the following:
• Among the Kentucky counties with NMP monitoring sites, emissions (for pollutants
with cancer UREs) are highest in Fayette County (LEKY) and Marshall County
(ATKY, TVKY) and lowest in Livingston County (BLKY) and Carter County
(GLKY).
• Benzene, formaldehyde, acetaldehyde are the highest emitted pollutants with cancer
UREs in Boyd County, where the Ashland sites are located. Nickel, hexavalent
chromium, and 2,4-dinitrotoluene are the pollutants with the highest toxicity-
weighted emissions (of the pollutants with cancer UREs) for Boyd County. Seven of
the highest emitted pollutants also have the highest toxicity-weighted emissions for
Boyd County.
• Benzene has the highest cancer risk approximations for ASKY and appears on both
emissions-based lists (ranking highest for quantity emitted and fourth for its toxicity-
weighted emissions). 1,3-Butadiene also appears on all three lists (ranking seventh
highest for quantity emitted and eighth for its toxicity-weighted emissions). Carbon
tetrachloride, 1,2-dichloroethane, and hexachloro-1,3-butadiene are also pollutants of
interest for ASKY but appear on neither emissions-based list for Boyd County.
• Nickel is the only pollutant of interest for ASKY-M to appear on both emissions-
based lists for Boyd County (ranking tenth highest for quantity emitted and highest
for its toxicity-weighted emissions). While cadmium ranks ninth in Boyd County for
its toxicity-weighted emissions, it is not among the highest emitted (ranking 19th).
Arsenic, which has the highest cancer risk approximations for ASKY-M, appears on
neither emissions-based list (ranking 24th for total emissions and 15th for toxicity-
weighted emissions).
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Carter County, where GLKY is located. Formaldehyde, benzene, and
naphthalene are the pollutants with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for this county. Eight of the highest emitted pollutants
also have the highest toxicity-weighted emissions for Carter County.
• Formaldehyde has the highest cancer risk approximations for GLKY, ranks first for
its toxicity-weighted emissions, and ranks second for its total emissions in Carter
County, as shown in Table 12-6. Benzene and acetaldehyde also appear on all three
lists. Carbon tetrachloride and arsenic appear on neither emissions-based list for
Carter County.
• Two POM Groups appear among the highest emitted pollutants in Carter County
(POM, Groups 2b and 2d) and four POM Groups appear among the pollutants with
the highest toxicity-weighted emissions (POM, Groups 2b, 2d, 5a, and 6). Many of
the PAHs sampled using Method TO-13A are part of POM, Groups 2b, 5a, and 6.
However, none of these pollutants failed screens for GLKY. Naphthalene is also
sampled for with Method TO-13 A; this pollutant appears on both emissions-based
lists for Carter County. Naphthalene failed screens for GLKY but was not identified
as a pollutant of interest for this site.
12-117
-------
Benzene, formaldehyde, and naphthalene are the highest emitted pollutants with
cancer UREs in Henderson County, where BAKY is located. Naphthalene,
formaldehyde, and nickel are the pollutants with the highest toxicity-weighted
emissions (of the pollutants with cancer UREs) for this county. Seven of the highest
emitted pollutants also have the highest toxicity-weighted emissions for Henderson
County.
Arsenic is the only pollutant of interest for BAKY. Arsenic appears on neither
emissions-based list for Henderson County (arsenic ranks 24th for total emissions and
13th for toxicity-weighted emissions).
Benzene, ethylbenzene, and formaldehyde are the highest emitted pollutants with
cancer UREs in Marshall County, where two of the three Calvert City sites are
located. Benzene, formaldehyde, and arsenic are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for this county.
Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Marshall County.
Marshall County is the only county with an NMP site for which vinyl chloride
appears among the highest emitted pollutants. The quantity of vinyl chloride emitted
in Marshall County (27.36 tpy) is the highest emissions for this pollutant among NMP
counties and is considerably higher than the next highest emissions of this pollutant
(9.62 tpy in Los Angeles County, California). Marshall County is also the only county
with an NMP site for which carbon tetrachloride appears among the highest emitted
pollutants. Marshall County is the only county with an NMP site that has carbon
tetrachloride emissions greater than 1 tpy (1.43 tpy). Marshall County is also the only
county with an NMP site for which 1,2-dichloroethane appears among the highest
emitted pollutants. The quantity of 1,2-dichloroethane emitted in Marshall County
(9.01 tpy) again ranks highest, with Monroe County, New York the next closest
(4.02 tpy).
Marshall County is the only county with an NMP site for which vinyl chloride and
1,2-dichloroethane appear among the pollutants with the highest toxicity-weighted
emissions.
Most of the VOC pollutants of interest for the Calvert City sites in Marshall County
appear on both emissions-based lists. Carbon tetrachloride is the exception, as this
pollutant appears among the highest emitted but not among those with the highest
toxicity-weighted emissions (ranking 18th).
Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Livingston County, where BLKY is located. Formaldehyde, benzene,
and 1,3-butadiene are the pollutants with the highest toxicity-weighted emissions (of
the pollutants with cancer UREs) for this county. Eight of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Livingston County.
Benzene and 1,3-butadiene are the only two pollutants of interest for BLKY that
appear among the pollutants on the emissions-based lists for Livingston County.
12-118
-------
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Fayette County, where LEKY is located. Formaldehyde, benzene,
and naphthalene are the pollutants with the highest toxicity-weighted emissions (of
the pollutants with cancer UREs) for this county. Eight of the highest emitted
pollutants also have the highest toxicity-weighted emissions for Fayette County.
• Benzene has the highest cancer risk approximation among LEKY's pollutants of
interest. Benzene and 1,3-butadiene are the only pollutants of interest for LEKY to
appear among the highest emitted pollutants in Fayette County and appear among
those with the highest toxicity-weighted emissions.
Observations from Table 12-7 include the following:
• Among the Kentucky counties with monitoring sites, emissions (for pollutants with
noncancer RfCs) are highest in Marshall County (ATKY, BLKY) and Fayette County
(LEKY) and lowest in Carter County (GLKY) and Livingston County (BLKY).
• Toluene, methanol, and xylenes are the highest emitted pollutants with noncancer
RfCs in Boyd County. Acrolein, chlorine, and manganese are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with noncancer RfCs) for Boyd
County. Two of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Boyd County.
• Although acrolein was sampled for at ASKY, this pollutant was excluded from the
pollutants of interest designation, and thus, subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Boyd
County's highest emitted pollutants.
• Of the pollutants of interest for ASKY, none appear on both emissions-based lists.
Benzene is among the highest emitted in Boyd County but is not among those with
the highest toxicity-weighted emissions. 1,3-Butadiene is among those with the
highest toxicity-weighted emissions but is not among the highest emitted. The
remaining pollutants of interest for ASKY appear on neither emissions-based list in
Table 12-7.
• Manganese ranks tenth for its total emissions and has the third highest toxicity-
weighted emissions for Boyd County. Nickel, lead, and cadmium also appear among
those pollutants with the highest toxicity-weighted emissions in Boyd County,
although they do not appear among the highest emitted in Boyd County. Boyd County
is the only county with an NMP sites for which all four of these metals are among
those with the highest toxicity-weighted emissions. Arsenic is the only pollutant of
interest for ASKY-M that does not appear in either emissions-based list.
• Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
RfCs in Carter County. Acrolein, 2,4-toluene diisocyanate, and formaldehyde are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for Carter County. Five of the highest emitted pollutants also have
the highest toxicity-weighted emissions for Carter County.
12-119
-------
Although acrolein was sampled for at GLKY, this pollutant was excluded from the
pollutants of interest designation, and thus, subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Carter
County's highest emitted pollutants.
Formaldehyde, acetaldehyde, and benzene appear on all three lists for GLKY.
1,3-Butadiene is among the pollutants with the highest toxicity-weighted emissions
but is not among the highest emitted in Carter County (though its emissions rank
11th). Arsenic, the only other pollutants of interest shown in Table 12-7 for GLKY,
appears on neither emissions-based list for Carter County (ranking 37th for total
emissions and 12th for its toxicity-weighted emissions).
Carbonyl sulfide, hydrofluoric acid, and toluene are the highest emitted pollutants
with noncancer RfCs in Henderson County. Henderson County is the only county
with an NMP site for which carbonyl sulfide appears among the 10 highest emitted
pollutants. Acrolein, hydrofluoric acid, and manganese are the pollutants with the
highest toxicity-weighted emissions (of the pollutants with noncancer RfCs) for this
county. Three of the highest emitted pollutants also have the highest toxicity-
weighted emissions for Henderson County.
Arsenic is the only pollutant of interest for BAKY. Arsenic appears on neither
emissions-based list (ranking 44th for total emissions and 20th for toxicity-weighted
emissions). Several other metals, including manganese, nickel, and cadmium, which
were sampled for at BAKY but were not identified as pollutants of interest, appear
among those with the highest toxicity-weighted emissions for Henderson County (of
those with noncancer RfCs).
Methanol, toluene, and xylenes are the highest emitted pollutants with noncancer
RfCs in Marshall County. Chlorine, acrolein, and hydrochloric acid are the pollutants
with the highest toxicity-weighted emissions (of the pollutants with noncancer RfCs)
for this county. Marshall County is one of two counties with an NMP site for which
acrolein was not the pollutant with the highest toxicity-weighted emissions (though it
ranks second for both). Four of the highest emitted pollutants also have the highest
toxicity-weighted emissions for Marshall County.
Benzene is the only pollutant of interest for ATKY and TVKY to appear on all three
lists. 1,3-Butadiene is among the pollutants with the highest noncancer hazard
approximations for the sites located in Marshall County and has the fifth highest
toxicity-weighted emissions but is not among the highest emitted (ranking 15th).
None of the other VOC pollutants of interest for ATKY or TVKY appear on either
emissions-based list for Marshall County.
Toluene, xylenes, and benzene are the highest emitted pollutants with noncancer
RfCs in Livingston County. Acrolein, formaldehyde, and 1,3-butadiene are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for this county. Five of the highest emitted pollutants also have the
highest toxicity-weighted emissions for Livingston County.
12-120
-------
• Although acrolein was sampled for at BLKY, this pollutant was excluded from the
pollutants of interest designation, and thus, subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among
Livingston County's highest emitted pollutants.
• Arsenic has the highest noncancer hazard approximations for BLKY but does not
appear on either emissions-based list (ranking 32nd for total emissions and 12th for
toxicity-weighted emissions). 1,3-Butadiene and benzene are also among BLKY's
pollutants of interest with the highest noncancer hazard approximations. These two
pollutants appear on both emissions-based lists for Livingston County.
• Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in Fayette County. Acrolein, 2,4-toluene diisocyanate, and formaldehyde are the
pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for this county. Five of the highest emitted pollutants also have the
highest toxicity-weighted emissions for Fayette County.
• Although acrolein was sampled for at LEKY, this pollutant was excluded from the
pollutants of interest designation, and thus, subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2. Acrolein does not appear among Fayette
County's highest emitted pollutants.
• Benzene is the only pollutant of interest for LEKY to appear on all three lists in
Table 12-7. 1,3-Butadiene ranks fourth for its toxicity-weighted emissions but is not
among the highest emitted in Fayette County. The remaining pollutants of interest for
LEKY do not appear on either emissions-based list for Fayette County
12.5 Summary of the 2015-2016 Monitoring Data for the Kentucky Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Six monitoring sites sampledfor VOCs; five monitoring sites sampledfor PMio
metals; one monitoring site sampledfor carbonyl compounds and PAHs. VOC
sampling at the LEKY site was discontinued at the end of July 2016.
~~~ The number ofpollutants failing screens for the Kentucky sites varies from five
(ASKY-M, BAKY) to 13 (four sites).
~~~ ASKY-M has the highest annual average concentrations of arsenic among NMP sites
sampling PMio metals, similar to 2013 and 2014. BAKY and LEKY are also among
NMP sites with the highest annual average concentrations of arsenic.
~~~ Some of the highest concentrations of VOCs were measured at the Calvert City sites,
particularly vinyl chloride, carbon tetrachloride, 1,3-butadiene, and
1.2-dichloroethane. TVKY has the highest annual average concentration of
1.3-butadiene, carbon tetrachloride, 1,2-dichloroethane, and 1,1,2-trichloroethane
among NMP sites sampling VOCs. Further, the annual averages for all three Calvert
12-121
-------
City sites rank among the highest calculatedfor carbon tetrachloride and
1.2-dichloroethane.
The most notable trends in concentrations are shown for GLKY. Concentrations of
1.3-butadiene and 1,2-dichloroethane measured at GLKY exhibit a decreasing trend
while concentrations of formaldehyde exhibit an increasing trend at this site.
Concentrations of 1,3-butadiene measured at BLKY also exhibit a decreasing trend.
Benzene and 1,3-butadiene concentrations measured at some Kentucky sites also
appear to have a decreasing trend, but the concentrations measured over the years
have been highly variable.
The cancer risk approximation for 1,2-dichloroethane for TVKY is the second highest
cancer risk approximation calculated among site-specific pollutants of interest.
12-122
-------
13.0 Site in Massachusetts
This section summarizes those data from samples collected at the NATTS site in
Massachusetts and generated by ERG, EPA's contract laboratory for the NMP, over the 2015
and 2016 monitoring efforts. This section also examines
the spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
Sections 1 through 4 for detailed discussions and
definitions regarding the various data analyses presented below.
13.1 Site Characterization
This section characterizes the BOMA monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient measurements.
The BOMA monitoring site is located in Boston. Figure 13-1 presents a composite
satellite image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 13-2 identifies nearby point source emissions locations by source category,
as reported in the 2014 NEI for point sources, version 1. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 13-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference but have been grayed out to emphasize
emissions sources within the boundary. Table 13-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates. Each figure and table
is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. ^
13-1
-------
Figure 13-1. Boston, Massachusetts (BOMA) Monitoring Site
Boston Medical Center
g i rial d'Lewis;
AthletiejjjP
Maditon
ParV
High
School
. John D
Obryant
.SCHL
Math &
Science
Malcolm X Blvd
fflpxbucy
William Eustis Playground
Highland / N
Park /
^PiaasgS''
US'GS ^ # £
Spur«, NASA, NGA^USGS
-© 2 008 Mffios^ffcorp
* John J
Connolly
-------
Figure 13-2. NEI Point Sources Located Within 10 Miles of BOMA
70355'0"W
County
Boston
Harbor \
V Middlesex \
\ County ^
I- ^ - " "4 *
O A \
*E ^ A
„*\ * ° \ +1°
Norfolk
OL- Metals Processing/Fabrication Facility (2)
A Military Base/National Security Facility (1)
X Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility (32)
[Ml Municipal Waste Combustor (1)
• Oil and/or Gas Production (2)
Paint and Coating Manufacturing Facility (3)
R Plastic, Resin, or Rubber Products Plant (2)
10 Port and Harbor Operations (1)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (2)
Ship/Boat Manufacturing or Repair Facility (1)
TT Telecommunications/Radio Facility (3)
Truck/Bus/Transportation Operations (1)
* Wastewater Treatment Facility (2)
6 Water Treatment Facility (1)
W Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
13-3
-------
Table 13-1. Geographical Information for the Massachusetts Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
BOMA
25-025-0042
Boston
Suffolk
Boston-
Cambridge-
Newton, MA-NH
42.329500,
-71.082600
Commercial
Urban/City
Center
27,654
Melnea Cass Blvd near
Shawmut Ave
'AADT reflects 2010 data (MA DOT, 2010)
BOLD ITALICS = EPA-designated NATTS Site
U>
-L
-------
The BOMA monitoring site is located at Dudley Square in Roxbury, a southwest
neighborhood of Boston, and is the Roxbury NATTS site. The surrounding area is commercial as
well as residential, as shown in Figure 13-1. Immediately to the east of the monitoring site are
town homes, to the north is a parking lot and to the west are commercial properties. The original
purpose for the location of this site was to measure population exposure to emissions from a city
bus terminal located another block west of the monitoring site, though buses servicing the area
have since been converted to compressed natural gas (CNG). The monitoring site is 1.3 miles
south of 1-90 and 1 mile southwest of 1-93.
As Figure 13-2 shows, BOMA is located near a large number of point sources, with a
high density of sources located a few miles to the west, northwest, and north of the site. The
source category with the highest number of emissions sources surrounding BOMA is the
institutions category, which includes schools, hospitals, and prisons. There are also numerous
airport and airport support operations, which include airports and related operations as well as
small runways and heliports, such as those associated with hospitals or television stations.
Sources located within 1 mile of BOMA include a hospital, a heliport at a hospital, a university,
a high school, a dry cleaning facility, and a large apartment building. Figure 13-2 shows that
BOMA is located less than 2 miles from the shoreline (Dorchester Bay).
In addition to providing city, county, CBS A, and land use/location setting information,
Table 13-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near BOMA is greater than 27,000 and in the middle of the range compared
to other NMP sites. The traffic estimate provided is for Melnea Cass Boulevard near Shawmut
Avenue.
13.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 13-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
13-5
-------
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 13-2. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. PMio metals and PAHs were sampled for at BOMA.
Table 13-2. 2015-2016 Risk-Based Screening Results for the Massachusetts Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
% of Total
Failures
Cumulative
%
Contribution
Boston, Massachusetts - BOMA
Arsenic (PMio)
0.00023
94
118
79.66
48.45
48.45
Naphthalene
0.029
79
121
65.29
40.72
89.18
Nickel (PMio)
0.0021
11
118
9.32
5.67
94.85
Cadmium (PMio)
0.00056
3
118
2.54
1.55
96.39
Benzo(a)pyrene
0.00057
2
117
1.71
1.03
97.42
Lead (PMio)
0.015
2
118
1.69
1.03
98.45
Acenaphthene
0.011
1
102
0.98
0.52
98.97
Fluoranthene
0.011
1
121
0.83
0.52
99.48
Fluorene
0.011
1
83
1.20
0.52
100.00
Total
194
1,016
19.09
Observations from Table 13-2 include the following:
• Concentrations of nine pollutants failed at least one screen for BOMA; approximately
19 percent of concentrations for these nine pollutants were greater than their
associated risk screening value (or failed screens). Of these nine pollutants, six failed
three screens or less.
• Four pollutants contributed to 95 percent of failed screens for BOMA and therefore
were identified as pollutants of interest for this site. These include three PMio metals
(arsenic, nickel, and cadmium) and one PAH (naphthalene).
• Together, arsenic and naphthalene account for nearly 90 percent of the total failed
screens for BOMA while nickel and cadmium account for 7 percent of the total failed
screens. BOMA is one of only two NMP sites for which cadmium is a pollutant of
interest.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
13-6
-------
13.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Massachusetts monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically to illustrate how each site's concentrations compare to the program-level
averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at the site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at BOMA are provided
in Appendices N and O.
13.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for BOMA, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for BOMA are
presented in Table 13-3, where applicable. Note that if a pollutant was not detected in a given
calendar quarter, the quarterly average simply reflects "0" because only zeros substituted for
non-detects were factored into the quarterly average concentration.
13-7
-------
Table 13-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Massachusetts Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Boston, Massachusetts - BOMA
Arsenic (PMio)
60/56/60
0.38
±0.09
0.84
±0.75
0.63
±0.23
0.36
±0.11
0.55
±0.19
58/58/58
0.40
±0.13
0.55
±0.22
0.47
±0.12
0.55
±0.18
0.48
±0.08
Cadmium (PMio)
60/60/60
0.11
±0.03
0.08
±0.02
0.08
±0.02
0.08
±0.02
0.09
±0.01
58/58/58
0.08
±0.02
0.26
±0.38
0.06
±0.01
0.52
±0.83
0.22
±0.21
Naphthalene
60/60/60
41.85
±9.91
35.58
±6.96
50.17
±8.37
37.70
±9.41
41.12
±4.36
61/61/61
29.91
±9.78
29.11
±4.41
35.34
±7.37
38.06
±8.39
33.05
±3.80
Nickel (PMio)
60/60/60
1.79
±0.58
1.01
±0.15
0.85
±0.13
0.68
±0.19
1.08
±0.19
58/58/58
0.98
±0.36
1.04
±0.42
1.31
±0.39
6.93
± 10.04
2.52
±2.36
-------
Observations for BOMA from Table 13-3 include the following:
• Naphthalene is the pollutant with the highest annual average concentrations among
BOMA's pollutants of interest. Among the metals, nickel has the highest annual
average concentrations, followed by arsenic and then cadmium.
• Concentrations of naphthalene measured at BOMA range from 12.8 ng/m3 to
93.9 ng/m3. Concentrations of naphthalene appear higher in 2015 than in 2016, based
on the quarterly and annual average concentration shown in Table 13-3. A review of
the data shows that the number of naphthalene concentrations greater than 50 ng/m3 is
considerably less for 2016 (3) than 2015 (17). However, the confidence intervals
indicate that the difference is not statistically significant.
• Concentrations of arsenic measured at BOMA range from 0.001 ng/m3 to 5.83 ng/m3,
which is the third highest arsenic concentration measured across the program. The
highest and second highest (1.83 ng/m3) arsenic concentrations measured at BOMA
were measured on back-to-back sample days in 2015 (June 29 and July 5). The
effects of the maximum concentration are evident in the second quarter average
concentration for 2015 (0.84 ± 0.75 ng/m3). The confidence interval for this quarterly
average is nearly the same magnitude as the concentration itself, indicating a high
level of variability, outliers, or both. Both the minimum and maximum arsenic
concentrations were measured at BOMA during this calendar quarter. Further,
BOMA's minimum arsenic concentration is the lowest measured detection of this
pollutant across the program (and only three arsenic measurements across the
program are less than 0.01 ng/m3, plus the five non-detects).
• Concentrations of nickel measured at BOMA range from 0.314 ng/m3 to 69.5 ng/m3,
which is the maximum nickel concentration measured across the program. This
measurement is more than three times greater than the next highest concentration
measured at a NMP site sampling PMio metals and seven times greater than the next
highest concentration measured at BOMA (both of which were measured in
November 2016). The effects of this outlier on the quarterly and annual average
concentrations is evident in Table 13-3. The confidence interval for the fourth quarter
average for 2016 is greater than the average itself and the confidence interval for the
annual average is similar in magnitude to the average itself.
• Concentrations of cadmium measured at BOMA range from 0.021 ng/m3 to
5.70 ng/m3, which is the second highest cadmium concentration measured across the
program. The fourth highest cadmium concentration across the program was also
measured at BOMA (2.43 ng/m3). BOMA is one of only three NMP sites at which
cadmium concentrations greater than 2 ng/m3 were measured. The calendar quarters
in which these higher concentrations were measured can be seen in the quarterly
averages in Table 13-3. The quarterly and annual average concentrations of cadmium
for 2015 are all around 0.10 ng/m3. This is also true for the first and third quarter
averages for 2016, while the second and fourth quarter averages for 2016 are several
times higher and have confidence intervals similar to or greater than the average itself
(and the confidence interval for the annual average for 2016 is also similar in
magnitude to the average itself). Of the 11 nickel concentrations measured at BOMA
greater than 2 ng/m3, eight were measured in 2016.
13-9
-------
• Table 4-12 presents the NMP sites with the 10 highest annual average concentrations
for the only program-level PAH pollutants of interest and Table 4-13 presents the
NMP sites with the 10 highest annual average concentrations for the only program-
level speciated metals pollutant of interest (arsenic). BOMA does not appear in either
table for naphthalene or arsenic.
13.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for each of the pollutants
listed in Table 13-3 for BOMA. Figures 13-3 through 13-6 overlay the site's minimum, annual
average, and maximum concentrations for each year onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
Figure 13-3. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
BOMA
012345678
Concentration (ng/m3)
&
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 13-3 presents the box plot for arsenic for BOMA and shows the following:
• The range of arsenic concentrations measured in 2015 appears considerably larger
than the range of arsenic concentrations measured in 2016. If the maximum
concentration of arsenic measured at BOMA in 2015 is excluded, the range for 2015
would still be larger than the range measured in 2016, but the difference between the
two would be much smaller (the second highest concentration measured in 2015 is
1.83 ng/m3). The maximum arsenic concentration measured at BOMA is the third
highest arsenic concentration measured across the program.
• Both annual average arsenic (PMio) concentrations for BOMA are less than the
program-level average concentration (0.70 ng/m3); the annual average for 2015 is the
same as the program-level median concentration (0.55 ng/m3), while the annual
average for 2016 is slightly less than the program-level median.
13-10
-------
Figure 13-4. Program vs. Site-Specific Average Cadmium (PMio) Concentrations
BOMA
I-
Program Max Concentration =5.99 ng/m3
1
. n
*
1 1
-
y i
r "i
BOMA 2016 Max = 5.70 ng/m3
i j
0.25
0.5
0.75 1
Concentration (ng/m3]
1.25
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Averag
¦ ~ ~ ~ i
Site: 2015 Average 2016 Average Concentration Range, 2015 & 2016
o
Figure 13-4 presents the box plot for cadmium for BOMA and shows the following:
• The program4evel maximum cadmium concentration (5.99 ng/m3) is not shown
directly on the box plot because the scale of the box plot would be too large to readily
observe data points at the lower end of the concentration range. Thus, the scale of the
box plot has been reduced to 2 ng/m3. Since the maximum cadmium concentration
measured at BOMA is also greater than the scale of the box plot, this concentration is
also labeled in the box plot.
• While not the maximum concentration measured across the program, the maximum
cadmium concentration measured at BOMA is just slightly less, and is the second
highest cadmium concentration measured across the program. The second highest
cadmium concentration measured at BOMA (2.43 ng/m3) was also measured in 2016,
and also exceeds the scale of the box plot.
• BOMA's annual average concentrations of cadmium for 2015 is less than the
program4evel average concentration (0.12 ng/m3), while the annual average
concentrations for 2016 is nearly twice the program-level average.
Figure 13-5. Program vs. Site-Specific Average Naphthalene Concentrations
50
100
150 200 250 300
Concentration (ng/m3)
350
400
450
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
13-11
-------
Figure 13-5 presents the box plot for naphthalene for BOMA and shows the following:
• The range of naphthalene concentrations measured at BOMA in 2015 is fairly similar
to the range of naphthalene concentrations measured at BOMA in 2016, both of
which are considerably smaller than the range of concentrations measured across the
program.
• The annual average naphthalene concentration for 2015 is just slight higher than the
annual average naphthalene concentration for 2016, and both are less than the
program-level average concentration (61.23 ng/m3) and the program-level median
concentration (48.90 ng/m3).
Figure 13-6. Program vs. Site-Specific Average Nickel (PMio) Concentrations
r
! Program Max Concentration = 69.5 ng/m3 i
¦ X
11
, ,
BOMA 2016 Max Concentrationn = 69.5 ng/m3
1 1
0 3 6 9 12 15 18 21 24
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 13-6 presents the box plot for nickel for BOMA and shows the following:
• Similar to cadmium, the program-level maximum nickel concentration (69.5 ng/m3) is
not shown directly on the box plot because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plot has been reduced. Since the maximum nickel concentration
measured at BOMA is the maximum nickel concentration measured across the
program, this is also labeled in the box plot.
• BOMA's maximum nickel concentration of 69.5 ng/m3 is the highest nickel
concentration measured across the program. The next two highest nickel
concentrations measured at BOMA (8.40 ng/m3 and 7.20 ng/m3, also measured in
2016), are an order of magnitude less.
• The annual average concentration of nickel for 2015 is similar to the program-level
average concentration of 1.09 ng/m3. The annual average concentration for 2016 is
more than twice the program-level average concentration.
13-12
-------
13.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
BOMA has sampled PMio metals under the NMP since 2003 and PAHs since 2008. Thus,
Figures 13-7 through 13-10 present the 1-year statistical metrics for each of the pollutants of
interest for BOMA. The statistical metrics presented for assessing trends include the substitution
of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of sampling is
required for inclusion in the trends analysis; in these cases, a 1-year average concentration is not
provided, although the range and percentiles are still presented. Note that the concentrations
from collocated analyses are averaged together for this trends analysis, as indicated in
Section 3.4.2.2, and may result in some differences from the previous sections. This is
particularly true for BOMA metals, where collocated samples account for more than half of the
sampling events at this site.
Figure 13-7. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
BOMA
I
X
I
X
_
4.
LjJ
•o
o-
2
-O"
*
•o-
2004
2005
2006
2007
2008
2009
2010
Year
2011
2012
X
£•
Lrl
i
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
13-13
-------
Observations from Figure 13-7 for arsenic concentrations measured at BOMA include the
following:
• Although sampling for PMio metals under the NMP began in 2003, data from that
year were excluded from this analysis because sampling did not begin until October.
In addition, samples were not collected during portions of April, May, September,
and October 2004. Because a full year's worth of data is not available for 2004, a
1-year average concentration is not presented, although the range of measurements is
provided.
• The maximum arsenic concentration shown was measured on July 5, 2008
(5.45 ng/m3). Three additional concentrations greater than 2 ng/m3 have been
measured at BOMA, one each in 2004, 2006, and 2015.
• The 1-year average concentrations of arsenic have fluctuated over the years, ranging
from 0.36 ng/m3 (2010) to 0.61 ng/m3 (2008). For 2008, the maximum concentration
is driving the 1-year average upward, which is evident from the median
concentration, which hardly changed between 2007 and 2008, even though the
smallest range of measurements was collected in 2007. If the maximum concentration
for 2008 was removed from the dataset, the 1-year average concentration for 2008
would decrease from 0.61 ng/m3 to 0.53 ng/m3, making the changes in the 1-year
averages between 2007 and 2009 more subtle.
• All of the statistical metrics exhibit a decrease from 2008 to 2009 and again for 2010,
when both central tendency parameters are at a minimum. Conversely, all of the
statistical metrics exhibit an increase from 2010 to 2011 and again for 2012.
• For 2013, a higher number of concentrations at the lower end of the concentration
range were measured while concentrations at the top of the range changed little. The
number of arsenic concentrations less than 0.25 ng/m3 increased from one in 2012 to
16 for 2013. This is explains the considerable decrease in the minimum and 5th
percentile shown for 2013, as well as the slight decreases in the 1-year average and
median concentrations. The number of arsenic concentrations less than 0.25 ng/m3
also increased for 2014 (18) and a non-detect was measured for the first (and only)
time. The maximum arsenic concentration measured at BOMA in 2014 is less than
1 ng/m3 for the first time since the onset of sampling.
• Most of the statistical parameters exhibit increases for 2015, although the median did
not change. Even the minimum concentration increased, as there were no non-detects
measured in 2015, although this is difficult to discern, as this is the lowest measured
detection of arsenic.
• The differences in the statistical parameters for 2015 and 2016 are greatly reduced if
the minimum and maximum concentrations measured in 2015 are removed from the
dataset.
13-14
-------
Figure 13-8. Yearly Statistical Metrics for Cadmium (PMio) Concentrations Measured at
BOMA
£
Maximum Averaged
Concentration for
2016 is 6.87 ng/m3
o.
I
ri-,
¦o
& 8 4 a 4-
[£i
2010 2011
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 13-8 for cadmium concentrations measured at BOMA include
the following:
• Cadmium concentrations of higher magnitude were measured often during the early
years of sampling. Cadmium concentrations greater than 0.5 ng/m3 account for more
than 30 percent of measurements in 2004 and 2005, and account for approximately
five percent or less for each of the following years, including several years when none
were measured.
• A significant decreasing trend is shown between 2005 and 2007, which is followed an
increasing trend through 2009. A second decreasing trend is shown afterward and,
with the exception of 2012, continued through 2014. If the maximum concentration
measured in 2012 was removed from the dataset, the 1-year average concentration
would have a continuous decreasing trend over this 5-year period. Only slight
changes are shown for 2015.
• While the maximum concentration measured in 2016 is driving the 1-year average
concentration upward, it is not the sole reason for the increases shown, as the 95th
percentile is at its highest since 2010. Two of the three highest cadmium
concentrations were measured at BOMA in 2016. However, the median cadmium
concentration is at a minimum over the period of sampling. The number of cadmium
concentrations less than 0.06 ng/m3 has increased at BOMA in recent years; three or
13-15
-------
fewer of these concentrations were measured prior to 2013. By 2016, concentrations
less than 0.06 ng/m3 account for 45 percent of the measurements.
Figure 13-9. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
BOMA
300
"E
c
m
c
O
u
100
50
—
o-
L-2-
LjJ
L2-
I
L2-
L-2—
L-1—1
2008 1 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile •••~•••Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2008.
Observations from Figure 13-9 for naphthalene concentrations measured at BOMA
include the following:
• BOMA began sampling PAHs under the NMP in May 2008. Because a full year's
worth of data is not available for 2008, a 1-year average concentration is not
presented, although the range of measurements is provided.
• The maximum naphthalene concentration (251 ng/m3) was measured at BOMA on
the very first sample day (May 6, 2008), although a similar measurement was also
collected in 2012 (235 ng/m3). Two additional concentrations greater than 200 ng/m3
have been measured at BOMA (another in 2008 and one in 2009).
• The median concentration exhibits a considerable decrease from 2008 to 2009, which
is the first full-year of sampling. This is due to the increase of naphthalene
concentrations at the lower end of the concentration range. The number of
naphthalene concentrations less than 50 ng/m3 increased from six measured in 2008
to 23 measured in 2009, accounting for nearly 40 percent of the samples collected in
2009 (compared to 16 percent for 2008).
13-16
-------
• Beginning with the first full-year of sampling, the difference between the 5th and
95th percentiles (the range of concentrations within which 90 percent of the
measurements lie) decreased each year through 2011. The range increased somewhat
for 2012 and is more similar to the range shown for 2010, before decreasing further
each year afterward and reaching a minimum in 2016.
• Naphthalene concentrations have a significant decreasing trend at BOMA. Excluding
2012, the 1-year average concentration has decreased each year at BOMA, decreasing
from 70.33 ng/m3 for 2009 to 33.05 ng/m3 for 2016. (If the maximum concentration
measured in 2012 was excluded from the dataset, the 1-year average concentration
would exhibit little change from 2011 to 2012.)
Figure 13-10. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at
BOMA
Maximum Averaged
Concentration for 2016
is 67.8 ng/m3
£
L-J-l
I
I
"O-
o
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because there were breaks in sampling during portions of 2004.
Observations from Figure 13-10 for nickel concentrations measured at BOMA include
the following:
• The maximum nickel concentration measured at BOMA prior to 2016 is 17.2 ng/m3
(measured in 2004). Nickel concentrations two and four times greater than this
measurement were measured at BOMA in 2016.
13-17
-------
• A steady decreasing trend in nickel concentrations measured at BOMA is shown
through 2010. Concentrations for 2011 increased just slightly, returning to 2009
levels. Even with the higher concentrations measured in 2012 and 2013, the 1-year
average concentration did not change significantly from 2011. Between 2009 and
2013, less than 0.2 ng/m3 separates the 1-year average concentrations.
• Considerable increases, however, are shown for 2014, as all of the statistical
parameters exhibit increases, except the minimum concentration, which did not
change. Three nickel concentrations measured in 2014 are greater than the maximum
nickel concentration measured in 2013. Further, the number nickel concentrations
greater than 2 ng/m3 measured at BOMA in 2014 is at its highest since 2008.
• All of the statistical parameters exhibit decreases for 2015, when the 1-year average
concentration is at a minimum. The number of nickel concentrations greater than
2 ng/m3 measured at BOMA in 2015 is at a minimum over the years of sampling.
• All of the statistical parameters exhibit increases for 2016, when the 1-year average
concentration is at a maximum. The maximum nickel concentrations increased by an
order of magnitude from 2015 to 2016 and the 1-year average and 95th percentile
more than doubled. Yet, the median concentration changed little, indicating that the
measurements on the higher end of the concentration range are driving some of these
changes. If the two highest concentrations were removed from the dataset, the
statistical parameters would still exhibit increases for 2016, though they would be
more subtle.
13.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the BOMA monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
13-18
-------
13.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BOMA, risk was examined by calculating cancer risk
and noncancer hazard approximations for each year annual average concentrations could be
calculated. These approximations can be used as risk estimates for cancer and noncancer effects
attributable to the pollutants of interest. Although the use of these approximations is limited, they
may help identify where policy-makers want to shift their air-monitoring priorities. Refer to
Section 3.4.2.3 for an explanation of how cancer risk and noncancer hazard approximations are
calculated and what limitations are associated with them. Annual averages, cancer UREs and/or
noncancer RfCs, and cancer risk and noncancer hazard approximations are presented in
Table 13-4, where applicable. Cancer risk approximations are presented as probabilities while
the noncancer hazard approximations are ratios and thus, unitless values.
Observations for BOMA from Table 13-4 include the following:
• Among the pollutants of interest for BOMA, naphthalene has the highest annual
average concentrations while cadmium has the lowest annual average concentrations.
• Arsenic has the highest cancer risk approximations among the pollutants of interest
for BOMA, both of which are greater than 2 in-a-million (2.38 in-a-million for 2015
and 2.08 in-a-million for 2016).
• None of the pollutants of interest for BOMA have noncancer hazard approximations
greater than 1.0; in fact, none of the pollutants of interest have noncancer hazard
approximations greater than 0.05. This indicates that adverse noncancer health effects
are not expected due to these individual pollutants.
13-19
-------
Table 13-4. Risk Approximations for the Massachusetts Monitoring Site
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Boston, Massachusetts - BOMA
Arsenic (PMio)
0.0043
0.000015
60/60
0.55
±0.19
2.38
0.04
58/58
0.48
±0.08
2.08
0.03
Cadmium (PMio)
0.0018
0.00001
60/60
0.09
±0.01
0.16
0.01
58/58
0.22
±0.21
0.39
0.02
Naphthalene
0.000034
0.003
60/60
41.12
±4.36
1.40
0.01
61/61
33.05
±3.80
1.12
0.01
Nickel (PMio)
0.00048
0.00009
60/60
1.08
±0.19
0.52
0.01
58/58
2.52
±2.36
1.21
0.03
-------
13.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 13-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 13-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 13-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
BOMA, as presented in Table 13-4. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 13-5. Table 13-6 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 13.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
13-21
-------
Table 13-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Massachusetts Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Boston, Massachusetts (Suffolk County) - BOMA
Formaldehyde
128.17
Formaldehyde
1.67E-03
Arsenic
2.38
Benzene
114.93
Benzene
8.96E-04
Arsenic
2.08
Acetaldehyde
59.29
1,3-Butadiene
6.17E-04
Naphthalene
1.40
Ethylbenzene
54.59
Naphthalene
3.32E-04
Nickel
1.21
1.3 -Butadiene
20.57
POM, Group 2b
2.10E-04
Naphthalene
1.12
Naphthalene
9.77
Ethylene oxide
1.84E-04
Nickel
0.52
POM, Group 2b
2.39
Arsenic, PM
1.70E-04
Cadmium
0.39
Trichloroethylene
1.79
Ethylbenzene
1.36E-04
Cadmium
0.16
POM, Group 2d
1.50
POM, Group 2d
1.32E-04
Nickel, PM
0.23
Acetaldehyde
1.30E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 13-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Massachusetts Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Boston, Massachusetts (Suffolk County) - BOMA
Toluene
455.19
Acrolein
586,790.86
Arsenic
0.04
Methanol
340.95
Formaldehyde
13,078.15
Arsenic
0.03
Xylenes
198.99
1.3 -Butadiene
10,286.36
Nickel
0.03
Formaldehyde
128.17
Acetaldehyde
6,588.00
Cadmium
0.02
Benzene
114.93
Benzene
3,831.01
Naphthalene
0.01
Hexane
88.55
Naphthalene
3,256.67
Nickel
0.01
Acetaldehyde
59.29
Arsenic, PM
2,639.31
Naphthalene
0.01
Ethylbenzene
54.59
Nickel, PM
2,598.79
Cadmium
0.01
Ethylene glycol
53.76
Cadmium, PM
2,447.43
Glycol ethers, gas
28.14
Xylenes
1,989.86
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 13-5 include the following:
• Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Suffolk County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.
• Eight of the highest emitted pollutants in Suffolk County also have the highest
toxicity-weighted emissions, with formaldehyde and benzene ranking highest on both
emissions-based lists.
• Naphthalene is the only pollutant of interest for BOMA to appear on both emissions-
based lists, ranking sixth for quantity emitted and fourth for its toxicity-weighted
emissions.
• Among the metal pollutants of interest, nickel ranks 10th highest for its total
emissions and arsenic ranks seventh for its toxicity-weighted emissions. Cadmium
appears on neither list (ranking 21st for total emissions and 14th for its toxicity-
weighted emissions).
• POM, Group 2b appears on both emissions-based lists in Table 13-5. POM, Group 2b
includes several PAHs sampled for at BOMA including acenaphthene and fluorene,
both of which failed screens. POM, Group 2d appears on both emissions-based lists;
POM, Group 2d does not include any PAHs sampled for at BOMA.
Observations from Table 13-6 include the following:
• Toluene, methanol, and xylenes are the highest emitted pollutants with noncancer
RfCs in Suffolk County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
• Four of the highest emitted pollutants in Suffolk County also have the highest
toxicity-weighted emissions.
• All four of BOMA's pollutants of interest appear among the pollutants with the
highest toxicity-weighted emissions for Suffolk County, although none of these
appear among the highest emitted pollutants (with noncancer RfCs).
13.5 Summary of the 2015-2016 Monitoring Data for BOMA
Results from several of the data analyses described in this section include the following:
~~~ Nine pollutants failed screens for BOMA, with arsenic and naphthalene accounting
for a majority of the failed screens.
~~~ Among the pollutants of interest for BOMA, naphthalene had the highest annual
average concentration for both 2015 and 2016.
13-24
-------
Some of the highest concentrations of arsenic, nickel, and cadmium across the
program were measured at BOMA, including the maximum nickel concentration. For
nickel and cadmium, these concentrations were the highest measured since the onset
of sampling.
Naphthalene concentrations have significantly decreased in magnitude at BOMA.
Arsenic has the highest cancer risk approximations among the pollutants of interest
for BOMA. None of the pollutants of interest for BOMA have noncancer hazard
approximations greater than an HQ of 1.0.
13-25
-------
14.0 Site in Michigan
This section summarizes those data from samples collected at the NATTS site in
Michigan and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and
2016 monitoring efforts. This section also examines the
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
Sections 1 through 4 for detailed discussions and
definitions regarding the various data analyses presented below.
14.1 Site Characterization
This section characterizes the Michigan monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient measurements.
The DEMI monitoring site is located in the Detroit-Warren-Dearborn, Michigan CBSA.
Figure 14-1 presents the composite satellite image retrieved from ArcGIS Explorer showing the
monitoring site and its immediate surroundings. Figure 14-2 identifies nearby point source
emissions locations by source category, as reported in the 2014 NEI for point sources, version 1.
Note that only sources within 10 miles of the site are included in the facility counts provided in
Figure 14-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference but have been
grayed out to emphasize the emissions sources within the boundary. Table 14-1 provides
supplemental geographical information such as land use, location setting, and locational
coordinates. Each figure and table is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. r
14-1
-------
Figure 14-1. Dearborn, Michigan (DEMI) Monitoring Site
Patton Memorial Park
Lapeer Pi
ISource: USGS^a <£
Source: NASA, NGA, USGS
: © 2008 Microsoft Corp..
4^
to
-------
Figure 14-2. NEI Point Sources Located Within 10 Miles of DEMI
Macomb
County
Oakland
County
Wayne
County
DEMI NATTS site O 10 mile radius County boundary
Source Category Group (No. of Facilities)
"t* Airport/Airline/Airport Support Operations (10)
$ Asphalt Production/Hot Mix Asphalt Plant (3)
0 Auto Body Shop/Painters/Automotive Stores (1)
« Automobile/Truck Manufacturing Facility (3)
B Bulk Terminal/Bulk Plant (9)
C Chemical Manufacturing Facility (5)
K Coke Battery (1)
1 Compressor Station (1)
# Electricity Generation Facility (7)
E Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (4)
F Food Processing/Agriculture Facility (2)
I Foundry, Iron and Steel (1)
Industrial Machinery or Equipment Plant (2)
O Institution (school, hospital, prison, etc.) (10)
(¦) Metal Can, Box, and Other Metal Container Manufacturing Facility (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (5)
<•> Metals Processing/Fabrication Facility (12)
X Mine/Quarry/Mineral Processing Facility (12)
? Miscellaneous Commercial/Industrial Facility (1)
m Municipal Waste Combustor (1)
~ Paint and Coating Manufacturing Facility (1)
-$• Petroleum Products Manufacturing Facility (1)
2 Petroleum Refinery (1)
<=> Pharmaceutical Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (1)
P Printing/Publishing/Paper Product Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (1)
V Steel Mill (1)
^ Testing Laboratory (1)
1 Wastewater Treatment Facility (1)
W Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
CANADA
83°10'0"W 83"5'0"W 83 WW 82"55,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Lake
St. Clair
i 1
83"25'0"W 83"20'0"W
Legend
'p ^>co
14-3
-------
Table 14-1. Geographical Information for the Michigan Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average
Daily
Traffic1
Intersection
Used for
Traffic Data
DEMI
26-163-0033
Dearborn
Wayne
Detroit-Warren-
Dearborn, MI
42.306674,
-83.148754
Industrial
Suburban
86,600
1-94 between Ford Plant and
Rotunda Dr
1AADT reflects 2015 data (MI DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
-------
DEMI is located in the parking lot of Salina Elementary School in Dearborn, just
southwest of Detroit, and is the Detroit NATTS site. The surrounding area is both suburban and
industrial in nature. Figure 14-1 shows that a freight yard is located just west of the site and a
residential neighborhood is located to the east. Industrial sources such as automobile and steel
manufacturing facilities are also located in the vicinity. The monitoring site lies between two
heavily traveled roadways, 1-75 (1.4 miles to the east) and 1-94 (1.2 miles to the west).
Figure 14-2 shows that DEMI is surrounded by numerous point sources. A cluster of
sources is located just west of DEMI. Another cluster of sources is located farther south. The
source categories with the most point sources within 10 miles of the site include the airport
source category, which includes airports and related operations as well as small runways and
heliports, such as those associated with hospitals or television stations; bulk terminals and bulk
plants; mines, quarries, and mineral processing facilities; and institutional facilities (schools,
prisons, and/or hospitals). Although difficult to discern in Figure 14-2, the closest sources to
DEMI are just west of the site and include a steel mill and an automobile/truck manufacturing
facility, part of which can be seen in the left-hand side of Figure 14-1, as well as a facility
generating electricity via combustion, a metal coatings facility, and a rail yard. DEMI is located
approximately 3 miles from the Canadian border; emission sources information is not provided
for Canada.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 14-1 also contains traffic volume information for DEMI as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume near DEMI is 86,600 and ranks 13th highest among NMP sites. Traffic for DEMI is
provided for 1-94, between the Ford Plant and Rotunda Drive.
14-5
-------
14.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 14-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 14-2. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. VOCs, carbonyl compounds, and PAHs were sampled for at DEMI.
Table 14-2. 2015-2016 Risk-Based Screening Results for the Michigan Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Dearborn, Michigan - DEMI
Naphthalene
0.029
122
123
99.19
12.72
12.72
Acetaldehyde
0.45
121
121
100.00
12.62
25.34
Formaldehyde
0.077
121
121
100.00
12.62
37.96
Benzene
0.13
120
120
100.00
12.51
50.47
Carbon Tetrachloride
0.17
120
120
100.00
12.51
62.98
1.3 -Butadiene
0.03
119
119
100.00
12.41
75.39
1,2-Dichloroethane
0.038
111
112
99.11
11.57
86.97
Fluorene
0.011
36
113
31.86
3.75
90.72
Acenaphthene
0.011
35
115
30.43
3.65
94.37
Ethylbenzene
0.4
35
120
29.17
3.65
98.02
Fluoranthene
0.011
12
123
9.76
1.25
99.27
Benzo(a)pyrene
0.00057
2
122
1.64
0.21
99.48
p-Dichlorobenzene
0.091
2
28
7.14
0.21
99.69
Hexachloro-1,3 -butadiene
0.045
1
2
50.00
0.10
99.79
Propionaldehyde
0.8
1
121
0.83
0.10
99.90
Xylenes
10
1
120
0.83
0.10
100.00
Total
959
1,700
56.41
14-6
-------
Observations from Table 14-2 for DEMI include the following:
• Concentrations of 16 pollutants failed at least one screen for DEMI; 56 percent of
concentrations for these 16 pollutants were greater than their associated risk screening
value (or failed screens).
• Ten pollutants contributed to 95 percent of failed screens for DEMI and therefore
were identified as pollutants of interest for this site. These 10 include two carbonyl
compounds, five VOCs, and three PAHs.
• Five pollutants listed in Table 14-2 failed 100 percent of screens (acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, and 1,3-butadiene), with two additional
pollutants failing 99 percent of screens (naphthalene and 1,2-dichloroethane). Each of
these pollutants contributed between 12 percent and 13 percent to the total number of
failed screens; together these seven pollutants account for nearly 87 of the total failed
screens.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
14.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Michigan monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at DEMI are provided
in Appendices J, M, and N.
14-7
-------
14.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for the Michigan monitoring site, as described in Section 3.1. The quarterly
average concentration of a particular pollutant is simply the average concentration of the
preprocessed daily measurements over a given calendar quarter. Quarterly average
concentrations include the substitution of zeros for all non-detects. A site must have a minimum
of 75 percent valid samples compared to the total number of samples possible within a given
calendar quarter for a quarterly average to be calculated. An annual average concentration
includes all measured detections and substituted zeros for non-detects for an entire year of
sampling. Annual averages were calculated for pollutants where three valid quarterly averages
could be calculated for a given year and where method completeness was greater than or equal to
85 percent, as presented in Section 2.4. Quarterly and annual average concentrations for the
pollutants of interest for the Michigan monitoring site are presented in Table 14-3, where
applicable. Note that concentrations of the PAHs are presented in ng/m3 for ease of viewing.
Also note that if a pollutant was not detected in a given calendar quarter, the quarterly average
simply reflects "0" because only zeros substituted for non-detects were factored into the
quarterly average concentration.
14-8
-------
Table 14-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Michigan Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
Oig/m3)
Q4
Avg
Oig/m3)
Annual
Average
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Dearborn, Michi
£an - DEMI
Acetaldehyde
60/60/60
1.74
±0.26
1.87
±0.27
2.02
±0.26
1.44
±0.32
1.77
±0.14
61/61/61
1.54
±0.15
2.53
±0.41
2.31
±0.26
1.48
±0.29
1.96
±0.18
Benzene
60/60/60
0.77
±0.12
0.71
±0.13
1.02
±0.20
0.69
±0.15
0.80
±0.08
60/60/60
0.65
±0.10
0.65
±0.19
0.79
±0.16
0.65
±0.19
0.68
±0.08
1.3 -Butadiene
60/60/60
0.08
±0.01
0.10
±0.02
0.10
±0.03
0.09
±0.03
0.09
±0.01
59/44/60
0.08
±0.02
0.09
±0.04
0.08
±0.02
0.08
±0.03
0.08
±0.01
Carbon Tetrachloride
60/60/60
0.66
±0.05
0.67
±0.03
0.69
±0.04
0.65
±0.04
0.67
±0.02
60/60/60
0.65
±0.04
0.72
±0.04
0.62
±0.08
0.65
±0.03
0.66
±0.03
1,2 -Dichloroethane
58/53/60
0.08
±<0.01
0.08
±0.01
0.06
±0.01
0.06
±0.01
0.07
±0.01
54/48/60
0.07
±0.01
0.07
±0.01
0.04
±0.02
0.06
±0.01
0.06
±0.01
Ethylbenzene
60/59/60
0.18
±0.04
0.39
±0.11
0.66
±0.28
0.26
±0.07
0.37
±0.09
60/59/60
0.19
±0.06
0.28
±0.08
0.66
±0.40
0.33
±0.13
0.36
±0.11
Formaldehyde
60/60/60
2.75
±0.42
3.83
±0.53
4.25
±0.61
2.49
±0.52
3.33
±0.31
61/61/61
2.17
±0.37
4.24
± 1.00
5.34
±0.82
2.51
±0.46
3.54
±0.47
Acenaphthene3
54/54/62
1.12
±0.74
12.16
±4.06
16.59
±4.30
6.55
±3.25
8.86
±2.14
61/61/61
3.77
±2.07
17.15
±9.10
16.80
±5.46
3.89
± 1.90
10.30
±3.05
Fluorene3
59/59/62
2.17
±0.61
10.52
±2.73
15.46
±3.82
4.74
± 1.53
7.93
± 1.69
54/54/61
2.92
± 1.62
15.13
±7.89
16.26
±4.68
4.41
± 1.86
9.57
±2.69
Naphthalene1
62/62/62
69.79
±9.00
112.68
± 13.75
166.07
± 39.28
118.94
±33.71
116.18
± 15.46
61/61/61
78.67
± 19.91
120.67
±31.54
141.91
±30.98
88.67
±27.52
107.01
± 14.51
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Observations for DEMI from Table 14-3 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde and
acetaldehyde; all other annual average concentrations are less than 1.0 |ig/m3.
• Concentrations of formaldehyde measured at DEMI range from 0.981 |ig/m3 (which
is the only one less than 1 |ig/m3) to 9.08 |ig/m3. The quarterly average concentrations
for DEMI show that formaldehyde concentrations tended to be higher during the
warmer months of the year. All but one of the 18 formaldehyde concentrations greater
than 5 |ig/m3 were measured at DEMI between April and September of either year.
Conversely, only two of the 20 formaldehyde concentrations less than 2 |ig/m3 were
measured during the first or fourth quarters of the year (with the two exceptions
measured in April).
• Concentrations of acetaldehyde measured at DEMI range from 0.760 |ig/m3 to
4.78 |ig/m3. Concentrations of acetaldehyde exhibit a similar pattern in seasonal
tendency as formaldehyde, although the difference is less pronounced, particularly for
2015 compared to 2016.
• Benzene has the highest annual average concentrations among the VOC pollutants of
interest for DEMI. Concentrations of benzene measured at DEMI range from
0.211 |ig/m3 to 1.95 |ig/m3. Most of the quarterly average concentrations of benzene
fall within a relatively small range, with the exception of the third quarter average
concentration for 2015 (1.02 ± 0.20 |ig/m3). A review of the data shows that the
maximum concentration of benzene was measured during this calendar quarter.
Further, seven benzene concentrations greater than 1 |ig/m3 were measured during
this quarter, with three or less measured during each of the other calendar quarters.
• Both third quarter average concentrations of ethylbenzene are approximately twice
the other quarterly averages and have the largest confidence intervals associated with
them. A similar observation was made in the 2014 NMP report. Concentrations of
ethylbenzene measured at DEMI span two orders of magnitude, ranging from
0.0348 |ig/m3 to 3.07 |ig/m3, which is the maximum ethylbenzene concentration
measured across the program. The second highest ethylbenzene concentration
measured at DEMI (2.16 |ig/m3) is the third highest measured across the program.
Both of these concentrations were measured in August, but in different years. The
five highest ethylbenzene concentrations were measured at DEMI in July, August, or
September (three in 2015, two in 2016). In 2015, 11 ethylbenzene concentrations
greater than or equal to 0.5 |ig/m3 were measured, with more than half (7) measured
during the third quarter. In 2016, 13 ethylbenzene concentrations greater than or equal
to 0.5 |ig/m3 were measured, with nearly half (6) measured during the third quarter
(with five measured on back-to-back sample days in late August and September).
Five more concentrations of this magnitude were measured during November of
2016.
• The remaining VOCs exhibit relatively little variability in their quarterly and annual
average concentrations.
14-10
-------
• Based on the quarterly average concentrations shown, concentrations of naphthalene
were significantly lower during the first quarter of 2015. This is also true for 2016,
although to a lesser extent. Concentrations of naphthalene measured at DEMI range
from to 19.4 ng/m3 to 312 ng/m3. All but three of the 30 naphthalene concentrations
less than 70 ng/m3 were measured during the first or fourth quarters of either year.
For 2015, all 11 were measured during these calendar quarters, seven during the first
and four during the fourth. For 2016, there was 19 of these, nine measured during the
first quarter, seven during the fourth, with the three exceptions measured in April.
Only one naphthalene concentration greater than 100 ng/m3 was measured at DEMI
during the first quarter of 2015; the number measured ranged from nine to 11 for the
remaining calendar quarters. For 2016, seven naphthalene concentrations greater than
100 ng/m3 were measured during the first (four) and fourth (three) quarters, while 23
were measured during the second (10) and third (13) quarters.
• Concentrations of fluorene were higher during the warmer months of both sampling
years, as indicated by the quarterly average concentrations. Concentrations of
fluorene measured at DEMI range from 0.770 ng/m3 to 43.8 ng/m3, and include 10
non-detects. Of the 25 fluorene concentrations greater than or equal to 15 ng/m3
measured at DEMI, only one was measured outside the second or third calendar
quarters. Conversely, all but two of the 37 fluorene concentrations less than 4 ng/m3
were measured during the first or fourth quarters of either year. Further, nine of the
10 non-detects were measured during the first or fourth calendar quarters; two were
measured during the first quarter of 2015, with the remainder measured between late
December 2015 and mid-April 2016.
• Quarterly average concentrations of acenaphthene exhibit a similar seasonal
tendency, although these quarterly average concentrations have even larger
confidence intervals associated with them. Concentrations of acenaphthene measured
at DEMI range from 0.673 ng/m3 to 50.0 ng/m3, and include eight non-detects. Note
the relatively low quarterly average concentration for the first quarter of 2015; all
eight non-detects were measured between January and March 2015, with seven on
back-to-back sample days.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for DEMI from
those tables include the following:
• DEMI appears in Table 4-10 for VOCs twice. This site has the ninth and tenth highest
annual average concentrations of carbon tetrachloride among NMP sites sampling this
pollutant; however, with few exceptions, the difference among the annual average
concentrations of this pollutant vary little across the sites.
• DEMI does not appear in Table 4-11 for its annual average concentrations of the
carbonyl compounds.
• DEMI has the highest (2015) and third highest (2016) annual average concentrations
of naphthalene among NMP sites sampling PAHs, as shown in Table 4-12. DEMI is
one of only two sites with an annual average concentration of naphthalene greater
14-11
-------
than 100 ng/m3. DEMI's annual average concentrations of acenaphthene and fluorene
both rank fifth (2016) and sixth (2015) among NMP sites sampling PAHs.
14.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for each of the pollutants
listed in Table 14-3. Figures 14-3 through 14-12 overlay the Michigan site's minimum, annual
average, and maximum concentrations for each year onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
Figure 14-3. Program vs. Site-Specific Average Acenaphthene Concentrations
r "¦1
Proeram Max Concentration = 108 ne/m3 ¦
1 1
1 II
L
I
1
-
1 1
0 15 30 45 60 75
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
e o —
Figure 14-3 presents the box plot for acenaphthene for DEMI and shows the following:
• The program-level maximum concentration (108 ng/m3) of acenaphthene is not
shown directly on the box plot because the scale of the box plot would be too large to
readily observe data points at the lower end of the concentration range. Thus, the
scale has been reduced.
• The maximum acenaphthene concentration measured at DEMI (50 ng/m3) is roughly
half the magnitude of the maximum concentration measured across the program.
• The range of concentrations measured in 2016 at DEMI is larger than the range of
concentrations measured in 2015.
• All eight non-detects of acenaphthene were measured at DEMI in 2015.
• Although the annual average for 2016 is slightly higher than the annual average for
2015, both are more than twice the program-level average concentration (4.36 ng/m3).
14-12
-------
Figure 14-4. Program vs. Site-Specific Average Acetaldehyde Concentrations
^ ZZ
0 2 4 6 8 10 12 14 16 18
Concentration (]ug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 14-4 presents the box plot for acetaldehyde for DEMI and shows the following:
• The range of acetaldehyde concentrations measured at DEMI in 2015 is smaller than
the range measured in 2016, although both are relatively small compared to the range
of concentrations measured across the program.
• Both of DEMI's annual average concentrations of acetaldehyde fall between the
program-level average concentration and the program-level third quartile.
Figure 14-5. Program vs. Site-Specific Average Benzene Concentrations
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
Figure 14-5 presents the box plot for benzene for DEMI and shows the following:
• All benzene concentrations measured at DEMI in 2015 and 2016 are less than
2 |ig/m3.
• The annual average benzene concentrations for this site fall on either side of the
program-level average concentration (0.72 |ig/m3).
14-13
-------
Figure 14-6. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
1
h—
i
| Program
Max Concentration = 3.90 (.ig/m3 !
I
r
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (jug/rn3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 14-6 presents the box plot for 1,3-butadiene for DEMI and shows the following:
• The program4evel maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plot in Figure 14-6 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plot has been reduced.
• The range of 1,3-butadiene concentrations measured at DEMI in 2015 is smaller than
the range measured in 2016, although both are relatively small compared to the range
of concentrations measured across the program. At the low end of the concentration
range, the difference between the two years is a result of the single non-detect
measured in 2016. Excluding this measurement, the minimum concentration for each
year would be similar.
• The annual average concentrations for DEMI are similar to each other, differing by
only 0.01 |ig/m3. These annual averages are is similar to the program-level average
concentration.
Figure 14-7. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
J >
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (|ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
14-14
-------
Figure 14-7 presents the box plot for carbon tetrachloride for DEMI and shows the
following:
• The range of carbon tetrachloride concentrations measured at DEMI is considerably
smaller than the range measured across the program, though fairly similar to the range
measured at the majority of NMP sites during the 2015 and 2016 monitoring efforts.
• DEMI's annual average concentration of carbon tetrachloride for 2015 is similar to
annual average concentration for 2016; both annual averages fall between the
program-level average concentration (0.64 |ig/m3) and the program-level third
quartile (0.69 |ig/m3).
Figure 14-8. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1
m-
,
Program Max Concentration = 45.8 jxg/m3 i
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
Figure 14-8 presents the box plot for 1,2-dichloroethane for DEMI and shows the
following:
• The scale of the box plot in Figure 14-8 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements. Note that the program-level
average is being driven by the measurements at the upper end of the concentration
range, as this average is nearly three times greater than the program-level third
quartile.
• The range of 1,2-dichloroethane concentrations measured at DEMI in 2015 is similar
to the range measured in 2016, with all measurements less than 0.12 |ig/m3, and only
three concentrations greater than the program-level third quartile.
• Both annual average concentrations of 1,2-dichloroethane for DEMI are less than the
program-level median concentration, with the annual average for 2016 also less than
the program-level first quartile.
14-15
-------
Figure 14-9. Program vs. Site-Specific Average Ethylbenzene Concentrations
i-n
0.5
1.5 2
Concentration (|ig/m3)
2.5
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 14-9 presents the box plot for ethylbenzene for DEMI and shows the following:
• The maximum ethylbenzene concentration measured at DEMI in 2016 is the
maximum concentration measured across the program. While the maximum
ethylbenzene concentration measured at DEMI in 2015 is les, it is still the third
highest ethylbenzene concentration measured across the program.
• The annual average concentrations of ethylbenzene for DEMI are similar to each
other and both are greater than both the program-level average and third quartile.
Figure 14-10. Program vs. Site-Specific Average Fluorene Concentrations
1
Program Max Concentration = 105 ng/m3
u
10
20
30
Concentration (ng/m3]
40
50
60
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 14-10 presents the box plot for fluorene for DEMI and shows the following:
• The scale of the box plot has also been reduced to allow for the observation of data
points at the lower end of the concentration range. Note that the program-level first
quartile of fluorene is zero and therefore not visible on the box plot.
• The maximum fluorene concentration measured at DEMI is considerably less than the
maximum concentration measured across the program. The range of concentrations
measured at DEMI in 2016 is larger than the range of concentrations measured in
2015, with the six highest fluorene concentrations measured at DEMI in 2016.
14-16
-------
• Both annual average concentrations of fluorene for DEMI are greater than the
program-level average concentration (4.36 ng/m3) and third quartile; the annual
average for 2016 is more than the twice the program-level average concentration and
ranks fifth highest among NMP sites sampling this pollutant (the annual average for
2015 ranks sixth highest).
Figure 14-11. Program vs. Site-Specific Average Formaldehyde Concentrations
ZD
0 3 6 9 12 15 18 21 24 27
Concentration (]ug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 14-11 presents the box plot for formaldehyde for DEMI and shows the following:
• The range of formaldehyde concentrations measured at DEMI in 2015 is smaller than
the range measured in 2016, although both are relatively small compared to the range
of concentrations measured across the program.
• Both of DEMI's annual average concentrations of formaldehyde fall between the
program-level average concentration and the program-level third quartile.
Figure 14-12. Program vs. Site-Specific Average Naphthalene Concentrations
50
100
150
200 250
Concentration (ng/m3)
300
350
400
450
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Avera
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o
Figure 14-12 presents the box plot for naphthalene for DEMI and shows the following:
• DEMI is one of only three NMP sites at which naphthalene concentrations greater
than 300 ng/m3 were measured.
14-17
-------
• Both annual average concentrations of naphthalene for DEMI are greater than
100 ng/m3; DEMI is the only site for which this is true. DEMI has the highest and
third highest annual averages of naphthalene across the program. DEMI's annual
average concentration for 2015 is just less than twice the program-level average
concentration (61.23 ng/m3).
14.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
DEMI has sampled VOCs and carbonyl compounds under the NMP since 2003 and PAHs since
2008. Thus, Figures 14-13 through 14-22 present the 1-year statistical metrics for each of the
pollutants of interest for DEMI. The statistical metrics presented for assessing trends include the
substitution of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of
sampling is required for inclusion in the trends analysis; in these cases, a 1-year average
concentration is not provided, although the range and percentiles are still presented.
Figure 14-13. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
DEMI
T
fl h ft 1
<
>•
2008 1 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
14-18
-------
Observations from Figure 14-13 for acenaphthene concentrations measured at DEMI
include the following:
• DEMI began sampling PAHs under the NMP in April 2008. Because a full year's
worth of data is not available for 2008, a 1-year average concentration is not
presented, although the range of measurements is provided.
• The maximum acenaphthene concentration (175 ng/m3) was measured at DEMI on
August 18, 2010. Four additional measurements greater than 100 ng/m3 have been
measured at DEMI (two in 2008, another in 2010, and one in 2011).
• Higher concentrations of acenaphthene tended to be measured during the warmer
months of the year. Of the 38 acenaphthene concentrations greater than 30 ng/m3, 32
were measured in June, July, or August of a given year, with all of these
concentrations measured at DEMI between May and September.
• Although most of the statistical metrics increased (at least slightly) from 2009 to
2010, the 1-year average concentration is being driven by the two highest
concentrations measured in 2010 (both greater than 100 ng/m3). The next highest
concentration measured in 2010 is considerably less (55.1 ng/m3). If the two highest
concentrations were excluded from the calculation, the 1-year average concentration
for 2010 would decrease from 13.73 ng/m3 to 9.01 ng/m3, which is similar to the
1-year average concentration for 2009.
• The 95th percentile increased steadily between 2009 and 2011, indicating that
concentrations of "higher" magnitude accounted for an increasing number of
measurements. The number of concentrations greater than 25 ng/m3 increased from
three to four to 10 during this period.
• The maximum, 95th percentile, and 1-year average concentration of acenaphthene
exhibit decreases between 2011 and 2014. Both the 1-year average and median
concentrations are at a minimum for 2014 and the first non-detect was measured in
2014. The 1-year average concentrations have decreased from 13.44 ng/m3 (2011) to
7.66 ng/m3 (2014) during this time. But with relatively large confidence intervals
associated with the 1-year average concentrations, these changes shown are not
statistically significant.
• Most of the changes shown in the years that follow are related to the statistical
parameters representing the concentrations on the upper end of the concentration
range. The minimum and 5th percentile for 2015 are both zero, as the number of non-
detects increased to eight. Non-detects were not measured during any other years of
sampling.
14-19
-------
Figure 14-14. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
DEMI
o
T
t
o.
Lj-I
LjJ
r^i
a
T
T
1
.. v-
L^J
I
p5-|
T
T
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because data from March 2007 to March 2008 were invalidated.
Observations from Figure 14-14 for acetaldehyde concentrations measured at DEMI
include the following:
• Carbonyl compounds have been sampled continuously at DEMI under the NMP since
2003, beginning with a l-in-12 day schedule in 2003 then changing to a l-in-6 day
schedule in the spring of 2004.
• Carbonyl compound samples collected on the primary collection system between
March 13, 2007 and March 25, 2008 were invalidated by the state of Michigan due to
a leak in the sample line. With only 12 valid samples in 2007, no statistical metrics
are provided. Because less than 75 percent of the samples were valid in 2008, a
1-year average is not presented for 2008, although the range of measurements is
provided.
• The maximum acetaldehyde concentration was measured at DEMI in 2004
(7.84 |ig/m3). In total, six concentrations greater than 5 |ig/m3 have been measured at
DEMI, three in 2004, two in 2005, and one in 2006 (and none in the years that
follow).
• The 1-year average concentration exhibits a decreasing trend after 2004 that
continues through 2006. A 1-year average concentration is not available for 2007 or
2008, although the median concentration, which is available for 2008, changed little
from 2006 to 2008, then decreased slightly for 2009. Both the 1-year average and
14-20
-------
median concentrations are at a minimum for 2009 (1.44 |ig/m3 and 1.30 |ig/m3,
respectively).
• The 1-year average concentration has a nearly continuous increasing trend after 2009
through 2014; a slight decrease is shown for 2015, which is followed by an additional
increase for 2016. The 1-year average concentration for 2016 is at its highest since
2005. The median concentration does not follow this exact pattern but is also at its
highest since 2005.
Figure 14-15. Yearly Statistical Metrics for Benzene Concentrations Measured at DEMI
o.
T
T
O
T
o
<
'-a
~
£
1
-8"
^ r
2
£
j"--y g
'-a-
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-15 for benzene concentrations measured at DEMI include
the following:
• VOCs have been sampled continuously at DEMI under the NMP since 2003.
However, the l-in-12 day schedule combined with a number of invalid samples
resulted in low completeness in 2003; as a result, a 1-year average concentration is
not presented for 2003.
• The three highest benzene concentrations were measured at DEMI in 2004 and range
from 5.44 |ig/m3 to 7.62 |ig/m3. Two other concentrations greater than 5 |ig/m3 have
been measured at DEMI, one in 2003 and one in 2007.
14-21
-------
• Both the 1-year average and median concentrations exhibit a steady decreasing trend
between 2004 and 2009. Between 2009 and 2012, the 1-year average concentration
has an undulating pattern and fluctuated between 0.81 |ig/m3 (2009) and 0.94 |ig/m3
(2010).
• A significant decrease in benzene concentrations is shown for 2013, as the smallest
range of benzene concentrations was measured at DEMI in 2013 and all of the
statistical metrics decreased except the minimum concentration. Both the 1-year
average (0.64 |ig/m3) and median (0.59 |ig/m3) concentrations are at a minimum for
2013.
• A slight increasing trend in benzene concentrations is shown between 2013 and 2015,
before decreasing again for 2016. However, the central tendency parameters
calculated for each of these years are still less than those calculated for the years 2012
and prior.
Figure 14-16. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
DEMI
I
S
2.
1
i
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-16 for 1,3-butadiene concentrations measured at DEMI
include the following:
• The maximum 1,3-butadiene concentration (1.04 |ig/m3) was measured on
October 18, 2004 and is the only 1,3-butadiene concentration greater than 1 |ig/m3
measured at DEMI, although additional concentrations greater than 0.90 |ig/m3 were
measured in 2004 and 2006.
14-22
-------
For 2004, the minimum, 5th percentile, and median concentrations are all zero,
indicating that at least half of the measurements were non-detects. Yet, two of the
three highest concentrations were also measured at DEMI in 2004; in addition, the
95th percentile is at a maximum for 2004. This indicates there is a high level of
variability within these measurements.
Fewer non-detects were measured in 2005 and 2006, as indicated by the increase in
the median concentration, and even fewer in most of the years that follow, as
indicated by the increase in the 5th percentile. The percentage of non-detects
decreased from a high of 60 percent in 2004 to 2 percent in 2008, then fluctuated
between 2 percent and 8 percent for the years that follow until 2014, when non-
detects were not measured. Non-detects were not measured in 2015 either, and a
single non-detect was measured in 2016.
Even as the number of non-detects decreased (and thus, the number of zeros factored
into the calculation decreased), the 1-year average concentration decreased by almost
half between 2006 and 2009. This was followed by an increasing trend between 2009
and 2012, returning the 1-year average concentration to near 2006 levels.
The 1-year average concentration decreased significantly from 2012 to 2013, as did
the median, both of which are at their lowest since 2009. The number of
1,3-butadiene concentrations greater than 0.1 |ig/m3 decreased considerably between
the two years, from 27 measured in 2012 to 10 in 2013.
All of the statistical metrics exhibit increases for 2014. Although decreases in the
1-year average concentration are shown between 2014 and 2016, confidence intervals
calculated indicate that the changes exhibited over the last several years of sampling
are not statistically significant.
14-23
-------
Figure 14-17. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at DEMI
I
o..
I
UjJ
t
LjJ
T
rh
i
I
2003 1 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-17 for carbon tetrachloride concentrations measured at
DEMI include the following:
• In 2003, measured detections ranged from 0.25 |ig/m3 to 0.76 |ig/m3, plus two non-
detects. This is the only year of sampling for which nearly half the measurements
were less than 0.6 |ig/m3.
• The range of concentrations measured in 2004 doubled from 2003 levels. The number
of measurements greater than 1 |ig/m3 increased from none in 2003 to 12 for 2004.
• The 1-year average concentration decreased by more than 0.1 |ig/m3 from 2004 to
2005, as the range of concentrations measured decreased substantially. Little change
in the 1-year average concentration is shown from 2005 to 2007, despite the
differences in the ranges of concentrations measured.
• With the exception of the 5th percentile, all of the statistical metrics increased for
2008, with the 1-year average and median concentrations for 2008 similar to the 95th
percentile for 2007.
• A steady decreasing trend in the 1-year average concentration is shown between 2008
and 2011. Between these years, the majority of concentrations fell within a tighter
concentration range, as indicated by the difference between the 5th and 95th
percentiles.
14-24
-------
• The difference between the 5th and 95th percentiles for 2012 is less than 0.25 |ig/m3,
which is the smallest difference up to this point, yet an increase in the 1-year average
and median concentrations is shown for 2012. The number of carbon tetrachloride
concentrations falling between 0.7 |ig/m3 and 0.9 |ig/m3 more than doubled from
2011 (13) to 2012 (32), accounting for more than half of the measurements for 2012.
All of the statistical parameters exhibit a slight decrease from 2012 to 2013, as the
number of carbon tetrachloride concentrations falling between 0.7 |ig/m3 and
0.9 |ig/m3 decreased by half (16).
• The smallest range of carbon tetrachloride concentrations was measured in 2014,
spanning just over 0.25 |ig/m3. In addition, the majority of concentrations measured
in 2014 fall into the tightest range of concentrations measured, as indicated by the
difference between the 5th and 95th percentiles. Despite this tightening of
measurements, little change is shown in the central tendency statistics for 2014. In
fact, the central tendency parameters changed little over the last four years of
sampling.
Figure 14-18. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at DEMI
0.20
0.18
0.16
0.14
0.12
J
c
o
2 0.10
c
CD
c
o
U
0.08
0.06
0.04
0.02
0.00
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented due to low completeness for 2003.
Observations from Figure 14-18 for 1,2-dichloroethane concentrations measured at
DEMI include the following:
• There were no measured detections of 1,2-dichloroethane in 2003, 2004, 2007, or
2008. Through 2011, the median concentration is zero for all years, indicating that at
least half of the measurements are non-detects: there was only one measured
Maxi mum Concentration
for 2006 is 3.44 |j.g/m3
6
.
T
o
w
V
¦ o -
/
•<>
0
2003 2004 2005 2006 2007 2008 2009 Year 2010 2011 2012 2013 2014 2015 2016
14-25
-------
detection in 2005, three in 2006, four in 2009, 12 in 2010, and 11 in 2011. The
number of measured detections increased by a factor of five for 2012, with a similar
number for each year after.
• The maximum 1,2-dichloroethane concentration was measured in 2006 (3.44 |ig/m3);
no other 1,2-dichloroethane concentration measured at DEMI is greater than
0.2 |ig/m3. The magnitude of this outlier explains why the 1-year average
concentration for 2006 is five times greater than the 95th percentile (there were only
three measured detections in 2006).
• As the number of measured detections increase, so do each of the corresponding
statistical metrics shown in Figure 14-18. The 1-year average concentration increased
significantly from 2011 to 2012, when the number of non-detects fell from 50 to 10.
This number decreased by half again for 2013, when the 1-year average and median
concentrations were at a maximum.
• A slight decreasing trend is shown in the central tendency parameters between 2013
and 2016.
Figure 14-19. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
DEMI
5.0
4.5
20031 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented due to low completeness for 2003.
14-26
-------
Observations from Figure 14-19 for ethylbenzene concentrations measured at DEMI
include the following:
• The maximum ethylbenzene concentration was measured at DEMI in September
2004 (4.35 |ig/m3). Three additional ethylbenzene concentrations greater than
3 |ig/m3 have been measured at DEMI (one each in 2011, 2012, and 2016). In total,
13 concentrations greater than 2 |ig/m3 have been measured at DEMI.
• A steady decreasing trend in the 1-year average concentration is shown after 2004,
although the rate of decrease levels out after 2006, with the 1-year average reaching a
minimum for 2008 (0.30 |ig/m3). Little change is shown for 2009.
• Increasingly higher maximum ethylbenzene concentrations were measured at DEMI
after 2008 and continues through 2012. The 1-year average concentration increases
significantly through 2011, then decreases slightly for 2012, despite the large range of
concentrations measured. While the maximum concentration increased for 2012, the
minimum concentration decreased (and one non-detect was measured). The number
of ethylbenzene concentrations at the lower end of the concentration range (those less
than 0.25 |ig/m3) nearly doubled from 2011 to 2012 (up from 11 to 19), resulting in
the slight decreases shown in the central tendency statistics for 2012, despite a few
higher concentrations measured.
• This decreasing trend continued through 2014. Concentrations less than 0.25 |ig/m3
account for an even greater percentage of the measurements in 2013 and 2014,
accounting for more than 40 percent of the measurements for 2013 and more than half
for 2014.
• Relatively little change is shown in the central tendency parameters between 2014
and 2016, aside from the increasing maximum concentration. Ethylbenzene
concentrations less than 0.25 |ig/m3 continued to account for around half of the
measurements during this time. Between 2013 and 2016, each of the 1-year average
ethylbenzene concentrations fell between 0.35 |ig/m3 and 0.40 |ig/m3.
14-27
-------
Figure 14-20. Yearly Statistical Metrics for Fluorene Concentrations Measured at DEMI
2012
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 14-20 for fluorene concentrations measured at DEMI include
the following:
• The trends graph for fluorene resembles the trends graph for acenaphthene
(Figure 14-13) in a number of ways.
• The maximum fluorene concentration (152 ng/m3) was measured at DEMI on
August 18, 2010, the same day that the maximum acenaphthene concentration was
measured. Two additional measurements greater than 100 ng/m3 have been measured
at DEMI, with all three measured during the month of August (one in 2008 and
another in 2010). All eight concentrations greater than 50 ng/m3 were measured in
June, July, or August of a given year and all 59 concentrations greater than or equal to
20 ng/m3 were measured at DEMI between May and September.
• Although all of the statistical metrics increased (at least slightly) from 2009 to 2010,
the 1-year average concentration is being driven by the two highest concentrations
measured in 2010 (both greater than 100 ng/m3). The next highest concentration
measured in 2010 is considerably less (44.8 ng/m3). If the two highest concentrations
were excluded from the calculation, the 1-year average concentration for 2010 would
decrease from 12.62 ng/m3 to 8.40 ng/m3, which is less than a 1 ng/m3 increase from
2009.
• The 95th percentile increased steadily between 2009 and 2011. The number of
concentrations greater than 25 ng/m3 increased from one to three to seven during this
14-28
-------
period. There were also seven concentrations greater than 25 ng/m3 measured in
2012, even though the 95th percentile exhibits a slight decrease.
• All of the statistical parameters exhibit decreases from 2012 to 2013 and again for
2014 (except the minimum concentration, which did not change). Both the 1-year
average and median concentrations are at a minimum for 2014. These central
tendency parameters both exhibit slight increases for 2015 and 2016.
• Between 2011 and 2016, the median concentrations have varied by less than 1 ng/m3,
ranging from 4.58 ng/m3 (2014) to 5.42 ng/m3 (2012). The 1-year average
concentrations exhibit more variability, ranging from 6.93 ng/m3 (2014) to
11.32 ng/m3 (2012). With the relatively large confidence intervals associated with
these 1-year average concentrations, the changes shown are not statistically
significant.
Figure 14-21. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
DEMI
I
rh
£
1-2^
I
•o
2
d*=
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because data from March 2007 to March 2008 was invalidated.
Observations from Figure 14-21 for formaldehyde concentrations measured at DEMI
include the following:
• Carbonyl compounds have been sampled continuously at DEMI under the NMP since
2003 but due to a leak in the sample line, samples collected between March 13, 2007
through March 25, 2008 were invalidated. With only 12 valid samples in 2007, no
statistical metrics are provided. With most samples collected during the first quarter
14-29
-------
of 2008 invalidated, a 1-year average concentration is not presented for 2008,
although the range of measurements is provided.
The five highest formaldehyde concentrations measured at DEMI were measured in
2005 and ranged from 13.3 |ig/m3 to 33.1 |ig/m3. All nine formaldehyde
concentrations greater than 9 |ig/m3 were measured during the first 3 years of
sampling.
All of the statistical parameters exhibit decreases from 2005 to 2006, with a
significant decrease in the 1-year average concentration, which decreased from
5.35 |ig/m3 to 2.92 |ig/m3. Even though a 1-year average could not be calculated for
2008, the concentration profile changed little from 2006 to 2008.
The range of formaldehyde concentrations measured is at a minimum for 2009; the
majority of formaldehyde concentrations, as indicated by the difference between the
5th and 95th percentiles, fell within the tightest range in 2009, indicating reduced
variability within the measurements in 2009.
Between 2006 and 2011, less than 0.5 |ig/m3 separates the 1-year average
concentrations, which ranged from 2.46 |ig/m3 (2009) and 2.92 |ig/m3 (2006).
All of the statistical parameters exhibit increases for 2012, with the 1-year average
concentration greater than 3 |ig/m3 for the first time since 2005. This is also true for
the median concentration.
The 1-year average formaldehyde concentration has varied between 3 |ig/m3 and
3.5 |ig/m3 over the last five years of sampling at DEMI.
14-30
-------
Figure 14-22. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
DEMI
o
o
2009 2010
2012
Year
O 5th Percentile
2014 2015
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 14-22 for naphthalene concentrations measured at DEMI
include the following:
• The maximum naphthalene concentration was measured at DEMI in July 2011
(473 ng/m3); five additional measurements greater than 400 ng/m3 have been
measured at DEMI (at least one in each of the first five years of sampling).
• The 95th percentile is at a maximum for the first year of sampling. While at least one
naphthalene concentration greater than 300 ng/m3 has been measured during each
year of sampling, these concentrations account for the highest percentage (10 percent)
of the measurements in 2008. Only one of these concentrations was measured in
2009, when the 95th percentile decreased by more than 100 ng/m3. Note that the
median concentration changed little between 2008 and 2009.
• With the exception of the maximum concentration, all of the statistical parameters
exhibit increases from 2009 to 2010. Relatively little change is shown in the
naphthalene concentrations measured at DEMI between 2010 and 2012.
• The smallest range of naphthalene concentrations was measured in 2013, with all of
the statistical parameters exhibiting decreases except the minimum concentration.
Both the 1-year average and median concentrations are at a minimum for 2013, with
the median concentration less than 100 ng/m3 for the first time.
14-31
-------
• All of the statistical parameters exhibit increases for 2014, with most exhibiting
additional increases for 2015. All of the statistical parameters exhibit decreases for
2016; the minimum, maximum, and 95th percentiles are at a minimum for 2016.
14.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the Michigan monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
14.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Michigan site, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 14-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 14-4 include the following:
• Formaldehyde has the highest annual average concentrations for DEMI, followed by
acetaldehyde, benzene, and carbon tetrachloride.
• Formaldehyde has the highest cancer risk approximations for DEMI (43.31 in-a-
million for 2015 and 46.02 in-a-million for 2016), with all other cancer risk
approximations an order of magnitude lower. Benzene is the only other pollutant of
interest with cancer risk approximations greater than 5 in-a-million.
• None of the pollutants of interest for DEMI have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximation for DEMI is formaldehyde (0.34 for 2015 and 0.36 for 2016).
14-32
-------
Table 14-4. Risk Approximations for the Michigan Monitoring Site
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Dearborn, Michigan - DEMI
Acetaldehyde
0.0000022
0.009
60/60
1.77
±0.14
3.89
0.20
61/61
1.96
±0.18
4.31
0.22
Benzene
0.0000078
0.03
60/60
0.80
±0.08
6.23
0.03
60/60
0.68
±0.08
5.33
0.02
1.3 -Butadiene
0.00003
0.002
60/60
0.09
±0.01
2.78
0.05
59/60
0.08
±0.01
2.47
0.04
Carbon Tetrachloride
0.000006
0.1
60/60
0.67
±0.02
4.01
0.01
60/60
0.66
±0.03
3.96
0.01
1,2 -Dichloroethane
0.000026
2.4
58/60
0.07
±0.01
1.78
<0.01
54/60
0.06
±0.01
1.57
<0.01
Ethylbenzene
0.0000025
1
60/60
0.37
±0.09
0.92
<0.01
60/60
0.36
±0.11
0.89
<0.01
Formaldehyde
0.000013
0.0098
60/60
3.33
±0.31
43.31
0.34
61/61
3.54
±0.47
46.02
0.36
Acenaphthene3
0.000088
54/62
8.86
±2.14
0.78
61/61
10.30
±3.05
0.91
Fluorene3
0.000088
59/62
7.93
± 1.69
0.70
54/61
9.57
±2.69
0.84
Naphthalene1
0.000034
0.003
62/62
116.18
± 15.46
3.95
0.04
61/61
107.01
± 14.51
3.64
0.04
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line are presented in ng/m3 for ease of viewing.
-------
14.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 14-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 14-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 14-5 provides the 10 pollutants of interest with the highest cancer risk approximations (in-
a-million) for DEMI, as presented in Table 14-4. The emissions, toxicity-weighted emissions,
and cancer risk approximations are shown in descending order in Table 14-5. Cancer risk
approximations for 2015 are presented in green while approximations for 2016 are in white.
Table 14-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 14.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
14-34
-------
Table 14-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the Michigan Monitoring Site
Top 10 Total Emissions for Pollutants
with Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Dearborn, Michigan (Wayne County) - DEMI
Benzene
531.30
Coke Oven Emissions, PM
2.09E-02
Formaldehyde
46.02
Formaldehyde
482.82
Formaldehyde
6.28E-03
Formaldehyde
43.31
Ethylbenzene
282.77
Benzene
4.14E-03
Benzene
6.23
Acetaldehyde
256.11
POM, Group 5a
3.09E-03
Benzene
5.33
1.3 -Butadiene
78.42
1,3-Butadiene
2.35E-03
Acetaldehyde
4.31
Naphthalene
44.94
Naphthalene
1.53E-03
Carbon Tetrachloride
4.01
Trichloroethylene
26.11
POM, Group 3
1.33E-03
Carbon Tetrachloride
3.96
Coke Oven Emissions, PM
21.14
Hexavalent Chromium PM
8.93E-04
Naphthalene
3.95
POM, Group 2b
9.55
POM, Group 2b
8.41E-04
Acetaldehyde
3.89
POM, Group 2d
7.18
Ethylbenzene
7.07E-04
Naphthalene
3.64
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 14-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the Michigan Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Dearborn, Michigan (Wayne County) - DEMI
Hydrochloric acid
3,196.06
Acrolein
1,592,409.10
Formaldehyde
0.36
Toluene
1,736.71
Hydrochloric acid
159,802.76
Formaldehyde
0.34
Xylenes
994.97
Formaldehyde
49,267.35
Acetaldehyde
0.22
Methanol
830.42
1,3-Butadiene
39,210.31
Acetaldehyde
0.20
Benzene
531.30
Acetaldehyde
28,456.78
1,3-Butadiene
0.05
Formaldehyde
482.82
Cyanide Compounds, PM
24,633.50
1,3-Butadiene
0.04
Hexane
354.50
Benzene
17,710.14
Naphthalene
0.04
Ethylene glycol
309.96
Manganese, PM
15,439.09
Naphthalene
0.04
Ethylbenzene
282.77
Naphthalene
14,978.94
Benzene
0.03
Acetaldehyde
256.11
T richloroethylene
13,057.09
Benzene
0.02
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 14-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Wayne County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Wayne County are coke oven emissions, formaldehyde, and
benzene.
• Seven of the highest emitted pollutants in Wayne County also have the highest
toxicity-weighted emissions. Wayne County is one of only two counties with an NMP
site for which coke oven emissions appear on both emissions-based lists.
• Formaldehyde has the highest cancer risk approximations for DEMI. This pollutant
also appears on both emissions-based lists, ranking second for both quantity emitted
and toxicity-weighted emissions. Benzene and naphthalene are also pollutants of
interest for DEMI that appear on both emissions-based lists.
• Acetaldehyde is another pollutant of interest whose cancer risk approximations
appear in Table 14-5 for DEMI. Acetaldehyde is one of the highest emitted pollutants
in Wayne County but does not appear among those with the highest toxicity-weighted
emissions (it ranks 12th).
• Carbon tetrachloride, the remaining pollutant of interest shown in Table 14-5, does
not appear on either emissions-based list.
Observations from Table 14-6 include the following:
• Hydrochloric acid, toluene, and xylenes are the highest emitted pollutants with
noncancer RfCs in Wayne County. The quantity of emissions for the highest-ranking
pollutants in Table 14-6 is an order of magnitude higher than the quantity of
emissions for the highest-ranking pollutants in Table 14-5.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) for Wayne County are acrolein, hydrochloric acid, and
formaldehyde. Although acrolein was sampled for at DEMI, this pollutant was
excluded from the pollutants of interest designation and thus, subsequent risk-based
screening evaluations due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2.
• Four of the highest emitted pollutants in Wayne County also have the highest
toxicity-weighted emissions.
• Formaldehyde has the highest noncancer hazard approximations for DEMI (although
none of the pollutants of interest have associated noncancer hazard approximations
greater than 1.0). Formaldehyde emissions rank sixth highest for Wayne County
while the toxicity-weighted emissions rank third (among the pollutants with
noncancer RfCs). Acetaldehyde and benzene also appear on all three lists for DEMI.
14-37
-------
• Naphthalene and 1,3-butadiene, the two remaining pollutants of interest for DEMI,
both appear among those with the toxicity-weighted emissions in Wayne County, but
do not appear among the highest emitted.
14.5 Summary of the 2015-2016 Monitoring Data for DEMI
Results from several of the data analyses described in this section include the following:
~~~ Sixteen pollutants failed screens for DEMI, including three carbonyl compounds,
eight VOCs, andfive PAHs.
~~~ Of the site-specific pollutants of interest, formaldehyde and acetaldehyde had the
highest annual average concentrations for DEMI. None of the other site-specific
pollutants of interest had annual average concentrations greater than 1 ng/m3.
~~~ The maximum ethylbenzene concentration measured across the program was
measured at DEMI in 2016.
~~~ DEMI's annual average concentration of naphthalene for 2015 is the highest annual
average concentration among NMP sites sampling PAHs.
~~~ A significant decrease in benzene concentrations occurred at DEMI for many years,
although concentrations have leveled off in recent years. Concentrations of
acetaldehyde have a slow, steady increasing trend over the last several years of
sampling.
~~~ Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for DEMI. None of the pollutants of interest for DEMI have noncancer
hazard approximations greater than an HQ of 1.0.
14-38
-------
15.0 Site in Missouri
This section summarizes those data from samples collected at the NATTS site in
Missouri and generated by ERG, EPA's contract laboratory
2016 monitoring efforts. This section also examines the
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
Sections 1 through 4 for detailed discussions and
definitions regarding the various data analyses presented below.
15.1 Site Characterization
This section characterizes the S4MO monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the site and assist in the interpretation of the ambient measurements.
The S4MO monitoring site is located in the St. Louis, MO-IL CBSA. Figure 15-1
presents a composite satellite image retrieved from ArcGIS Explorer showing the monitoring site
and its immediate surroundings. Figure 15-2 identifies nearby point source emissions locations
by source category, as reported in the 2014 NEI for point sources, version 1. Note that only
sources within 10 miles of the site are included in the facility counts provided in Figure 15-2. A
10-mile boundary was chosen to give the reader an indication of which emissions sources and
emissions source categories could potentially have a direct effect on the air quality at the
monitoring site. Further, this boundary provides both the proximity of emissions sources to the
monitoring site as well as the quantity of such sources within a given distance of the site. Sources
outside the 10-mile boundary are still visible on the map for reference but have been grayed out
to emphasize emissions sources within the boundary. Table 15-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates. Each
figure and table is discussed in detail in the paragraphs that follow.
for the NMP, over the 2015 and
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. ?
15-1
-------
Figure 15-1. St. Louis, Missouri (S4MO) Monitoring Site
A , SouTc e: UJ S ^
Sour cc":. N AS A^NG A •US G S
2'otTb^icro£,oft Corp.^
Ul
to
-------
Figure 15-2. NEI Point Sources Located Within 10 Miles of S4MO
Missouri
River
ILLINOIS
Mississippi
River ,
St. Louis
City
St. Louis
County
Legend
~ S4MO NATTS site
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
10 mile radius
County boundary
Source Category Group (No. ol
TAirport/Airline/Airport Support Operations (16)
Asphalt Production/Hot Mix Asphalt Plant (5)
T Brewery/Distillery/Winery (1)
#Building/Construction (1)
B Bulk Terminal/Bulk Plant (9)
C Chemical Manufacturing Facility (22)
SCoke Battery (1)
i Compressor Station (2)
AConcrete Batch Plant (5)
IXtrematory (1)
0Dry Cleaning Facility (3)
©Electrical Equipment Manufacturing Facility (1)
f Electricity Generation Facility (3)
£ Electroplating, Plating, Polishing, Anodizing,
and Coloring Facility (2)
^Ethanol Biorefinery (2)
Facilities)
'^'Fertilizer Plant (1)
F Food Processing/Agriculture Facility (9)
I Foundry, Iron and Steel (2)
AFoundry, non-ferrous (1)
If Gasoline/Diesel Service Station (1)
^Industrial Machinery or Equipment Plant (3)
O Institution (school, hospital, prison, etc.) (9)
¦ Landfill (3)
©Leather and Leather Products Facility (1)
* Metal Coating, Engraving, and Allied Services
" to Manufacturers (5)
©Metals Processing/Fabrication Facility (11)
X Mine/Quarry/Mineral Processing Facility (10)
9 Miscellaneous Commercial/Industrial Facility
" (28)
(]Paint and Coating Manufacturing Facility (5)
< Pesticide Manufacturing Plant (2)
•^Petroleum Products Manufacturing Facility (3)
czPharmaceutical Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (4)
Port and Harbor Operations (8)
p Printing/Publishing/Paper Product
Manufacturing Facility (4)
X Rail Yard/Rail Line Operations (15)
qRailroad Engines/Parts Manufacturing Facility
(1)
-------
Table 15-1. Geographical Information for the Missouri Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for Traffic Data
S4MO
29-510-0085
St. Louis
St. Louis
City
St. Louis, MO-IL
38.656498,
-90.198646
Residential
Urban/City
Center
57,558
1-70 nearRte 115/Salisbury St
lAADT reflects 2015 data (MO DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
-------
S4M0 is located in central St. Louis. Figure 15-1 shows that the S4MO monitoring site is
located less than one-quarter mile west of 1-70. The Mississippi River, which separates Missouri
and Illinois, is less than 1 mile east of the site. Although the area directly around the monitoring
site is primarily residential, industrial facilities are located nearby, primarily just on the other
side of 1-70. Figure 15-2 shows that a large number of point sources are located within 10 miles
of S4MO, particularly on the east side of the Missouri/Illinois border. The source categories with
the greatest number of point sources surrounding S4MO include chemical manufacturing
facilities; airport and airport support operations, which include airports and related operations as
well as small runways and heliports, such as those associated with hospitals or television
stations; rail yard/rail line operations; metals processing/fabrication facilities; and mines,
quarries, and mineral processing facilities. Within 1 mile of S4MO are a pharmaceutical
manufacturing facility, a printing and publishing facility, a leather products facility, a fertilizer
plant, and a metals processing/fabrication facility.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 15-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near S4MO is nearly 58,000 and ranks 17th highest among other NMP sites,
which falls in the upper third of traffic volumes compared to other NMP sites. The traffic
estimate provided is for 1-70 near Route 115/Salisbury Street.
15.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for the S4MO
monitoring site to identify site-specific "pollutants of interest," which allows analysts and
readers to focus on a subset of pollutants through the context of risk. Each pollutant's
preprocessed daily measurement was compared to its associated risk screening value. If the
concentration was greater than the risk screening value, then the concentration "failed the
screen." The site-specific results of this risk-based screening process are presented in Table 15-2
and incorporate measurements from both 2015 and 2016. Pollutants of interest are those for
which the individual pollutant's total failed screens contribute to the top 95 percent of the site's
total failed screens and are shaded in gray in Table 15-2. It is important to note which pollutants
were sampled for at each site when reviewing the results of this analysis. VOCs, PAHs, carbonyl
compounds, and metals (PMio) were sampled for at S4MO.
15-5
-------
Table 15-2. 2015-2016 Risk-Based Screening Results for the Missouri Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
St. Louis, Missouri - S4MO
Acetaldehyde
0.45
120
120
100.00
10.90
10.90
Arsenic (PMio)
0.00023
120
121
99.17
10.90
21.80
Benzene
0.13
120
120
100.00
10.90
32.70
Carbon Tetrachloride
0.17
120
120
100.00
10.90
43.60
Formaldehyde
0.077
120
120
100.00
10.90
54.50
1.3 -Butadiene
0.03
119
119
100.00
10.81
65.30
1,2 -Dichloroethane
0.038
114
114
100.00
10.35
75.66
Naphthalene
0.029
107
118
90.68
9.72
85.38
p-Dichlorobenzene
0.091
50
97
51.55
4.54
89.92
Fluorene
0.011
27
100
27.00
2.45
92.37
Acenaphthene
0.011
23
107
21.50
2.09
94.46
Ethylbenzene
0.4
16
120
13.33
1.45
95.91
Hexachloro-1,3 -butadiene
0.045
10
12
83.33
0.91
96.82
Manganese (PMio)
0.03
10
121
8.26
0.91
97.73
Nickel (PMio)
0.0021
6
121
4.96
0.54
98.27
Benzo(a)pyrene
0.00057
5
114
4.39
0.45
98.73
1,2 -Dibromoethane
0.0017
5
5
100.00
0.45
99.18
Lead (PMio)
0.015
3
121
2.48
0.27
99.46
Propionaldehyde
0.8
3
117
2.56
0.27
99.73
Cadmium (PMio)
0.00056
1
121
0.83
0.09
99.82
Chloroprene
0.0021
1
1
100.00
0.09
99.91
Fluoranthene
0.011
1
118
0.85
0.09
100.00
Total
1,101
2,227
49.44
15-6
-------
Observations from Table 15-2 include the following:
• Concentrations of 22 pollutants failed at least one screen for S4MO; approximately
49 percent of concentrations for these 22 pollutants were greater than their associated
risk screening value (or failed screens).
• S4MO has the highest number of concentrations failing screens (1,101) among all
NMP sites, as shown in Table 4-9 of Section 4.2. The failure rate for S4MO, when
incorporating all pollutants with screening values, is approximately 20 percent. This
is due primarily to the relatively large number of pollutants sampled for at this site, as
discussed in Section 4.2.
• Twelve pollutants contributed to 95 percent of failed screens for S4MO and therefore
were identified as pollutants of interest for this site. These 12 pollutants include two
carbonyl compounds, six VOCs, one PMio metal, and three PAHs. S4MO ties with
two other sites for the greatest number of pollutants of interest among NMP sites.
• Acetaldehyde, formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene and
1,2-dichloroethane failed 100 percent of screens for S4MO and were detected in all or
nearly all the samples collected. 1,2-Dibromoethane and chloroprene also failed
100 percent of screens but were detected in few VOC samples collected (five and
one, respectively), and are not pollutants of interest for S4MO.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
15.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Missouri monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of monitoring.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
15-7
-------
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at S4MO are provided
in Appendices J, M, N, and O.
15.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for the Missouri site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where at least three valid quarterly averages could be calculated for a given year
and where method completeness was greater than or equal to 85 percent, as presented in
Section 2.4. Quarterly and annual average concentrations for the pollutants of interest for S4MO
are presented in Table 15-3, where applicable. Note that concentrations of the PAHs and metals
are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not detected in a
given calendar quarter, the quarterly average simply reflects "0" because only zeros substituted
for non-detects were factored into the quarterly average concentration.
15-8
-------
Table 15-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Missouri Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
Oig/m3)
Q4
Avg
Oig/m3)
Annual
Average
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
St. Louis, Missouri - S4MO
Acetaldehyde
60/60/60
1.34
±0.21
1.89
±0.29
1.85
±0.19
1.56
±0.35
1.66
±0.14
60/60/60
1.12
±0.15
1.85
±0.45
1.86
±0.18
1.58
±0.36
1.59
±0.16
Benzene
60/60/60
0.80
±0.08
0.47
±0.07
0.71
±0.24
0.67
±0.16
0.66
±0.08
60/60/60
0.65
±0.11
0.46
±0.08
0.56
±0.09
0.78
±0.24
0.62
±0.07
1.3 -Butadiene
59/59/60
0.08
±0.02
0.06
±0.02
0.07
±0.01
0.09
±0.04
0.08
±0.01
60/41/60
0.08
±0.02
0.08
±0.02
0.08
±0.02
0.13
±0.07
0.09
±0.02
Carbon Tetrachloride
60/60/60
0.63
±0.04
0.62
±0.02
0.66
±0.03
0.60
±0.03
0.63
±0.02
60/60/60
0.63
±0.03
0.71
±0.04
0.63
±0.07
0.62
±0.05
0.65
±0.02
p-Dichlorobenzene
49/15/60
0.09
±0.06
0.13
±0.07
0.18
±0.09
0.17
±0.13
0.14
±0.04
48/19/60
0.07
±0.04
0.17
±0.12
0.33
±0.28
0.31
±0.26
0.22
±0.10
1,2 -Dichloroethane
60/55/60
0.09
±0.01
0.08
±0.01
0.06
±0.01
0.07
±0.01
0.07
±<0.01
54/52/60
0.08
±0.01
0.09
±0.01
0.06
±0.02
0.06
±0.02
0.07
±0.01
Ethylbenzene
60/56/60
0.18
±0.04
0.16
±0.03
0.23
±0.05
0.24
±0.09
0.20
±0.03
60/60/60
0.18
±0.06
0.19
±0.05
0.28
±0.06
0.34
±0.15
0.25
±0.05
Formaldehyde
60/60/60
2.20
±0.42
3.58
±0.93
4.46
±0.89
2.14
±0.50
3.09
±0.42
60/60/60
2.07
±0.24
4.32
± 1.34
4.33
±0.73
2.04
±0.44
3.19
±0.48
Acenaphthene3
55/55/58
1.93
±0.80
7.54
±2.76
13.12
±3.04
4.07
± 1.31
6.51
± 1.47
52/52/60
1.56
± 1.04
6.82
±3.44
12.81
±4.10
3.68
± 1.71
6.13
± 1.71
Arsenic (PMi0)a
60/60/60
0.67
±0.15
0.80
±0.20
1.14
±0.26
0.89
±0.33
0.88
±0.12
61/61/61
0.69
±0.13
0.92
±0.28
1.00
±0.18
0.99
±0.41
0.90
±0.13
Fluorene3
53/53/58
2.65
±0.65
7.28
±2.23
12.71
±2.52
3.66
± 1.29
6.40
± 1.31
47/47/60
1.22
±0.93
8.09
±3.76
14.03
±3.86
3.80
± 1.60
6.67
± 1.81
Naphthalene1
58/58/58
66.81
± 12.67
62.93
± 15.35
82.25
± 14.83
97.69
± 34.90
78.32
± 11.77
60/59/60
49.09
± 14.01
71.87
±35.37
74.39
±22.30
100.08
±37.28
73.48
± 14.07
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Observations for S4MO from Table 15-3 include the following:
• The pollutants with the highest annual average concentrations are formaldehyde and
acetaldehyde. These are the only pollutants of interest with annual averages greater
than 1 |ig/m3.
• Concentrations of formaldehyde measured at S4MO range from 0.862 |ig/m3 to
10.7 |ig/m3. The two highest concentrations measured at S4MO were both measured
on June 11th (10.7 |ig/m3 on June 11, 2016 and 9.23 |ig/m3 on June 11, 2015).
July 17th was also a sample day with relatively high concentrations for both years
(8.25 |ig/m3 on July 17, 2015 and 7.04 |ig/m3 on July 17, 2016). The quarterly
average concentrations of formaldehyde exhibit a seasonal trend, with the second and
third quarter averages significantly higher than the remaining quarterly averages for
both years, indicating that higher formaldehyde concentrations tended to be measured
during the warmer months of the year at S4MO. These quarterly averages also exhibit
more variability, as indicated by the larger confidence intervals. All but two of the 13
formaldehyde concentrations greater than 4 |ig/m3 measured at S4MO in 2015 were
measured during the second or third quarters of the year; for 2016, all 16
formaldehyde concentrations greater than 4 |ig/m3 were measured at S4MO during
the second or third quarters. At the lower end of the concentration range,
formaldehyde concentrations less than 2 |ig/m3 were measured at S4MO primarily
during the first or fourth quarters of the year; of the 29 formaldehyde concentrations
less than 2 |ig/m3 measured at S4MO, only two were measured outside the first or
fourth quarters, with one measured each year.
• Concentrations of acetaldehyde measured at S4MO range from 0.547 |ig/m3 to
3.68 |ig/m3. Concentrations of acetaldehyde appear lowest during the first quarter of
each year, although the quarterly average concentrations are not statistically different.
All 15 acetaldehyde concentrations less than 1 |ig/m3 were measured at S4MO during
the first or fourth quarters of either year. Five concentrations less than 1 |ig/m3 were
measured during 2015, two during the first quarter and three during the fourth (all in
December); 10 concentrations less than 1 |ig/m3 were measured during 2016, seven
during the first quarter and three during the fourth (again, all in December).
• Among the VOCs, benzene and carbon tetrachloride have the highest annual average
concentrations, the former for 2015 and the latter for 2016, although the annual
averages for these two pollutants are similar to each other. Benzene concentrations
are more variable, though. Concentrations of benzene measured at S4MO span an
order of magnitude, ranging from 0.282 |ig/m3 to 2.18 |ig/m3. The second quarter
average concentrations of benzene for both years appear lower than the quarterly
averages shown for the remaining quarterly averages. A similar observation was
made in the 2014 NMP report. None of the 10 benzene concentrations greater than
1 |ig/m3 were measured during the second quarter of either year (this is also true for
the third quarter of 2016).
• Concentrations of carbon tetrachloride measured at S4MO range from 0.347 |ig/m3 to
0.857 |ig/m3, with most concentrations falling between 0.50 |ig/m3 and 0.75 |ig/m3.
The seven highest carbon tetrachloride concentrations (those greater than 0.75 |ig/m3)
were measured in 2016.
15-10
-------
The quarterly average concentrations of 1,3-butadiene exhibit relatively little
variability in 2015. This is mostly true for 2016, with the first, second, and third
quarterly average concentrations appearing nearly identical to each other; the fourth
quarter average concentration, however, does not resemble the others
(0.13 ± 0.07 |ig/m3), A review of the data shows that 1,3-butadiene concentrations
measured at S4MO range from 0.031 |ig/m3 to 0.419 |ig/m3, plus a single non-detect.
Four of the five highest 1,3-butadiene concentrations measured at S4MO (those
greater than 0.20 |ig/m3) were measured between October and December 2016 (with
the exception in November 2015).
The quarterly average concentrations of />dichlorobenzene exhibit considerable
variability, particularly for 2016, with the quarterly averages for the second half of
the year considerably higher than those for the first half of the year, although the
confidence intervals shown are relatively large and nearly the same magnitude as the
averages themselves. Concentrations of />dichlorobenzene measured at S4MO range
from 0.0241 |ig/m3 to 1.80 |ig/m3 and include 23 non-detects. Although the maximum
/;-dichlorobenzene concentration measured across the program was not measured at
S4MO, the next five highest concentrations were measured at this site. S4MO is the
only NMP site at which more than one />dichlorobenzene concentration greater than
1 |ig/m3 was measured (four, which were all measured during the second half of
2016); /;-dichlorobenzene concentrations measured at S4MO account for nine of the
19 /;-dichlorobenzene concentrations greater than 0.5 |ig/m3 measured across the
program, the most among all sites sampling VOCs (the next highest site has four).
Concentrations of ethylbenzene appear higher during the second half of each year,
although the differences are not statistically significant. The three highest
ethylbenzene concentrations measured at S4MO were measured during the fourth
quarter of 2016, including the only concentration greater than 1 |ig/m3 (1,01 |ig/m3).
Five of the six ethylbenzene concentrations greater than 0.5 |ig/m3 measured at
S4MO were measured in October, November, or December of either year.
Arsenic is the only PMio metal pollutant of interest for S4MO. For both years, the
quarterly average concentrations of arsenic for the third quarter are greater than
1 ng/m3. A review of the data shows that arsenic concentrations measured at S4MO
range from 0.107 ng/m3 to 3.62 ng/m3. Thirty-five arsenic concentrations greater than
1 ng/m3 were measured at S4MO, with 17 measured in 2015 and 18 measured in
2016. The third quarter is the calendar quarter in which the highest number of these
concentrations were measured. At the other end of the concentration range, few
arsenic concentrations less than 0.5 ng/m3 were measured at S4MO during the third
calendar quarters; of the 24 arsenic concentrations less than 0.5 ng/m3 measured at
this site, only one was measured during the third quarters (none in 2015 and one in
2016).
Among the three PAHs identified as pollutants of interest for S4MO, naphthalene has
the highest annual average concentrations. The concentrations of naphthalene
measured at S4MO are highly variable, as indicated by the quarterly averages shown
in Table 15-3, each with relatively large confidence intervals. For both years, the
fourth quarter average is the highest, nearly 100 ng/m3 for 2015 and just surpassing
this for 2016. Concentrations of naphthalene measured at S4MO range from
15-11
-------
0.660 ng/m3 to 243 ng/m3. Of the 25 naphthalene concentrations greater than
100 ng/m3 measured at S4MO, most were measured during the fourth quarter of
either year. The three naphthalene concentrations less than 1 ng/m3 measured at
S4MO are among the lowest naphthalene concentrations measured at an NMP in both
2015 and 2016 (only one site has a lower naphthalene concentration, and S4MO is the
only site with more than one of these). By comparison, the next lowest naphthalene
concentration measured at S4MO is 20.7 ng/m3.
• Concentrations of acenaphthene and fluorene appear to be highest during the warmer
months of the year, based on the quarterly average concentrations, although each of
the quarterly averages exhibits a considerable level of variability. A review of the
data shows that the 25 highest concentrations of fluorene were measured at S4MO
during the second or third quarters of either year; the 23 highest concentrations of
acenaphthene were measured at S4MO during the second or third quarters of either
year. Higher concentrations of these two PAHs were often measured on the same
days. For example, the highest fluorene concentration (27.6 ng/m3) was measured on
August 10, 2016, the same day the second highest acenaphthene concentration
(28.0 ng/m3) was measured; the highest acenaphthene concentration (28.2 ng/m3) was
measured on September 24, 2016, the same day the second highest fluorene
concentration (25.4 ng/m3) was measured.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for S4MO from
those tables include the following:
• S4MO appears in Tables 4-10 through 4-13 a total of nine times, appearing in each
table except Table 4-11 for carbonyl compounds.
• S4MO has the third highest (2016) and sixth highest (2015) annual average
concentrations of />dichlorobenzene among NMP sites sampling VOCs, as shown in
Table 4-10.
• Both of S4MO's annual average concentrations of acenaphthene and fluorene appear
in Table 4-12, though ranking in the lower half of the table. S4MO's annual average
concentration of naphthalene for 2015 ranks ninth highest among NMP sites sampling
this pollutant.
• S4MO has the sixth (2016) and seventh (2015) highest annual average concentrations
of arsenic among NMP sites sampling PMio metals, as shown in Table 4-13.
15.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 15-3 for S4MO. Figures 15-3 through 15-14 overlay the site's minimum, annual average,
and maximum concentrations for each year onto the program-level minimum, first quartile,
15-12
-------
median, average, third quartile, and maximum concentrations, as described in Section 3.4.2.1,
and are discussed below. If an annual average concentration could not be calculated, the range of
concentrations are still provided in the figures that follow.
Figure 15-3. Program vs. Site-Specific Average Acenaphthene Concentrations
S4MO
1
1 A
Program MaxConcentration = 108 ng/m3
1 1
"
¦ 1
1 1
30 45
Concentration (ng/m3)
60
Program: 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Ave rag
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
Figure 15-3 presents the box plot for acenaphthene for S4MO and shows the following:
• The program-level maximum acenaphthene concentration (108 ng/m3) is not shown
directly on the box plot in Figure 15-3 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plot has been reduced.
• The maximum acenaphthene concentration measured at S4MO (28.2 ng/m3) is about
one-fourth the magnitude of the maximum acenaphthene concentration measured
across the program. Four acenaphthene concentrations measured in 2016 are higher
than the maximum concentration measured in 2015 (20.2 ng/m3). Yet, S4MO's
annual average concentrations for both years are greater than the program-level
average concentration (4.36 ng/m3).
• There were no non-detects of acenaphthene measured at S4MO while non-detects
account for 18 percent of the measurements at the program level.
Figure 15-4. Program vs. Site-Specific Average Acetaldehyde Concentrations
it:
8 10 12
Concentration (jug/rn3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Averag
¦ ~ ~ ~
Site: 2015Average 2016 Average Concentration Range, 2015 & 2016
a o —
15-13
-------
Figure 15-4 presents the box plot for acetaldehyde for S4MO and shows the following:
• The entire range of acetaldehyde concentrations measured at S4MO in 2015 is less
than 3 |ig/m3; two concentrations measured in 2016 (3.30 |ig/m3 and 3.68 |ig/m3) are
greater than the maximum concentration measured in 2015 (2.87 |ig/m3).
• S4MO's annual average concentration of acetaldehyde for 2015 is similar to the
program-level average concentration (1.67 |ig/m3), with the annual average for 2016
just slightly less than these.
• The lowest concentrations of acetaldehyde measured at S4MO were measured on the
same dates each year: December 26th and December 8th.
Figure 15-5. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
S4MO
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 15-5 presents the box plot for arsenic (PMio) for S4MO and shows the following:
• The maximum arsenic (PMio) concentration measured at S4MO (3.62 ng/m3) is about
one-half the maximum concentration measured across the program. The next highest
arsenic concentration measured at S4MO is considerably less (the maximum
concentration shown for 2015, 2.17 ng/m3); several arsenic concentrations of this
magnitude were measured at S4MO both years.
• S4MO's annual average concentration of arsenic (PMio) for 2015 is similar to the
annual average concentration for 2016. Both annual averages are greater than the
program-level average concentration and similar to the program-level third quartile.
This site has the sixth and seventh highest annual average concentrations of arsenic
among NMP sites sampling PMio metals.
15-14
-------
Figure 15-6. Program vs. Site-Specific Average Benzene Concentrations
n
012345678
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 15-6 presents the box plot for benzene for S4MO and shows the following:
• The range of benzene concentrations measured at S4MO in 2016 is slightly smaller
than the range of concentrations measured in 2016, though both are relatively small
compared to the range measured at the program-level.
• Both annual average concentrations of benzene for S4MO lie between the program-
level median concentration (0.59 |ig/m3) and the program-level average concentration
(0.72 |ig/m3).
Figure 15-7. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
i
! Program Max Concentration = 3.90 jxg/m3 n
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (|ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 15-7 presents the box plot for 1,3-butadiene for S4MO and shows the following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plot as the scale has also been reduced to allow for the observation
of data points at the lower end of the concentration range.
• If the minimum concentration measured in 2015 (non-detect) and maximum
concentration measured in 2016 (0.419 |ig/m3) were excluded, the range of
1,3-butadiene concentrations measured at S4MO would be nearly identical.
15-15
-------
• The annual average concentration of 1,3-butadiene for 2015 is just less than the
program-level average concentration while the annual average for 2016 is just greater
than the program-level average.
Figure 15-8. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
!
¦¦I
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jjg/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
Figure 15-8 presents the box plot for carbon tetrachloride for S4MO and shows the
following:
• The program-level median and average concentrations are similar to each other in
magnitude (both approximately 0.64 |ig/m3) and thus, are plotted nearly on top of
each other.
• The range of carbon tetrachloride concentrations measured at S4MO in 2015 is
smaller than the range measured in 2016, although the entire range of concentrations
spans just over 0.5 |ig/m3.
• The annual average concentrations of carbon tetrachloride for S4MO are similar to
both the program-level median and average concentrations.
Figure 15-9. Program vs. Site-Specific Average />-Dichlorobenzene Concentrations
S4MO
-
A
Program Max Concentration
= 2.78 |ig/m3
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
15-16
-------
Figure 15-9 presents the box plot for p-dichlorobenzene for S4MO and shows the
following:
• The program-level maximum /;-dichlorobenzene concentration (2.78 |ig/m3) is not
shown directly on the box plot as the scale has been reduced to allow for the
observation of data points at the lower end of the concentration range. In addition, the
first and second quartiles for p-dichlorobenzene are zero and therefore not visible on
the graph due to the large number of non-detects for this pollutant (more than 50
percent of the measurements are non-detects for /;-dichlorobenzene). Twenty-three
non-detects were measured at S4MO.
• While the maximum p-dichlorobenzene concentration measured across the program
was not measured at S4MO, the second, third, fourth, fifth, and sixth highest
concentrations of p-dichlorobenzene were measured at this site. The nine highest
/;-dichlorobenzene concentrations measured across the program (those greater than
0.75 |ig/m3) were split between S4MO and one other NMP site (NBIL). All four
/;-dichlorobenzene concentrations measured at S4MO were measured in 2016.
• Both annual average concentrations for S4MO are greater than the program-level
average concentration (0.05 |ig/m3). This site has the third highest (2016) and sixth
highest (2015) annual average concentrations of p-dichlorobenzene among NMP sites
sampling VOCs.
Figure 15-10. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1
m-
, ,
| Program Max Concentration = 45.8 (.ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (|ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 15-10 presents the box plot for 1,2-dichloroethane for S4MO and shows the
following:
• The scale of the box plot in Figure 15-10 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• All concentrations of 1,2-dichloroethane measured at S4MO are less than 0.15 |ig/m3,
or half the program-level average concentration of 0.30 |ig/m3, which is being driven
by the measurements at the upper end of the concentration range.
15-17
-------
• The annual average concentrations for S4MO are both less than the program-level
median concentration (0.081 |ig/m3).
Figure 15-11. Program vs. Site-Specific Average Ethylbenzene Concentrations
tr-
1.5 2
Concentration (|jg/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Aveage Concentration Range, 2015 & 2016
o
Figure 15-11 presents the box plot for ethylbenzene for S4MO and shows the following:
• A single ethylbenzene concentration greater than 1 |ig/m3 was measured at S4MO,
which is about one-third the magnitude of the maximum ethylbenzene concentration
measured across the program.
• The annual average concentrations of ethylbenzene for S4MO fall between the
program-level median (0.17 |ig/m3) and average (0.26 |ig/m3) concentrations.
Figure 15-12. Program vs. Site-Specific Average Fluorene Concentrations
_
"
! Program Max Concentration
105 ng/m3
1
S4MO
30
Concentration (ng/m3]
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
Figure 15-12 presents the box plot for fluorene for S4MO and shows the following:
• The program-level maximum fluorene concentration (105 ng/m3) is not shown
directly on the box plot in Figure 15-12 as the scale of the box plot has been reduced.
Note that the first quartile is zero and therefore not visible on the box plot due to the
number of non-detects.
• All of the fluorene concentrations measured at S4MO are considerably less than the
maximum fluorene concentration measured across the program. Five fluorene
15-18
-------
concentrations measured in 2016 are greater than the maximum fluorene
concentration measured at S4MO in 2015 (18.5 ng/m3). Yet, the annual average
concentrations for both years are similar to each other and are greater than the
program-level average concentration (4.36 ng/m3).
• Eighteen non-detects of fluorene were measured at S4MO, accounting for 15 percent
of the measurements, while non-detects account for 29 percent of the measurements
at the program-level.
Figure 15-13. Program vs. Site-Specific Average Formaldehyde Concentrations
m
12 15 18 21 24 27
Concentration (jug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 15-13 presents the box plot for formaldehyde for S4MO and shows the following:
• One formaldehyde concentration greater than 10 |ig/m3 was measured at S4MO; this
concentration is less than half the maximum concentration measured across the
program.
• The annual average concentrations of formaldehyde for S4MO are fairly similar to
the program-level average concentration (3.05 |ig/m3), particularly 2015.
15-19
-------
Figure 15-14. Program vs. Site-Specific Average Naphthalene Concentrations
50 100
150 200 250
Concentration (ng/m3)
300 350 400 450
Program: IstQuartile 2ndQuartile 3rdQuartile 4-thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 15-14 presents the box plot for naphthalene for S4MO and shows the following:
• Non-detects of naphthalene were not measured in 2015 or 2016, across the program
or at S4MO, despite the low minimum concentrations shown on the box plot. As
previously mentioned, some of the lowest concentrations of naphthalene across the
program were measured at S4MO. This is an anomaly, for both this site and for the
program.
• The range of naphthalene concentrations measured at S4MO in 2015 is similar to the
range of concentrations measured at this site in 2016.
• Both annual average concentrations of naphthalene for S4MO fall between the
program-level average concentration (61.23 ng/m3) and the program-level third
quartile (82.15 ng/m3).
15.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
S4MO has sampled VOCs and carbonyl compounds under the NMP since 2002, PMio metals
since 2003, and PAHs since 2008. Thus, Figures 15-15 through 15-26 present the 1-year
statistical metrics for each of the pollutants of interest for S4MO. The statistical metrics
presented for assessing trends include the substitution of zeros for non-detects. If sampling began
mid-year, a minimum of 6 months of sampling is required for inclusion in the trends analysis; in
these cases, a 1-year average concentration is not provided, although the range and percentiles
are still presented.
15-20
-------
Figure 15-15. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
S4MO
0
2012
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-15 for acenaphthene concentrations measured at S4MO
include the following:
• S4MO began sampling PAHs under the NMP in April 2008. Because a full year's
worth of data is not available, a 1-year average concentration for 2008 is not
presented, although the range of measurements is provided.
• Three measurements greater than 30 ng/m3 have been measured at S4MO, two in
September 2008 and another in July 2011.
• All of the statistical parameters shown exhibit decreases from 2008 to 2009. In all,
11 concentrations measured in 2008 are greater than the maximum concentration
measured in 2009. In addition, acenaphthene concentrations less than 5 ng/m3
accounted for more than twice the percentage of samples collected in 2009
(64 percent) compared to 2008 (32 percent).
• Although the range of concentrations measured increased from 2009 to 2010 and
again for 2011, the median concentration decreased slightly each year.
• Between 2011 and 2014, the 1-year average concentration has a fluctuating pattern, as
years with lower averages alternate with years with higher averages. The median
concentration has a similar pattern, although the increase shown for 2014 is smaller
than the increase shown in the 1-year average concentration.
15-21
-------
• Little change in the 1-year average concentration is shown between 2014 and 2016,
despite an increasing number of non-detects (from none, to three, to eight over the
3-year period) and fluctuations in the magnitude of concentrations measured,
predominantly at the lower end of the concentration range, though a few higher
concentrations were also measured in 2016.
Figure 15-16. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
S4MO
I
L£J
I
X
... o-..
2
i
&
pi-
•O-
•O
r£i r^l
Tj! j£j
2013
2014
2015 2016
O 5th Percentile
— Minimum
O 95th Percentile
Observations from Figure 15-16 for acetaldehyde concentrations measured at S4MO
include the following:
• Because carbonyl compound sampling under the NMP did not begin at S4MO until
December 2002, data from 2002 were excluded from this analysis.
• The maximum acetaldehyde concentration was measured in 2004 (32.5 |ig/m3) and is
more than twice the next highest concentration (15.5 |ig/m3, measured in 2007).
• Even with the maximum concentration measured in 2004, nearly all of the statistical
parameters decreased from 2003 to 2004. The maximum concentration measured in
2004 is nearly six times higher than the next highest concentration measured that year
(5.72 |ig/m3). Excluding the outlier, five concentrations measured in 2003 are higher
than the second highest concentration measured in 2004. At the other end of the
concentration range, the number of acetaldehyde concentrations less than 3 |ig/m3
increased from 22 to 34, and thus, accounted for more than half of the measurements
in 2004. Additional decreases are shown for 2005.
15-22
-------
• Between 2003 and 2012, the 1-year average concentrations have an undulating
pattern, with a few years of a decreasing trend followed by a few years of an
increasing trend. During this time, the 1-year average concentrations varied between
1.83 |ig/m3 (2008) and 4.10 |ig/m3 (2010).
• After the significant decrease shown between 2010 and 2012, the changes shown are
more subtle in nature. The concentrations measured during the 3-year period from
2012 to 2014 exhibit the least year-to-year variability in concentrations measured
since the onset of sampling. Additional decreases are shown after 2014, with several
of the parameters at a minimum for 2016, including both the 1-year average and
median concentrations. The median concentration for 2016 is less than 1.50 |ig/m3 for
the first time since the onset of sampling.
Figure 15-17. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
S4MO
Maximum
Concentrati on for
2013 is 44.1 ng/m3
5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2003.
Observations from Figure 15-17 for arsenic concentrations measured at S4MO include
the following:
• S4MO began sampling metals under the NMP in July 2003. Because a full year's
worth of data is not available, a 1-year average concentration is not presented,
although the range of measurements is provided.
• The maximum arsenic concentration was measured at S4MO on December 26, 2007
(44.1 ng/m3). Five additional arsenic concentrations greater than 10 ng/m3 have been
measured at S4MO (three in 2005 and one each in 2003 and 2009).
15-23
-------
• This figure shows that years with little variability in the measurements alternate with
years with significant variability, particularly between 2004 and 2010. Less
measurement variability is shown in the years that follow.
• Most of the statistical parameters are at a minimum for 2013. The range of
measurements, the difference between the 5th and 95th percentiles, and the difference
between the median and 1-year average concentrations are also at a minimum for
2013.
• With the exception of the 5th percentile, increases are shown for each of the
parameters for 2014, although some are slight (the median increased by less than
0.1 ng/m3) while others are relatively large (the maximum concentration doubled
from 2013 to 2014).
• Despite a few changes in concentrations at the upper and/or lower end of the
concentration range, relatively little change is shown in the central tendency
parameters between 2014 and 2016.
Figure 15-18. Yearly Statistical Metrics for Benzene Concentrations Measured at S4MO
o
T
T
r5!
~
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
Observations from Figure 15-18 for benzene concentrations measured at S4MO include
the following:
• Because VOC sampling under the NMP did not begin at S4MO until December 2002,
2002 data was excluded from this analysis.
15-24
-------
One benzene concentration greater than 5 |ig/m3 has been measured at S4MO (2003).
Three additional benzene concentrations greater than 4 |ig/m3 have also been
measured (one each in 2003, 2006, and 2008).
The 1-year average concentration exhibits a steady decreasing trend through 2007,
representing a 44 percent decrease. In the years between 2007 and 2011, the 1-year
average concentrations have a slight undulating pattern, with the 1-year average
varying between 0.80 |ig/m3 (2011) and 1.03 |ig/m3 (2010).
From 2011 to 2012, the statistical parameters representing the upper end of the
concentration range (the maximum and 95th percentile) increased while the statistical
parameters representing the lower end of the concentration range (the minimum and
5th percentile) decreased, indicating a widening of concentrations measured. Yet, the
1-year average concentration did not change and the median decreased. The number
of concentrations greater than 1 |ig/m3 doubled (from six in 2011 to 12 in 2012) while
the number of concentrations less than 0.5 |ig/m3 increased from two in 2011 to 12 in
2012.
Several of the statistical parameters are at a minimum for 2013, including the 1-year
average concentration (0.61 |ig/m3), when a single benzene concentration greater than
1 |ig/m3 was measured.
All of the statistical parameters exhibit slight increases for 2014. Despite a few
changes at the upper end of the concentration range, the central tendency parameters
both exhibit additional decreases for 2015 and 2016, with the median concentration at
a minimum for 2016.
15-25
-------
Figure 15-19. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
S4MO
I
x
• 1^..
rh
X
A
I
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 15-19 for 1,3-butadiene concentrations measured at S4MO
include the following:
• The maximum 1,3-butadiene concentration was measured at S4MO in 2003, although
a similar concentration was also measured in 2008. These are the only two
1,3-butadiene concentrations greater than 1.0 [j,g/m3 that have been measured at
S4MO.
• The minimum, 5th percentile, and median concentrations are all zero for 2003 and
2004, indicating that at least 50 percent of the measurements were non-detects. The
number of non-detects decreased after 2004, from a maximum of 43 non-detects in
2004 to a minimum of zero in 2010 (also 2012 and 2016). After 2006, no more than
five non-detects of 1,3-butadiene have been measured at S4MO in any given year.
• Between 2004 and 2008, the 1-year average concentration changed little, ranging
from 0.079 [j,g/m3 (2005) to 0.095 [j,g/m3 (2006). Greater fluctuations are shown in the
years that follow. Years with a higher number of non-detects, as indicated by a
minimum and 5th percentile of zero, such as 2009, 2011, and 2013, alternate with
years without any non-detects (2010 and 2012) and concentrations that are higher in
magnitude, as indicated by the 95th percentile and maximum concentration. This
pattern ends with 2014, as two non-detects were measured in 2014, but the undulating
pattern in the 1-year average concentration continues.
15-26
-------
Figure 15-20. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at S4MO
r-5n
I
A
|-5-
'O-
O
T
-a-
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
Observations from Figure 15-20 for carbon tetrachloride concentrations measured at
S4MO include the following:
• Twenty of the 21 non-detects of carbon tetrachloride were measured at S4MO in
2003, 2004, or 2005, with a single non-detect measured in 2007.
• A steady increasing trend in the 1-year average concentration is shown through 2006.
Although the maximum concentration decreased substantially from 2006 to 2007 and
a non-detect was measured, the decrease in the 1-year average concentration is not
statistically significant and the median concentration did not change at all. In fact, the
median concentration is steady at 0.57 [j,g/m3 between 2004 and 2007.
• All of the statistical parameters exhibit increases from 2007 to 2008. Twenty
concentrations, or nearly one-third of the concentrations, measured in 2008 are
greater than the maximum concentration measured in 2007.
• Both the median and 1-year average concentrations have a decreasing trend between
2008 and 2010, with 1-year average concentration returning to near 2007 levels.
• Between 2010 and 2012, the 1-year average concentration has a significant increasing
trend even as the majority of concentrations measured are falling into a tighter range,
as indicated by the decreasing difference between the 5th and 95th percentiles for
these years.
15-27
-------
• Nearly all of the statistical parameters exhibit decreases for 2013 and again for 2014.
A larger number of concentrations at the lower end of the concentration range was
measured each year, while fewer concentrations at the upper end of the concentration
range were measured. The number of concentrations less than 0.65 [j,g/m3 increased
between 2012 and 2014, from 20 in 2012 to 34 in 2013 and 35 in 2014. At the other
end of the concentration range, fewer concentrations greater than 0.8 [j,g/m3 have been
measured each year, with the fewest measured in 2014 (one).
• The tightest range of carbon tetrachloride concentrations was measured at S4MO in
2015, with less than 0.3 [j,g/m3 separating the minimum and maximum concentrations
measured. The concentration profile widens for 2016, though. Even though 1-year
average concentration exhibits an increase from 2015 to 2016, the change is not
statistically significant. In fact, less than 0.03 [j,g/m3 separates the 1-year average
concentrations between 2013 and 2016.
Figure 15-21. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
S4MO
O 5th Percentile
O 95th Percentile
Observations from Figure 15-21 for /;-dichlorobenzene concentrations measured at
S4MO include the following:
• The minimum, 5th percentile, and median concentrations are all zero for 2003, 2004,
and 2005, indicating that at least 50 percent of the measurements were non-detects.
The percentage of non-detects was at a maximum in 2003 (90 percent), after which
the percentage decreased, reaching a minimum of 5 percent for 2009. The percentage
of non-detects varies between 10 percent (2012) and 25 percent (2014) each year
following 2009.
15-28
-------
After little change in the early years, the 1-year average and median concentrations
increased steadily between 2005 and 2008. However, the relatively large number of
non-detects (zeros) combined with the range of measured detections result in a
relatively high level of variability, based on the confidence intervals calculated for the
1-year averages. This is particularly true for 2008, when the maximum
/;-dichlorobenzene concentration was measured (6.08 (J,g/m3). If the maximum
concentration for 2008 was excluded from the dataset, the concentration profile for
2008 would more closely resemble the concentration profile for 2007.
The concentrations measured decreased considerably from 2008 to 2009 then
increased again in 2010. The 95th percentile decreased by almost half from 2008 to
2009, then increased by a factor of four for 2010. The number of concentrations
greater than 1 [j.g/m3 decreased from three in 2008 to one in 2009, then increased to
six for 2010, which is the most across the years of sampling. At the same time, the
number of non-detects decreased from eight in 2008 to three 2009, then returned to
eight in 2010.
Although the range of concentrations measured in 2011 is similar to the range of
concentrations measured in 2010, the 95th percentile and 1-year average
concentration decreased considerably. Further decreases are shown for these
parameters for 2012 and again for 2013. Several of the statistical parameters are at a
minimum for 2013, including the 1-year average concentration, which is less than
0.1 [j,g/m3 for the first time. This year has the smallest range of concentrations
measured by a considerable margin.
Several of the statistical parameters exhibit increases each year after 2013, including
the 95th percentile, which in 2016 is greater than 1 [j,g/m3 for the first time since
2010, and the 1-year average concentration, which in 2016 is greater than 0.2 [j,g/m3
for the first time in five years.
15-29
-------
Figure 15-22. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at S4MO
O V-
I
rin
1
X
o
¦O"
,0'
r^i
-y>-
LjJ
I
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
Observations from Figure 15-22 for 1,2-dichloroethane concentrations measured at
S4MO include the following:
• The minimum, 5th percentile, and median concentrations are all zero through 2011,
indicating that at least 50 percent of the measurements were non-detects. There were
no measured detections of 1,2-dichloroethane in 2003, 2004, or 2007, one measured
detection in 2005, and two each in 2006 and 2008. The number of measured
detections increased steadily afterward, reaching a maximum of 60 for 2015, when
non-detects were not measured. Non-detects accounted for 10 percent of the
measurements in 2016, the most since 2011.
• As the number of measured detections increased in the later years of sampling, each
of the corresponding statistical metrics also increased. The 5th percentile and median
concentrations are greater than zero beginning with 2012, when measured detections
accounted for a majority of the measurements for the first time. The central tendency
parameters are both at a maximum for 2013, when they both are approximately
0.09 ^ig/m3.
• The 1-year average concentration decreases slightly each year after 2013, although
the change over the 4-year period is less than 0.015 [j,g/m3.
15-30
-------
Figure 15-23. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
S4MO
o.
o.
o.
i
o
o
1
I I
|X
<>¦
o
-2-
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 15-23 for ethylbenzene concentrations measured at S4MO
include the following:
• Both ethylbenzene concentrations greater than 2.5 [j,g/m3 were measured at S4MO in
2003 (2.85 [j,g/m3 and 2.63 (j,g/m3). Most of the 18 ethylbenzene concentrations
greater than 1.5 [j,g/m3 were measured in 2008 or earlier, with two exceptions
measured in 2010.
• Concentrations of ethylbenzene exhibit a significant decreasing trend between 2003
and 2009, when most of the statistical parameters are a minimum.
• With the exception of the minimum concentration, all of the statistical parameters
exhibit increases for 2010, in some cases doubling (1-year average and median),
tripling (95th percentile) or increasing by an even higher amount (maximum). The
1-year average concentration for 2010 is nearly equivalent to the maximum
concentration measured in 2009. Fifteen concentrations measured in 2010 are greater
than the maximum concentration measured in 2009.
• A steady decreasing trend in the ethylbenzene concentrations measured at S4MO is
shown again after 2010. The 1-year average concentration decreases by more than
half between 2010 and 2015.
All of the statistical parameters exhibit slight increases for 2016.
15-31
-------
Figure 15-24. Yearly Statistical Metrics for Fluorene Concentrations Measured at S4MO
y
T
X
I
I
2~
20081 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-24 for fluorene concentrations measured at S4MO include
the following:
• The box and whisker plots for fluorene measurements resemble the plots for
acenaphthene presented in Figure 15-15.
• Two concentrations greater than 30 ng/m3 have been measured at S4MO, one on
July 2, 2011 (31.4 ng/m3) and one on July 2, 2012 (31.3 ng/m3). The highest fluorene
concentrations tended to be measured during the warmer months of the year. Of the
48 fluorene concentrations greater than 15 ng/m3, 35 were measured at S4MO
between June and August of any given year and none were measured in January,
February, March, or December.
• Despite fluctuations in the measurements at the upper end of the concentration scale
and little change at lower end of the concentration scale, the median concentration
decreases each year through 2011, with the largest change shown between 2008 and
2009. The percentage of fluorene concentrations less than 5 ng/m3 nearly doubled
over this 4-year period, accounting for 32 percent of the measurements in 2008 and
more than 60 percent of the measurements in 2011.
• With the exception of the maximum concentration, which virtually did not change,
the statistical parameters exhibit increases from 2011 to 2012. This is because the
number of measurements at the upper end of the range increased while the number of
measurements at the lower end of the concentration range decreased. The number of
15-32
-------
concentrations greater than 10 ng/m3 increased from 13 in 2011 to 22 in 2012;
conversely, the number of concentrations less than 2 ng/m3 decreased from 11 in
2011 to three in 2012.
• All of the statistical parameters exhibit decreases for 2013.
• The first non-detects (four) of fluorene were measured at S4MO in 2014, as indicated
by the minimum and 5th percentile decreasing to zero. The number of non-detects
increased to five in 2015 and 13 in 2016, accounting for more than one-fifth of the
measurements in 2016. During this 3-year period, little change is shown in the 1-year
average concentrations of fluorene.
Figure 15-25. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
S4MO
50.0 |
45.0
40.0
35.0 U
sr 30.0
E
1
| 25.0
C
m
c
u 20.0
15.0 U
rL
T
. I _
T
o
L2t
r
&
£
£
i-5-i
•O-
£
1
rL
I
-Qr
o
k| jaj y
4
£
2
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
Observations from Figure 15-25 for formaldehyde concentrations measured at S4MO
include the following:
• The maximum formaldehyde concentration (43.8 (J,g/m3) was measured in 2004 on
the same day that the maximum acetaldehyde concentration was measured
(August 31, 2004). This concentration is more than twice the next highest
concentration (17.8 (j,g/m3), which was measured in 2011. The six highest
concentrations of formaldehyde were measured in 2004 (2) or 2011 (4).
• The 1-year average concentration has a decreasing trend between 2004 and 2006.
After the increase shown for 2007, the decreasing trend resumed through 2009, when
the 1-year average was at a minimum (2.46 |ig/m3). The 1-year average concentration
15-33
-------
did not change significantly between 2009 and 2010, even though the smallest range
of concentrations was measured in 2010.
• Most of the statistical parameters exhibit considerable increases from 2010 to 2011.
Eleven concentrations of formaldehyde measured in 2011 are greater than the
maximum concentration measured in 2010.
• Most of the statistical parameters exhibit decreases from 2011 to 2012.
• The central tendency statistics exhibit little change between 2012 and 2016.
Figure 15-26. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
S4MO
2012
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2008.
Observations from Figure 15-26 for naphthalene concentrations measured at S4MO
include the following:
• Naphthalene concentrations measured at S4MO exhibit considerable variability,
spanning three orders of magnitude and ranging from 0.660 ng/m3 (2016) to
784 ng/m3 (2010). Naphthalene concentrations less than 10 ng/m3 have only been
measured in 2014, 2015, and 2016.
• The years when relatively high concentrations were measured alternate with years
when the highest concentrations are considerably less, resulting in the 1-year average
(and median) concentrations having an undulating pattern. The difference decreases,
though, in the later years of sampling.
15-34
-------
15.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the S4MO monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
15.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for S4MO, risk was examined by calculating cancer risk and
noncancer hazard approximations for each year annual average concentrations could be
calculated. These approximations can be used as risk estimates for cancer and noncancer effects
attributable to the pollutants of interest. Although the use of these approximations is limited, they
may help identify where policy-makers want to shift their air monitoring priorities. Refer to
Section 3.4.2.3 for an explanation of how cancer risk and noncancer hazard approximations are
calculated and what limitations are associated with them. Annual averages, cancer UREs and/or
noncancer RfCs, and cancer risk and noncancer hazard approximations are presented in
Table 15-4, where applicable. Cancer risk approximations are presented as probabilities while
the noncancer hazard approximations are ratios and thus, unitless values.
Observations for S4MO from Table 15-4 include the following:
• The pollutants with the highest annual average concentrations for S4MO are
formaldehyde and acetaldehyde; these are the only pollutants of interest with annual
average concentrations greater than 1 [j,g/m3.
• Formaldehyde has the highest cancer risk approximations for S4MO (40.19 in-a-
million for 2015 and 41.49 in-a-million for 2016). Formaldehyde's cancer risk
approximations are at least an order of magnitude higher than the cancer risk
approximations for the other pollutants of interest. Benzene has the next highest
cancer risk approximations for S4MO (5.17 in-a-million for 2015 and 4.81 in-a-
million for 2016).
• None of the pollutants of interest for S4MO have noncancer hazard approximations
greater than 1.0, indicating that no adverse noncancer health effects are expected from
these individual pollutants. The pollutant with the highest noncancer hazard
approximations is formaldehyde (0.32 for 2015 and 0.33 for 2016).
15-35
-------
Table 15-4. Risk Approximations for the Missouri Monitoring Site
2015
2016
# of
Measured
Risk Approximations
# of
Measured
Risk Approximations
Cancer
Noncancer
Detections
Annual
Detections
Annual
URE
RfC
vs. # of
Average
Cancer
Noncancer
vs. # of
Average
Cancer
Noncancer
Pollutant
frig/m3)1
(mg/m3)
Samples
frig/m3)
(in-a-million)
(HQ)
Samples
frig/m3)
(in-a-million)
(HQ)
St. Louis, Missouri - S4MO
1.66
1.59
Acetaldehyde
0.0000022
0.009
60/60
±0.14
3.65
0.18
60/60
±0.16
3.51
0.18
0.66
0.62
Benzene
0.0000078
0.03
60/60
±0.08
5.17
0.02
60/60
±0.07
4.81
0.02
0.08
0.09
1,3-Butadiene
0.00003
0.002
59/60
±0.01
2.30
0.04
60/60
±0.02
2.73
0.05
0.63
0.65
Carbon Tetrachloride
0.000006
0.1
60/60
±0.02
3.77
0.01
60/60
±0.02
3.88
0.01
0.14
0.22
p-Dichlorobcnzcnc
0.000011
0.8
49/60
±0.04
1.57
<0.01
48/60
±0.10
2.44
<0.01
0.07
0.07
1,2 -Dichloroethane
0.000026
2.4
60/60
±<0.01
1.91
<0.01
54/60
±0.01
1.89
<0.01
0.20
0.25
Ethylbenzene
0.0000025
1
60/60
±0.03
0.51
<0.01
60/60
±0.05
0.62
<0.01
3.09
3.19
Formaldehyde
0.000013
0.0098
60/60
±0.42
40.19
0.32
60/60
±0.48
41.49
0.33
6.51
6.13
Acenaphthene3
0.000088
—
55/58
± 1.47
0.57
—
52/60
± 1.71
0.54
—
0.88
0.90
Arsenic (PMi0)a
0.0043
0.000015
60/60
±0.12
3.77
0.06
61/61
±0.13
3.86
0.06
6.40
6.67
Fluorene3
0.000088
—
53/58
± 1.31
0.56
—
47/60
± 1.81
0.59
—
78.32
73.48
Naphthalene1
0.000034
0.003
58/58
± 11.77
2.66
0.03
60/60
± 14.07
2.50
0.02
- = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
-------
15.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 15-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 15-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 15-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for S4MO, as presented in Table 15-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 15-5.
Table 15-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 15.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
15-37
-------
Table 15-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Missouri Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
St. Louis, Missouri (St. Louis City + County) - S4MO
Formaldehyde
428.58
Formaldehyde
5.57E-03
Formaldehyde
41.49
Benzene
409.22
Naphthalene
3.38E-03
Formaldehyde
40.19
Acetaldehyde
250.69
Benzene
3.19E-03
Benzene
5.17
Ethylbenzene
238.48
1,3-Butadiene
1.96E-03
Benzene
4.81
Naphthalene
99.40
Hexavalent Chromium PM
1.10E-03
Carbon Tetrachloride
3.88
1.3 -Butadiene
65.33
POM, Group 2b
7.74E-04
Arsenic
3.86
Bis(2-ethylhexyl) phthalate, gas
63.69
POM, Group 2d
6.31E-04
Arsenic
3.77
Trichloroethylene
32.97
Arsenic, PM
6.18E-04
Carbon Tetrachloride
3.77
POM, Group 2b
8.79
Ethylbenzene
5.96E-04
Acetaldehyde
3.65
POM, Group 2d
7.17
Acetaldehyde
5.52E-04
Acetaldehyde
3.51
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 15-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Missouri Monitoring Site
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
St. Louis, Missouri (St. Louis City + County) - S4MO
Toluene
1,563.52
Acrolein
1,485,460.58
Formaldehyde
0.33
Xylenes
875.34
2,4-Toluene diisocyanate
87,183.14
Formaldehyde
0.32
Methanol
694.65
Formaldehyde
43,732.45
Acetaldehyde
0.18
Formaldehyde
428.58
Naphthalene
33,133.17
Acetaldehyde
0.18
Benzene
409.22
1.3 -Butadiene
32,666.95
Arsenic
0.06
Hexane
280.51
Acetaldehyde
27,854.05
Arsenic
0.06
Acetaldehyde
250.69
T richloroethylene
16,482.55
1,3-Butadiene
0.05
Ethylbenzene
238.48
Benzene
13,640.63
1,3-Butadiene
0.04
Naphthalene
99.40
Cadmium, PM
12,122.02
Naphthalene
0.03
Ethylene glycol
85.32
Arsenic, PM
9,577.54
Naphthalene
0.02
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 15-5 include the following:
• Emissions and toxicity-weighted emissions for S4MO are presented as the city-based
(FIPS 29-189) emissions summed with the county-based emissions (29-510).
• Formaldehyde, benzene, and acetaldehyde are the highest emitted pollutants with
cancer UREs in St. Louis.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, naphthalene, and benzene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions.
• Formaldehyde tops all three lists, with the highest quantity emitted, the highest
toxicity-weighted emissions, and the highest cancer risk approximations. Benzene,
and acetaldehyde also appear on all three lists.
• Arsenic has the sixth and seventh highest cancer risk approximations for S4MO.
While arsenic is not one of the highest emitted pollutants in St. Louis, it ranks eighth
for its toxicity-weighted emissions. Carbon tetrachloride also appears among the
pollutants of interest with the highest cancer risk approximations for S4MO but
appears on neither emissions-based list.
• POM, Group 2b is the ninth highest emitted "pollutant" in St. Louis and ranks sixth
for toxicity-weighted emissions. POM, Group 2b includes several PAHs sampled for
at S4MO including acenaphthene and fluorene, which are pollutants of interest for
S4MO. These pollutants are not among those with the highest cancer risk
approximations for S4MO.
Observations from Table 15-6 include the following:
• Emissions and toxicity-weighted emissions for S4MO are presented as the city-based
(FIPS 29-189) emissions summed with the county-based emissions (29-510).
• Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in St. Louis.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 2,4-toluene diisocyanate, and formaldehyde. Although
acrolein was sampled for at S4MO, this pollutant was excluded from the pollutants of
interest designation, and thus, subsequent risk-based screening evaluations, due to
questions about the consistency and reliability of the measurements, as discussed in
Section 3.2.
• Four of the highest emitted pollutants in St. Louis also have the highest toxicity-
weighted emissions.
• Formaldehyde, the pollutant with highest noncancer hazard approximations for
S4MO, has the third highest toxicity-weighted emissions and the fourth highest total
15-40
-------
emissions (of the pollutants with noncancer RfCs). Acetaldehyde and naphthalene
also appear on all three lists.
• Arsenic and 1,3-butadiene, both pollutants of interest for S4MO, appear among the
pollutants with the highest toxicity-weighted emissions, but are not among the highest
emitted.
15.5 Summary of the 2015-2016 Monitoring Data for S4MO
Results from several of the data analyses described in this section include the following:
~~~ Twenty-two pollutants failed screens for S4MO. S4MO failed the highest number of
screens among all NMP sites.
~~~ Formaldehyde and acetaldehyde have the highest annual average concentrations for
S4MO. These are the only pollutants of interest with annual averages greater than
1 ng/m3.
~~~ S4MO has the third and sixth highest annual average concentrations of
p-dichlorobenzene among NMP sites sampling VOCs. S4MO also has some of the
highest annual average concentrations of arsenic, acenaphthene, fluorene, and
naphthalene among NMP sites sampling these pollutants.
~~~ Concentrations of acetaldehyde, benzene, and ethylbenzene have decreased
significantly since sampling began at S4MO.
~~~ Formaldehyde has the highest cancer risk approximations of the pollutants of interest
for S4MO. None of the pollutants of interest for S4MO have noncancer hazard
approximations greater than an HQ of 1.0.
15-41
-------
16.0 Sites in New Jersey
This section summarizes those data from samples collected at UATMP sites in New
Jersey and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial
through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
16.1 Site Characterization
This section characterizes the New Jersey monitoring sites by providing a description of
the nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
One New Jersey monitoring site (CSNJ) is located in the Philadelphia-Camden-
Wilmington, PA-NJ-DE-MD CBSA while the other four New Jersey sites (CHNJ, ELNJ, NBNJ,
and NRNJ) are located within the New York-Newark-Jersey City, NY-NJ-PA CBSA.
Figure 16-1 presents a composite satellite image retrieved from ArcGIS Explorer showing the
CSNJ monitoring site and its immediate surroundings. Figure 16-2 identifies nearby point source
emissions locations by source category, as reported in the 2014 NEI for point sources, version 1.
Note that only sources within 10 miles of CSNJ are included in the facility counts provided in
Figure 16-2. A 10-mile boundary was chosen to give the reader an indication of which emissions
sources and emissions source categories could potentially have a direct effect on the air quality at
the monitoring site. Further, this boundary provides both the proximity of emissions sources to
the monitoring site as well as the quantity of such sources within a given distance of the site.
Sources outside the 10-mile boundary are still visible on the map for reference but have been
grayed out to emphasize emissions sources within the boundary. Figures 16-3 through 16-8
present the composite satellite maps and emissions source maps for CHNJ, ELNJ, NBNJ, and
NRNJ. Table 16-1 provides supplemental geographical information for each site, such as land
use, location setting, and locational coordinates. Each figure and table is discussed in detail in the
paragraphs that follow.
and temporal characteristics of the ambient monitoring
concentrations and reviews them through the context of
risk. Readers are encouraged to refer to Sections 1
IXiUi ijcncmlcd h> sources oilier
lluin LR(.i. LIJ.\'s conliiicl
kiliomioiA I'orihc \\ll\ ciiv noi
included m I lie d^ila aiuil> ses
conliiined in llus report
16-1
-------
Figure 16-1. Camden, New Jersey (CSNJ) Monitoring Site
ON
IJ
'W^-ngWsL I
,Walnut-St|
Avel"1,
Mechanic^t
,Unsdowne'
-------
Figure 16-2. NEI Point Sources Located Within 10 Miles of CSNJ
Burlington \-ft
Iff County \ ^
Camden
County
i Gloucester
\ County
Delaware
pRiver
Legend
CSNJ UATMP site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
Airport/Airline/Airport Support Operations (33)
$ Asphalt Production/Hot Mix Asphalt Plant (2)
s Automobile/Truck Manufacturing Facility (1)
B Bulk Terminal/Bulk Plant (13)
C Chemical Manufacturing Facility (5)
1 Compressor Station (5)
f Electricity Generation Facility (12)
F Food Processing/Agriculture Facility (3)
O Institution (school, hospital, prison, etc.) (36)
A Landfill (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (2)
<•> Metals Processing/Fabrication Facility (8)
A Military Base/National Security Facility (3)
X Mine/Quarry/Mineral Processing Facility (3)
? Miscellaneous Commercial/Industrial Facility (22)
M Municipal Waste Combustor (2)
• Oil and/or Gas Production (2)
'[] Paint and Coating Manufacturing Facility (2)
0 Petroleum Products Manufacturing Facility (2)
3 Petroleum Refinery (5)
cz> Pharmaceutical Manufacturing Facility (2)
R Plastic, Resin, or Rubber Products Plant (9)
V Port and Harbor Operations (1)
P Printing/Publishing/Paper Product Manufacturing Facility (9)
E Pulp and Paper Plant (2)
ii, Ship/Boat Manufacturing or Repair Facility (1)
TT Telecommunications/Radio Facility (1)
** Truck/Bus/Transportation Operations (2)
* Wastewater Treatment Facility (4)
?_
\
\
J -
. - " " \
PENNSYLVANIA
\
\
\
1 I 1
75°10'0"W 75°5'0"W 75°0'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
16-3
-------
Figure 16-3. Chester, New Jersey (CHNJ) Monitoring Site
-------
Figure 16-4. NEI Point Sources Located Within 10 Miles of CHN.I
Sussex
County
Morris
County
Somerset
County
Hunterdon \
County i
Legend
~ CHNJ UATMP site
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
O 10 mile radius | | County boundary
\ Warren
\ County
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (1)
f Airport/Airline/Airport Support Operations (12)
c Chemical Manufacturing Facility (2)
e Electrical Equipment Manufacturing Facility (3)
f Food Processing/Agriculture Facility (1)
? Miscellaneous Commercial/Industrial Facility (1)
w Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
16-5
-------
Figure 16-5. Elizabeth, New Jersey (ELNJ) Monitoring Site
Sth Ward Park
f Source: USGS
irA'/NASA, NGA, USGS
2-00'8 Microsoft Corp.
-------
Figure 16-6. North Brunswick, New Jersey (NBNJ) Monitoring Site
MDudley Kd>
Rutgers Unrversity-Coofc-Douglass Camput f
S O u r c e: U S G S
Stfifrfce; fN ASA" NG A, USGS
¦.. I _ &
Westons Mill
DrtnW
-------
Figure 16-7. East Brunswick, New Jersey (NRNJ) Monitoring Site
Westons Mill Pond
>a >» .
Bernard
VVhiteh*11 Rd ,
Sullivan Way Sullivan Way
Salt Meadow Rd
G\
00
-------
Figure 16-8. NEI Point Sources Located Within 10 Miles of ELNJ, NBNJ, and NRNJ
HudsonJ-
County
1 Essex
\ County
Morris
County
0 NEW »
\ YORK $
\ Richmond
¦ \ County
Union \
County i
Somerset
County
Raritan
Bay
Monmouth \
County \
Middlesex \
County \
_) 10 mile radius | County boundary
Source Category Group (No. of Facilities)
"t" Airport/Airline/Airport Support Operations (49)
£ Asphalt Production/Hot Mix Asphalt Plant (7)
0 Auto Body Shop/Painters/Automotive Stores (1)
Y Brewery/Distillery/Winery (1)
# Building/Construction (1)
B Bulk Terminal/Bulk Plant (23)
C Chemical Manufacturing Facility (28)
1 Compressor Station (3)
6 Electrical Equipment Manufacturing Facility (3)
f Electricity Generation Facility (21)
E Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (2)
F Food Processing/Agriculture Facility (8)
A Foundry, non-ferrous (1)
^ Industrial Machinery or Equipment Plant (2)
O Institution (school, hospital, prison, etc.) (13)
A Landfill (4)
(¦) Metal Can, Box, and Other Metal Container Manufacturing Facility (3)
A Metal Coating, Engraving, and Allied Services to Manufacturers (7)
® Metals Processing/Fabrication Facility (11)
X Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility (18)
M Municipal Waste Combustor (2)
Q Paint and Coating Manufacturing Facility (7)
H Petroleum Refinery (3)
cd Pharmaceutical Manufacturing Facility (8)
R Plastic, Resin, or Rubber Products Plant (14)
^ Port and Harbor Operations (2)
P Printing/Publishing/Paper Product Manufacturing Facility (11)
E3 Pulp and Paper Plant (2)
V Steel Mill (2)
© Testing Laboratory (2)
Truck/Bus/Transportation Operations (1)
* V\testewater Treatment Facility (7)
6 V\&ter Treatment Facility (1)
W Woodwork, Furniture, Millwork and Wood Preserving Facility (1)
LfiflSnd Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
~ ELNJ UATM P site ~ NBNJ UATMP site NRNJ UATMP site
16-9
-------
Table 16-1. Geographical Information for the New Jersey Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic3
Intersection
Used for Traffic Data
CSNJ
34-007-0002
Camden
Camden
Philadelphia-Camden-
Wilmington PA-NJ-DE-
MD
39.934446,
-75.125291
Industrial
Urban/City
Center
3,231
S 2nd St, south of Walnut St
CHNJ
34-027-3001
Chester
Morris
New York-Newark-
Jersey City, NY-NJ-PA
40.787628,
-74.676301
Agricultural
Rural
11,215
MendhamRd (510/24), east of Fox
Chase Rd
ELNJ
34-039-0004
Elizabeth
Union
New York-Newark-
Jersey City, NY-NJ-PA
40.641440,
-74.208365
Industrial
Suburban
250,000
1-95, between Exits 13 & 13 A
NBNJ
34-023-0006
North
Brunswick
Middlesex
New York-Newark-
Jersey City, NY-NJ-PA
40.472825,
-74.422403
Agricultural
Rural
114,322
US-1, E of Ryders Ln/617
NRNJ
34-023-0011
East
Brunswick
Middlesex
New York-Newark-
Jersey City, NY-NJ-PA
40.462182,
-74.429439
Agricultural
Rural
22,297
Ryders Ln/617, between 1-95 and
Blueberry Dr
3AADT for ELNJ reflects 2006 data from NJ Department of Treasury (NJ DOTr, 2008); AADT reflects 2010 data for NBNJ, 2012 data for CSNJ and CHNJ, and 2014 data
for NRNJ from the NJ DOT (NJ DOT, 2016).
-------
The CSNJ monitoring site is located in the city of Camden in southwest New Jersey, just
across the state line from Philadelphia, Pennsylvania. The monitoring site is located in an
industrial area a few blocks east of the Delaware River, as shown in Figure 16-1. Residential
areas are located to the east between the site and 1-676. Figure 16-2 shows that the large number
of point sources located within 10 miles of CSNJ are involved in a variety of industries. The
source categories with the largest number of facilities within 10 miles of CSNJ include
institutions (such as schools, hospitals, and prisons); airports and airport support operations,
which include airports and related operations as well as small runways and heliports, such as
those associated with hospitals or television stations; bulk terminals and bulk plants; and
electricity generating facilities. The sources closest to CSNJ include a metals processing and
fabrication facility; a mine/quarry/minerals processing facility; a wastewater treatment facility;
and an airport/airport operations facility.
CHNJ is located in northern New Jersey, in the town of Chester, west of the New York
City metropolitan area. Figure 16-3 shows that CHNJ is located in an open area near Building 1
of the Department of Public Works, off Routes 513 (North Road) and 510 (Main Street). The
surrounding area is rural and agricultural, with a rolling topography, but surrounded by small
neighborhoods. Two schools are located on the other site of Route 510 to the south-southwest of
CHNJ. Although the location is considered part of the New York City metro area, the site's
location is outside most of the urbanized areas. Figure 16-4 shows that few sources are located
within a few miles of CHNJ. The source category with the greatest number of emissions sources
within 10 miles of CHNJ is the airport source category. The sources closest to CHNJ include a
privately-owned heliport to the south and a woodwork, furniture, millwork, and wood preserving
facility to the west.
ELNJ is located in the city of Elizabeth, which lies just south of Newark and west of
Newark Bay and Staten Island, New York. As Figure 16-5 shows, the monitoring site is located
near the toll plaza just off Exit 13 of the New Jersey Turnpike (1-95). Interstate-278 intersects the
Turnpike here as well. The surrounding area is highly industrialized, with an oil refinery located
just southwest of the site. Additional industry is located to the southwest, west, and on the east
side of the Turnpike, while a school and park are located to the north and residential
neighborhoods are located to the northwest. The New Jersey/New York border can be seen in the
lower right-hand corner of Figure 16-5, with Staten Island on the east side of the Arthur Kill
channel.
16-11
-------
NBNJ is located in North Brunswick, approximately 16 miles southwest of Elizabeth.
The monitoring site is located on the property of Rutgers University's Cook-Douglass campus,
on a horticultural farm. The surrounding area is agricultural and rural, although residential
neighborhoods are located to the east, across a branch of the Raritan River, as shown in
Figure 16-6. County Road 617 (Ryders Lane) and US-1 intersect just west of the site and the
New Jersey Turnpike/I-95 runs northeast-southwest less than 1 mile east of the site.
The sampling equipment at NBNJ was moved to a new location (NRNJ) for 2016. NRNJ
is located just less than 1 mile farther south down Ryders Lane. This site is still located on
Rutgers University, but moved to horticultural farm number 3, the fields of which are a
prominent feature in Figure 16-7. The area is classified as agricultural, with residential properties
located to the east and south. Weston's Mill Pond, a section of the Lawrence Brook, a tributary
of the Raritan River, runs to the north and west of the farm (LBWP, 2013) and separates the farm
from a window and door manufacturing facility located about one half-mile to the west. Both
NBNJ and NRNJ are wedged between 1-95 to the east, which can be seen in the lower right-hand
corner of Figure 16-7, and US-1 to the northwest, shown in the upper left-hand corner, with
NRNJ roughly one-half mile from each.
Figure 16-8 provides the emissions sources for ELNJ, NBNJ, and NRNJ on one map, as
the outer portions of the 10-mile boundaries for these three sites intersect; NBNJ and NRNJ are
16 miles and 17 miles southwest of ELNJ, respectively. Many emissions sources surround these
sites. The majority of the emissions sources are located in the northern half of Middlesex County
and northeastward toward New York City and northern New Jersey. The source categories with
the greatest number of emissions sources in the vicinity of these sites include airport operations,
chemical manufacturing, bulk terminals and bulk plants, and electricity generation via
combustion. The emissions sources in closest proximity to the ELNJ monitoring site are in the
wastewater treatment, chemical manufacturing, bulk terminals/bulk plant, petroleum refining,
miscellaneous commercial/industrial facility, and electricity generation via combustion source
categories. The emissions sources in closest proximity to the NBNJ and NRNJ monitoring sites
are involved in airport and airport support operations, pharmaceutical manufacturing, and plastic,
resin, or rubber products manufacturing.
16-12
-------
In addition to providing city, county, CBS A, and land use/location setting information,
Table 16-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site. ELNJ
experiences the highest traffic volume among the New Jersey sites (250,000); this is not
unexpected given ELNJ's location along the New Jersey Turnpike/I-95, between Exit 13 and
13A. ELNJ's traffic volume is also the highest among all NMP sites. The traffic volume for
NBNJ is about half that experienced near ELNJ, with the volumes near the remaining sites
considerably less. Traffic data for NBNJ are provided for US-1, east of State Road 617 (Ryders
Lane); traffic data for NRNJ are provided for Ryders Lane, between Blueberry Drive and 1-95;
traffic data for CHNJ are provided for Route 510, east of Fox Chase Road; and traffic data for
CSNJ are provided for South 2nd Street, south of Walnut Street.
16.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 16-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 16-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. VOCs and carbonyl compounds were
sampled for at each of the New Jersey sites.
16-13
-------
Table 16-2. 2015-2016 Risk-Based Screening Results for the New Jersey Monitoring Sites
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Camden, New Jersey - CSNJ
Formaldehyde
0.077
119
119
100.00
14.53
14.53
Acetaldehyde
0.45
118
119
99.16
14.41
28.94
Benzene
0.13
116
116
100.00
14.16
43.10
Carbon Tetrachloride
0.17
116
116
100.00
14.16
57.26
1.3 -Butadiene
0.03
111
114
97.37
13.55
70.82
1,2-Dichloroethane
0.038
109
109
100.00
13.31
84.13
Ethylbenzene
0.4
53
116
45.69
6.47
90.60
Bro mo methane
0.5
23
116
19.83
2.81
93.41
p-Dichlorobenzene
0.091
20
63
31.75
2.44
95.85
Hexachloro-1,3 -butadiene
0.045
16
17
94.12
1.95
97.80
Trichloroethylene
0.2
8
54
14.81
0.98
98.78
1,2-Dibromoethane
0.0017
5
5
100.00
0.61
99.39
Propionaldehyde
0.8
3
119
2.52
0.37
99.76
Dichloromethane
60
1
116
0.86
0.12
99.88
Vinyl chloride
0.11
1
64
1.56
0.12
100.00
Total
819
1,363
60.09
Chester, New Jersey - CHNJ
Carbon Tetrachloride
0.17
112
113
99.12
20.14
20.14
Benzene
0.13
111
113
98.23
19.96
40.11
1,2-Dichloroethane
0.038
103
103
100.00
18.53
58.63
Formaldehyde
0.077
94
94
100.00
16.91
75.54
Acetaldehyde
0.45
89
94
94.68
16.01
91.55
1.3 -Butadiene
0.03
30
72
41.67
5.40
96.94
Hexachloro-1,3 -butadiene
0.045
12
13
92.31
2.16
99.10
1,2-Dibromoethane
0.0017
3
3
100.00
0.54
99.64
p-Dichlorobenzene
0.091
2
25
8.00
0.36
100.00
Total
556
630
88.25
16-14
-------
Table 16-2. 2015-2016 Risk-Based Screening Results for the New Jersey Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.45
121
121
100.00
15.59
15.59
Formaldehyde
0.077
121
121
100.00
15.59
31.19
Benzene
0.13
120
120
100.00
15.46
46.65
Carbon Tetrachloride
0.17
120
120
100.00
15.46
62.11
1.3 -Butadiene
0.03
117
120
97.50
15.08
77.19
1,2-Dichloroethane
0.038
112
112
100.00
14.43
91.62
Ethylbenzene
0.4
27
120
22.50
3.48
95.10
p-Dichlorobenzene
0.091
14
65
21.54
1.80
96.91
Propionaldehyde
0.8
13
121
10.74
1.68
98.58
Hexachloro-1,3 -butadiene
0.045
7
9
77.78
0.90
99.48
1,2-Dibromoethane
0.0017
4
4
100.00
0.52
100.00
Total
776
1033
75.12
North Brunswick, New Jersey - NBNJ
Acetaldehyde
0.45
59
59
100.00
17.15
17.15
Benzene
0.13
59
59
100.00
17.15
34.30
Carbon Tetrachloride
0.17
59
59
100.00
17.15
51.45
1,2-Dichloroethane
0.038
59
59
100.00
17.15
68.60
Formaldehyde
0.077
59
59
100.00
17.15
85.76
1.3 -Butadiene
0.03
41
54
75.93
11.92
97.67
Hexachloro-1,3 -butadiene
0.045
3
3
100.00
0.87
98.55
1,2-Dibromoethane
0.0017
2
2
100.00
0.58
99.13
Propionaldehyde
0.8
2
59
3.39
0.58
99.71
p-Dichlorobenzene
0.091
1
22
4.55
0.29
100.00
Total
344
435
79.08
East Brunswick, New Jersey - NRNJ
Acetaldehyde
0.45
60
60
100.00
16.04
16.04
Benzene
0.13
60
60
100.00
16.04
32.09
Carbon Tetrachloride
0.17
60
60
100.00
16.04
48.13
Formaldehyde
0.077
60
60
100.00
16.04
64.17
1,2-Dichloroethane
0.038
57
57
100.00
15.24
79.41
1,3-Butadiene
0.03
47
52
90.38
12.57
91.98
Hexachloro-1,3 -butadiene
0.045
16
17
94.12
4.28
96.26
Ethylbenzene
0.4
6
60
10.00
1.60
97.86
1,2-Dibromoethane
0.0017
4
4
100.00
1.07
98.93
p-Dichlorobenzene
0.091
4
30
13.33
1.07
100.00
Total
374
460
81.30
16-15
-------
Observations from Table 16-2 include the following:
• Concentrations of 15 pollutants failed at least one screen for CSNJ; 60 percent of
concentrations for these 15 pollutants were greater than their associated risk screening
value (or failed screens).
• Nine pollutants contributed to 95 percent of failed screens for CSNJ and therefore
were identified as pollutants of interest for this site. These nine include two carbonyl
compounds and seven VOCs. CSNJ is the only NMP site for which bromomethane is
a pollutant of interest.
• Concentrations of nine pollutants failed at least one screen for CHNJ; 88 percent of
concentrations for these nine pollutants were greater than their associated risk
screening value (or failed screens).
• Six pollutants contributed to 95 percent of failed screens for CHNJ and therefore
were identified as pollutants of interest for this site. These six include two carbonyl
compounds and four VOCs.
• Concentrations of 11 pollutants failed at least one screen for ELNJ, with 75 percent of
concentrations for these 11 pollutants greater than their associated risk screening
value (or failing screens).
• Seven pollutants contributed to 95 percent of failed screens for ELNJ and therefore
were identified as pollutants of interest for this site. These seven include two carbonyl
compounds and five VOCs.
• Concentrations of 10 pollutants failed at least one screen for NBNJ, with 79 percent
of concentrations for these 10 pollutants greater than their associated risk screening
value (or failing screens).
• Six pollutants contributed to 95 percent of failed screens for NBNJ and therefore
were identified as pollutants of interest for this site. These six include two carbonyl
compounds and four VOCs.
• Concentrations of 10 pollutants failed at least one screen for NRNJ, with 81 percent
of concentrations for these 10 pollutants greater than their associated risk screening
value (or failing screens).
• Seven pollutants contributed to 95 percent of failed screens for NRNJ and therefore
were identified as pollutants of interest for this site. These seven include two carbonyl
compounds and five VOCs.
• The New Jersey sites have six pollutants of interest in common: acetaldehyde,
formaldehyde, benzene, carbon tetrachloride, 1,3-butadiene, and 1,2-dichloroethane.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
16-16
-------
16.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the New Jersey monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of monitoring.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at the New Jersey
monitoring sites are provided in Appendices J and M.
16.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each New Jersey site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the New Jersey
monitoring sites are presented in Table 16-3, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration. As the
instrumentation at NBNJ was moved to NRNJ for 2016, only one year of data is provided for
each site.
16-17
-------
Table 16-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the New Jersey Monitoring Sites
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
Pollutant
# >MDL/
# Samples
Avg
frig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Average
frig/m3)
# >MDL/
# Samples
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Average
frig/m3)
Camden, New Jersey - CSNJ
1.52
2.00
2.69
1.90
2.03
2.29
2.35
3.61
2.45
2.65
Acetaldehyde
59/59/59
±0.25
±0.33
±0.72
±0.50
±0.25
60/60/60
±0.55
±0.54
± 1.22
±0.54
±0.37
0.97
0.66
0.58
1.16
0.86
0.87
0.55
0.54
0.97
0.74
Benzene
61/61/61
±0.20
±0.10
±0.10
±0.41
±0.14
55/55/55
±0.25
±0.12
±0.14
±0.27
±0.11
1.15
2.46
0.07
0.30
0.93
0.95
0.35
0.08
0.10
0.41
Bromomethane
61/60/61
±0.74
±3.15
±0.03
±0.41
±0.74
55/22/55
±0.74
±0.32
±0.01
±0.04
±0.24
0.10
0.07
0.05
0.14
0.09
0.11
0.06
0.05
0.11
0.08
1.3 -Butadiene
61/59/61
±0.02
±0.02
±0.01
±0.05
±0.02
53/36/55
±0.05
±0.02
±0.01
±0.04
±0.02
0.57
0.64
0.64
0.64
0.62
0.57
0.72
0.60
0.64
0.63
Carbon Tetrachloride
61/61/61
±0.08
±0.03
±0.04
±0.03
±0.02
55/55/55
±0.07
±0.03
±0.08
±0.04
±0.03
0.02
0.04
0.05
0.07
0.05
0.04
0.09
0.04
0.07
0.06
p-Dichlorobenzene
34/2/61
±0.01
±0.03
±0.02
±0.03
±0.01
29/5/55
±0.03
±0.07
±0.03
±0.06
±0.02
0.09
0.11
0.06
0.09
0.09
0.09
0.10
0.04
0.07
0.08
1,2 -Dichloroethane
60/55/61
±0.01
±0.03
±0.01
±0.02
±0.01
49/49/55
±0.01
±0.02
±0.02
±0.02
±0.01
0.30
0.43
0.40
0.54
0.42
0.30
0.32
0.69
0.64
0.47
Ethylbenzene
61/61/61
±0.06
±0.08
±0.07
±0.15
±0.06
55/55/55
±0.12
±0.13
±0.25
±0.19
±0.09
2.05
3.60
5.48
3.14
3.57
3.62
4.34
4.57
3.78
4.06
Formaldehyde
59/59/59
±0.26
±0.90
±0.77
±0.68
±0.46
60/60/60
±0.84
± 1.00
± 1.00
±0.70
±0.43
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 16-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the New Jersey Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
Chester, New Jersey - CHNJ
1.08
0.98
0.91
1.08
1.05
1.61
1.16
Acetaldehyde
34/34/34
±0.22
NA
NA
±0.29
NA
60/60/60
±0.22
±0.32
±0.16
±0.39
±0.15
0.54
0.30
0.46
0.40
0.47
0.25
0.21
0.41
0.34
Benzene
55/55/55
±0.10
±0.05
NA
±0.08
±0.04
58/58/58
±0.07
±0.03
±0.03
±0.06
±0.04
0.02
0.01
0.03
0.02
0.02
0.01
0.01
0.02
0.02
1.3 -Butadiene
38/16/55
±0.02
±0.01
NA
±0.01
±0.01
34/2/58
±0.01
±0.01
±0.01
±0.01
± <0.01
0.64
0.65
0.63
0.64
0.54
0.69
0.62
0.61
0.62
Carbon Tetrachloride
55/55/55
±0.03
±0.02
NA
±0.05
±0.02
58/58/58
±0.08
±0.04
±0.07
±0.03
±0.03
0.08
0.07
0.07
0.07
0.08
0.08
0.04
0.05
0.06
1,2 -Dichloroethane
54/43/55
±0.01
±0.01
NA
±0.01
±0.01
49/46/58
±0.01
±0.01
±0.02
±0.02
±0.01
1.77
1.37
1.56
2.52
3.53
1.18
2.16
Formaldehyde
34/34/34
±0.33
NA
NA
±0.38
NA
60/60/60
±0.40
±0.89
±0.93
±0.30
±0.39
Elizabeth, New Jersey - ELNJ
1.78
2.43
3.11
2.66
2.49
2.07
2.82
2.89
2.23
2.50
Acetaldehyde
60/60/60
±0.36
±0.58
±0.55
±0.78
±0.31
61/61/61
±0.53
±0.65
±0.34
±0.50
±0.26
0.93
0.67
0.72
0.95
0.82
0.94
0.70
0.69
0.99
0.83
Benzene
60/60/60
±0.17
±0.09
±0.11
±0.22
±0.08
60/60/60
±0.23
±0.14
±0.12
±0.20
±0.09
0.13
0.11
0.08
0.15
0.12
0.14
0.11
0.10
0.15
0.12
1.3 -Butadiene
60/59/60
±0.04
±0.02
±0.01
±0.05
±0.02
60/52/60
±0.04
±0.02
±0.03
±0.05
±0.02
0.54
0.62
0.66
0.64
0.62
0.57
0.67
0.63
0.61
0.62
Carbon Tetrachloride
60/60/60
±0.08
±0.02
±0.04
±0.02
±0.03
60/60/60
±0.07
±0.04
±0.06
±0.04
±0.03
0.08
0.08
0.06
0.08
0.07
0.09
0.09
0.05
0.07
0.08
1,2 -Dichloroethane
57/51/60
±0.01
±0.01
±0.01
±0.01
±0.01
55/52/60
±0.01
±0.01
±0.02
±0.02
±0.01
0.21
0.27
0.26
0.36
0.28
0.28
0.28
0.35
0.38
0.32
Ethylbenzene
60/60/60
±0.05
±0.05
±0.05
±0.11
±0.04
60/57/60
±0.11
±0.07
±0.09
±0.08
±0.04
2.32
4.07
6.94
4.31
4.38
3.80
4.81
5.36
3.81
4.43
Formaldehyde
60/60/60
±0.38
± 1.33
± 1.41
± 1.13
±0.68
61/61/61
±0.72
±0.95
±0.76
±0.69
±0.41
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 16-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the New Jersey Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
North Brunswick, New Jersey - NBNJ
Acetaldehyde
59/59/59
1.38
±0.20
2.12
±0.43
2.49
±0.33
2.16
±0.46
2.03
±0.20
NS
NS
NS
NS
NS
NS
Benzene
59/59/59
0.75
±0.12
0.37
±0.06
0.44
±0.15
0.55
±0.13
0.53
±0.07
NS
NS
NS
NS
NS
NS
1.3 -Butadiene
54/41/59
0.07
±0.02
0.03
±0.01
0.03
±0.01
0.06
±0.02
0.05
±0.01
NS
NS
NS
NS
NS
NS
Carbon Tetrachloride
59/59/59
0.61
±0.04
0.64
±0.03
0.65
±0.04
0.66
±0.02
0.64
±0.02
NS
NS
NS
NS
NS
NS
1,2 -Dichloroethane
59/51/59
0.08
±0.01
0.07
±0.01
0.06
±0.01
0.08
±0.01
0.07
±<0.01
NS
NS
NS
NS
NS
NS
Formaldehyde
59/59/59
2.34
±0.47
3.94
± 1.12
5.34
± 1.10
1.41
±0.31
3.29
±0.56
NS
NS
NS
NS
NS
NS
East Brunswick, New Jersey - NRNJ
Acetaldehyde
NS
NS
NS
NS
NS
NS
60/60/60
1.13
±0.34
1.64
±0.37
2.05
±0.33
1.59
±0.43
1.59
±0.19
Benzene
NS
NS
NS
NS
NS
NS
60/60/60
0.71
±0.16
0.37
±0.06
0.35
±0.08
0.59
±0.13
0.51
±0.07
1.3 -Butadiene
NS
NS
NS
NS
NS
NS
52/25/60
0.09
±0.03
0.05
±0.01
0.03
±0.02
0.07
±0.02
0.06
±0.01
Carbon Tetrachloride
NS
NS
NS
NS
NS
NS
60/60/60
0.59
±0.07
0.71
±0.04
0.59
±0.07
0.60
±0.06
0.62
±0.03
1,2 -Dichloroethane
NS
NS
NS
NS
NS
NS
57/55/60
0.09
±0.01
0.10
±0.01
0.06
±0.01
0.07
±0.01
0.08
±0.01
Formaldehyde
NS
NS
NS
NS
NS
NS
60/60/60
1.47
±0.30
2.83
±0.66
3.52
±0.43
2.03
±0.42
2.44
±0.30
Hexachloro-1,3 -butadiene
NS
NS
NS
NS
NS
NS
17/0/60
0.01
±0.02
0.04
±0.03
0.03
±0.03
0.03
±0.03
0.03
±0.01
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
-------
Observations for CSNJ from Table 16-3 include the following:
• The pollutants of interest with the highest annual average concentrations are
formaldehyde and acetaldehyde; these are the only two pollutants with annual
average concentrations greater than 1 |ig/m3. Of the VOCs, bromomethane has the
highest annual average concentration for 2015 (0.93 ± 0.74 |ig/m3) while benzene has
the highest annual average for 2016 (0.74 ± 0.11 |ig/m3).
• The quarterly average concentrations of formaldehyde calculated for 2015 (which
range from 2.05 ± 0.26 |ig/m3 for the first quarter to 5.48 ± 0.77 |ig/m3 for the third
quarter) span a wider range than those calculated for 2016 (which range from
3.62 ± 0.84 |ig/m3 for the first quarter to 4.57 ± 1.00 |ig/m3 for the third quarter). The
first quarter average concentration for 2015 is significantly less than all of the other
quarterly average concentrations shown. Formaldehyde concentrations measured at
CSNJ range from 0.954 |ig/m3 to 8.81 |ig/m3. Formaldehyde concentrations greater
than 4 |ig/m3 were not measured at CSNJ during the first quarter of 2015, the only
calendar quarter for which this is true. Further, only one formaldehyde concentration
greater than 3 |ig/m3 was measured at CSNJ between January and March 2015; for
most other calendar quarters, formaldehyde concentrations greater than 3 |ig/m3
account for the majority of measurements (the exception is the fourth quarter of 2015,
although concentrations of this magnitude still accounted for more than 40 percent of
the measurements).
• Concentrations of acetaldehyde appear higher in 2016 than in 2015, based on the
quarterly and annual average concentrations shown in Table 16-3, although there is
considerable variability within the measurements. Acetaldehyde concentrations
measured at CSNJ range from 0.287 |ig/m3 to 8.18 |ig/m3. Of the 13 acetaldehyde
concentrations greater than 3.5 |ig/m3, only two were measured in 2015, compared to
11 in 2016. The maximum acetaldehyde concentration was measured on
July 17, 2016 and another concentration of similar magnitude was also measured on
the next sample day. CSNJ is one of four NMP sites at which an acetaldehyde
concentration greater than 8 |ig/m3 was measured. The minimum acetaldehyde
concentration was also measured at CSNJ in July 2016. This explains the large
confidence interval associated with the third quarter average concentration for 2016.
• Among the VOC pollutants of interest, bromomethane has the highest annual average
concentration for 2015 (0.93 ± 0.74 |ig/m3). The annual average for 2016 is less
(0.41 ± 0.24 |ig/m3), although both exhibit a relatively high-level of variability
(particularly 2015). Concentrations of bromomethane measured at CSNJ range from
0.0311 |ig/m3 to 20.0 |ig/m3. Seventeen of the 18 bromomethane concentrations
greater than 1 |ig/m3 measured across the program were measured at CSNJ, with all
10 concentrations greater than 2 |ig/m3 measured at CSNJ. The two highest
concentrations (20.0 |ig/m3 and 10.2 |ig/m3) were both measured in April 2015,
explaining at least in part, the high quarterly average and even higher confidence
interval for the second quarter of 2015. In fact, of the 17 bromomethane
concentrations greater than 1 |ig/m3 measured at this site, all but one was measured
during January, February, March, or April of either year, with most of these measured
in March or April. Bromomethane concentrations greater than 0.4 |ig/m3 were not
measured at CSNJ during the second half of either year.
16-21
-------
• Benzene has the highest annual average concentration for 2016 among the VOC
pollutants of interest. Concentrations of benzene measured at CSNJ range from
0.263 |ig/m3 to 3.97 |ig/m3. Of the 23 benzene concentrations greater than 1 |ig/m3
measured at CSNJ, all but one were measured during the first or fourth quarters of
either year. Conversely, 19 of the 20 lowest benzene concentrations measured at
CSNJ were measured during the second or third quarters. This explains the
differences shown in the quarterly average benzene concentrations, even though the
differences are not statistically significant.
• A similar tendency is shown for 1,3-butadiene. Of the 28 1,3-butadiene
concentrations greater than 0.1 |ig/m3 measured at CSNJ, only four were measured
outside the first or fourth quarters (two each year).
• Ethylbenzene concentrations measured at CSNJ range from 0.104 |ig/m3 to
1.67 |ig/m3, which is among the highest ethylbenzene concentrations measured across
the program. For 2016, the third and fourth quarter average concentrations of
ethylbenzene are twice the first and second quarter average concentrations. A review
of CSNJ's ethylbenzene data shows that four of the five ethylbenzene concentrations
greater than 1 |ig/m3 were measured at CSNJ during the second half of 2016 (with the
exception measured during the second half of 2015). Conversely, only one
ethylbenzene concentration less than 0.2 |ig/m3 was measured during the second half
of 2016, compared to 13 during the first half of the year.
Observations for CHNJ from Table 16-3 include the following:
• As described in Section 2.4, carbonyl compound samples collected at CHNJ between
March 31, 2015 and September 3, 2015 were invalidated due to contamination issues
with the collection system. As a result, second quarter, third quarter, and annual
average concentrations for 2015 could not be calculated. Formaldehyde and
acetaldehyde have the highest available quarterly averages shown for 2015, though,
based on available quarterly averages, and have the highest quarterly and annual
averages for 2016, among the pollutants of interest for CHNJ.
• Based on 2016 data, "higher" formaldehyde concentrations were more often
measured during the warmer months of the year. Concentrations of formaldehyde
measured at CHNJ in 2016 range from 0.174 |ig/m3 to 7.27 |ig/m3, with all 11
concentrations of at least 3.5 |ig/m3 measured at CHNJ between May and August.
• Concentrations of acetaldehyde measured at CHNJ in 2016 range from 0.146 |ig/m3
to 3.30 |ig/m3. The three highest acetaldehyde concentrations measured at CHNJ were
measured on sample days in November. Further, half of the 14 acetaldehyde
concentrations greater than 1.5 |ig/m3 were measured in either November or
December. This helps explain the higher quarterly average calculated for the fourth
quarter of 2016 compared to the other quarterly average concentrations for 2016.
• A number of intermittent issues with the collection system experienced throughout
the third quarter of 2015 resulted in too many invalidated VOC samples for a
quarterly average concentration to be calculated.
16-22
-------
• Among the VOCs, carbon tetrachloride has the highest annual average
concentrations. Concentrations of carbon tetrachloride measured at CHNJ range from
0.126 |ig/m3 to 0.851 |ig/m3, although most concentrations fell between 0.5 |ig/m3
and 0.75 |ig/m3. The minimum carbon tetrachloride concentration measured at CHNJ
ties with another concentration of the same magnitude for the second lowest carbon
tetrachloride concentration measured at an NMP site in 2015 and 2016.
• CHNJ is the only NMP site at which at a benzene concentration of at least 1 |ig/m3
was not measured. CHNJ has the lowest quarterly average concentration of benzene
(0.21 ± 0.03 |ig/m3, third quarter 2016) and the lowest annual average concentration
of benzene (0.34 ± 0.04 |ig/m3, 2016) among NMP sites sampling this pollutant.
Observations for ELNJ from Table 16-3 include the following:
• The annual average concentrations for the pollutants of interest for ELNJ are similar
across the two years of sampling.
• The pollutants of interest with the highest annual average concentrations are
formaldehyde and acetaldehyde. These are the only two pollutants with annual
average concentrations greater than 1 |ig/m3. Of the VOCs, benzene has the highest
annual average concentrations.
• The first quarter average concentration of formaldehyde for 2015 is significantly less
than the other quarterly averages for 2015, each of which has a relatively large
confidence interval. The first quarter average concentration for 2015 is also
significantly less than each of the quarterly average concentrations for 2016. A
review of the data shows that formaldehyde concentrations measured at ELNJ range
from 0.932 |ig/m3 to 12.6 |ig/m3. The number of formaldehyde concentrations less
than 2 |ig/m3 measured during the first quarter of 2015 (8) is twice the number
measured during the rest of the calendar quarters combined (4). Further, none of the
61 formaldehyde concentrations greater than 4 |ig/m3 were measured during the first
quarter of 2015, but range in number from five (the second quarter of 2015) to 13
(third quarter of both years) for the remaining calendar quarters.
• The first quarter average concentration of acetaldehyde for 2015 is also lower than
the other quarterly averages calculated for this site, although the difference is less
pronounced. Acetaldehyde concentrations measured at ELNJ span an order of
magnitude, ranging from 0.650 |ig/m3 to 6.07 |ig/m3. None of the 35 formaldehyde
concentrations greater than or equal to 3 |ig/m3 were measured during the first quarter
of 2015, but range in number from three (the second quarter of 2015) to seven (third
quarter of both years) for the remaining calendar quarters.
• Benzene concentrations measured at ELNJ range from 0.320 |ig/m3 to 2.06 |ig/m3.
Benzene concentrations measured at ELNJ appear to be higher during the colder
months of the year and lower during the warmer months of the year, based on the
quarterly average concentrations. Few of the benzene concentrations greater than
1 |ig/m3 measured at ELNJ were measured outside the first or fourth quarters of the
year; of the 29 benzene concentrations greater than 1 |ig/m3, only five were not
measured between January and March or October and December. However,
16-23
-------
confidence intervals indicate that the difference among the quarterly average
concentrations is not statistically significant.
Observations for NBNJ and NRNJ from Table 16-3 include the following:
• The instrumentation at NBNJ was moved to NRNJ at the end of 2015, where
sampling resumed. Thus, quarterly and annual average concentrations are available
for NBNJ for 2015 and for NRNJ for 2016.
• For the VOCs, the annual average concentrations are similar between the two
sampling locations. Carbonyl compounds exhibit considerable differences,
particularly acetaldehyde.
• Concentrations of acetaldehyde measured at NBNJ in 2015 range from 0.753 |ig/m3
to 4.03 |ig/m3, while concentrations of acetaldehyde measured at NRNJ in 2016 range
from 0.660 |ig/m3 to 5.26 |ig/m3. Even though a wider range of concentrations was
measured at NRNJ, acetaldehyde concentrations greater than 2 |ig/m3 account for
nearly half of the measurements at NBNJ, but account for less than one-third of the
measurements at NRNJ. At the other end of the concentration range, the number of
acetaldehyde concentrations less than 1 |ig/m3 measured at NRNJ (15) is three times
the number measured at NBNJ (5).
• Concentrations of formaldehyde measured at NBNJ in 2015 range from 0.530 |ig/m3
to 10.1 |ig/m3, while concentrations of formaldehyde measured at NRNJ in 2016
range from 0.491 |ig/m3 to 3.57 |ig/m3. Eight formaldehyde concentrations greater
than 5 |ig/m3 were measured at NBNJ, while only two were measured at NRNJ.
• The VOCs with the highest annual average concentrations for NBNJ and NRNJ are
the same: carbon tetrachloride and benzene.
• The first and fourth quarter average concentrations of benzene for NRNJ are
significantly higher than the other quarterly averages. Benzene concentrations
measured at NRNJ range from 0.214 |ig/m3 to 1.54 |ig/m3. A review of the data
shows that all but one of the 10 benzene concentrations greater than 0.75 |ig/m3 were
measured at NBNJ between January and March or October and December, including
the three greater than 1 |ig/m3. A similar observation can be made for NBNJ,
although the differences among the quarter averages is not statistically significant.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the New
Jersey sites from those tables include the following:
• The New Jersey sites appear in Table 4-10 for VOCs a total of several times (CSNJ,
four times; ELNJ, twice; and NRNJ, once).
• Two New Jersey sites appear in Table 4-10 for hexachloro-1,3-butadiene, with the
2016 annual averages for NRNJ and CSNJ ranking fourth and ninth, respectively, for
this pollutant. Note that annual average concentrations of this pollutant vary by about
0.04 |ig/m3 across the program. (Much of the 2015 hexachloro-1,3-butadiene data
16-24
-------
was invalidated due to contamination of a laboratory internal standard, as described in
Section 2.4, and thus, there are no 2015 annual averages for this pollutant.)
• CSNJ's annual average concentrations of ethylbenzene both appear among those sites
with the highest annual averages of this pollutant, ranking sixth (2016) and eighth
(2015). This site also has one of the highest annual averages of />dichlorobenzene,
ranking ninth for its 2016 annual average.
• ELNJ's annual average concentrations of 1,3-butadiene both appear among those
sites with the highest annual averages of this pollutant, ranking seventh (2016) and
ninth (2015).
• CSNJ, ELNJ, and NBNJ both appear in Table 4-11 for the carbonyl compounds.
ELNJ has the sixth (2016) and seventh (2015) highest annual average concentrations
of acetaldehyde and formaldehyde among NMP sites sampling these pollutants.
CSNJ's 2016 annual averages of these two pollutants also appear in Table 4-11,
ranking fourth for acetaldehyde and eighth for formaldehyde. In addition, NBNJ's
annual average concentration of acetaldehyde for 2015 ranks tenth among NMP sites
sampling this pollutant.
16.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for each of the pollutants
listed in Table 16-3 for each of the New Jersey sites. Figures 16-9 through 16-18 overlay the
sites' minimum, annual average, and maximum concentrations for each year onto the program-
level minimum, first quartile, median, average, third quartile, and maximum concentrations, as
described in Section 3.4.2.1, and are discussed below. If an annual average concentration could
not be calculated, the range of concentrations are still provided in the figures that follow.
16-25
-------
Figure 16-9. Program vs. Site-Specific Average Acetaldehyde Concentrations
CSNJ
CHNJ
ELNJ
NBNJ
6
Ell
1 1 1 1 1 1 1 1
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
O
o
Figure 16-9 presents the box plots for acetaldehyde for the New Jersey sites and shows
the following:
• All of the acetaldehyde concentrations measured at the New Jersey sites are
considerably less than the program-level maximum concentration (17.2 |ig/m3).
Among the New Jersey sites, the range of acetaldehyde concentrations is largest for
CSNJ (2016) and smallest for CHNJ (2015), although half of this site's measurements
were invalidated.
• The annual average concentrations for CSNJ and ELNJ are greater than the program-
level average concentration (1.67 |ig/m3). This is also true for NBNJ's annual average
concentration for 2015. NRNJ's annual average for 2016 is similar to the program-
16-26
-------
level average while the annual average for CHNJ is less than the program-level
average and median concentrations.
• An annual average concentration for CHNJ for 2015 could not be calculated. The site
relocation from NBNJ to NRNJ results in one annual average for each of these sites.
Figure 16-10. Program vs. Site-Specific Average Benzene Concentrations
CSNJ
m
CHNJ
Sf
ELNJ
NBNJ
E
4
Concentration ([ig/m3]
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
Figure 16-10 presents the box plots for benzene for the New Jersey sites and shows the
following:
• The maximum benzene concentration measured at a New Jersey site was measured at
CSNJ (3.97 |ig/m3), which is among the higher benzene concentrations measured
across the program.
16-27
-------
• The range of benzene concentrations measured was largest for CSNJ and smallest for
CHNJ. The minimum benzene concentration measured across the program was
measured at CHNJ and the entire range of benzene concentrations measured at this
site is less than the program-level third quartile (0.85 |ig/m3).
• The annual average concentrations of benzene for ELNJ and CSNJ are similar to or
greater than the program-level average concentration (0.72 |ig/m3), while the
available annual averages for the remaining New Jersey sites are less than the
program-level average and median concentrations.
Figure 16-11. Program vs. Site-Specific Average Bromomethane Concentrations
Program Max Concentration =20.0 ne/m3 1
1
1 A _____ __ ______ __ _J
1
I" 1
[ CSNJ 2015 Max = 20.0 (ig/rrr; CSNJ 2016 Max = 4.98 |ig/m3 j
i Y ¦ ¦
0 0.25 0.5 0.75 1 1.25 1.5
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 16-11 presents the box plot for bromomethane for CSNJ and shows the following:
• This figure presents the box plot for bromomethane for CSNJ, the only New Jersey
site for which this pollutant was identified as a pollutant of interest. CSNJ is the only
NMP site for which bromomethane was identified as a pollutant of interest.
• The program-level maximum bromomethane concentration (20.0 |ig/m3) is not shown
directly on the box plot in Figure 16-11 because the scale of the box plot would be
too large to readily observe data points at the lower end of the concentration range.
Thus, the scale of the box plot has been reduced to 1.5 |ig/m3. Also, since the
maximum bromomethane concentration measured each year at CSNJ is greater than
the scale of the box plot, the site-specific maximum concentrations are labeled.
• Nineteen of the 20 highest concentrations of bromomethane measured across the
program were measured at CSNJ.
• Although the annual average concentration of bromomethane for 2015 is more than
twice the annual average for 2016, both are considerably greater than program-level
average concentration (0.087 |ig/m3). CSNJ is the only NMP site with an annual
average concentration of bromomethane greater than 0.1 |ig/m3.
16-28
-------
Figure 16-12. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
l_x
| T | Program MaxConcentration = 3.90 (.ig/m3
¦ r
r
.
Program Max Concentration = 3.90 (.ig/m3
¦ •
Program Max Concentration = 3.90 (.ig/m3
, r -
Program MaxConcentration = 3.90 (.ig/m3
1
P,
i
Program
Vlax Concentration = 3.90 |ig/m3
¦
1'
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 16-12 presents the box plots for 1,3-butadiene for the New Jersey sites and shows
the following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plots in Figure 16-12 as the scale has also been reduced to allow
for the observation of data points at the lower end of the concentration range.
• All of the 1,3-butadiene concentrations measured at the New Jersey sites are less than
0.45 |ig/m3, A few non-detects were measured at each site, with the exception of
ELNJ. The minimum concentration measured each year at ELNJ is just less than the
program-level first quartile (0.03 |ig/m3).
16-29
-------
• Both annual average concentrations of 1,3-butadiene for ELNJ are greater than the
program-level average concentration (0.086 |ig/m3), while the annual averages for
CSNJ are similar to the program-level average. The available annual averages for
NBNJ and NRNJ are similar to the program-level median concentration while both
annual averages for CHNJ are less than the program-level first quartile. CHNJ's 2016
annual average concentration is the lowest annual average of 1,3-butadiene among all
NMP sites sampling VOCs.
Figure 16-13. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
rr
CHNJ
ELNJ
NBNJ
NRNJ
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
16-30
-------
Figure 16-13 presents the box plots for carbon tetrachloride for the New Jersey sites and
shows the following:
• Carbon tetrachloride concentrations measured at these sites range from 0.126 |ig/m3
and 0.870 |ig/m3, though relatively few measurements outside the 0.25 |ig/m3 to
0.80 |ig/m3 range were measured. Though the maximum concentration measured at
NBNJ is not much different than the other sites, it's the minimum concentration that
gives this site its compressed range. Only one concentration less than 0.5 |ig/m3 was
measured at NBNJ, with the number ranging from seven to 11 for the other New
Jersey sites.
• The annual average concentrations of carbon tetrachloride for the New Jersey sites
are similar to each other, ranging from 0.62 |ig/m3 and 0.64 |ig/m3, with each falling
on either side of the program-level average concentration of 0.63 |ig/m3.
Figure 16-14. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
Program Max Concentration =2.78 (ig/rn3 i
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (jug/rn3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 16-14 presents the box plot forp-dichlorobenzene for CSNJ and shows the
following:
• CSNJ is the only New Jersey site for which p-dichlorobenzene was identified as a
pollutant of interest.
• Similar to other pollutants, the program-level maximum p-dichlorobenzene
concentration (2.78 |ig/m3) is not shown directly on the box plot as the scale of the
box plot has been reduced to allow for the observation of data points at the lower end
of the concentration range. Note that the program-level first and second quartiles are
zero for this pollutant, indicating that at least 50 percent of the measurements across
the program are non-detects and thus, are not visible on the box plot.
• The maximum p-dichlorobenzene concentration measured at CSNJ (0.470 |ig/m3) is
considerably less than the maximum concentration measured across the program.
• CSNJ's annual average concentration for 2015 is similar to the program-level
average, while the annual average for 2016 is slightly higher.
16-31
-------
Figure 16-15. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
CSNJ
1
~
| Program MaxConcentration =45.8 (.ig/m3 i
m
S"
Program Max Concentration = 45.8 (.ig/m3 i
A
br
1 -J
| Program Max Concentration = 45.8 (.ig/m3
H-
\
t
Program Max Concentration = 45.8 (.ig/m3 i
NRNJ
¦
I,
Program Max Concentration = 45.8 |ig/m3 !
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
Figure 16-15 presents the box plots for 1,2-dichloroethane for the New Jersey sites and
shows the following:
• The scale of the box plots in Figure 16-15 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• The maximum concentration of 1,2-dichloroethane measured at a New Jersey site is
0.251 |ig/m3, which is less than the program-level average concentration of
0.30 |ig/m3, which is being driven by the measurements at the upper end of the
concentration range.
16-32
-------
• The annual average concentrations for the New Jersey sites are less than 0.1 |ig/m3,
with most of them less than the program-level median concentration (0.081 |ig/m3).
Figure 16-16. Program vs. Site-Specific Average Ethylbenzene Concentrations
§t^ ;
¦
r<
m
0 0.5 1 1.5 2 2.5 3
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 16-16 presents the box plots for ethylbenzene for CSNJ and ELNJ and shows the
following:
• This figure presents the box plots for CSNJ and ELNJ, the New Jersey sites for which
ethylbenzene was identified as a pollutant of interest.
• The range of ethylbenzene concentrations was larger for CSNJ than ELNJ.
• Ethylbenzene concentrations less than 0.1 |ig/m3 were not measured at CSNJ. The
minimum ethylbenzene concentration measured at CSNJ in 2015 and 2016 are both
greater than the program-level first quartile.
• The annual average concentrations for CSNJ are both greater than the program-level
average (0.26 |ig/m3) and third quartile (0.32 |ig/m3). The annual average
concentrations for ELNJ are less than those calculated CSNJ, with the annual average
for 2015 similar to the program-level average and the annual average for 2016 similar
to the program-level average third quartile.
16-33
-------
Figure 16-17. Program vs. Site-Specific Average Formaldehyde Concentrations
I
m-
-• ¦
¦H
1
¦=
—0 ¦
¦ ^
¦
1
1
1
0 3 6 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 16-17 presents the box plots for formaldehyde for the New Jersey sites and shows
the following:
• Among the New Jersey sites, the maximum formaldehyde concentration was
measured at ELNJ, although this measurement is half the magnitude of the program-
level maximum concentration.
• The annual average concentrations of formaldehyde for ELNJ are similar to each
other and both are greater than the program-level average and third quartile. The
annual averages for CSNJ fall on either side of the program-level third quartile
(3.78 |ig/m3), NBNJ's annual average concentration for 2015 falls between the
program-level average (3.05 |ig/m3) and third quartile. The available annual averages
for CHNJ and NRNJ are both less than the program-level average concentration.
16-34
-------
Figure 16-18. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
NRNJ
-
,
| Program Max Concentration = 1.02 (ig/m3 ¦
0.00 0.05 0.10 0.15 0.20 0.25
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 16-18 presents the box plots for hexachloro-1,3-butadiene for the New Jersey sites
and shows the following:
• NRNJ is the only New Jersey site for which hexachloro-1,3-butadiene was identified
as a pollutant of interest.
• The scale of this box plot has also been reduced, as the program-level maximum
hexachl oro-1,3-butadiene concentration (1.02 |ig/m3) is an order of magnitude higher
than the next highest concentration measured (0.150 |ig/m3). Note that the first,
second, and third quartiles for hexachloro-l,3-butadiene are zero at the program-level
(and therefore not visible on the box plot) due to the large number of non-detects.
• Hexachl oro-1,3-butadiene was detected in just over one-quarter of the samples
collected at NRNJ in 2016, although none of these were above the MDL for this
pollutant.
• The annual average hexachl oro-1,3-butadiene concentration for NRNJ is just slightly
greater than the program-level average concentration (0.017 |ig/m3). NRNJ has the
fourth highest annual average concentration of hexachl oro-1,3-butadiene among
NMP sites sampling this pollutant.
16.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
CHNJ, ELNJ, and NBNJ have sampled VOCs and carbonyl compounds under the NMP for
many years. ELNJ has sampled under the NMP since 2000 and CHNJ and NBNJ since 2001.
Thus, Figures 16-19 through 16-37 present the 1-year statistical metrics for each of the pollutants
of interest first for CHNJ, then for ELNJ and NBNJ (through 2015 for this site). The statistical
metrics presented for assessing trends include the substitution of zeros for non-detects. If
sampling began mid-year, a minimum of 6 months of sampling is required for inclusion in the
16-35
-------
trends analysis; in these cases, a 1-year average concentration is not provided, although the range
and percentiles are still presented. CSNJ began sampling under the NMP is 2013; thus, a trends
analysis was not performed for this site. Similarly, a trends analysis was not performed for
NRNJ.
Figure 16-19. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
CHNJ
I
i
Maximum Concentration
for 2004 is 29.1 p.g/m3
&
I
•O-
LgJ
i
i
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015" 2016
Year
O 5th Percentile
O 95th Percentile ....... Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2A 1-year average is not presented due to a contamination issue with the collection system in 2015.
Observations from Figure 16-19 for acetaldehyde concentrations measured at CHNJ
include the following:
• Sampling for carbonyl compounds under the NMP began at CHNJ in May 2001.
Because a full year's worth of data is not available for 2001, a 1-year average
concentration is not presented, although the range of measurements is provided. In
addition, a 1-year average concentration could not be calculated for 2015 due to
collection system contamination issues resulting in the invalidation of a large number
of samples collected in 2015.
• The two highest acetaldehyde concentrations were measured at CHNJ in 2004
(29.1 |ig/m3 and 11.5 |ig/m3). All other concentrations measured in 2004 were less
than 3 |ig/m3. Two additional acetaldehyde concentrations greater than 5 |ig/m3 have
been measured at CHNJ, one in 2005 (8.38 |ig/m3) and one in 2012 (5.38 |ig/m3).
16-36
-------
• An overall decreasing trend in the 1-year average concentrations is shown though
2006, with the exception of 2004, when the maximum concentrations were measured.
Between 2006 and 2010 the 1-year average concentrations changed relatively little,
varying by less than 0.25 |ig/m3 over these years.
• All of the statistical metrics exhibit an increase from 2010 to 2011. Although the
maximum concentration increased again for 2012, all of the other statistical
parameters decreased at least slightly. The 95th percentile decreased by nearly
1 |ig/m3 from 2011 to 2012, indicating that fewer concentrations at the upper end of
the range were measured in 2012. The second highest concentration measured in
2012 is half the magnitude of the maximum concentration for 2012.
• A decreasing trend is shown after 2012 and continues through 2014. Additional
decreases are shown for each of the statistical parameters for 2015, although a 1-year
average concentration could not be calculated. The median acetaldehyde
concentration for 2015 is less than 1 |ig/m3 for the first time and at a minimum for the
period of sampling. However, only roughly half of the samples remained valid
following the invalidation related to the contamination of the collection system.
• Although the range of acetaldehyde concentrations measured at CHNJ expanded
slightly for 2016, the 1-year average concentration is at a minimum for 2016.
Figure 16-20. Yearly Statistical Metrics for Benzene Concentrations Measured at CHNJ
3.5
I 2.0
3
c
O
c
m
c -1--3
o
u
1.0
I
1 T T
0.5
-J
J
2-1 ^
fc....
r
o_
T
>
£
-2-
r
£
i
4
L <
~ *
'-s
r
>
' ^
t—'
>....
^ s
h Lg
ft a
] * *
20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2 A 1-year average is not presented due to low completeness in 2005.
16-37
-------
Observations from Figure 16-20 for benzene concentrations measured at CHNJ include
the following:
• Similar to carbonyl compounds, sampling for VOCs under the NMP began at CHNJ
in May 2001. Because a full year's worth of data is not available, a 1-year average
concentration is not presented, although the range of measurements is provided. In
addition, a 1-year average concentration for 2005 is not provided due to low
completeness.
• The maximum benzene concentration measured at CHNJ was measured on
September 13, 2013 (2.88 |ig/m3). In total, eight benzene concentrations greater than
2 |ig/m3 have been measured at CHNJ since the onset of sampling (one was measured
in 2001, two in 2008, two in 2009, and one each in 2011, 2012, and 2013).
• The 1-year average and median concentrations exhibit a significant decreasing trend
between 2002 and 2007, although a 1-year average concentration is not provided for
2005; 2007 is the first year that both the 1-year average and median concentrations
are less than 0.5 |ig/m3.
• Even though an increase in the 1-year average concentration is shown from 2007 to
2008, this increase is being driven less by the two measurements greater than 2 |ig/m3
and more by the measurements in the mid- to upper-end of the concentration range.
This is evident from the increase shown in the median concentration. The number of
concentrations between 0.5 |ig/m3 and 1 |ig/m3 nearly doubled from 2007 to 2008
(from 15 to 28).
• The difference between the 5th and 95th percentiles, or the range within which the
majority of concentrations fall, increased from 2008 to 2009, indicating that the
majority of concentrations fell into a larger range and an increase in variability of the
concentrations measured, despite the decreases shown in the 1-year average and
median concentrations. Conversely, the difference between the 5th and 95th
percentiles for 2010 is at its smallest since the onset of sampling.
• An increase in the 1-year average, median, 95th percentile, and maximum
concentrations is shown from 2010 to 2011 and again for 2012. The 1-year average
concentration increased each year during this period and in 2012 is at its highest since
2004.
• The 1-year average concentration has a significant decreasing trend after 2012.
Although the range of concentrations measured is at its largest for 2013, both central
tendency parameters exhibit decreases for 2013. The number of benzene
concentrations greater than 0.5 |ig/m3 decreased from 43 measured in 2012 to 19
measured in 2013. Despite the differences in the minimum and maximum
concentrations measured in 2013 and 2014, the 5th percentile, 95th percentile, 1-year
average, and median concentrations exhibit little change from 2013 to 2014.
• Each of the statistical parameters exhibit decreases after 2014, with nearly all of them
at a minimum for 2016.
16-38
-------
Figure 16-21. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
CHNJ
Q
r-J-i
^
... • O..
I
I
' -o
20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2 A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 16-21 for 1,3-butadiene concentrations measured at CHNJ
include the following:
• The maximum 1,3-butadiene concentration was measured in 2003 (0.58 |ig/m3) and is
the only concentration greater than 0.5 |ig/m3 measured at CHNJ. In total, five
1,3-butadiene concentrations measured at CHNJ are greater than 0.2 |ig/m3.
• For 2001 and 2004, the minimum, 5th percentile, median, and 95th percentile are
zero. This is because the percentage of non-detects was greater than 95 percent for
these years. More than 50 percent of the measurements were non-detects between
2001 and 2005 (as well as 2010), as indicated by the median concentration. The
percentage of non-detects decreased steadily between 2004 (96 percent) and 2008
(17 percent). After 2008, the percentage of non-detects varies considerably, from
fewer than 10 percent (2014) to greater than 70 percent (2010).
• The 1-year average and median concentrations have a decreasing trend from 2008
through 2010, which is followed by an increasing trend in the years immediately
after. While these changes do correspond with the changes in non-detects discussed
above, the measurement of concentrations on the higher end of the concentration
range became more frequent over the years, particularly in 2014. The number of
1,3-butadiene concentrations greater than 0.05 |ig/m3 ranged from five to 10 between
2006 and 2008, decreased to four for 2009 and two for 2010, then increased each year
16-39
-------
afterward, reaching a maximum of 40 in 2014, and accounting for more than half of
the measurements for the first time. Thus, the increasing trend in the 1-year average
and median concentrations shown between 2008 and 2014 are only partially
explained by changes in the number of non-detects from year-to-year.
• Concentrations of 1,3-butadiene decreased significantly at CHNJ after 2014. The
maximum concentration measured in 2015 is similar to the 95th percentile for 2014;
likewise, the maximum concentration measured in 2016 is similar to the 95th
percentile for 2015.
Figure 16-22. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at CHNJ
r-J-l
l
l
l
T
I
, O-
1
I
I
o
T
2001 2002 2003 2004 2005^ 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2 A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 16-22 for carbon tetrachloride concentrations measured at
CHNJ include the following:
• The range of carbon tetrachloride concentrations measured appears to expand
considerably from 2001 to 2002, with the measurement ranges changing only slightly
through 2005. While a larger range of concentrations was measured during these
years compared to 2001, the measurement of a few non-detects each year during this
period contributes to degree of increase in the range shown. After 2005, only one
non-detect was reported (2007).
• All of the statistical parameters exhibit an increase from 2007 to 2008. The 95th
percentile for 2007 is just greater than the 1-year average and median concentrations
16-40
-------
calculated for 2008. Thirteen concentrations measured in 2008 were greater than the
maximum concentration measured in 2007, including seven greater than 1 |ig/m3. The
number of measurements greater than 0.6 |ig/m3 nearly doubled from 2007 (21) to
2008 (39). A similar number of concentrations greater than 0.6 |ig/m3 was measured
in 2009 and the minimum concentration increased by an order of magnitude from
2008. Yet the 1-year average increased only slightly and the median concentration
decreased slightly.
• All of the statistical parameters exhibit decreases from 2009 to 2010, with little
change shown for 2011, except for the maximum concentration. Between 2010 and
2016, the majority of carbon tetrachloride concentrations measured fell between
0.4 |ig/m3 and 0.8 |ig/m3. The 1-year average concentrations for the years 2010
through 2016 vary by less than 0.07 |ig/m3.
Figure 16-23. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at CHNJ
! Maximum Concentration!
| for 2008 is 1.27 ng/m3 J
1
T
T rS-i
° -0- -0- -0-
. I i---:
I
:--U
r5-
©•
<
. o
>¦'"
-O
20011 2002 2003 2004 20052 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2 A 1-year average is not presented due to low completeness in 2005.
Observations from Figure 16-23 for 1,2-dichloroethane concentrations measured at CHNJ
include the following:
• There were no measured detections of 1,2-dichloroethane between 2001 and 2004.
There were one or two measured detections each year between 2005 and 2008. After
2008, the percentage of measured detections increased significantly, from 7 percent in
2009, to 25 percent for 2010, 30 percent in 2011, and 95 percent for 2012. This
explains the significant increase in the 1-year average concentrations shown for the
16-41
-------
later years of sampling. Measured detections account for at least 85 percent of
concentrations measured during each of the last 5 years.
• The 1-year average and median concentrations vary little between 2012 and 2014,
with each falling between 0.07 |ig/m3 and 0.08 |ig/m3. The 1-year average
concentration decreases slightly for 2015 and again for 2016, although the changes
are not statistically significant. The median concentration also decreases slightly for
2015 but increases slightly for 2016.
Figure 16-24. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
CHNJ
1
Maximum Concentration
for 2004 is57.2 ng/m3
!
I
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015z 2016
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2A 1-year average is not presented due to a contamination issue with the collection system in 2015.
Observations from Figure 16-24 for formaldehyde concentrations measured at CHNJ
include the following:
• The two highest formaldehyde concentrations were measured on the same days in
2004 as the two highest concentrations of acetaldehyde. The maximum concentration
of formaldehyde (57.2 |ig/m3) is nearly twice the second highest concentration
(30.4 |ig/m3) and an order of magnitude higher than the third highest concentration
measured in 2004 (5.21 |ig/m3).
• With the exception of 2004, a decreasing trend in the 1-year average and median
formaldehyde concentrations is shown though 2006. Slight increases in these
parameters are shown for 2007, after which the 1-year average and median
16-42
-------
concentrations changed little through 2009. Roughly 0.5 |ig/m3 separates the 1-year
average concentrations calculated for the period between 2006 and 2009.
The 1-year and median concentrations decreased significantly for 2010, the only year
in which both statistical parameters are less than 2 |ig/m3. This is due primarily to the
measurements at the lower end of the concentration range. The number of
formaldehyde concentrations less than 1 |ig/m3 increased from two in 2009 to 21 in
2010.
Similar to acetaldehyde, all of the statistical metrics calculated for formaldehyde
exhibit an increase from 2010 to 2011, including the 95th percentile, which is greater
than the maximum concentration measured in 2010. Four formaldehyde
concentrations measured in 2011 are greater than the maximum concentration
measured in 2010 and the number of measurements greater than 2 |ig/m3 nearly
doubled, from 13 in 2010 to 25 in 2011.
Although the range of measurements decreased for 2012, little change is shown in the
1-year average concentration and the median continued to increase. This is primarily
due to decreases in the number of concentrations at the lower end of the concentration
range. The number of formaldehyde measurements less than 1 |ig/m3 decreased from
19 in 2011 to five in 2012.
With the exception of the minimum concentration, all of the statistical parameters
exhibit decreases for 2013, albeit slight ones. Relatively little change is shown in the
range of concentrations measured in 2014 compared to 2013.
While the smallest range of formaldehyde concentrations was measured at CHNJ in
2015, only roughly half of the samples remained valid following the invalidation
related to a collection system contamination issue.
The concentration profile for 2016 resembles the concentration profile for 2014. The
1-year average concentrations for 2013, 2014, and 2016 vary by less than 0.1 |ig/m3.
16-43
-------
Figure 16-25. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
ELNJ
T
—
y
T
T
T
a
-a-
I
£
LjJ
I
-l&l
£
LjJ
-A,...
LjJ
1
...<>
2
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-25 for acetaldehyde concentrations measured at ELNJ
include the following:
• ELNJ is the longest running NMP site. Carbonyl compound sampling under the NMP
began at ELNJ in January 2000. However, sporadic sampling at the beginning of
2000 combined with a l-in-12 day sampling schedule led to completeness less than
85 percent. Thus, a 1-year average concentration is not presented for 2000, although
the range of measurements is provided.
• The maximum acetaldehyde concentration was measured at ELNJ in 2007
(15.5 |ig/m3), although a concentration of similar magnitude was also measured in
2005. In total, 22 acetaldehyde concentrations greater than 10 |ig/m3 have been
measured at ELNJ, all of which were measured between 2003 and 2007.
• The range of acetaldehyde concentrations measured between 2003 and 2007 is
considerably higher than those collected during the first 3 years of sampling. The
1-year average concentration increased significantly from 2002 to 2003. This
increasing trend continued through 2007, although the rate of change slowed over the
years. A significant decrease in the measurements is shown from 2007 to 2008, where
the maximum concentration measured in 2008 is less than the 1-year average
calculated for 2007. The range of concentrations measured in 2008 is more similar to
the range shown before 2003.
16-44
-------
• Although an increasing trend is also shown between 2008 and 2011, the 1-year
average concentrations are roughly half the magnitude of those shown before 2008.
• All of the statistical parameters exhibit decreases from 2011 to 2012, with additional
decreases shown for some of the parameters for 2013. Despite variations in the range
of acetaldehyde concentrations measured over the last five years of sampling at
ELNJ, little change is shown in the central tendency parameters. Less than
0.30 |ig/m3 separates the 1-year average concentrations shown between 2012 and
2016; less than 0.25 |ig/m3 separates the median concentrations for these years.
Figure 16-26. Yearly Statistical Metrics for Benzene Concentrations Measured at ELNJ
4
i
£
Maximum concentration!
for 2008 is 34.3 ng/m3 !
I
o
-o-
r-O-
-o-
Oi
&
L2"
LoJ
I
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-26 for benzene concentrations measured at ELNJ include
the following:
• VOC sampling under the NMP also began at ELNJ in January 2000. However, a
1-year average concentration is not presented for 2000 due to low completeness,
although the range of measurements is provided.
• The maximum benzene concentration (34.3 |ig/m3) was measured in 2008 and is
more than four times higher than the next highest concentration (8.00 |ig/m3), which
was measured in 2009. A total of five benzene concentrations greater than 5 |ig/m3
have been measured at ELNJ.
• A fairly steady decreasing trend in the 1-year average and median benzene
concentrations is shown through 2007.
16-45
-------
All of the statistical parameters exhibit at least a slight increase for 2008. If the
maximum concentration for 2008 was removed from the data set, the 1-year average
concentration would exhibit only a slight increase for 2008. Thus, it is this single
concentration that is primarily driving the change in the 1-year average concentration.
The median concentration is influenced less by outliers, as this statistical parameter
represents the midpoint of a data set. The median increased by less than 0.03 |ig/m3
between 2007 and 2008, further indicating that this outlier is the primary driver
pulling the 1-year average concentration upward. However, the minimum
concentration increased by a factor of three from 2007 to 2008, with eight
concentrations measured in 2007 less than the minimum concentration measured in
2008, indicating that the outlier is not the only factor.
Even though two of the three highest benzene concentrations were measured at ELNJ
in 2009, the 1-year average concentration decreased from 2008 to 2009, likely a result
of the magnitude of the outlier affecting the 2008 calculations. If the maximum
concentration measured in 2008 was removed from the dataset, the 1-year average
concentrations would exhibit a slight increasing trend between 2007 and 2009,
although 2009 would then have the largest confidence interval among the years
shown.
Benzene concentrations measured in 2010, 2011, and 2012 were fairly consistent. The
difference in the 1-year average concentrations for these years is less than 0.04 |ig/m3,
with each hovering around 1 |ig/m3.
Additional decreases are shown for 2013, as benzene concentrations greater than
2 |ig/m3 were not measured in 2013, and the 1-year average concentration is less than
1 |ig/m3 for the first time. Although a few higher concentrations were measured in the
years that follow, the 1-year average benzene concentration changed little between
2013 and 2016.
16-46
-------
Figure 16-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
ELNJ
I
I
¦©¦
T
I
Maximum Concentration
for 2009 is2.57 ng/m3
I
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-27 for 1,3-butadiene concentrations measured at ELNJ
include the following:
• The maximum concentration of 1,3-butadiene (2.57 |ig/m3) was measured at ELNJ in
2009 and is nearly two and a half times the next highest concentration (1.04 |ig/m3,
measured in 2001). These are the only concentrations of 1,3-butadiene measured at
ELNJ that are greater than 1 |ig/m3. A total of 16 concentrations measured at ELNJ
are greater than 0.5 |ig/m3, all of which were measured in 2009 or earlier.
• The minimum and 5th percentile are zero for the first 6 years of sampling, indicating
that at least 5 percent of the measurements were non-detects. For 2004, the median
concentration is also zero, indicating that at least half of the measurements were non-
detects. Between 2000 and 2005, the percentage of non-detects ranged from
10 percent (2001) to 57 percent (2004). After 2005, only five non-detects of
1,3-butadiene have been measured at ELNJ (three in 2006 and two in 2011).
• A decreasing trend in the 1-year average concentration is shown through 2004, when
the 1-year average is at a minimum. The 1-year average concentration is fairly static
in the years that follow. Even with the higher concentration measured in 2009, the
1-year average concentration for 2009 is similar to the 1-year average concentration
for 2008. Between 2004 and 2016, the 1-year average concentration has ranged from
0.11 |ig/m3 (2004 and 2013) to 0.16 |ig/m3 (2006 and 2009).
16-47
-------
Figure 16-28. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at ELNJ
I
I
rln
A I
I
y
T
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-28 for carbon tetrachloride concentrations measured at
ELNJ include the following:
• The minimum and 5th percentile are zero for five of the first 6 years of sampling,
indicating that at least 5 percent of the measurements were non-detects (2001 being
the exception). After 2005, only one non-detect has been reported (2010).
• The 1-year average carbon tetrachloride concentrations vary by approximately
0.1 |ig/m3 during the period from 2001 to 2007, even though the range of
concentrations measured varies. All of the statistical parameters exhibit an increase in
magnitude from 2007 to 2008, which is the first year that the 1-year average
concentration is greater than 0.6 |ig/m3, with additional, albeit slight, increases for
several of the parameters for 2009. All of the 1-year average concentrations between
2008 and 2014 are greater than 0.6 |ig/m3.
• The difference between the 5th percentile and 95th percentile, or the range within
which the majority of measurements fall, has an overall decreasing trend after 2005
and is at a minimum for 2013 (less than 0.25 |ig/m3 separates the 5th percentile and
95th percentile for 2013). A slight widening of the range is shown over the last few
years of sampling.
16-48
-------
Figure 16-29. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at ELNJ
I
I
I
-©- -0- -0- J>-
<¦ £
.o
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Minimum — Median — Maximum O 95th Percentile Avera
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-29 for 1,2-dichloroethane concentrations measured at ELNJ
include the following:
• There were no measured detections of 1,2-dichloroethane between 2000 and 2004.
Between one and three measured detections were measured between 2005 and 2007,
after which there were no measured detections in 2008. After 2008, the number of
measured detections increased significantly, from five in 2009, to 11 for 2010, 16 in
2011, and 55 for 2012. With the exception of 2013, non-detects account for fewer
than 10 percent of the measurements after 2011.
• 2012 is the first year that the median concentration is greater than zero, as only six
non-detects were measured. The range of concentrations measured in 2012 is
relatively small, ranging from 0.061 |ig/m3 to 0.150 |ig/m3. For 2013, the number of
non-detects more than doubled (from six in 2012 to 14 in 2013) while the number of
1,2-dichloroethane concentrations greater than 0.1 |ig/m3 measured at ELNJ increased
from eight in 2012 to 20 in 2013.
• 2014 and 2015 are the only years that the 5th percentile is greater than zero; a
combined five non-detects were measured at ELNJ in 2014 and 2015. The 5th
percentile returned to zero for 2016, when five non-detects were measured.
• Between 2012 and 2016, the 1-year average concentrations have ranged between
0.074 |ig/m3 (2013 and 2015) and 0.086 |ig/m3 (2014).
16-49
-------
Figure 16-30. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
ELNJ
6.0
E
3
c
¦° 3 0
c
m
c
O
u
1
H
-4
»-
<
i-..
<
1
~ <
J ^
>...
_ "S
^
^ L,
<
n ¦¦¦'
I
-¦¦¦ r
<
r u
"1 r"
r
^ ^
J Lj-I L;
A d ^ A
g H M n Q B
^ X ® *
20001 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-30 for ethylbenzene concentrations measured at ELNJ
include the following:
• ELNJ's concentration profiles for ethylbenzene resemble ELNJ's concentration
profiles for benzene.
• There is a significant decreasing trend in the 1-year average and median
concentrations of ethylbenzene between 2003 and 2007.
• A significant increase in the statistical parameters is shown for 2008. The maximum
concentration measured in 2008 is more than twice the magnitude of the maximum
concentration measured in 2007; further, 1-year average and median concentrations
for 2008 are greater than the 95th percentile for 2007. The median concentration for
2008 is 0.76 |ig/m3, indicating that half of the concentrations measured at ELNJ in
2008 are greater than this concentration. By comparison, only three concentrations
measured in 2007 are greater than the median for 2008.
• The concentrations profile for 2009 more closely resembles the concentration profile
for 2007, with the exception of the maximum concentration measured.
• The despite the differences in the range of ethylbenzene concentrations measured
each year between 2009 and 2012, less than 0.1 |ig/m3 separates the 1-year average
concentrations.
16-50
-------
• After 2012, the range of concentrations measured compresses slightly each year
through 2015, when less than 0.65 |ig/m3 separates the minimum and maximum
concentrations measured. Both central tendency parameters are also at a minimum for
2015. Relatively little change is shown for 2016,
Figure 16-31. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
ELNJ
E
io.o
i
T
t
T
t
T
¥
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
Maximum o 95th Percentile
1 A 1-year average is not presented due to low completeness in 2000.
Observations from Figure 16-31 for formaldehyde concentrations measured at ELNJ
include the following:
• The maximum formaldehyde concentration was measured at ELNJ in 2013
(15.9 |ig/m3). A total of 18 concentrations greater than 10 |ig/m3 have been measured
at ELNJ, with the most measured in 2007 and 2015 (three each year).
• After decreasing by more than 1 |ig/m3 from 2000 to 2002, the median concentration
increased by more than 2 |ig/m3 for 2003. The 1-year average concentration also
exhibits a significant increase for 2003, with additional increases in both parameters
shown for 2004 and 2005. The number of formaldehyde concentrations greater than
4 |ig/m3 nearly tripled from 2002 to 2003 (from 9 to 25), and continued increasing
through 2005, with concentrations greater than 4 |ig/m3 accounting for at least half of
the concentrations measured each year through 2007.
• Similar to acetaldehyde, the 1-year average and median concentrations of
formaldehyde decreased significantly between 2007 and 2008, as the magnitude of
concentrations measured decreased considerably. Afterward, an increasing trend is
16-51
-------
shown through 2010, followed by a decrease for 2011, then another round of
increasing. The 1-year average concentration of formaldehyde for 2013 (4.90 |ig/m3)
is the highest 1-year average calculated since the onset of sampling.
• A slight decrease is shown in all of the statistical parameters for 2014 except the
minimum concentration (which is at a maximum for 2014). The 1-year average
concentration changed little over the last three years of sampling, despite differences
in the magnitude of concentrations measured, hovering around 4.40 |ig/m3 for each
year.
Figure 16-32. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
NBNJ
25.0
20.0
" 15.0
E
1
c
o
c
dj
c
u 10.0
5.0
0.0
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2A 1-year average is not presented due to a contamination issue with the collection system in 2014.
Observations from Figure 16-32 for acetaldehyde concentrations measured at NBNJ
include the following:
• Sampling for carbonyl compounds under the NMP began at NBNJ in May 2001.
Because a full year's worth of data is not available for 2001, a 1-year average
concentration is not presented, although the range of measurements is provided. In
addition, a contamination issue with the collection system resulted in the invalidation
of carbonyl compound data from May 2014 through the end of the year; thus, a 1-year
average concentration is not provided for 2014, although the range of concentrations
is provided.
; Maximum Concentration !
| for 2004 is 111 M-g/m3
I
I
o
o.
X
I
a"""
O
-Q-
T
T
&
LgJ
'"¦io,""'
£
pL
J LgJ
&
L^J
«
rh r*i
iqI
LpJ '-a-
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20142 2015
Year
16-52
-------
The maximum acetaldehyde concentration was measured in 2004 (111 |ig/m3). This
concentration is nearly seven times higher than the next highest concentration
(16.2 |ig/m3, measured in 2005).
All 17 acetaldehyde concentrations greater than 10 |ig/m were measured in either
2004 or 2005. This, along with the outlier concentration measured in 2004, explains
the significant increase in the statistical metrics shown from 2003 to 2004. Even
without an outlier for 2005, most of the statistical metrics for 2005 exhibit slight
increases from 2004 levels. The 1-year average concentration, however, does not. If
the outlier was removed from the data set for 2004, the 1-year average concentration
for 2004 would be less than the 1-year average concentration for 2005.
Each of the statistical parameters decreases significantly between 2005 and 2007,
with the 1-year average concentration decreasing from 6.21 |ig/m3 to 1.56 |ig/m3.
This is followed by a significant increase in the concentrations measured for 2008,
with the range of concentrations measured doubling.
Between 2008 and 2011, the 1-year average concentrations have an undulating
pattern, fluctuating between 2 |ig/m3 and 3 |ig/m3.
The acetaldehyde concentrations measured at NBNJ decreased significantly for 2012,
with both the 1-year average and median concentrations at a minimum (1.41 |ig/m3
and 1.36 |ig/m3, respectively). The smallest range of acetaldehyde concentrations was
measured at NBNJ in 2013, although slight increases are shown for the 1-year
average and median concentrations.
For 2014, a contamination issue with the collection system resulted in the invalidation
of carbonyl compound data from May 2014 through the end of the year. Thus, the
concentration profile for 2014 includes samples collected during the first four months
of the year. The minimum concentration measured in 2014 is greater than the 1-year
average concentration for 2013 and the median concentration for 2014 is similar to
the maximum concentration measured in 2013.
Each of the available statistical parameters exhibit a decrease from 2014 to 2015,
after which sampling at NBNJ was discontinued and moved to NRNJ.
16-53
-------
Figure 16-33. Yearly Statistical Metrics for Benzene Concentrations Measured at NBNJ
I
LjJ
l
T
o
2
1
T
\L
•a
i
T
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Year
O 5th Percentile — Minimum — Median
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-33 for benzene concentrations measured at NBNJ include
the following:
• Sampling for VOCs under the NMP also began at NBNJ in May 2001. Because a full
year's worth of data is not available for 2001, a 1-year average concentration is not
presented, although the range of measurements is provided.
• The maximum benzene concentration was measured in 2012 (4.00 |ig/m3); three
additional concentrations of benzene greater than 3 |ig/m3 have been measured at
NBNJ.
• The 1-year average concentration decreases slightly between 2002 and 2004, exhibits
little change for 2005 before decreasing significantly for 2006 and 2007. The median
concentration for 2007 is less than 0.5 |ig/m3 for the first time since the onset of
sampling.
• With the exception of the maximum concentration, all of the statistical parameters
exhibit an increase for 2008, representing a return to 2006 levels for most of the
parameters.
• Between 2008 and 2011, the 1-year average concentration changes little, ranging
from 0.65 |ig/m3 (2010) to 0.71 |ig/m3 (2011), even though there is considerable
fluctuation in the range of benzene concentrations measured.
16-54
-------
• The 1-year average benzene concentration exhibits an increase from 2011 to 2012, as
did many of the statistical parameters, even though the majority of the measurements
fell into a smaller range for 2012 than 2011, as indicated by the 5th and 95th
percentiles. The minimum and 5th percentile increased considerably for 2012; 17
benzene concentrations measured in 2011 are less than the minimum concentration
measured in 2012 (0.49 |ig/m3). In addition, the number of measurements at the
upper-end of the concentration range increased substantially for 2012. In addition to a
higher maximum concentration, the number of benzene measurements greater than
0.75 |ig/m3 increased from 11 in 2011 to 31 in 2012, accounting for more than half of
the concentrations measured in 2012.
• A significant decrease in the 1-year average concentrations is shown after 2012. The
1-year average concentration is at a minimum for 2015 (though there is little
difference between these parameters for 2014 and 2015).
Figure 16-34. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
NBNJ
<
>•
*.
X
o
... /
s.
*- ...
<
mj
> <
r L
t
a
r L
A
> ¦
r —
r v
J 1
>• *
1
t *
r .
2001 1 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-34 for 1,3-butadiene concentrations measured at NBNJ
include the following:
• The maximum 1,3-butadiene concentration was measured at NBNJ in 2005
(0.47 |ig/m3) and is the only measurement greater than 0.35 |ig/m3 measured at
NBNJ.
16-55
-------
The minimum, 5th percentile, and median concentrations are zero for 2002 through
2004. This indicates that at least half of the measurements were non-detects for these
years, and for 2004, non-detects accounted for all but four of the measurements. The
median concentration increased from 2004 to 2005, with the number of non-detects
decreasing by half. The minimum and 5th percentile are still zero for 2005 through
2007. Further decreases in the number of non-detects are indicated by the 5th
percentile increasing for 2008 through 2010, when the number of non-detects ranged
from one (2008) to three (2009). The number of non-detects increased considerably
for 2011 (17), an increase that is evident from the return of the 5th percentile to zero.
There were no non-detects measured in 2012, as indicating by the minimum
concentration, which is greater than zero for the first time. Between three and five
non-detects were measured each year between 2013 and 2015.
The 1-year average concentration of 1,3-butadiene decreased significantly from 2003
to 2004. This is primarily a result in the number of non-detects, which increased from
35 in 2003 to 56 in 2004. Thus, many zeros were substituted into this average. The
increase in the 1-year average concentration shown from 2004 to 2005 results from a
combination of fewer non-detects and a larger range of concentrations measured. The
number of non-detects decreased to 27 for 2005, accounting for fewer than half of the
measurements for the first time.
The 1-year average concentration exhibits little change between 2005 and 2011,
ranging from 0.047 |ig/m3 (2009) to 0.057 |ig/m3 (2008), even as the range within
which the majority of the concentrations are measured tightened each year through
2010.
The 1-year average concentration increases significantly from 2011 to 2012.
Increases are also exhibited by each of the other statistical parameters. This is largely
due to the decrease in non-detects (and thus, the number of zeroes substituted for non-
detects in the calculations) from 17 non-detects in 2011 to zero for 2012. The number
of concentrations at the upper end of the concentration range increased as well; the
number of measurements greater than 0.1 |ig/m3 more than doubled, increasing from
eight in 2011 to 18 in 2012.
A decreasing trend in 1,3-butadiene concentrations is shown after 2012, and by 2015,
have returned to levels measured prior to 2012.
16-56
-------
Figure 16-35. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at NBNJ
|-S-
T
X
1
I
,-5-
£
o"
o
£
'
£
•:
[I
u
rh
T
-
-
¦-
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Year
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-35 for carbon tetrachloride concentrations measured at
NBNJ include the following:
• The range of carbon tetrachloride concentrations measured in 2001 was considerably
smaller than those collected in the years immediately following. The considerable
decrease in the minimum concentration shown for 2002 to 2005 is due to non-detects,
which account for at least 5 percent of the concentrations measured for each year
during this time frame.
• The 1-year average concentration changed little between 2002 and 2005, ranging
from 0.49 |ig/m3 to 0.53 |ig/m3. An increase in the 1-year average concentration is
shown from 2005 to 2006, although the change is not statistically significant. This is a
result of higher concentrations at the upper end of the concentration range combined
with the loss of non-detects. A slight decrease in the 1-year average is shown from
2006 to 2007, as the majority of measurements fell into a tighter concentration range.
Between 2004 and 2007, the median concentration varied by only 0.003 |ig/m3,
despite variations in the range of concentrations measured.
• All of the statistical parameters exhibit increases for 2008. The minimum
concentration measured increased considerably from 2007 to 2008. In addition, 20
concentrations measured in 2008 were greater than the maximum concentration
measured in 2007.
16-57
-------
• Each of the statistical parameters exhibits a decrease after 2008 that continues
through 2010. This is followed by an increase in most of the statistical parameters
through 2012.
• The range of carbon tetrachloride concentrations measured at NBNJ decreased over
the last several years of sampling, which is particularly evident for 2014 and 2015,
when the smallest range of concentrations was measured.
Figure 16-36. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at NBNJ
^
,
•*" mm 'Ctor*
I
-o- m
o
I
O
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Year
O 5th Percentile
— Minimum
— Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
Observations from Figure 16-36 for 1,2-dichloroethane concentrations measured at NBNJ
include the following:
• There were no measured detections of 1,2-dichloroethane between 2001 and 2004.
Between 2005 and 2007, fewer than five measured detections were measured year
year, after which there were no measured detections in 2008. After 2008, the number
of measured detections increased significantly, from a total of three in 2009, to 11 for
2010, 18 in 2011, and 58 for 2012. The detection rate was at or above 95 percent for
each of the last four years of sampling, including 2015, when there are no non-
detects. This increase in the number of measured detections is very similar to what
was exhibited by the measurements at CHNJ and ELNJ. This also explains the
significant increase in the 1-year average concentrations shown, particularly for the
later years of sampling.
16-58
-------
• Excluding non-detects, the range of 1,2-dichloroethane concentrations measured in
2012 is relatively small, ranging from 0.053 |ig/m3 to 0.146 |ig/m3. Although a
relatively similar range was measured in 2013, the central tendency parameters both
exhibit increases. The number of 1,2-dichloroethane concentrations greater than
0.1 |ig/m3 increased by a factor of three from 2012 (6) to 2013 (18).
• A slight decreasing trend in 1,2-dichloroethane concentrations is shown through
2015.
Figure 16-37. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
NBNJ
60.0
50.0
40.0
1E
1
1 300
c
CD
c
o
U
20.0
10.0
0.0
20011 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 20142 2015
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2001.
2A 1-year average is not presented due to a contamination issue with the collection system in 2014.
Observations from Figure 16-37 for formaldehyde concentrations measured at NBNJ
include the following:
• The maximum formaldehyde concentration (96.1 |ig/m3) was measured at NBNJ on
the same day in 2004 that the highest acetaldehyde concentration was measured
(August 31, 2004). This concentration of formaldehyde is more than three times
greater than the next highest concentration (27.7 |ig/m3, measured in 2011). A total of
37 formaldehyde concentrations greater than 10 |ig/m3 have been measured at NBNJ,
with at least one measured during 10 of the 15 years of sampling.
• After little change between 2002 and 2003, each of the statistical metrics exhibit
increases from 2003 to 2004. This is due in part to the outlying concentration
measured in 2004; however, concentrations were higher overall in 2004 compared to
Maximum Concentration
for 2004 is 96.1 ng/m3
A
"""¦9 s &
A
16-59
-------
2003. If the maximum concentration was excluded from the calculations for 2004, the
1-year average concentration for 2004 would still exhibit more than a 1 |ig/m3
increase. This is also true for the median concentration. The number of formaldehyde
concentrations greater than 3 |ig/m3 more than doubled from 2003 to 2004, from 16 to
34. Excluding the outlier, the range of concentrations measured in 2004 is similar to
the range measured in 2005, yet the median concentration exhibited another 1 |ig/m3
increase. The number of formaldehyde concentrations greater than 3 |ig/m3 increased
further for 2015, accounting for more than 75 percent of the measurements in 2005.
After 2005, concentrations of formaldehyde measured at NBNJ decreased
significantly, with the 1-year average and median concentrations decreasing each year
and reaching a minimum for 2008. This year also has the smallest range of
formaldehyde concentrations measured, although a similar range was also measured
in 2010.
Between 2008 and 2012, a year with more variability in the measurements alternates
with a year with less variability.
Though the range of concentrations measured in 2013 is fairly similar to the range
measured in 2012, both central tendency parameters are greater than 2 |ig/m3 for the
first time since 2006. The number of formaldehyde concentrations greater than
3 |ig/m3, which had not changed much between 2010 and 2012, doubled from 2012
(7) to 2013 (14).
A 1-year average concentration is not provided for 2014, as a collection system
contamination issue resulted in the invalidation of carbonyl compound data from May
2014 through the end of the year. The statistical metrics shown for formaldehyde for
2014 resemble those shown for acetaldehyde in Figure 16-32. Nine formaldehyde
concentrations measured in 2014 are greater than the maximum concentration
measured in 2013 and the median concentration for 2014 is greater than the 95th
percentile shown for 2013.
All of the statistical parameters exhibit decreases from 2014 to 2015. However, if
2014 data is excluded, concentrations of formaldehyde exhibit an increasing trend,
with the 1-year average concentration nearly doubling between 2012 (1.83 |ig/m3)
and 2015 (3.29 |ig/m3).
16-60
-------
16.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each New Jersey monitoring site. Refer to Sections 3.2,
3.4.2.3, and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time
frames, and calculations associated with these risk-based screenings.
16.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New Jersey sites, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 16-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
16-61
-------
Table 16-4. Risk Approximations for the New Jersey Monitoring Sites
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Camden, New Jersey - CSNJ
Acetaldehyde
0.0000022
0.009
59/59
2.03
±0.25
4.47
0.23
60/60
2.65
±0.37
5.84
0.29
Benzene
0.0000078
0.03
61/61
0.86
±0.14
6.70
0.03
55/55
0.74
±0.11
5.80
0.02
Bromomethane
0.005
61/61
0.93
±0.74
0.19
55/55
0.41
±0.24
0.08
1,3-Butadiene
0.00003
0.002
61/61
0.09
±0.02
2.68
0.04
53/55
0.08
±0.02
2.47
0.04
Carbon Tetrachloride
0.000006
0.1
61/61
0.62
±0.02
3.74
0.01
55/55
0.63
±0.03
3.78
0.01
p-Dichlorobcnzcnc
0.000011
0.8
34/61
0.05
±0.01
0.51
<0.01
29/55
0.06
±0.02
0.65
<0.01
1,2 -Dichloroethane
0.000026
2.4
60/61
0.09
±0.01
2.22
<0.01
49/55
0.08
±0.01
2.01
<0.01
Ethylbenzene
0.0000025
1
61/61
0.42
±0.06
1.06
<0.01
55/55
0.47
±0.09
1.17
<0.01
Formaldehyde
0.000013
0.0098
59/59
3.57
±0.46
46.45
0.36
60/60
4.06
±0.43
52.82
0.41
- = A Cancer URE or Noncancer RfC is not available.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 16-4. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Chester, New Jersey - CHNJ
Acetaldehyde
0.0000022
0.009
34/34
NA
NA
NA
60/60
1.16
±0.15
2.56
0.13
Benzene
0.0000078
0.03
55/55
0.40
±0.04
3.10
0.01
58/58
0.34
±0.04
2.66
0.01
1,3-Butadiene
0.00003
0.002
38/55
0.02
±0.01
0.66
0.01
34/58
0.02
± <0.01
0.51
0.01
Carbon Tetrachloride
0.000006
0.1
55/55
0.64
±0.02
3.85
0.01
58/58
0.62
±0.03
3.69
0.01
1,2 -Dichloroethane
0.000026
2.4
54/55
0.07
±0.01
1.78
<0.01
49/58
0.06
±0.01
1.66
<0.01
Formaldehyde
0.000013
0.0098
34/34
NA
NA
NA
60/60
2.16
±0.39
28.12
0.22
Elizabeth, New Jersey - ELNJ
Acetaldehyde
0.0000022
0.009
60/60
2.49
±0.31
5.48
0.28
61/61
2.50
±0.26
5.49
0.28
Benzene
0.0000078
0.03
60/60
0.82
±0.08
6.39
0.03
60/60
0.83
±0.09
6.47
0.03
1,3-Butadiene
0.00003
0.002
60/60
0.12
±0.02
3.53
0.06
60/60
0.12
±0.02
3.72
0.06
Carbon Tetrachloride
0.000006
0.1
60/60
0.62
±0.03
3.70
0.01
60/60
0.62
±0.03
3.73
0.01
1,2 -Dichloroethane
0.000026
2.4
57/60
0.07
±0.01
1.94
<0.01
55/60
0.08
±0.01
1.95
<0.01
Ethylbenzene
0.0000025
1
60/60
0.28
±0.04
0.69
<0.01
60/60
0.32
±0.04
0.80
<0.01
Formaldehyde
0.000013
0.0098
60/60
4.38
±0.68
56.99
0.45
61/61
4.43
±0.41
57.61
0.45
- = A Cancer URE or Noncancer RfC is not available.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Table 16-4. Risk Approximations for the New Jersey Monitoring Sites (Continued)
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
North Brunswick, New Jersey - NBNJ
Acetaldehyde
0.0000022
0.009
59/59
2.03
±0.20
4.48
0.23
NS
NS
NS
NS
Benzene
0.0000078
0.03
59/59
0.53
±0.07
4.13
0.02
NS
NS
NS
NS
1,3-Butadiene
0.00003
0.002
54/59
0.05
±0.01
1.46
0.02
NS
NS
NS
NS
Carbon Tetrachloride
0.000006
0.1
59/59
0.64
±0.02
3.85
0.01
NS
NS
NS
NS
1,2 -Dichloroethane
0.000026
2.4
59/59
0.07
±<0.01
1.89
<0.01
NS
NS
NS
NS
Formaldehyde
0.000013
0.0098
59/59
3.29
±0.56
42.77
0.34
NS
NS
NS
NS
East Brunswick, New Jersey - NRNJ
Acetaldehyde
0.0000022
0.009
NS
NS
NS
NS
60/60
1.59
±0.19
3.51
0.18
Benzene
0.0000078
0.03
NS
NS
NS
NS
60/60
0.51
±0.07
4.00
0.02
1,3-Butadiene
0.00003
0.002
NS
NS
NS
NS
52/60
0.06
±0.01
1.76
0.03
Carbon Tetrachloride
0.000006
0.1
NS
NS
NS
NS
60/60
0.62
±0.03
3.73
0.01
1,2 -Dichloroethane
0.000026
2.4
NS
NS
NS
NS
57/60
0.08
±0.01
2.09
<0.01
Formaldehyde
0.000013
0.0098
NS
NS
NS
NS
60/60
2.44
±0.30
31.70
0.25
Hexachloro-1,3 -butadiene
0.000022
0.09
NS
NS
NS
NS
17/60
0.03
±0.01
0.60
<0.01
- = A Cancer URE or Noncancer RfC is not available.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
-------
Observations from Table 16-4 include the following:
• For CSNJ, the pollutants of interest with the highest annual average concentrations
are formaldehyde and acetaldehyde. Formaldehyde has the highest cancer risk
approximations for this site (46.45 in-a-million for 2015 and 52.82 in-a-million for
2015). The cancer risk approximations for formaldehyde are at least an order of
magnitude higher than the cancer risk approximations for the other pollutants of
interest for CSNJ. None of the pollutants of interest for CSNJ have noncancer hazard
approximations greater than 1.0, indicating that adverse noncancer health effects are
not expected from these individual pollutants. Formaldehyde is the pollutant with the
highest noncancer hazard approximations for CSNJ (0.36 for 2015 and 0.41 for
2016).
• For CHNJ, the pollutants with the highest annual average concentrations for 2016 are
formaldehyde, acetaldehyde, and carbon tetrachloride. Formaldehyde has the highest
cancer risk approximation for 2016 (28.12 in-a-million). The 2016 cancer risk
approximation for formaldehyde is at least an order of magnitude higher than the
cancer risk approximations for the other pollutants of interest for CHNJ. None of the
pollutants of interest for 2016 have noncancer hazard approximations greater than
1.0, indicating that adverse noncancer health effects are not expected from these
individual pollutants. Formaldehyde is the pollutant with the highest noncancer
hazard approximation for CHNJ (0.22). Annual average concentrations could not be
calculated for the carbonyl compounds for 2015, and thus, only the VOCs have
annual averages for 2015 in Table 16-4. The risk approximations for the VOCs for
2015 are similar to or slightly higher than the risk approximations for 2015.
• For ELNJ, the pollutants with the highest annual average concentrations are
formaldehyde, acetaldehyde, and benzene. These three pollutants also have the
highest cancer risk approximations for this site, although the cancer risk
approximations for benzene are greater than the cancer risk approximations for
acetaldehyde. ELNJ's cancer risk approximations for formaldehyde (56.99 in-a-
million for 2015 and 57.61 in-a-million for 2016) are the highest cancer risk
approximations calculated among the pollutants of interest for the New Jersey sites.
None of the pollutants of interest for ELNJ have noncancer hazard approximations
greater than 1.0, indicating that adverse noncancer health effects are not expected
from these individual pollutants. Formaldehyde is the pollutant with the highest
noncancer hazard approximations for ELNJ (0.45 for both years).
• For both NBNJ and NRNJ, the pollutants with the highest annual average
concentrations are formaldehyde, acetaldehyde, carbon tetrachloride, and benzene.
Formaldehyde has the highest cancer risk approximation for each site (42.77 in-a-
million for NBNJ in 2015 and 31.70 in-a-million for NRNJ in 2016); the cancer risk
approximations for the remaining pollutants of interest are at least an order of
magnitude lower. None of the pollutants of interest for either site have noncancer
hazard approximations greater than 1.0, indicating that adverse noncancer health
effects are not expected from these individual pollutants. Formaldehyde is the
pollutant with the highest noncancer hazard approximations for NBNJ and NRNJ
(0.34 and 0.25, respectively).
16-65
-------
16.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 16-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 16-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 16-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each New Jersey site, as presented in Table 16-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 16-5.
Table 16-6 presents similar information but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 16.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
16-66
-------
Table 16-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the New Jersey Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Camden, New Jersey (Camden County) - CSNJ
Benzene
99.22
Formaldehyde
1.04E-03
Formaldehyde
52.82
Formaldehyde
79.81
Benzene
7.74E-04
Formaldehyde
46.45
Ethylbenzene
50.34
1,3-Butadiene
4.90E-04
Benzene
6.70
Acetaldehyde
44.10
Naphthalene
2.73E-04
Acetaldehyde
5.84
1.3 -Butadiene
16.32
POM, Group 2b
1.75E-04
Benzene
5.80
Naphthalene
8.03
POM, Group 5a
1.35E-04
Acetaldehyde
4.47
POM, Group 2b
1.99
POM, Group 2d
1.33E-04
Carbon Tetrachloride
3.78
POM, Group 2d
1.51
Ethylbenzene
1.26E-04
Carbon Tetrachloride
3.74
Trichloroethylene
0.68
Acetaldehyde
9.70E-05
1,3-Butadiene
2.68
Acrylonitrile
0.22
Arsenic, PM
8.84E-05
1,3-Butadiene
2.47
Chester, New Jersey (Morris County) - CHNJ
Benzene
146.18
Formaldehyde
1.70E-03
Formaldehyde
28.12
Formaldehyde
130.73
Benzene
1.14E-03
Carbon Tetrachloride
3.85
Ethylbenzene
73.50
1,3-Butadiene
7.28E-04
Carbon Tetrachloride
3.69
Acetaldehyde
67.44
Naphthalene
4.42E-04
Benzene
3.10
1,3-Butadiene
24.26
POM, Group 2b
2.72E-04
Benzene
2.66
Naphthalene
12.99
POM, Group 5a
2.32E-04
Acetaldehyde
2.56
Dichloromethane
4.47
POM, Group 2d
2.09E-04
1,2-Dichloroethane
1.78
POM, Group 2b
3.09
Ethylbenzene
1.84E-04
1,2-Dichloroethane
1.66
POM, Group 2d
2.37
Acetaldehyde
1.48E-04
1,3-Butadiene
0.66
Trichloroethylene
1.10
Arsenic, PM
1.29E-04
1,3-Butadiene
0.51
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 16-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Elizabeth, New Jersey (Union County) - ELNJ
Formaldehyde
124.94
Formaldehyde
1.62E-03
Formaldehyde
57.61
Benzene
111.35
Benzene
8.69E-04
Formaldehyde
56.99
Acetaldehyde
57.68
1,3-Butadiene
5.42E-04
Benzene
6.47
Ethylbenzene
57.15
Naphthalene
3.70E-04
Benzene
6.39
1.3 -Butadiene
18.08
POM, Group 2b
1.94E-04
Acetaldehyde
5.49
Naphthalene
10.89
Nickel, PM
1.84E-04
Acetaldehyde
5.48
POM, Group 2b
2.20
Arsenic, PM
1.72E-04
Carbon Tetrachloride
3.73
POM, Group 2d
1.77
POM, Group 2d
1.56E-04
1,3-Butadiene
3.72
Dichloromethane
1.36
POM, Group 5a
1.55E-04
Carbon Tetrachloride
3.70
Trichloroethylene
1.09
Ethylbenzene
1.43E-04
1,3-Butadiene
3.53
North Brunswick, New Jersey (Middlesex County) - NBNJ
Formaldehyde
191.17
Formaldehyde
2.49E-03
Formaldehyde
42.77
Benzene
181.78
Ethylene oxide
1.85E-03
Acetaldehyde
4.48
Acetaldehyde
90.56
Benzene
1.42E-03
Benzene
4.13
Ethylbenzene
89.72
1,3-Butadiene
8.49E-04
Carbon Tetrachloride
3.85
1,3-Butadiene
28.31
Naphthalene
6.13E-04
1,2-Dichloroethane
1.89
Naphthalene
18.04
POM, Group 2b
3.31E-04
1,3-Butadiene
1.46
Dichloromethane
10.35
POM, Group 5a
2.74E-04
POM, Group 2b
3.76
POM, Group 2d
2.65E-04
POM, Group 2d
3.01
Ethylbenzene
2.24E-04
Trichloroethylene
1.83
Acetaldehyde
1.99E-04
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 16-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
East Brunswick, New Jersey (Middlesex County) - NRNJ
Formaldehyde
191.17
Formaldehyde
2.49E-03
Formaldehyde
31.70
Benzene
181.78
Ethylene oxide
1.85E-03
Benzene
4.00
Acetaldehyde
90.56
Benzene
1.42E-03
Carbon Tetrachloride
3.73
Ethylbenzene
89.72
1,3-Butadiene
8.49E-04
Acetaldehyde
3.51
1.3 -Butadiene
28.31
Naphthalene
6.13E-04
1,2-Dichloroethane
2.09
Naphthalene
18.04
POM, Group 2b
3.31E-04
1,3-Butadiene
1.76
Dichloromethane
10.35
POM, Group 5a
2.74E-04
Hexachloro-1,3 -butadiene
0.60
POM, Group 2b
3.76
POM, Group 2d
2.65E-04
POM, Group 2d
3.01
Ethylbenzene
2.24E-04
Trichloroethylene
1.83
Acetaldehyde
1.99E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 16-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the New Jersey Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Camden, New Jersey (Camden County) - CSNJ
Toluene
374.94
Acrolein
300,819.57
Formaldehyde
0.41
Methanol
363.96
1,3-Butadiene
8,161.97
Formaldehyde
0.36
Xylenes
181.81
Formaldehyde
8,143.65
Acetaldehyde
0.29
Benzene
99.22
Acetaldehyde
4,900.03
Acetaldehyde
0.23
Formaldehyde
79.81
Benzene
3,307.25
Bromo methane
0.19
Hexane
66.08
Naphthalene
2,677.40
Bromo methane
0.08
Ethylbenzene
50.34
Cadmium, PM
2,533.67
1,3-Butadiene
0.04
Acetaldehyde
44.10
Xylenes
1,818.11
1,3-Butadiene
0.04
Hydrochloric acid
32.85
Hydrochloric acid
1,642.29
Benzene
0.03
Ethylene glycol
27.10
Lead, PM
1,517.16
Benzene
0.02
Chester, New Jersey (Morris County) - CHNJ
Toluene
501.59
Acrolein
412,708.27
Formaldehyde
0.22
Methanol
353.16
Formaldehyde
13,339.43
Acetaldehyde
0.13
Xylenes
267.79
1,3-Butadiene
12,129.01
Benzene
0.01
Benzene
146.18
Acetaldehyde
7,493.84
Benzene
0.01
Formaldehyde
130.73
Benzene
4,872.78
1,3-Butadiene
0.01
Hexane
87.15
Naphthalene
4,329.86
1,3-Butadiene
0.01
Ethylbenzene
73.50
Xylenes
2,677.85
Carbon Tetrachloride
0.01
Acetaldehyde
67.44
Nickel, PM
2,090.44
Carbon Tetrachloride
0.01
Ethylene glycol
28.12
Arsenic, PM
1,996.26
1,2-Dichloroethane
<0.01
1.3 -Butadiene
24.26
Lead, PM
1,989.59
1,2-Dichloroethane
<0.01
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 16-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Elizabeth, New Jersey (Union County) - ELNJ
Toluene
429.29
Acrolein
324,893.75
Formaldehyde
0.45
Methanol
395.60
Cyanide Compounds, PM
38,375.00
Formaldehyde
0.45
Xylenes
209.50
Formaldehyde
12,749.01
Acetaldehyde
0.28
Formaldehyde
124.94
1,3-Butadiene
9,038.25
Acetaldehyde
0.28
Benzene
111.35
Acetaldehyde
6,408.71
1,3-Butadiene
0.06
Hexane
75.26
Chlorine
4,376.67
1,3-Butadiene
0.06
Acetaldehyde
57.68
Nickel, PM
4,269.72
Benzene
0.03
Ethylbenzene
57.15
Benzene
3,711.66
Benzene
0.03
Ethylene glycol
31.63
Naphthalene
3,628.65
Carbon Tetrachloride
0.01
Cyanide Compounds, PM
30.70
Arsenic, PM
2,670.68
Carbon Tetrachloride
0.01
North Brunswick, New Jersey (Middlesex County) - NBNJ
Toluene
662.03
Acrolein
554,943.51
Formaldehyde
0.34
Methanol
597.44
Formaldehyde
19,506.78
Acetaldehyde
0.23
Xylenes
333.05
1,3-Butadiene
14,156.30
1,3-Butadiene
0.02
Formaldehyde
191.17
Acetaldehyde
10,062.20
Benzene
0.02
Benzene
181.78
Benzene
6,059.50
Carbon Tetrachloride
0.01
Hexane
130.27
Naphthalene
6,012.36
1,2-Dichloroethane
<0.01
Acetaldehyde
90.56
Titanium tetrachloride
5,095.00
Ethylbenzene
89.72
Lead, PM
3,737.36
Ethylene glycol
45.93
Cadmium, PM
3,669.89
Glycol ethers, gas
41.32
Xylenes
3,330.51
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 16-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the New Jersey Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
East Brunswick, New Jersey (Middlesex County) - NRNJ
Toluene
662.03
Acrolein
554,943.51
Formaldehyde
0.25
Methanol
597.44
Formaldehyde
19,506.78
Acetaldehyde
0.18
Xylenes
333.05
1.3 -Butadiene
14,156.30
1,3-Butadiene
0.03
Formaldehyde
191.17
Acetaldehyde
10,062.20
Benzene
0.02
Benzene
181.78
Benzene
6,059.50
Carbon Tetrachloride
0.01
Hexane
130.27
Naphthalene
6,012.36
Hexachloro-1,3 -butadiene
<0.01
Acetaldehyde
90.56
Titanium tetrachloride
5,095.00
1,2-Dichloroethane
<0.01
Ethylbenzene
89.72
Lead, PM
3,737.36
Ethylene glycol
45.93
Cadmium, PM
3,669.89
Glycol ethers, gas
41.32
Xylenes
3,330.51
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 16-5 include the following:
• Benzene, formaldehyde, ethylbenzene, and acetaldehyde are the highest emitted
pollutants with cancer UREs in all four New Jersey counties with NMP sites,
although the order varies. In fact, these four counties have eight pollutants in common
among the highest pollutants emitted.
• Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for three of the four
New Jersey counties, with Middlesex County (NBNJ and NRNJ) as the exception. In
Middlesex County, formaldehyde also has the highest toxicity-weighted emissions,
followed by ethylene oxide prior to benzene and 1,3-butadiene.
• Eight of the 10 highest emitted pollutants in Camden, Morris, and Middlesex
Counties also have the highest toxicity-weighted emissions. Seven of the highest
emitted pollutants in Union County also have the highest toxicity-weighted
emissions.
• Formaldehyde, benzene, acetaldehyde, and 1,3-butadiene are among the pollutants of
interest with the highest cancer risk approximations for CSNJ and also appear on both
emissions-based lists. Carbon tetrachloride is also among the pollutants with the
highest cancer risk approximations for CSNJ, although this pollutant appears on
neither emissions-based list for Camden County.
• Formaldehyde, benzene, acetaldehyde, and 1,3-butadiene are among the pollutants of
interest with the highest cancer risk approximations for CHNJ and also appear on
both emissions-based lists. Carbon tetrachloride and 1,2-dichloroethane are also
among the pollutants with the highest cancer risk approximations for CHNJ, although
these pollutants appear on neither emissions-based list for Morris County.
• Formaldehyde is the pollutant with the highest emissions in Union County, the
highest toxicity-weighted emissions, and has the highest cancer risk approximations
for ELNJ. In addition to formaldehyde, benzene and 1,3-butadiene also appear on all
three lists. Acetaldehyde is also among the pollutants with the highest cancer risk
approximations for ELNJ, and is among the highest emitted, but is not among those
with the highest toxicity-weighted emissions. Carbon tetrachloride is also among the
pollutants of interest with the highest cancer risk approximations for ELNJ, although
this pollutant appears on neither emissions-based list for Union County.
• Formaldehyde also tops all three lists for Middlesex County and NBNJ and NRNJ. In
addition to formaldehyde, benzene, acetaldehyde, and 1,3-butadiene are among the
pollutants with the highest cancer risk approximations for both of these sites and also
appear on both emissions-based lists for Middlesex County. Carbon tetrachloride and
1,2-dichloroethane are also among the pollutants of interest with the highest cancer
risk approximations for NBNJ and NRNJ, although these pollutants appear on neither
emissions-based list for Middlesex County. This is also true for hexachloro-1,3-
butadiene for NRNJ.
16-73
-------
Observations from Table 16-6 include the following:
• Toluene, methanol, and xylenes are the highest emitted pollutants with noncancer
RfCs in all four New Jersey counties with NMP sites.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for all four New Jersey counties but is not among the
highest emitted pollutants for any of the New Jersey counties (acrolein ranks between
15th and 17th for these counties). Although acrolein was sampled for at all five sites,
this pollutant was excluded from the pollutant of interest designation, and thus,
subsequent risk-based screening evaluations, due to questions about the consistency
and reliability of the measurements, as discussed in Section 3.2. Formaldehyde and
1,3-butadiene are the pollutants with the second and third highest toxicity-weighted
emissions in three of the four counties. For Union County (ELNJ), cyanide
compounds rank higher than these two pollutants for this county's toxicity-weighted
emissions.
• Between four and five of the 10 highest emitted pollutants also have the highest
toxicity-weighted emissions for each of the New Jersey counties.
• Formaldehyde has the highest noncancer hazard approximations for each of the New
Jersey sites and appears on both emissions-based lists for each county.
• In addition to formaldehyde, acetaldehyde and benzene are pollutants of interest for
CSNJ that also appear on both emissions-based lists for Camden County.
1,3-Butadiene is another pollutant of interest for CSNJ that appears among those with
the highest toxicity-weighted emissions but is not among the highest emitted in
Camden County (of the pollutants with noncancer RfCs). Bromomethane is a
pollutant of interest for CSNJ that appears on neither emissions-based list.
• In addition to formaldehyde, benzene, acetaldehyde, and 1,3-butadiene are among the
pollutants of interest with the highest noncancer hazard approximations for CHNJ and
also appear on both emissions-based lists. Carbon tetrachloride and 1,2-
dichloroethane are also among the pollutants with the highest noncancer hazard
approximations for CHNJ, although these pollutants appear on neither emissions-
based list for Morris County.
• In addition to formaldehyde, acetaldehyde and benzene are pollutants of interest for
ELNJ that also appear on both emissions-based lists for Union County. 1,3-Butadiene
is another pollutant of interest for ELNJ that appears among those with the highest
toxicity-weighted emissions but is not among the highest emitted in Union County (of
the pollutants with noncancer RfCs). Carbon tetrachloride is a pollutant of interest for
ELNJ that appears on neither emissions-based list.
• In addition to formaldehyde, acetaldehyde and benzene are pollutants of interest for
NBNJ and NRNJ that also appear on both emissions-based lists for Middlesex
County. 1,3-Butadiene is another pollutant of interest for these two sites that appears
among those with the highest toxicity-weighted emissions but is not among the
highest emitted in Middlesex County (of the pollutants with noncancer RfCs). Carbon
16-74
-------
tetrachloride and 1,2-dichloroethane are also pollutants of interest for NBNJ and
NRNJ but appear on neither emissions-based list. This is also true for
hexachloro-1,3-butadiene for NRNJ.
16.5 Summary of the 2015-2016 Monitoring Data for the New Jersey Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Concentrations of 15 pollutants failed at least one screen for CSNJ; nine pollutants
failed screens for CHNJ; 11 pollutants failed screens for ELNJ; 10 pollutants failed
screens for NBNJ; and 10 pollutants failed screens for NRNJ
~~~ Formaldehyde and acetaldehyde had the highest annual average concentrations for
each of the New Jersey sites, where they could be calculated. Among the VOCs,
benzene and carbon tetrachloride had the highest annual average concentrations for
each site, with one exception. Bromomethane had the highest annual average
concentration for CSNJ in 2015. The highest bromomethane concentrations across
the program were measured at CSNJ.
~~~ ELNJ is the longest running NMP site participating under the NMP. Concentrations
of benzene have decreased significantly at this site since the onset of sampling. This is
also true of ethylbenzene, although concentrations have leveled out in the last few
years. Concentrations of benzene also have decreasing trends at CHNJ and, to a
lesser extent, NBNJ.
~~~ Formaldehyde has the highest cancer risk approximations among the New Jersey
sites 'pollutants of interest, where they could be calculated. None of the pollutants of
interest for these sites have noncancer hazard approximations greater than an HQ of
1.0.
16-75
-------
17.0 Sites in New York
This section summarizes those data from samples collected at the NATTS sites in New
York and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial
and temporal characteristics of the ambient monitoring
concentrations and reviews them through the context of
risk. Readers are encouraged to refer to Sections 1
through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
17.1 Site Characterization
This section characterizes the New York monitoring sites by providing a description of
the nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient monitoring measurements.
One New York monitoring site is located in the Bronx Borough of New York City
(BXNY) and one is located in Rochester (ROCH). Figure 17-1 presents a composite satellite
image retrieved from ArcGIS Explorer showing the BXNY monitoring site and its immediate
surroundings. Figure 17-2 identifies nearby point source emissions locations by source category,
as reported in the 2014 NEI for point sources, version 1. Note that only sources within 10 miles
of BXNY are included in the facility counts provided in Figure 17-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference but have been grayed out to emphasize
emissions sources within the boundary. Figures 17-3 and 17-4 are the composite satellite image
and emissions sources map for ROCH. Table 17-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates. Each figure and table
is discussed in detail in the paragraphs that follow.
IXiUi jjcncmlct.1 h> sources oilier
lluin LR(.i. LIJ.\'s conliiicl
kiliomioiA I'orihc \\ll\ ciiv noi
included m I lie d^ila aiuil> ses
conliiined in llns rcf>oi'l
17-1
-------
Figure 17-1. Bronx, New York (BXNY) Monitoring Site
ITfrJf
-------
Figure 17-2. NEI Point Sources Located Within 10 Miles of BXNY
Hudson
River
> Westchester
\ County
Bronx
County
1 Nassau
\ County
East River
Queens
County
Kings \
County \
- \
\
Long .
Island
Sound ^
V
BXNY NATTS site Q 10 mile radius County boundary
Source Category Group (No. of Facilities)
*t" Airport/Airline/Airport Support Operations (24)
$ Asphalt Production/Hot Mix Asphalt Plant (1)
» Automobile/Truck Manufacturing Facility (1)
B Bulk Terminal/Bulk Plant (4)
C Chemical Manufacturing Facility (3)
1 Compressor Station (2)
f Electricity Generation Facility (17)
^ Electroplating, Plating, Polishing, Anodizing, and
Coloring Facility (1)
F Food Processing/Agriculture Facility (4)
S Glass Plant (1)
> Hotels/Motels/Lodging (1)
O Institution (school, hospital, prison, etc.) (23)
. Metal Coating, Engraving, and Allied Services to
Manufacturers (1)
® Metals Processing/Fabrication Facility (3)
X Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility (25)
• Oil and/or Gas Production (2)
Q Paint and Coating Manufacturing Facility (2)
<—> Pharmaceutical Manufacturing Facility (1)
R Plastic, Resin, or Rubber Products Plant (2)
Printing/Publishing/Paper Product Manufacturing
Facility (11)
* Wastewater Treatment Facility (5)
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
^ 1
74U10'0"W 74"5'0"W
Legend
17-3
-------
Figure 17-3. Rochester, Mew York (ROCH) Monitoring Site
Blossom Rd
Blossom Rd
Gale-Terrace
Srentwood St,t
-• Halstead Stl
f 8 4".
-e
•Hi/isideAv
.Beckwith'TerTace
4 | ¦ "p.'urce: USfeS /
| Source* NASA, NGA, USGS
, I© £.83 8 ;Microsoft/fco.rp.
.NundaBh/d.
.Nunda.Blvd
-------
Figure 17-4. NET Point Sources Located Within 10 Miles of ROCH
Lake
Ontario
Wayne
County
77"45'0"W
Source Category Group (No. of Facilities)
f
Airport/Airline/Airport Support Operations (7)
*
Industrial Machinery or Equipment Plant (1)
B
Bulk Terminal/Bulk Plant (3)
a
Landfill (1)
C
Chemical Manufacturing Facility (4)
©
Metals Processing/Fabrication Facility (3)
e
Electrical Equipment Manufacturing Facility (2)
?
Miscellaneous Commercial/Industrial Facility (1)
f
Electricity Generation Facility (1)
CD
Pharmaceutical Manufacturing Facility (1)
Glass Plant (1)
P
Printing/Publishing/Paper Product Manufacturing Facility (3)
Legend
~ ROCH NATTS site
O 10 mile radius
County boundary
\
\
\
v
\ Monroe
^ County
\
\
\
\
Ontario
^ County
. V- -
\
\
\
77°35'0"W 77"30'0"W 77"25'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
17-5
-------
Table 17-1. Geographical Information for the New York Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for Traffic Data
BXNY
36-005-0110
New York
Bronx
New York-Newark-
Jersey City,
NY-NJ-PA
40.816180,
-73.902000
Residential
Urban/City
Center
100,898
1-278 between 1-87 & 1-895
ROCH
36-055-1007
Rochester
Monroe
Rochester, NY
43.146180,
-77.548170
Residential
Urban/City
Center
85,833
1-490 at 1-590
1AADT reflects 2015 data (NYS DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
-------
BXNY is located on the property of Public School 52 (PS 52) in Bronx, New York,
northeast of Manhattan. The site was established in 1999 and is considered one of the premier
particulate sampling sites in New York City and is the Bronx (#1) NATTS site. The surrounding
area is urban and residential, as shown in Figure 17-1. The Bruckner Expressway (1-278) is
located a few blocks east of the monitoring site and other heavily traveled roadways are also
located within a few miles of the site. A freight yard and other industries lie on the southeast and
south side of 1-278, part of which can be seen in the lower right-hand side of Figure 17-1. BXNY
is just over one-half mile from the East River at its closest point.
Figure 17-2 shows the numerous point sources that are located within 10 miles of BXNY,
with a majority of the emissions sources located to the south and west of the site. The source
categories with the greatest number of emissions sources surrounding the site include institutions
such as hospitals, schools, and prisons; airport and airport support operations, which include
airports and related operations as well as small runways and heliports, such as those associated
with hospitals or television stations; electricity generation via combustion; and the miscellaneous
commercial and industrial facility source category. The point source closest to BXNY is a
compressor station.
ROCH is located at a power substation on the east side of Rochester, in western New
York. Rochester is approximately halfway between Syracuse and Buffalo, with Lake Ontario
situated to the north. Although the area north and west of the site is primarily residential, as
shown in Figure 17-3, a rail line transverses the area just south of the site, and 1-590 and 1-490
intersect farther south with commercial areas adjacent to this corridor. The site is used by
researchers from several universities for short-term air monitoring studies and is the Rochester
NATTS site.
As Figure 17-4 shows, the relatively few point sources within 10 miles of ROCH are
located primarily on the west half of the 10-mile boundary. The airport and airport support
operations source category is the source category with the greatest number of emissions sources
surrounding ROCH, although there are also bulk plants/bulk terminals, chemical manufacturers,
metals processors/fabricators, electrical equipment manufactures, and printing, publishing, and
paper product manufacturers nearby, to name a few. The closest sources to ROCH are a metals
processing and fabrication facility and an electrical equipment manufacturer.
17-7
-------
In addition to providing city, county, CBS A, and land use/location setting information,
Table 17-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site.
Traffic volume is higher near BXNY than ROCH, which rank 12th and 14th, respectively,
among NMP sites, with both in the upper third compared to other NMP sites. The traffic estimate
for BXNY is for 1-278 between 1-87 and 1-895; the traffic estimate for ROCH is provided for
1-490 at 1-590.
17.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 17-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 17-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. PAHs were sampled for at both New York
sites.
Observations from Table 17-3 include the following:
• Concentrations of six pollutants failed screens for BXNY; nearly 25 percent of
concentrations for these six pollutants were greater than their associated risk
screening value (or failed screens). BXNY is one of only two NMP sites at which a
concentration of acenaphthylene failed a screen.
• Four of the six PAHs that failed screens were identified as pollutants of interest for
BXNY.
• Concentrations of eight pollutants failed screens for ROCH; 23 percent of
concentrations for these eight pollutants were greater than their associated risk
screening value (or failed screens). ROCH is the only site for which
benzo(a)anthracene, benzo(b)fluoranthene, and dibenz(a,h)anthracene failed at least
one screen. For each of these three pollutants, the concentration that failed a screen
was measured on the same day: July 23, 2015.
17-8
-------
• Four of the eight pollutants contributed to 95 percent of failed screens for ROCH and
therefore were identified as pollutants of interest for this site.
• BXNY and ROCH have three pollutants of interest in common: naphthalene,
acenaphthene, and fluorene. Naphthalene failed the majority of screens for each site,
accounting for 72 percent of failed screens for BXNY and 40 percent of failed screens
for ROCH. Acenaphthene and fluorene together account for 33 failed screens for
BXNY and 100 failed screens for ROCH. Thus, the number of failed screens of
acenaphthene and fluorene is three times greater for ROCH than BXNY. A similar
observation was made in the 2014 NMP report. Concentrations of acenaphthene and
fluorene measured at ROCH failed the most screens of any NMP site sampling these
pollutants.
Table 17-2. 2015-2016 Risk-Based Screening Results for the New York Monitoring Sites
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bronx, New York - BXNY
Naphthalene
0.029
118
120
98.33
72.39
72.39
Fluorene
0.011
19
102
18.63
11.66
84.05
Acenaphthene
0.011
14
117
11.97
8.59
92.64
Benzo(a)pyrene
0.00057
7
117
5.98
4.29
96.93
Acenaphthylene
0.011
3
84
3.57
1.84
98.77
Fluoranthene
0.011
2
120
1.67
1.23
100.00
Total
163
660
24.70
Rochester, New York - ROCH
Naphthalene
0.029
82
116
70.69
39.61
39.61
Acenaphthene
0.011
51
108
47.22
24.64
64.25
Fluorene
0.011
49
104
47.12
23.67
87.92
Fluoranthene
0.011
20
116
17.24
9.66
97.58
Benzo(a)pyrene
0.00057
2
114
1.75
0.97
98.55
Benzo(a)anthracene
0.0057
1
114
0.88
0.48
99.03
Benzo(b)fluoranthene
0.0057
1
114
0.88
0.48
99.52
Dibenz(a,h)anthracene
0.00052
1
96
1.04
0.48
100.00
Total
207
882
23.47
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
17-9
-------
17.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the New York monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at BXNY and ROCH
are provided in Appendix N.
17.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each New York site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the New York
monitoring sites are presented in Table 17-3, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
17-10
-------
Table 17-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the New York Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Bronx, New York - BXNY
Acenaphthene
57/57/59
1.81
±0.64
7.22
±2.20
10.00
± 1.94
3.33
±0.96
5.65
± 1.12
60/60/61
1.92
±0.57
6.06
±2.05
9.44
±2.53
3.94
± 1.91
5.28
± 1.14
Benzo(a)pyrene
57/29/59
0.36
±0.16
0.14
±0.03
0.14
±0.08
0.21
±0.11
0.21
±0.05
60/42/61
0.27
±0.15
0.15
±0.09
0.12
±0.06
0.12
±0.06
0.17
±0.05
Fluorene
53/53/59
3.18
±0.82
7.78
± 1.89
12.02
±2.74
3.22
± 1.45
6.61
± 1.30
49/49/61
1.40
± 1.02
6.81
±2.62
11.54
±3.05
4.79
± 1.85
6.06
± 1.41
Naphthalene
59/59/59
110.32
±29.94
107.87
± 16.14
115.61
± 20.43
118.20
± 35.06
113.05
± 12.40
61/61/61
92.79
±26.38
105.85
± 22.24
88.48
± 17.44
86.08
± 24.25
93.29
± 10.96
Rochester, New York - ROCH
Acenaphthene
53/53/56
2.77
±2.17
24.99
±8.70
34.20
±3.93
5.74
± 1.77
17.37
±4.25
55/55/60
2.09
± 1.10
15.80
±7.38
33.12
±4.27
4.39
±2.33
13.81
±3.80
Fluoranthene
56/56/56
1.21
±0.53
8.51
±3.15
10.61
±2.90
1.99
±0.66
5.73
± 1.51
60/60/60
1.46
±0.49
7.01
±3.55
9.92
±2.59
1.78
±0.51
5.03
± 1.39
Fluorene
53/53/56
2.61
± 1.37
18.25
±5.73
25.76
±3.34
4.43
± 1.73
13.07
±3.08
51/51/60
1.83
± 1.07
13.86
±6.15
27.36
±3.77
3.91
± 1.50
11.71
±3.14
Naphthalene
56/56/56
37.52
±9.52
58.68
± 13.04
88.33
±8.08
47.40
± 10.19
58.54
±7.04
60/60/60
36.83
± 12.52
53.04
± 16.70
80.34
± 11.54
33.67
±9.79
51.02
±7.73
-------
Observations for BXNY from Table 17-3 include the following:
• Of the pollutants of interest for BXNY, naphthalene has the highest annual average
concentrations, benzo(a)pyrene has the lowest, and the annual averages for
acenaphthene and fluorene are similar to each other.
• Concentrations of naphthalene measured at BXNY range from 19.0 ng/m3 to
278 ng/m3, with five of the six naphthalene concentrations greater than 200 ng/m3
measured at BXNY in 2015. Concentrations of naphthalene measured in 2015 appear
higher than those measured in 2016, based on the annual average concentrations,
although the difference is not statistically significant. Each of the quarterly average
concentrations has considerable variability associated with it, especially those
calculated for the first and fourth quarters of each year.
• Concentrations of benzo(a)pyrene measured at BXNY span two orders of magnitude,
ranging from 0.0219 ng/m3 to 1.17 ng/m3, and include three non-detects. The
maximum benzo(a)pyrene concentration measured each year was measured during
the first quarter (1.17 ng/m3, in January 2016 and 1.08 ng/m3 in February 2015). For
both years, the first quarter average concentration is the highest among the four
quarterly averages for each year, although the difference is not statistically
significant. The number of benzo(a)pyrene concentrations greater than 0.1 ng/m3 is
highest for the first quarter for each year. Twelve benzo(a)pyrene concentrations
greater than 0.1 ng/m3 were measured at BNXY in 2016, compared to seven or fewer
during each of the other calendar quarters. A similar observation can be made for
2015.
• The annual average and quarterly average concentrations of acenaphthene and
fluorene are similar to each other. Concentrations of acenaphthene range from
0.688 ng/m3 to 20.2 ng/m3 and three non-detects while concentrations of fluorene
range from 1.07 ng/m3 to 20.5 ng/m3 plus 13 non-detects. For both pollutants, the
second and third quarter average concentrations for 2015 are significantly higher than
the first and fourth quarter averages, indicating that concentrations tended to be
higher during the warmer months of the year. For 2016, the differences among the
second, third, and fourth quarter averages is less pronounced.
Observations for ROCH from Table 17-3 include the following:
• Of the pollutants of interest for ROCH, naphthalene has the highest annual average
concentrations, fluoranthene has the lowest, and the annual averages for acenaphthene
and fluorene are similar to each other. The annual averages of naphthalene for ROCH
are roughly half the magnitude of the annual averages for BXNY; conversely, the
annual averages of acenaphthene and fluorene for ROCH are two to three times
greater than the annual averages for BXNY.
• For 2015, the third quarter average concentration of naphthalene is significantly
higher than each of the remaining quarterly average concentrations. Concentrations of
naphthalene measured at ROCH in 2015 range from 13.9 ng/m3 to 106 ng/m3, with all
four concentrations greater than 100 ng/m3 measured in July, August, and September.
Further, none of the 26 naphthalene concentrations less than 60 ng/m3 were measured
at ROCH during the third quarter. A similar observation can be made for 2016,
17-12
-------
although the difference is not statistically significant, as there is more variability in
the calendar quarters within which the "higher" naphthalene concentrations were
measured.
• Quarterly average concentrations of acenaphthene, fluorene, and fluoranthene for the
second and third quarters are significantly higher than the remaining quarterly
average concentrations of each year, indicating that higher concentrations were
measured during the warmer months of the year.
• Concentrations of acenaphthene span two orders of magnitude, ranging from
0.695 ng/m3 to 61.4 ng/m3 plus eight non-detects. All 47 acenaphthene concentrations
greater than or equal to 15 ng/m3 measured at ROCH were measured during the
second or third quarters of either year. None of the eight non-detects were measured
outside the first quarter of either year.
• Concentrations of fluorene measured at ROCH range from 0.877 ng/m3 to 41.4 ng/m3
plus 12 non-detects. All 35 fluorene concentrations greater than 20 ng/m3 measured at
ROCH were measured during the second or third quarters of the year. Only one of the
12 non-detects was measured outside the first or fourth quarter of either year.
• Concentrations of fluoranthene measured at ROCH range from 0.399 ng/m3 to
24.2 ng/m3. All 44 fluoranthene concentrations greater than 5 ng/m3 measured at
ROCH were measured during the second or third quarters of either year. Further,
none of the 17 fluoranthene concentrations less than 1 ng/m3 were measured outside
the first or fourth quarters of either year.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for BXNY and
ROCH from those tables include the following:
• BXNY has the second (2015) and fourth (2016) highest annual average
concentrations of naphthalene among NMP sites sampling PAHs, as shown in
Table 4-12. BXNY is one of only two NMP sites with an annual average naphthalene
concentration greater than 100 ng/m3. ROCH does not appear in Table 4-12 for
naphthalene.
• ROCH's annual average concentrations of acenaphthene and fluorene both rank third
(2015) and fourth (2016) highest among NMP sites sampling PAHs. BXNY does not
appear in Table 4-12 for its annual average concentrations of acenaphthene. BXNY's
annual average fluorene concentration for 2016 ranks ninth while this site's annual
average for 2015 does not appear in Table 4-12 (it ranks 11th).
17-13
-------
17.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 17-3 forBXNY andROCH. Figures 17-5 through 17-9 overlay the site's minimum, annual
average, and maximum concentrations for each year onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations for each pollutant, as
described in Section 3.4.2.1, and are discussed below. If an annual average concentration could
not be calculated, the range of concentrations are still provided in the figures that follow.
Figure 17-5. Program vs. Site-Specific Average Acenaphthene Concentrations
BXNY
j
"i
| Program MaxConcentration = 108 ng/m3
r
1 1 *
1 1
¦ n
i
i
i
i
i
i
i
>
i
i
i
i
i
i
i
— i
¦ 1
; Program Max concentration = Wis ng/rrr ¦
0 15 30 45 60 75
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
• o
Figure 17-5 presents the box plots for acenaphthene for BXNY and ROCH and shows the
following:
• The program-level maximum concentration (108 ng/m3) is not shown directly on the
box plots because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale has been reduced.
• The maximum acenaphthene concentration measured at ROCH is about three times
greater than the maximum concentration measured at BXNY. Although the highest
concentration measured at ROCH is half the magnitude of the maximum
acenaphthene concentration measured across the program, it is the ninth highest
concentration measured across the program.
17-14
-------
• The annual average concentrations for BXNY are similar to each other, and just
greater than the program-level average concentration (4.36 ng/m3), while the annual
averages for ROCH are three (2016) and four (2015) times greater the program-level
average concentration.
• ROCH has the third (2015) and fourth (2016) highest annual average concentrations
of acenaphthene among NMP sites sampling PAHs (behind only NBIL).
Figure 17-6. Program vs. Site-Specific Average Benzo(a)pyrene Concentrations
1 -j 1
Program Max Concentration = 5.82 ng/m3
¦ J* '
1 la
w
r 1
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o
Figure 17-6 presents the box plot for benzo(a)pyrene for BXNY and shows the following:
• The program-level maximum concentration (5.82 ng/m3) is not shown directly on the
box plot because the scale of the box plot would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale has also been
reduced.
• BXNY is one of only two sites for which benzo(a)pyrene is a pollutant of interest
(UNVT is the other).
• The maximum benzo(a)pyrene concentration measured at BXNY is not the maximum
concentration measured across the program, although it is the among the highest
measured. (The maximum benzo(a)pyrene concentration measured across the
program was actually measured at ROCH, although benzo(a)pyrene was not
identified as a pollutant of interest for ROCH).
• The annual average concentrations of benzo(a)pyrene for BXNY are both greater than
the program-level average concentration (0.09 ng/m3), with BXNY's annual average
for 2015 more than twice the magnitude.
17-15
-------
Figure 17-7. Program vs. Site-Specific Average Fluoranthene Concentrations
E 1
i
1 ;
Program Max Concentration = 57.3 ne/m3
i
i
i
r
0 5 10 15 20 25 30 35 40
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 17-7 presents the box plot for fluoranthene for ROCH and shows the following:
• The program-level maximum concentration (57.3 ng/m3) is not shown directly on the
box plot because the scale of the box plot would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale of this box plot has
also been reduced.
• ROCH is one of only three NMP sites for which fluoranthene is a pollutant of interest
(NBIL and GPCO are the others).
• The maximum fluoranthene concentration measured at ROCH is considerably less
than the maximum concentration measured across the program, although few NMP
sites have measurements greater than ROCH's maximum fluoranthene
concentrations.
• Both annual average concentrations for ROCH are more than twice the program-level
average concentration (2.39 ng/m3). Only one NMP site each year has an annual
average concentration greater than ROCH's (NBIL).
17-16
-------
Figure 17-8. Program vs. Site-Specific Average Fluorene Concentrations
¦
J
\ Program Max Concentration = 105 ng/m3
r
1
ROCH
Q 1
1
1 *
H 1
rv
1 1
Progra m Max Concentration = lub ng/rrr |
30
Concentration (ng/m3]
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
Figure 17-8 presents the box plots for fluorene for BXNY and ROCH and shows the
following:
• The program-level maximum concentration (105 ng/m3) is not shown directly on the
box plots because the scale of the box plots would be too large to readily observe data
points at the lower end of the concentration range. Thus, the scale has been reduced.
The program-level first quartile is zero for this pollutant, indicating that at least 25
percent of the measurements across the program are non-detects and thus, are not
visible on the box plot.
• The maximum fluorene concentration measured at ROCH is more than two times
greater than the maximum concentration measured at BXNY, although both sites'
maximum concentrations are considerably less than the highest concentration
measured across the program.
• The annual average concentrations for ROCH are roughly twice the annual averages
for BXNY, although both sites' annual averages are greater than the program-level
average concentration (4.36 ng/m3).
• ROCH has the third (2015) and fourth (2016) highest annual average concentrations
of fluorene among NMP sites sampling PAHs (behind only NBIL).
17-17
-------
Figure 17-9. Program vs. Site-Specific Average Naphthalene Concentrations
BXNY
ROCH i
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4-thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
0 Q
Figure 17-9 presents the box plots for naphthalene for BXNY and ROCH and shows the
following:
• In contrast to the box plots for the other pollutants of interest in common for the New
York sites, Figure 17-9 shows that the maximum naphthalene concentrations
measured at BXNY are considerably greater than those measured at ROCH.
• The minimum naphthalene concentrations measured each year at ROCH are fairly
similar to each other while the minimum concentrations measured at BXNY are
considerably different. The minimum concentration measured at BXNY in 2015 is
more than twice the minimum concentration measured in 2016. The five lowest
naphthalene concentrations measured at BXNY were measured in 2016. The
minimum concentration measured at BXNY in 2015 is higher than the program-level
first quartile and the highest minimum concentration measured among NMP sites
sampling naphthalene in 2015. A similar observation was made in the 2013 and 2014
NMP reports.
• The annual average naphthalene concentrations for ROCH fall between the program-
level median concentration (48.90 ng/m3) and the program-level average
(61.23 ng/m3) concentration. The annual average concentrations for BXNY roughly
twice the magnitude of ROCH's annual averages, and are greater than the program-
level average and third quartile. BXNY has the second (2015) and fourth (2016)
highest annual average concentrations of naphthalene among NMP sites sampling
PAHs.
17-18
-------
17.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
Sampling for PAHs at BXNY began in July 2008. However, in June 2010, the instrumentation at
BXNY was relocated to a new, temporary location due to roofing construction near the BXNY
site. Two years later, the instrumentation was returned to the BXNY site and sampling resumed
at this location in July 2012. The trends analysis was performed for BXNY but does not include
data from the temporary location. Figures 17-10 through 17-13 present the 1-year statistical
metrics for the pollutants of interest for BXNY.
Sampling for PAHs at ROCH also began in July 2008. However, a collection error was
discovered at the site, resulting in the invalidation of nearly one and one-half years' worth of
samples between July 2009 and December 2010. Thus, a 1-year average concentration for 2009
is not provided on the trends graph and no statistical metrics are provided for 2010. This,
combined with the mid-year start in 2008, results in the calculation of few 1-year average
concentrations for the ROCH monitoring site during the early years of sampling. Figure 17-14
through 17-17 present the 1-year statistical metrics for the pollutants of interest for ROCH.
17-19
-------
Figure 17-10. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
BXNY
I
o
o-
—Q—
—2r-
—Q-
2
I
20081 2009 20102 20112 20122 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented due to temporary site relocation from June 2010-July 2012.
Observations from Figure 17-10 for acenaphthene concentrations measured at BXNY
include the following:
• The maximum acenaphthene concentration was measured at BXNY in 2014
(23.4 ng/m3), although concentrations greater than 20 ng/m3 were also measured in
2010, 2012, and 2015.
• Prior to 2013, there is only one year (2009) with a full year's worth of samples for
BXNY. With the exception of the maximum concentration, the concentration profile
for 2009 resembles the concentration profile for 2013.
• Several of the statistical parameters exhibit decreases over the last several years of
sampling, with several of the statistical parameters at a minimum for 2016. The
median concentration decreased each year between 2012 and 2016, decreasing by
more than half during this period, from 7.52 ng/m3 (2012) to 3.32 ng/m3 (2016). The
first non-detects were measured in 2015 and 2016. The 1-year average concentration
decreased slightly each year between 2013 and 2016, although the confidence
intervals indicate the difference is not statistically significant as there is considerable
variability in the measurements.
17-20
-------
Figure 17-11. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at
BXNY
Maximum Concentration
is 22.4 ng/m3for 2010
2-
2009
2012
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented due to temporary site relocation from June 2010-July 2012.
Observations from Figure 17-11 for benzo(a)pyrene concentrations measured at BXNY
include the following:
• The maximum benzo(a)pyrene concentration was measured at BXNY in 2010
(22.4 ng/m3) and is an order of magnitude greater than the next highest
benzo(a)pyrene concentration measured at this site (2.37 ng/m3), which was measured
in 2014. No other concentrations greater than 2 ng/m3 have been measured at BXNY.
• Prior to 2013, there is only one year (2009) with a full year's worth of samples for
BXNY. If the outlier measured in 2010 was excluded from the dataset, the
concentration profile for 2010 would have the second widest range of measurements
(behind 2014), the largest difference between its 5th and 95th percentiles, and the
highest median concentration. By contrast, nearly all of the statistical parameters are
at a minimum for 2012.
• The median benzo(a)pyrene concentration varies by less than 0.05 ng/m3 over the last
several years of sampling, ranging from 0.10 ng/m3 (2013) to 0.14 ng/m3 (2015). The
1-year average concentration varies slightly more, though the differences are not
statistically significant.
17-21
-------
Figure 17-12. Yearly Statistical Metrics for Fluorene Concentrations Measured at BXNY
E
~Ss
Maximum Concentration
is 114 ng/m3for2010
2012
Year
O 5th Percentile
— Minimum
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented due to temporary site relocation from June 2010-July 2012.
Observations from Figure 17-12 for fluorene concentrations measured at BXNY include
the following:
• The maximum fluorene concentration (114 ng/m3) was measured on the same day the
maximum benzo(a)pyrene concentration was measured (January 14, 2010).
• Non-detects of fluorene were not measured until 2013. The number of non-detects
measured has increased each year since, from three in 2013 to 12 in 2016.
• The median fluorene concentration decreased by more than 1 ng/m3 from 2012 to
2013 and again for 2014. Little change is shown for 2015, before additional decreases
are shown for 2016. During this time, the median concentration decreased from
7.49 ng/m3 for 2012 to 4.44 ng/m3 for 2016. The 1-year average concentration varies
by less than 1 ng/m3 during this period, ranging from 7.01 ng/m3 for 2013 to
6.06 ng/m3 for 2016.
17-22
-------
Figure 17-13. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
BXNY
Maximum Concentration
is 1170 ng/m3 for 2010
2008 1 2009 2010
2012
Year
2014 2015
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented due to temporary site relocation from June 2010-July 2012.
Observations from Figure 17-13 for naphthalene concentrations measured at BXNY
include the following:
• The maximum naphthalene concentration (1,170 ng/m3) was measured on the same
day the maximum benzo(a)pyrene and fluorene concentrations were measured
(January 14, 2010). The next highest concentration, measured in 2009, was nearly
half as high (525 ng/m3). No additional naphthalene concentrations greater than
300 ng/m3 have been measured at BXNY.
• The central tendency parameters have an undulating pattern between 2013 and 2016.
With the exception of the maximum concentration, each of the statistical parameters
is at a minimum for 2016. Both the 1-year average and median concentrations are less
than 100 ng/m3 for the first time.
17-23
-------
Figure 17-14. Yearly Statistical Metrics for Acenaphthene Concentrations Measured at
ROCH
I
JL I
o...
...o-
—
2012
Year
O 5th Percentile
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 17-14 for acenaphthene concentrations measured at ROCH
include the following:
• The range of acenaphthene concentrations appears to have decreased by half from
2008 to 2009, although 2008 includes data from July through December while 2009
includes data from January through June.
• The concentrations measured in 2011 are similar to the concentrations measured in
2012.
• The range of concentrations measured increased from 2012 to 2013. The median
concentration nearly doubled from 2012 to 2013 while the 1-year average
concentration increased by 58 percent.
• The maximum concentration increased considerably for 2014, with three
acenaphthene concentrations measured in 2014 that are greater than the maximum
concentration measured in 2013. Despite these higher measurements, the 1-year
average and median concentrations decreased slightly.
• Nearly all of the statistical parameters exhibit decreases for 2015 and 2016. The only
non-detects measured at ROCH were measured in 2015 (3) and 2016 (5).
17-24
-------
Figure 17-15. Yearly Statistical Metrics for Fluoranthene Concentrations Measured at
ROCH
40 0
E
c
O
c
m
c
O
u
I
o
<>..
.o
a
o
2008
2009 2
2010 2
2011
2012
Year
2013
2014
2015
2016
1
0
5th Percentile
— Minimum
-
Median
Maxim
um
O
95th Percentile
Ave
rag
1
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 17-15 for fluoranthene concentrations measured at ROCH
include the following:
• Two fluoranthene concentrations greater than 30 ng/m3 have been measured at
ROCH, one in 2011 (33.7 ng/m3) and one in 2012 (30.4 ng/m3).
• The range of fluoranthene concentrations measured decreased each year between
2011 and 2014, when several of the statistical parameters, including the 1-year
average concentration, were at a minimum.
• With the exception of the minimum concentration, each of the statistical parameters
exhibits an increase for 2015. In addition to a few "higher" concentrations, the
number of fluoranthene concentrations greater than 10 ng/m3 tripled, increasing from
four measured in 2014 to 13 measured in 2015.
• Although the range of concentrations measured in 2015 is not much different than the
range measured in 2016, the central tendency parameters decreased, and the median
concentration is at a minimum for the period of sampling. The number of
fluoranthene concentrations less than 2 ng/m3 is at its highest in 2016, accounting for
nearly half of the measurements.
17-25
-------
• Excluding 2014, each of the 1-year average concentrations falls between 5 ng/m3 and
6 ng/m3.
Figure 17-16. Yearly Statistical Metrics for Fluorene Concentrations Measured at ROCH
2012
Year
O 5th Percentile
— Minimum
Maximum O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 17-16 for fluorene concentrations measured at ROCH include
the following:
• The concentration profiles for fluorene resemble the concentration profiles for
acenaphthene.
• The range of fluorene concentrations measured at ROCH decreased from 2008 to
2009, with the median concentration decreasing by more than half during this time.
However, each year's concentration profile includes half a year's worth of samples
(2008 includes data from July through December and 2009 includes data from
January through June).
• The concentrations measured in 2011 are similar to the concentrations measured in
2012.
Both central tendency parameters exhibit increases for 2013, as the range of
measurements increased from 2012 to 2013 (at both ends of the concentration range).
The median increased by 67 percent from 2012 to 2013 while the 1-year average
concentration increased by 35 percent.
17-26
-------
• Relatively little change in the 1-year average concentration is shown between 2013
and 2016, despite variations in the concentrations measured. The maximum
concentration of fluorene measured at ROCH was measured in 2014 (104 ng/m3).
Two additional concentrations greater than 50 ng/m3 have been measured at this site,
one in 2013 (53.4 ng/m3) and another in 2014 (51.4 ng/m3). Despite the higher
concentrations measured, only slight changes are shown in both the 1-year average
and median concentrations. This is due primarily to the increase in non-detects
measured in 2014 (seven). The first non-detects were measured at ROCH in 2013
(two), with several measured each year thereafter. The number of non-detects is at a
maximum of nine for 2016.
Figure 17-17. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
ROCH
•o...
2012
Year
O 5th Percentile
— Minimum
Maximum o 95th Percentile Averc
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
2 Some statistical metrics are not presented because data from July 2009 to Dec 2010 was invalidated.
Observations from Figure 17-17 for naphthalene concentrations measured at ROCH
include the following:
• Similar to the other pollutants of interest, the range of naphthalene concentrations
decreased from 2008 to 2009. However, each year's concentration profile includes
half a year's worth of samples.
17-27
-------
• Even though the maximum concentration increased each year between 2011 and
2013, the 1-year average naphthalene concentrations calculated for 2011, 2012, and
2013 exhibit little change, varying by less than 1 ng/m3 across these years. The range
of naphthalene concentrations measured at ROCH expanded further in 2014, though
the 1-year average for 2014 decreased slightly. This is also true for the median
concentration. Several of the lowest naphthalene concentrations were measured in
2014, including the most concentrations less than 20 ng/m3 (seven) since the onset of
sampling.
• The range of naphthalene concentrations measured decreased considerably in 2015.
Yet, the 1-year average changed little and the median increased by nearly 50 percent
and is greater than the 1-year average concentration. This results from an increase in
the number of measurements in the mid- to upper-half of the concentration range. The
number of naphthalene concentrations between 50 ng/m3 and 100 ng/m3 increased
from 12 in 2014 to 30 for 2015.
• Although the range of concentrations measured in 2016 is only slightly larger than the
range measured in 2015, both central tendency parameters exhibit decreases, with the
1-year average concentration at a minimum for 2016. A higher number of
naphthalene concentrations at the lower end of the concentration range were
measured in 2016; the number of concentrations less than 25 ng/m3 increased from
seven measured in 2015 to 16 in 2016.
17-28
-------
17.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the New York monitoring sites. Refer to Sections 3.2,
3.4.2.3, and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time
frames, and calculations associated with these risk-based screenings.
17.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the New York sites, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 17-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for the New York sites from Table 17-4 include the following:
• Naphthalene has the highest annual average concentrations among the pollutants of
interest for each site.
• Naphthalene has the highest cancer risk approximations for BXNY (both greater than
3 in-a-million). The cancer risk approximations for the other pollutants of interest for
BXNY are less than 1 in-a-million.
• For ROCH, naphthalene also has the highest cancer risk approximations, though the
difference among the cancer risk approximations among ROCH's pollutants of
interest is considerably less (three of the four pollutants have cancer risk
approximations between 1 in-a-million and 2 in-a-million).
• Naphthalene is the only site-specific pollutant of interest for either site that has a
noncancer RfC. The noncancer hazard approximations for naphthalene for each site
are less than 0.05, considerably less than 1.0, indicating that no adverse noncancer
health effects are expected from this individual pollutant.
17-29
-------
Table 17-4. Risk Approximations for the New York Monitoring Sites
Pollutant
Cancer
URE
(Ug/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Bronx, New York - BXNY
Acenaphthene
0.000088
57/59
5.65
± 1.12
0.50
60/61
5.28
± 1.14
0.46
Benzo(a)pyrene
0.00176
57/59
0.21
±0.05
0.37
60/61
0.17
±0.05
0.29
Fluorene
0.000088
53/59
6.61
± 1.30
0.58
49/61
6.06
± 1.41
0.53
Naphthalene
0.000034
0.003
59/59
113.05
± 12.40
3.84
0.04
61/61
93.29
± 10.96
3.17
0.03
Rochester, New York - ROCH
Acenaphthene
0.000088
53/56
17.37
±4.25
1.53
55/60
13.81
±3.80
1.22
Fluoranthene
0.000088
56/56
5.73
± 1.51
0.50
60/60
5.03
± 1.39
0.44
Fluorene
0.000088
53/56
13.07
±3.08
1.15
51/60
11.71
±3.14
1.03
Naphthalene
0.000034
0.003
56/56
58.54
±7.04
1.99
0.02
60/60
51.02
±7.73
1.73
0.02
— = A Cancer URE or Noncancer RfC is not available.
-------
17.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 17-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 17-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 17-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for each site, as presented in Table 17-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 17-5.
Table 17-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 17.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
17-31
-------
Table 17-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for
Pollutants with Cancer UREs for the New York Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bronx, New York (Bronx County) - BXNY
Benzene
140.01
Formaldehyde
1.73E-03
Naphthalene
3.84
Formaldehyde
133.13
Benzene
1.09E-03
Naphthalene
3.17
Dichloromethane
112.31
1,3-Butadiene
5.93E-04
Fluorene
0.58
Ethylbenzene
76.18
Naphthalene
5.87E-04
Fluorene
0.53
Acetaldehyde
68.23
POM, Group 2b
2.52E-04
Acenaphthene
0.50
1.3 -Butadiene
19.77
POM, Group 2d
2.45E-04
Acenaphthene
0.46
Naphthalene
17.28
Ethylene oxide
2.29E-04
Benzo(a)pyrene
0.37
Tetrachloroethylene
15.79
Arsenic, PM
1.95E-04
Benzo(a)pyrene
0.29
1,4-Dioxane
15.39
Ethylbenzene
1.90E-04
POM, Group 2b
2.86
POM, Group 5a
1.62E-04
Rochester, New York (Monroe County) - ROCH
p-Dichlorobenzene
314.57
p-Dichlorobcnzcnc
3.46E-03
Naphthalene
1.99
Benzene
223.73
Formaldehyde
2.34E-03
Naphthalene
1.73
Formaldehyde
180.29
Benzene
1.75E-03
Acenaphthene
1.53
Ethylbenzene
106.86
1,3-Butadiene
1.08E-03
Acenaphthene
1.22
Acetaldehyde
102.79
Naphthalene
8.00E-04
Fluorene
1.15
Tetrachloroethylene
81.75
Arsenic, PM
4.41E-04
Fluorene
1.03
1,3 -Dichloropropene
80.40
POM, Group 2b
4.27E-04
Fluoranthene
0.50
1.3 -Butadiene
35.96
POM, Group 2d
3.35E-04
Fluoranthene
0.44
Naphthalene
23.52
1,3 -Dichloropropene
3.22E-04
Dichloromethane
12.92
POM, Group 5a
3.20E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 17-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for
Pollutants with Noncancer RfCs for the New York Monitoring Sites
Top 10 Total Emissions for Pollutants
with Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Haza
on Annual Avera
(Site-S
rd Approximations Based
ge Concentrations
jecific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Bronx, New York (Bronx County) - BXNY
Toluene
416.80
Acrolein
290,569.07
Naphthalene
0.04
Hexane
298.98
Acrylic acid
26,156.73
Naphthalene
0.03
Xylenes
243.02
Formaldehyde
13,585.10
Benzene
140.01
1.3 -Butadiene
9,883.64
Formaldehyde
133.13
Acetaldehyde
7,581.21
Dichloromethane
112.31
Naphthalene
5,758.99
Ethylene glycol
77.61
Benzene
4,667.05
Ethylbenzene
76.18
Cadmium, PM
4,020.53
Acetaldehyde
68.23
Diethanolamine
3,889.87
Isophorone
61.68
Triethylainine
3,671.84
Rochester, New York (Monroe County) - ROCH
Chlorobenzene
1,466.95
Acrolein
541,813.81
Naphthalene
0.02
Toluene
600.97
Chlorine
30,337.40
Naphthalene
0.02
Xylenes
539.99
Triethylainine
28,821.23
p-Dichlorobenzene
314.57
Formaldehyde
18,396.73
Hexane
306.97
1,3-Butadiene
17,979.41
Hydrochloric acid
260.76
Hydrochloric acid
13,038.15
Benzene
223.73
Acetaldehyde
11,421.29
Triethylainine
201.75
Cadmium, PM
9,670.45
Formaldehyde
180.29
Naphthalene
7,840.69
Methanol
129.75
Benzene
7,457.62
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 17-5 include the following:
• Benzene, formaldehyde, and dichloromethane are the highest emitted pollutants with
cancer UREs in Bronx County. />Dichlorobenzene, benzene, and formaldehyde are
the highest emitted pollutants in Monroe County.
• Formaldehyde, benzene, and 1,3-butadiene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Bronx County.
/;-Dichlorobenzene, formaldehyde, and benzene are the pollutants with the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) for Monroe County.
• Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Bronx County; six of the highest emitted pollutants also have the
highest toxicity-weighted emissions for Monroe County.
• Naphthalene, which is a pollutant of interest for both sites and has the highest cancer
risk approximations for each site, appears on both emissions-based lists for Bronx and
Monroe Counties.
• Emissions of several POM Groups rank among the pollutants with the highest
toxicity-weighted emissions for Bronx County. POM, Group 2b appears on both
emissions-based lists for Bronx County and includes several PAHs sampled for at
BXNY, including acenaphthene, fluoranthene, and fluorene. POM, Group 5a also
appears among those with the highest toxicity-weighted emissions for Bronx County;
this group includes benzo(a)pyrene, which is also a pollutant of interest for BXNY.
POM, Group 2d also appears among those with the highest toxicity-weighted
emissions for Bronx County, although none of the PAHs sampled with Method
TO-13A are included in this group. POM, Groups 2b, 2d, and 5a also appear among
the pollutants with the highest toxicity-weighted emissions for Monroe County,
although none appear among the highest emitted pollutants for Monroe County.
Observations from Table 17-6 include the following:
• Toluene, hexane, and xylenes are the highest emitted pollutants with noncancer RfCs
in Bronx County. Chlorobenzene is the highest emitted pollutant with a noncancer
RfC in Monroe County, followed by toluene and xylenes.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) is acrolein for both counties.
• Three of the highest emitted pollutants in Bronx County are also among the pollutants
with the highest toxicity-weighted emissions; four of the highest emitted pollutants in
Monroe County are also among the pollutants with the highest toxicity-weighted
emissions.
• Naphthalene is the only pollutant of interest for each site for which a noncancer RfC
is available, and thus noncancer hazard approximations could be calculated.
Naphthalene is among the pollutants with the highest toxicity-weighted emissions for
each county, but is not among the highest emitted pollutants with a noncancer toxicity
factor for either county.
17-34
-------
17.5 Summary of the 2015-2016 Monitoring Data for BXNY and ROCH
Results from several of the data analyses described in this section include the following:
~~~ Six PAHs failed screens for BXNY, of which four were identified as pollutants of
interest. Eight PAHs failed screens for ROCH, of which four were identified as
pollutants of interest. Naphthalene, acenaphthene, andfluorene were identified as
pollutants of interest for both New York monitoring sites.
~~~ Naphthalene has the highest annual average concentrations for both sites, although
the annual averages for BXNY are nearly twice the annual averages for ROCH.
~~~ BXNY has some of the highest annual average concentrations of naphthalene among
NMP sites sampling PAHs. ROCH has some of the highest annual average
concentrations of acenaphthene andfluorene among NMP sites sampling PAHs.
~~~ For both BXNY and ROCH, the 1-year average concentrations of naphthalene for
2016 are at a minimum over the course of sampling.
~~~ Naphthalene has the highest cancer risk approximations among the pollutants of
interest for both BXNY and ROCH (although the differences among the pollutants of
interest for ROCH is relatively small). None of the pollutants of interest have
noncancer hazard approximations greater than an HQ of 1.0.
17-35
-------
18.0 Sites in Oklahoma
This section summarizes those data from samples collected at the UATMP sites in
Oklahoma and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and
2016 monitoring efforts. This section also examines the
context of risk. Readers are encouraged to refer to Sections 1 conumed in ihis ivpori
through 4 for detailed discussions and definitions regarding __
the various data analyses presented below.
18.1 Site Characterization
This section characterizes the Oklahoma monitoring sites by providing a description of
the nearby area surrounding each monitoring site; plotting emissions sources surrounding the
monitoring sites; and presenting traffic data and other characterizing information for each site.
This information is provided to give the reader insight regarding factors that may influence the
air quality near the sites and assist in the interpretation of the ambient measurements.
Seven monitoring sites are located in Oklahoma. Three sites (TOOK, TMOK, and
TROK) are located in Tulsa. Two additional monitoring sites are located in Oklahoma City
(OCOK and NROK), with another located in Yukon, just west of Oklahoma City (YUOK). The
final site (BROK) is located in the town of Bradley, 42 miles south-southwest of Oklahoma City.
Figures 18-1 through 18-3 present composite satellite images retrieved from ArcGIS Explorer
showing the Tulsa monitoring sites and their immediate surroundings. Figure 18-4 identifies
nearby point source emissions locations by source category, as reported in the 2014 NEI for
point sources, version 1. Note that only sources within 10 miles of the sites are included in the
facility counts provided in Figure 18-4. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring sites. Further, this boundary provides both the
proximity of emissions sources to the monitoring sites as well as the quantity of such sources
within a given distance of the sites. Sources outside the 10-mile boundaries are still visible on the
map for reference but have been grayed out to emphasize emissions sources within the
boundaries. Figures 18-5 through 18-10 are the composite satellite maps and emissions sources
maps for the remaining sites. Table 18-1 provides supplemental geographical information such
as land use, location setting, and locational coordinates for each site. Each figure and table is
discussed in detail in the paragraphs that follow.
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
IXiln jjcncmlcd h> sources oilier
lluin I !R( i. LIJ.\'s conliiicl
kiliomioi'\ lor iIk- \\ll\ ;iic noi
included in I he d^ila ;iiuil\ses
18-1
-------
Figure 18-1. Public Works, Tulsa, Oklahoma (TOOK) Monitoring Site
; Wood ward >Blvd«
.24th»St S
iE*25th-St-S'
-------
00
U>
;i3BtK-SrNlj
Figure 18-2. Fire Station, Tulsa, Oklahoma (TIY1QK) Monitoring Site
-------
Figure 18-3. Riverside, Tulsa, Oklahoma (TROK) Monitoring Site
— - - -W I
-W-Edison-St
W Edison-S'
W F.-IS50- t
.W.Easton.RI^
¦O.W.EastonaBl
V -
.WEastonSt-
¦W'Easton«St
W Easlon.St*—^
-WXameronrSt.
W Admiral Slvd
-.W.Brady St
00
-L
-W.ArcJ>er*PI '.«>
* -ft 7* -3 •
W Archer» St
¦pi
<9eBtvd
Newblock Park
Source: USGS
Source: NASA, NGA, USGS
© 2008 Microsoft Corp.
-------
Figure 18-4. NEI Point Sources Located Within 10 Miles of TMOK, TOOK, and TROK
I County
Rogers
County
Tulsa
County
Keystone
Lake
Arkansas
River
Creek
County
Miles
Legend Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
TMOK UATMP site TOOK UATMP site 'fe TROK UATMP site
0 10 mile radius | County boundary
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility
(4)
Airport/Airline/Airport Support Operations
(15)
Asphalt Production/Hot Mix Asphalt Plant
(1)
Automobile/Truck Manufacturing Facility
(2)
Brick, Structural Clay, or Clay Ceramics
Plant (1)
Bulk Terminal/Bulk Plant (1)
Chemical Manufacturing Facility (2)
Compressor Station (2)
Electricity Generation Facility (2)
Electroplating, Plating, Polishing,
Anodizing, and Coloring Facility (1)
I Foundry, Iron and Steel (1) [
& Glass Plant (1)
Industrial Machinery or Equipment Plant
(2)
Q Institution (school, hospital, prison, etc.) (1)
¦ Landfill (2)
Metal Can, Box, and Other Metal
Container Manufacturing Facility (2)
. Metal Coating, Engraving, and Allied
' Services to Manufacturers (2)
<•> Metals Processing/Fabrication Facility (4)
v Mine/Quarry/Mineral Processing Facility
(1)
Miscellaneous Commercial/Industrial
Facility (2)
Municipal Waste Combustor (1)
Oil and/or Gas Production (1)
Paint and Coating Manufacturing Facility
(2)
Petroleum Products Manufacturing Facility
(2)
Petroleum Refinery (2)
Plastic, Resin, or Rubber Products Plant
(4)
Portland Cement Manufacturing Facility (1)
Printing, Coating and Dyeing of Fabrics
Facility (1)
Rail Yard/Rail Line Operations (1)
Railroad Engines/Parts Manufacturing
Facility (1)
18-5
-------
Figure 18-5. Oklahoma City, Oklahoma (OCOK) Monitoring Site
Ave
fynn'Cirdel
Banner Ave
•Repperdme.^ve-
iPep'perdineA^©
Jpnaire;Dr*
Harding Ave-
Harding Pi - ^
SimsAve,
landon Dr
^nriiling-HillBlvdi
wSmilingiHjllBlvd'
Dixie Ln»
U3 -^ Old Hickory.Dr.
¦EvermayDra.
Admiral St-
Oklahoma Ctinman University
¦NE-I40th St
Colonial In
' 1
Nantucket Rd
E Memorial Rd
E Memorial Rd
E Memorial Rd
'GreyFoxRun-
Butternut-PI
Httt 7L ,, o!t>d«Elm
CSource: U S G S'
Sdurfe; NASA, NGA, U5G5 ^
© 2008 Microsoft Cor ex.?
-¦?/¦ Country PlaeeRd;
627ft
fr»o'n>PI iPinon-PI;
00
b\
-------
Figure 18-6. Near-road, Oklahoma City, Oklahoma (NROK) Monitoring Site
r-W Mi'*-
NW 36th St
NW 36th St
¦NW 36th St
32ni#St/
I f " 1
»StJGeoY9e
Will Rogen Park
11 rr.i
NW 3?nd'St-
i r "z^i.y.
,NW;31st,
NW-31st;St
nimi
NW,,3.0th St
'NW 30th St
iNW30th'St^
NW"30th'St.
iNWOIiver-Sl
NW 29th St |
NW-29th'St
INWi29th)St-
18
PL. 1 1 IT
NW 29th"St .
\ vi f ^ i11(
NW 28th St
00
<1
-------
Figure 18-7. Yukon, Oklahoma (YUOK) Monitoring Site
00
00
yklleyftospital
Kingl Canyon i
•V^parklandDr
E'PaYklah'd.Dr
1 rCarlsbad
Sequoia' Pa rtf.br.
irfowmassDr
S Olympicb'r ; ' I
lEMesaVerdeDr
' -is - ' ',< E Olympic Dr'
-Fiddler Rd
'NW'eth.st^i
—uJStfu'rce; USGS
Source: N^SA, N G A, USGS
© 2008 'Microsoft Corp.
NW 10th St
-------
Figure 18-8. NET Point Sources Located Within 10 Miles of NROK, OCOK, and YUOK
Lake
Hefner
North; Canadian
, River •
Lake
'Overholser
Canadian
"-^River
Source Category Group (No. of Facilities)
Airport/Airline/Airport Support Operations (23)
~ Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminal/Bulk Plant (1)
c Chemical Manufacturing Facility (2)
i Compressor Station (6)
f Electricity Generation Facility (2)
Electroplating, Plating, Polishing, Anodizing, and
Coloring Facility (1)
F Food Processing/Agriculture Facility (2)
Industrial Machinery or Equipment Plant (2)
o Institution (school, hospital, prison, etc.) (1)
> Landfill (2)
^ Metal Can, Box, and Other Metal Container
Manufacturing Facility (1)
© Metals Processing/Fabrication Facility (3)
? Miscellaneous Commercial/Industrial Facility (6)
• Oil and/or Gas Production (15)
"q Paint and Coating Manufacturing Facility (2)
R Plastic. Resin, or Rubber Products Plant (2)
Q Railroad Engines/Parts Manufacturing Facility (1)
I onflnrl Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
tV NROK UATMP site ^ OCOK UATMP site ^ YUOK UATMP site
Q 10 mile radius County boundary
Logan
I County
Kingfisher
County
Canadian
County
Cleveland
' County
ake Stanley
Draper
Grady
County
Oklahoma
County
18-9
-------
Figure 18-9. Bradley, Oklahoma (BROK) Monitoring Site
-------
Figure 18-10. NET Point Sources Located Within 10 Miles of BROK
Miles
I
97°55'0"W
Legend
Grady
County
McClain
County
Garvin
County
7°45'0"W 97°40'0"W 97°35"0"W 97"30'0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
BROK UATMP site O 10 mile radius [ County boundary
Source Category Group (No. of Facilities)
t Airport/Airline/Airport Support Operations (2)
: Compressor Station (2)
© Gas Plant (2)
¦ Landfill (1)
• Oil and/or Gas Production (14)
18-11
-------
Table 18-1. Geographical Information for the Oklahoma Monitoring Sites
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
TOOK
40-143-0235
Tulsa
Tulsa
Tulsa, OK
36.126945,
-95.998941
Industrial
Urban/City
Center
66,800
1-244, west side of loop
TMOK
40-143-1127
Tulsa
Tulsa
Tulsa, OK
36.204902,
-95.976537
Residential
Urban/City
Center
4,400
E 36th St N/11, west of US-75
TOOK
40-143-0179
Tulsa
Tulsa
Tulsa, OK
36.154830,
-96.015845
Industrial
Urban/City
Center
55,400
Hwy 64/51/412, west of 1-244
OCOK
40-109-1037
Oklahoma
City
Oklahoma
Oklahoma City,
OK
35.614131,
-97.475083
Residential
Suburban
52,500
US-77, north of 44 (Turnpike), before
bend
NROK
40-109-0097
Oklahoma
City
Oklahoma
Oklahoma City,
OK
35.502979,
-97.577661
Commercial
Urban/City
Center
167,600
1-44, south of NW 39th Expressway
YUOK
40-017-0101
Yukon
Canadian
Oklahoma City,
OK
35.479215,
-97.751503
Commercial
Suburban
42,900
1-40, west of Hwy 4
(east of Exit 132)
BROK
40-051-0065
Bradley
Grady
Oklahoma City,
OK
34.87696,
-97.70748
Residential
Rural
3,100
Hwy-19, between Alex and Bradley
lAADT reflects 2015 data (OK DOT, 2015)
-------
TOOK is located in West Tulsa, on the southwest side of the Arkansas River. The site is
located in the parking lot of the Public Works building. This location is between the Arkansas
River and 1-244, which runs parallel to Southwest Boulevard. The surrounding area is primarily
industrial, although residential areas are located immediately west of the site. The site is located
near the City of Tulsa West Maintenance Yard, which includes a public access CNG station. As
shown in Figure 18-1, an oil refinery is located on the south side of West 25th Street South,
south of TOOK. Another refinery is located to the northwest of the site, on the other side of
1-244, though not visible in the figure. A rail yard is also located on the west side of 1-244, which
can be seen on left-hand side of Figure 18-1.
TMOK is located in north Tulsa on the property of Fire Station Number 24. As shown in
Figure 18-2, the intersection of North Peoria Avenue (Highway 11) and East 36th Street North
lies just to the northeast of the site. The surrounding area is primarily residential, with wooded
areas just to the east, an early childhood education facility and an elementary school to the south,
and a park to the west.
The TROK monitoring site is located west of downtown Tulsa, less than one-half mile
north of the Arkansas River and north-northwest of the TOOK site. Although the area
surrounding the TROK monitoring site is classified as industrial, the site is immediately adjacent
to a residential dwelling, less than one-quarter mile south of Highway 412/51 (Sand Springs
Expressway). The site is elevated above the river, and a wooded area separates the residential
area from the industrial areas west of Newblock Park, as shown in Figure 18-3.
Figure 18-4 shows that the Tulsa sites are located approximately 5 miles apart, with
TMOK farthest north and TOOK farthest south. Many of the emissions sources are clustered
around TOOK and TROK, while there are no point sources within 2 miles of TMOK. A number
of the emissions sources are located along a diagonal line running northeast-southwest through
the center of Tulsa County. There are a variety of industries in the area although the source
category with the greatest number of sources surrounding the Tulsa sites is the airport source
category, which includes airports and related operations as well as small runways and heliports,
such as those associated with hospitals or television stations. Point sources located within one
mile of TOOK include a petroleum refinery; a petroleum products manufacturing facility; a rail
yard; a metal coating, engraving, and allied services to manufacturers facility; and a facility
generating electricity via combustion. The closest point source in the 2014 NEI is just over one
mile east of TROK; this point source is in the miscellaneous commercial/industrial facility
18-13
-------
category. Several of the facilities closest to TROK are also among the closest to TOOK.
However, several industrial facilities are located between the site and river but are not included
in the NEI for point sources. The two point sources within 3 miles of TMOK are involved in
metal coating, engraving, and allied services and metals processing and fabrication.
OCOK is located in northern Oklahoma City, on the property of Oklahoma Christian
University of Science and Arts. The site is located in the northwest corner of the University, near
the athletic fields. The areas surrounding the university are primarily residential. Heavily
traveled roadways such as 1-35 and 1-44 to the east and John Kilpatrick Turnpike to the south are
within a few miles of the site, although outside the boundaries of Figure 18-5.
The NROK site is located in northwest Oklahoma City. The site serves as a near-road site
and is located just off 1-44, near exit 123, where the highway crosses NW 32nd Street, as shown
in Figure 18-6. This location is just over a half-mile south of the 1-44 and NW 39th Expressway
junction. Residential areas are located on the east side of the highway and the Will Rogers Park
and Gardens complex is located to the west.
The YUOK site is located in Yukon, a town to the west of Oklahoma City and in
neighboring Canadian County. The monitoring site is located at the Integris water tower, just
south of 1-40. The site is located in a primarily commercial area, although the area north of 1-40
is highly residential and the area to the south is of mixed usage. An oil well pump jack is located
to the southwest of YUOK, which is shown in the middle of the green field to the southwest of
YUOK in Figure 18-7. Yukon is a rapidly growing area, with both commercial and residential
development.
Figure 18-8 shows the orientation of the Oklahoma City sites, with NROK located about
mid-way between YUOK to the west-southwest and OCOK to the northeast. Most of the point
sources located within 10 miles of these sites are located in the center of Oklahoma City (east
and south of NROK). The source categories with the greatest number of sources surrounding
these sites are the airport source category and the oil and gas production category. The point
source closest to OCOK is involved in metals processing and fabrication. The point source
closest to NROK is a heliport. The source closest to YUOK is an oil and gas production facility.
The BROK site is located in the town of Bradley, 42 miles south of Oklahoma City. The
site is located on the property of the Bradley Fire Department, behind the Post Office, just south
of Highway 19, as shown in Figure 18-9. The surrounding area is rural in nature, with mostly
18-14
-------
residential properties surrounding the site. A church and playground are located farther down
Parker Street. This site was established to monitor for air quality effects related to oil and gas
production, but related activities have decreased in the area (OK DEQ, 2017). Figure 18-10
shows that most of the point sources within 10 miles of BROK are involved in oil and gas
production, including the only two sources within a few miles of the site.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 18-1 also contains traffic volume information for each site as well as the location for
which the traffic volume was obtained. This information is provided because emissions from
motor vehicles can significantly affect concentrations measured at a given monitoring site.
Among the Tulsa sites, the traffic volume passing the TMOK site is considerably less than the
traffic volume near the other two Tulsa sites. For the Oklahoma City sites, the traffic volume
near NROK is significantly higher than the traffic near OCOK and YUOK. Not surprisingly, the
traffic volume near BROK is the lowest among the Oklahoma sites. The traffic data for NROK
ranks 5th highest among all NMP sites; the traffic data for TOOK, TROK, OCOK, and YUOK
rank between 16th and 21st highest among NMP sites; the traffic data for TMOK and BROK are
in the bottom third compared to other NMP sites.
18.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. For each site, each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 18-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 18-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. VOCs, carbonyl compounds, and metals
(TSP) were sampled for at all three Tulsa sites, OCOK, and YUOK. VOCs, SNMOCs, and
carbonyl compounds were sampled for at BROK and NROK. In addition, canister samples
collected at BROK and NROK were also analyzed for methane.
18-15
-------
Table 18-2. 2015-2016 Risk-Based Screening Results for the Oklahoma Monitoring Sites
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
0.45
121
121
100.00
12.22
12.22
Formaldehyde
0.077
121
121
100.00
12.22
24.44
Arsenic (TSP)
0.00023
120
121
99.17
12.12
36.57
Benzene
0.13
120
120
100.00
12.12
48.69
Carbon Tetrachloride
0.17
120
120
100.00
12.12
60.81
1.3 -Butadiene
0.03
109
117
93.16
11.01
71.82
1,2 -Dichloroethane
0.038
106
106
100.00
10.71
82.53
Ethylbenzene
0.4
49
120
40.83
4.95
87.47
Manganese (TSP)
0.03
39
121
32.23
3.94
91.41
Nickel (TSP)
0.0021
30
121
24.79
3.03
94.44
p-Dichlorobcnzcnc
0.091
16
79
20.25
1.62
96.06
Hexachloro-1,3 -butadiene
0.045
14
19
73.68
1.41
97.47
Propionaldehyde
0.8
10
120
8.33
1.01
98.48
Cadmium (TSP)
0.00056
6
121
4.96
0.61
99.09
1,2 -Dibromoethane
0.0017
4
4
100.00
0.40
99.49
Lead (TSP)
0.015
4
121
3.31
0.40
99.90
T richloroethylene
0.2
1
27
3.70
0.10
100.00
Total
990
1,679
58.96
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.45
120
120
100.00
12.78
12.78
Benzene
0.13
120
120
100.00
12.78
25.56
Carbon Tetrachloride
0.17
120
120
100.00
12.78
38.34
Formaldehyde
0.077
120
120
100.00
12.78
51.12
Arsenic (TSP)
0.00023
113
120
94.17
12.03
63.15
1,2 -Dichloroethane
0.038
111
111
100.00
11.82
74.97
1.3 -Butadiene
0.03
110
115
95.65
11.71
86.69
Ethylbenzene
0.4
42
120
35.00
4.47
91.16
p-Dichlorobenzene
0.091
33
83
39.76
3.51
94.68
Hexachloro-1,3 -butadiene
0.045
18
19
94.74
1.92
96.59
Manganese (TSP)
0.03
10
120
8.33
1.06
97.66
1,2 -Dibromoethane
0.0017
6
6
100.00
0.64
98.30
Nickel (TSP)
0.0021
6
120
5.00
0.64
98.94
Cadmium (TSP)
0.00056
5
120
4.17
0.53
99.47
Propionaldehyde
0.8
2
118
1.69
0.21
99.68
T richloroethylene
0.2
2
27
7.41
0.21
99.89
Lead (TSP)
0.015
1
120
0.83
0.11
100.00
Total
939
1,679
55.93
18-16
-------
Table 18-2. 2015-2016 Risk-Based Screening Results for the Oklahoma Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Riverside, Tulsa, Oklahoma - TROK
Benzene
0.13
121
121
100.00
12.90
12.90
Carbon Tetrachloride
0.17
121
121
100.00
12.90
25.80
Acetaldehyde
0.45
120
120
100.00
12.79
38.59
Formaldehyde
0.077
120
120
100.00
12.79
51.39
Arsenic (TSP)
0.00023
119
121
98.35
12.69
64.07
1.3 -Butadiene
0.03
110
119
92.44
11.73
75.80
1,2 -Dichloroethane
0.038
110
110
100.00
11.73
87.53
Ethylbenzene
0.4
39
121
32.23
4.16
91.68
Hexachloro-1,3 -butadiene
0.045
18
19
94.74
1.92
93.60
Nickel (TSP)
0.0021
16
121
13.22
1.71
95.31
p-Dichlorobenzene
0.091
14
67
20.90
1.49
96.80
Manganese (TSP)
0.03
10
121
8.26
1.07
97.87
1,2 -Dibromoethane
0.0017
9
9
100.00
0.96
98.83
Propionaldehyde
0.8
5
117
4.27
0.53
99.36
Cadmium (TSP)
0.00056
3
121
2.48
0.32
99.68
1,1,2-Trichloroethane
0.0625
2
4
50.00
0.21
99.89
Lead (TSP)
0.015
1
121
0.83
0.11
100.00
Total
938
1,653
56.75
Oklahoma City, Oklahoma - OCOK
Benzene
0.13
121
121
100.00
14.67
14.67
Acetaldehyde
0.45
120
120
100.00
14.55
29.21
Carbon Tetrachloride
0.17
120
121
99.17
14.55
43.76
Formaldehyde
0.077
120
120
100.00
14.55
58.30
1,2 -Dichloroethane
0.038
110
111
99.10
13.33
71.64
Arsenic (TSP)
0.00023
108
120
90.00
13.09
84.73
1.3 -Butadiene
0.03
77
104
74.04
9.33
94.06
Hexachloro-1,3 -butadiene
0.045
15
19
78.95
1.82
95.88
Manganese (TSP)
0.03
7
120
5.83
0.85
96.73
1,2 -Dibromoethane
0.0017
6
6
100.00
0.73
97.45
p-Dichlorobenzene
0.091
5
44
11.36
0.61
98.06
T richloroethylene
0.2
5
23
21.74
0.61
98.67
Ethylbenzene
0.4
4
121
3.31
0.48
99.15
Nickel (TSP)
0.0021
3
120
2.50
0.36
99.52
Propionaldehyde
0.8
3
120
2.50
0.36
99.88
Cadmium (TSP)
0.00056
1
120
0.83
0.12
100.00
Total
825
1,510
54.64
18-17
-------
Table 18-2. 2015-2016 Risk-Based Screening Results for the Oklahoma Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Near-road, Oklahoma City, Oklahoma - NROK
Acetaldehyde
0.45
38
38
100.00
13.77
13.77
Benzene
0.13
38
38
100.00
13.77
27.54
Carbon Tetrachloride
0.17
38
38
100.00
13.77
41.30
Formaldehyde
0.077
38
38
100.00
13.77
55.07
1.3 -Butadiene
0.03
37
38
97.37
13.41
68.48
1,2 -Dichloroethane
0.038
32
32
100.00
11.59
80.07
Ethylbenzene
0.4
28
38
73.68
10.14
90.22
p-Dichlorobenzene
0.091
16
32
50.00
5.80
96.01
Hexachloro-1,3 -butadiene
0.045
10
10
100.00
3.62
99.64
1,2 -Dibromoethane
0.0017
1
1
100.00
0.36
100.00
Total
276
303
91.09
Yukon, Oklahoma - YUOK
Acetaldehyde
0.45
120
120
100.00
14.67
14.67
Benzene
0.13
120
120
100.00
14.67
29.34
Carbon Tetrachloride
0.17
120
120
100.00
14.67
44.01
Formaldehyde
0.077
120
120
100.00
14.67
58.68
1,2 -Dichloroethane
0.038
113
114
99.12
13.81
72.49
Arsenic (TSP)
0.00023
106
120
88.33
12.96
85.45
1.3 -Butadiene
0.03
73
108
67.59
8.92
94.38
Hexachloro-1,3 -butadiene
0.045
13
15
86.67
1.59
95.97
Manganese (TSP)
0.03
9
120
7.50
1.10
97.07
1,2 -Dibromoethane
0.0017
8
8
100.00
0.98
98.04
Ethylbenzene
0.4
7
120
5.83
0.86
98.90
Nickel (TSP)
0.0021
4
120
3.33
0.49
99.39
p-Dichlorobcnzcnc
0.091
3
49
6.12
0.37
99.76
Cadmium (TSP)
0.00056
1
120
0.83
0.12
99.88
1,1,2-Trichloroethane
0.0625
1
5
20.00
0.12
100.00
Total
818
1,379
59.32
18-18
-------
Table 18-2. 2015-2016 Risk-Based Screening Results for the Oklahoma Monitoring Sites
(Continued)
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bradley, Oklahoma - BROK
Acetaldehyde
0.45
100
100
100.00
18.28
18.28
Formaldehyde
0.077
100
100
100.00
18.28
36.56
Benzene
0.13
98
98
100.00
17.92
54.48
Carbon Tetrachloride
0.17
97
98
98.98
17.73
72.21
1,2 -Dichloroethane
0.038
90
90
100.00
16.45
88.67
1.3 -Butadiene
0.03
29
74
39.19
5.30
93.97
Propionaldehyde
0.8
13
100
13.00
2.38
96.34
Hexachloro-1,3 -butadiene
0.045
9
9
100.00
1.65
97.99
p-Dichlorobcnzcnc
0.091
5
24
20.83
0.91
98.90
Ethylbenzene
0.4
4
97
4.12
0.73
99.63
1,2 -Dibromoethane
0.0017
2
2
100.00
0.37
100.00
Total
547
792
69.07
Observations from Table 18-2 include the following:
• Concentrations of 17 pollutants failed at least one screen for TOOK; nearly
59 percent of concentrations for these 17 pollutants were greater than their associated
risk screening value (or failed screens).
• Concentrations of 11 pollutants contributed to 95 percent of failed screens for TOOK
and therefore were identified as pollutants of interest for this site. These 11 include
two carbonyl compounds, six VOCs, and three TSP metals. TOOK is one of only two
NMP sites for which manganese was identified as a pollutant of interest.
• Concentrations of 17 pollutants failed at least one screen for TMOK; nearly
56 percent of concentrations for these 17 pollutants were greater than their associated
risk screening value (or failed screens).
• Concentrations of 10 pollutants contributed to 95 percent of failed screens for TMOK
and therefore were identified as pollutants of interest for this site. These 10 include
two carbonyl compounds, seven VOCs, and one TSP metal.
• Concentrations of 17 pollutants failed at least one screen for TROK; nearly
57 percent of concentrations for these 17 pollutants were greater than their associated
risk screening value (or failed screens).
• Concentrations of 10 pollutants contributed to 95 percent of failed screens for TROK
and therefore were identified as pollutants of interest for this site. These 10 include
two carbonyl compounds, six VOCs, and two TSP metals.
18-19
-------
Concentrations of 16 pollutants failed at least one screen for OCOK; nearly
55 percent of concentrations for these 16 pollutants were greater than their associated
risk screening value (or failed screens).
Concentrations of eight pollutants contributed to 95 percent of failed screens for
OCOK and therefore were identified as pollutants of interest for this site. These eight
include two carbonyl compounds, five VOCs, and one TSP metal.
Concentrations of 10 pollutants failed at least one screen for NROK; 91 percent of
concentrations for these 10 pollutants were greater than their associated risk screening
value (or failed screens).
Concentrations of eight pollutants contributed to 95 percent of failed screens for
NROK and therefore were identified as pollutants of interest for this site. These eight
include two carbonyl compounds and six VOCs.
Concentrations of 15 pollutants failed at least one screen for YUOK; 59 percent of
concentrations for these 15 pollutants were greater than their associated risk screening
value (or failed screens).
Concentrations of eight pollutants contributed to 95 percent of failed screens for
YUOK and therefore were identified as pollutants of interest for this site. These eight
include two carbonyl compounds, five VOCs, and one TSP metal.
Concentrations of 11 pollutants failed at least one screen for BROK; 69 percent of
concentrations for these 11 pollutants were greater than their associated risk screening
value (or failed screens).
Concentrations of seven pollutants contributed to 95 percent of failed screens for
BROK and therefore were identified as pollutants of interest for this site. These seven
include three carbonyl compounds and four VOCs. BROK is one of only two NMP
sites for which propionaldehyde was identified as a pollutant of interest.
The number of pollutants identified as pollutants of interest range from seven
(BROK) to 11 (TOOK) among the Oklahoma sites. The Tulsa sites each have at least
10 pollutants of interest while the Oklahoma City sites have eight, and BROK has
seven.
Note that sampling at BROK began in April 2015 and sampling at NROK began in
May 2016.
As described in Section 3.2, if a pollutant was measured by both the TO-15 and
SNMOC methods at the same site, the TO-15 results were used for the risk-based
screening process. As BROK and NROK sampled both VOCs (TO-15) and
SNMOCs, the TO-15 results were used for the 12 pollutants these methods have in
common.
18-20
-------
• The Oklahoma sites have six pollutants of interest in common: acetaldehyde,
benzene, 1,3-butadiene, carbon tetrachloride, 1,2-dichloroethane, and formaldehyde.
If the two sites not sampling metals are excluded (NROK and BROK), arsenic would
also be on this list.
• Concentrations measured at TOOK failed the fourth highest number of screens
among NMP sites, with other five Oklahoma sites that sampled over the full two
years ranking in the top third, as shown in Table 4-9.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
18.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Oklahoma monitoring sites. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
Methane is not a HAP, and therefore has no risk screening value. Thus, methane results
are automatically excluded from the sections that follow. However, Appendix D provides
individual methane measurements and Appendix L provides statistical summaries for the period
of sampling for BROK and NROK. Similarly, site-specific statistical summaries for all pollutants
sampled for at the Oklahoma sites are provided in Appendices J through M and O.
18-21
-------
18.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for each Oklahoma site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the Oklahoma
monitoring sites are presented in Table 18-3, where applicable. Note that concentrations of the
TSP metals are presented in ng/m3 for ease of viewing. Also note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
18-22
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
Oig/m3)
Q3
Avg
Oig/m3)
Q4
Avg
Oig/m3)
Annual
Average
frig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
60/60/60
1.52
±0.33
2.11
±0.41
2.56
±0.34
1.88
±0.51
2.02
±0.21
61/61/61
1.61
±0.30
1.92
±0.36
2.33
±0.23
2.06
±0.47
1.97
±0.18
Benzene
59/59/59
0.91
±0.22
0.90
±0.13
1.31
±0.21
1.23
±0.35
1.09
±0.12
61/61/61
1.23
±0.33
0.85
±0.18
0.90
±0.17
1.33
±0.29
1.08
±0.13
1.3 -Butadiene
56/53/59
0.07
±0.02
0.06
±0.01
0.05
±0.02
0.10
±0.05
0.07
±0.01
61/35/61
0.11
±0.03
0.06
±0.02
0.06
±0.02
0.11
±0.04
0.08
±0.01
Carbon Tetrachloride
59/59/59
0.58
±0.04
0.6
±0.02
0.66
±0.04
0.60
±0.02
0.61
±0.02
61/61/61
0.57
±0.04
0.67
±0.05
0.61
±0.05
0.61
±0.04
0.61
±0.02
p-Dichlorobenzene
48/2/59
0.05
±0.01
0.05
±0.01
0.04
±0.02
0.08
±0.04
0.06
±0.01
31/2/61
0.05
±0.02
0.04
±0.02
0.03
±0.02
0.07
±0.05
0.05
±0.02
1,2 -Dichloroethane
51/49/59
0.11
±0.01
0.06
±0.03
0.07
±0.02
0.12
±0.02
0.09
±0.01
55/55/61
0.12
±0.01
0.12
±0.02
0.06
±0.02
0.10
±0.02
0.10
±0.01
Ethylbenzene
59/59/59
0.30
±0.09
0.36
±0.06
0.41
±0.12
0.54
±0.18
0.40
±0.06
61/61/61
0.48
±0.18
0.37
±0.12
0.43
±0.12
0.60
±0.21
0.47
±0.08
Formaldehyde
60/60/60
1.78
±0.31
2.85
±0.84
4.53
±0.63
2.18
±0.45
2.84
±0.39
61/61/61
2.38
±0.55
3.79
±0.82
4.82
±0.59
2.41
±0.38
3.33
±0.39
Arsenic (TSP)a
60/60/60
0.67
±0.11
0.69
±0.12
0.95
±0.23
0.82
±0.18
0.78
±0.08
61/61/61
0.83
±0.21
0.76
±0.18
0.84
±0.11
1.14
±0.35
0.89
±0.11
Manganese (TSP)a
60/60/60
21.78
±5.35
24.05
±8.91
29.97
±7.00
28.23
± 10.47
26.01
±3.92
61/61/61
27.92
± 11.24
25.41
±9.23
29.04
±5.94
25.32
±7.36
26.94
±4.15
Nickel (TSP)a
60/59/60
2.49
± 1.31
1.68
±0.45
1.79
±0.23
1.85
±0.52
1.95
±0.36
61/61/61
2.35
±0.94
1.94
±0.97
2.93
±2.84
1.43
±0.34
2.17
±0.75
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
60/60/60
1.58
±0.36
2.25
±0.44
2.10
±0.24
1.42
±0.31
1.84
±0.19
60/60/60
1.42
±0.30
1.66
±0.25
2.21
±0.3
1.82
±0.37
1.77
±0.16
Benzene
60/60/60
0.85
±0.16
0.63
±0.14
0.92
±0.25
0.97
±0.21
0.84
±0.10
60/60/60
0.91
±0.25
0.74
±0.19
0.76
±0.21
1.21
±0.29
0.90
±0.12
1.3 -Butadiene
56/53/60
0.09
±0.04
0.07
±0.02
0.08
±0.03
0.10
±0.04
0.09
±0.02
59/41/60
0.13
±0.05
0.08
±0.03
0.08
±0.02
0.15
±0.06
0.11
±0.02
Carbon Tetrachloride
60/60/60
0.56
±0.04
0.63
±0.03
0.67
±0.04
0.61
±0.04
0.62
±0.02
60/60/60
0.60
±0.04
0.69
±0.04
0.63
±0.04
0.61
±0.06
0.63
±0.02
p-Dichlorobcnzcnc
42/3/60
0.04
±0.02
0.04
±0.02
0.05
±0.03
0.10
±0.03
0.06
±0.01
41/13/60
0.06
±0.04
0.06
±0.03
0.08
±0.04
0.11
±0.06
0.08
±0.02
1,2 -Dichloroethane
54/51/60
0.10
±0.02
0.06
±0.03
0.07
±0.02
0.11
±0.01
0.08
±0.01
57/57/60
0.11
±0.02
0.11
±0.02
0.08
±0.02
0.10
±0.03
0.10
±0.01
Ethylbenzene
60/59/60
0.25
±0.07
0.29
±0.07
0.36
±0.10
0.38
±0.10
0.32
±0.04
60/58/60
0.34
±0.16
0.34
±0.11
0.41
±0.12
0.70
±0.22
0.44
±0.08
Formaldehyde
60/60/60
1.94
±0.37
3.23
±0.75
3.88
±0.58
1.81
±0.43
2.71
±0.34
60/60/60
2.31
±0.65
3.44
±0.84
4.66
±0.77
2.23
±0.32
3.16
±0.41
Hexachloro-1,3 -butadiene
1/0/60
NR
NR
NR
NR
NR
18/0/60
0.02
±0.02
0.03
±0.03
0.03
±0.03
0.03
±0.03
0.03
±0.01
Arsenic (TSP)a
59/59/59
0.58
±0.18
0.56
±0.14
0.75
±0.15
0.68
±0.14
0.64
±0.07
61/61/61
0.53
±0.17
0.60
±0.13
0.70
±0.12
0.84
±0.21
0.67
±0.08
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
Riverside, Tulsa, Oklahoma - TROK
Acetaldehvde
59/59/59
1.55
±0.31
2.00
±0.44
2.45
±0.32
1.68
±0.34
1.92
±0.19
61/61/61
1.43
±0.27
2.06
±0.34
2.70
±0.30
1.91
±0.37
2.02
±0.19
Benzene
60/60/60
0.76
±0.11
0.72
±0.15
0.79
±0.09
0.88
±0.12
0.79
±0.06
61/61/61
0.84
±0.16
0.76
±0.17
0.98
±0.24
1.00
±0.20
0.89
±0.09
1.3 -Butadiene
58/52/60
0.08
±0.02
0.06
±0.01
0.06
±0.02
0.08
±0.03
0.07
±0.01
61/44/61
0.10
±0.03
0.06
±0.01
0.06
±0.02
0.10
±0.02
0.08
±0.01
Carbon Tetrachloride
60/60/60
0.57
±0.04
0.58
±0.04
0.66
±0.03
0.62
±0.05
0.61
±0.02
61/61/61
0.52
±0.07
0.67
±0.04
0.60
±0.04
0.61
±0.03
0.60
±0.03
1,2 -Dichloroethane
53/49/60
0.10
±0.01
0.06
±0.03
0.06
±0.02
0.09
±0.02
0.08
±0.01
57/57/61
0.10
±0.01
0.11
±0.02
0.07
±0.03
0.10
±0.01
0.10
±0.01
Ethvlbenzene
60/59/60
0.23
±0.05
0.31
±0.07
0.36
±0.05
0.34
±0.06
0.31
±0.03
61/61/61
0.25
±0.09
0.35
±0.12
0.58
±0.16
0.46
±0.13
0.41
±0.07
Formaldehyde
59/59/59
1.81
±0.41
2.83
±0.92
4.05
±0.70
2.05
±0.48
2.67
±0.38
61/61/61
2.07
±0.41
3.25
±0.77
4.33
±0.60
2.08
±0.35
2.92
±0.35
Hexachloro-1,3 -butadiene
1/0/60
NR
NR
NR
NR
NR
18/0/61
0.02
±0.02
0.05
±0.03
0.01
±0.02
0.03
±0.02
0.03
±0.01
Arsenic (TSP)a
60/60/60
0.77
±0.23
0.76
±0.25
1.00
±0.46
0.85
±0.22
0.85
±0.14
61/61/61
0.79
±0.29
0.79
±0.15
0.79
±0.15
1.43
±0.74
0.95
±0.20
Nickel (TSP)a
60/58/60
1.63
±0.51
1.41
±0.45
1.32
±0.28
1.41
±0.37
1.44
±0.19
61/60/61
1.22
±0.54
1.12
±0.24
1.32
±0.32
1.15
±0.35
1.20
±0.18
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
60/60/60
1.41
±0.38
1.96
±0.48
2.73
±0.33
1.60
±0.34
1.92
±0.22
60/60/60
1.21
±0.19
1.69
±0.29
1.92
±0.22
1.75
±0.35
1.63
±0.14
Benzene
60/60/60
0.64
±0.05
0.49
±0.10
0.58
±0.09
0.61
±0.08
0.58
±0.04
61/61/61
0.76
±0.42
0.44
±0.08
0.55
±0.09
0.68
±0.11
0.61
±0.11
1.3 -Butadiene
50/36/60
0.04
±0.01
0.03
±0.01
0.03
±0.01
0.05
±0.02
0.04
±0.01
54/14/61
0.05
±0.02
0.03
±0.01
0.03
±0.01
0.06
±0.02
0.05
±0.01
Carbon Tetrachloride
60/60/60
0.61
±0.04
0.62
±0.03
0.68
±0.03
0.63
±0.04
0.64
±0.02
61/61/61
0.54
±0.08
0.67
±0.03
0.62
±0.05
0.63
±0.06
0.61
±0.03
1,2 -Dichloroethane
56/50/60
0.10
±0.01
0.08
±0.01
0.05
±0.02
0.09
±0.01
0.08
±0.01
55/52/61
0.09
±0.01
0.08
±0.01
0.04
±0.02
0.07
±0.01
0.07
±0.01
Formaldehyde
60/60/60
1.90
±0.37
2.75
±0.76
5.71
±0.80
2.26
±0.70
3.16
±0.50
60/60/60
1.51
±0.25
2.96
±0.78
4.37
±0.63
2.24
±0.33
2.72
±0.37
Hexachloro-1,3 -butadiene
5/0/60
NR
NR
NR
NR
NR
14/0/61
0.01
±0.01
0.03
±0.02
0.01
±0.02
0.02
±0.02
0.02
±0.01
Arsenic (TSP)a
60/60/60
0.45
±0.10
0.57
±0.15
0.65
±0.16
0.56
±0.11
0.56
±0.07
60/60/60
0.41
±0.10
0.49
±0.10
0.59
±0.12
0.59
±0.18
0.51
±0.06
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
Near-road, Oklahoma City, Oklahoma - NROK
Acetaldehyde
NS
NS
NS
NS
NS
NS
38/38/38
NS
NR
2.16
±0.18
1.85
±0.41
NR
Benzene
NS
NS
NS
NS
NS
NS
38/38/38
NS
NR
1.17
±0.10
0.94
±0.23
NR
1.3 -Butadiene
NS
NS
NS
NS
NS
NS
38/36/38
NS
NR
0.17
±0.03
0.15
±0.04
NR
Carbon Tetrachloride
NS
NS
NS
NS
NS
NS
38/38/38
NS
NR
0.50
±0.10
0.50
±0.06
NR
p-Dichlorobenzene
NS
NS
NS
NS
NS
NS
32/10/38
NS
NR
0.12
±0.04
0.11
±0.05
NR
1,2 -Dichloroethane
NS
NS
NS
NS
NS
NS
32/30/38
NS
NR
0.05
±0.02
0.07
±0.02
NR
Ethylbenzene
NS
NS
NS
NS
NS
NS
38/37/38
NS
NR
0.80
±0.07
0.37
±0.13
NR
Formaldehyde
NS
NS
NS
NS
NS
NS
38/38/38
NS
NR
4.80
±0.55
2.45
±0.41
NR
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(Ug/m3)
Q2
Avg
(Ug/m3)
Q3
Avg
(Ug/m3)
Q4
Avg
(Ug/m3)
Annual
Average
(Ug/m3)
Yukon, Oklahoma - YUOK
Acetaldehyde
59/59/59
1.30
±0.34
1.64
±0.28
2.16
±0.26
1.49
±0.29
1.64
±0.16
61/61/61
1.16
±0.15
1.59
±0.27
1.93
±0.17
1.72
±0.30
1.60
±0.13
Benzene
59/59/59
0.65
±0.06
0.41
±0.07
0.60
±0.25
0.56
±0.04
0.56
±0.07
61/61/61
0.56
±0.10
0.41
±0.05
0.50
±0.11
0.61
±0.10
0.52
±0.05
1.3 -Butadiene
51/33/59
0.04
±0.01
0.04
±0.01
0.03
±0.02
0.04
±0.02
0.04
±0.01
57/10/61
0.04
±0.01
0.04
±0.01
0.04
±0.01
0.04
±0.01
0.04
±0.01
Carbon Tetrachloride
59/59/59
0.58
±0.06
0.64
±0.03
0.67
±0.03
0.58
±0.06
0.62
±0.03
61/61/61
0.58
±0.03
0.64
±0.03
0.61
±0.03
0.60
±0.08
0.61
±0.02
1,2 -Dichloroethane
57/47/59
0.09
±0.01
0.08
±0.01
0.06
±0.02
0.08
±0.01
0.08
±0.01
57/54/61
0.09
±0.01
0.09
±0.01
0.05
±0.02
0.08
±0.01
0.08
±0.01
Formaldehyde
59/59/59
1.76
±0.32
2.40
±0.54
4.57
±0.61
2.23
±0.81
2.71
±0.39
61/61/61
1.33
±0.22
2.83
±0.70
4.35
±0.50
2.13
±0.45
2.64
±0.37
Hexachloro-1,3 -butadiene
2/0/59
NR
NR
NR
NR
NR
13/0/61
<0.01
±0.01
0.03
±0.02
0.01
±0.02
0.02
±0.02
0.02
±0.01
Arsenic (TSP)a
59/59/59
0.46
±0.14
0.51
±0.09
0.62
±0.24
0.62
±0.21
0.55
±0.09
61/61/61
0.39
±0.09
0.54
±0.20
0.67
±0.15
0.58
±0.21
0.54
±0.08
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Table 18-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Oklahoma Monitoring Sites
(Continued)
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
# >MDL/
Avg
Avg
Avg
Avg
Average
# >MDL/
Avg
Avg
Avg
Avg
Average
Pollutant
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
# Samples
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
(Ug/m3)
Bradley, Oklahoma - BROK
1.50
6.23
4.23
4.06
1.08
1.73
1.69
1.40
1.46
Acetaldehyde
40/40/40
NS
±0.40
±3.78
± 1.61
± 1.42
60/60/60
±0.18
±0.29
±0.25
±0.26
±0.14
0.75
0.85
0.73
0.67
0.84
0.96
0.80
Benzene
39/39/39
NS
NA
±0.14
±0.20
NA
59/59/59
±0.19
±0.21
±0.20
±0.26
±0.10
0.02
0.02
0.02
0.02
0.02
0.02
0.02
1.3 -Butadiene
29/13/39
NS
NA
±0.01
±0.01
NA
45/2/59
±0.01
±0.01
±0.01
±0.01
± <0.01
0.66
0.56
0.56
0.70
0.57
0.54
0.59
Carbon Tetrachloride
39/39/39
NS
NA
±0.03
±0.05
NA
59/59/59
±0.07
±0.03
±0.09
±0.07
±0.04
0.07
0.08
0.10
0.12
0.06
0.10
0.09
1,2 -Dichloroethane
34/32/39
NS
NA
±0.02
±0.02
NA
56/56/59
±0.01
±0.02
±0.02
±0.01
±0.01
1.81
8.27
4.58
4.95
1.20
2.30
3.36
1.52
2.07
Formaldehyde
40/40/40
NS
±0.60
±3.86
± 1.89
± 1.59
60/60/60
±0.21
±0.50
±0.59
±0.33
±0.30
0.27
1.75
0.97
1.01
0.26
0.42
0.37
0.28
0.33
Propionaldehyde
40/40/40
NS
±0.08
± 1.16
±0.38
±0.42
60/60/60
±0.05
±0.08
±0.06
±0.05
±0.03
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Observations for the Tulsa sites from Table 18-4 include the following:
• The pollutants with the highest annual average concentrations for each of the Tulsa
sites are formaldehyde, acetaldehyde, and benzene. Annual averages for
formaldehyde and acetaldehyde are greater than 1 |ig/m3 for all three sites; TOOK's
annual average concentrations for benzene are also greater than 1 |ig/m3.
• Across the Tulsa sites, annual average concentrations of formaldehyde range from
2.67 ± 0.38 |ig/m3 (TROK, 2015) to 3.33 ± 0.39 |ig/m3 (TOOK, 2016). For each site,
the annual average for 2016 appears higher than the annual average for 2015, but the
difference is not statistically significant. The quarterly average concentrations for
TOOK, TMOK, and TROK indicate that higher concentrations were most often
measured during the warmer months of the year. Thirty-nine of the 43 concentrations
of formaldehyde greater than 5 |ig/m3 measured at these three sites were measured in
June, July, or August of either year. In 2015, all 16 formaldehyde concentrations
greater than 5 |ig/m3 were measured between June and August; for 2016, 23 of the 27
formaldehyde concentrations greater than 5 |ig/m3 were measured between June and
August (with the exceptions in February and September).
• Across the Tulsa sites, annual average concentrations of acetaldehyde range from
1.77 ± 0.16 |ig/m3 (TMOK, 2016) to 2.02 ± 0.21 |ig/m3 (TOOK, 2015), although a
similar average was calculated for TROK in 2016. For each site, the annual averages
for 2015 and 2016 vary by 0.1 |ig/m3 or less. While the highest acetaldehyde
concentrations were often measured at these sites between June and August, there is
more variability during the seasons in which the higher concentrations were measured
than formaldehyde.
• Across the Tulsa sites, annual average concentrations of benzene range from
0.79 ± 0.06 |ig/m3 (TROK, 2015) to 1.09 ± 0.12 |ig/m3 (TOOK, 2015). For each site,
the annual averages for 2015 and 2016 vary by 0.1 |ig/m3 or less, with the annual
averages for TOOK varying by 0.01 |ig/m3. Ten benzene concentrations greater than
2 |ig/m3 were measured at these three sites, with eight of the 10 measured at TOOK,
and one each measured at TROK and TMOK.
• Hexachloro-1,3-butadiene is a pollutant of interest for TMOK and TROK. Quarterly
and annual averages for hexachloro-l,3-butadiene for 2015 are not presented in
Table 18-3 due to the use of a contaminated internal standard at the laboratory for
Method TO-15, which resulted in the invalidation of the results from early
March 2015 through mid-December 2015, as described in Section 2.4. TMOK and
TROK's annual average concentrations of hexchloro-l,3-butadiene for 2016 are
similar to each other, with the quarterly averages exhibiting slightly more variability.
• Several of the pollutants of interest exhibited little variability across the three Tulsa
sites. For example, annual averages of carbon tetrachloride vary by 0.03 |ig/m3 across
the three sites, though this is expected given the ubiquitous nature of this pollutant.
1,2-Dichloroethane is another example, with annual averages varying by only
0.02 |ig/m3 across these sites, as is 1,3-butadiene, with annual averages varying by
only 0.04 |ig/m3.
18-30
-------
• Arsenic is the only TSP metal that is a pollutant of interest across the three Tulsa
sites. Annual average concentrations of arsenic range from 0.64 ± 0.07 ng/m3
(TMOK, 2015) to 0.95 ± 0.20 ng/m3 (TROK, 2016). Arsenic concentrations measured
at TROK in 2016 exhibit the most variability, based on the confidence intervals
shown. Three of the four quarterly average concentrations for 2016 are the same
(each is 0.79 ng/m3, though the confidence intervals vary); the fourth quarter average
concentration for 2016 is considerably higher (1.43 ± 0.74 ng/m3), with a much larger
confidence interval, indicating the potential for outliers. The maximum arsenic
concentration measured at TROK (5.65 ng/m3) was measured in October 2016 and is
the fourth highest arsenic concentration measured across the program. Across the
Tulsa sites, six of the seven highest arsenic concentrations (those greater than
2 ng/m3) were measured at TROK (five of which were measured in 2016).
• Concentrations of nickel measured at TOOK exhibit considerable variability,
particularly for 2016. Quarterly average concentrations of nickel for TOOK range
from 1.43 ± 0.34 ng/m3 (fourth quarter, 2016) to 2.93 ± 2.84 ng/m3 (third quarter,
2016). The maximum nickel concentration measured at TOOK (22.1 ng/m3) is more
than twice the next highest nickel concentration measured at TOOK (10.3 ng/m3).
The nine highest nickel concentrations measured across the Tulsa sites were
measured at TOOK. Nickel concentrations measured at TROK, the other Tulsa site
for which nickel is a pollutant of interest, exhibit less variability.
Observations for the Oklahoma City sites from Table 18-4 include the following:
• OCOK and NROK are located within Oklahoma City; YUOK is located just outside
the city in nearby Yukon. Although the town of Bradley is located well outside the
urban reach of Oklahoma City, Grady County is part of the Oklahoma City CBSA,
and thus, BROK is included in this section.
• Sampling at BROK began in April 2015, thus first quarter averages could not be
calculated for 2015. In addition, second quarter averages, and thus annual averages,
for the VOCs for 2015 could not be calculated for this site. Because sampling at
NROK began in May 2016, few quarterly averages and no annual averages could be
calculated for this site.
• The pollutants with the highest annual average concentrations for each of the
Oklahoma City sites, where they could be calculated, are formaldehyde and
acetaldehyde. These are the only pollutants with annual averages greater than
1 |ig/m3, with one exception (BROK's annual average concentration of
propionaldehyde for 2015).
• Across the Oklahoma City sites, annual average concentrations of formaldehyde
range from 2.07 ± 0.30 |ig/m3 (BROK, 2016) to 4.95 ± 1.59 |ig/m3 (BROK, 2015).
Both the highest and lowest annual averages were calculated for BROK, with the
annual average for 2015 more than twice the annual average for 2016. If the annual
averages for BROK are excluded, the annual averages across the remaining sites vary
by about 0.5 |ig/m3. The seven highest formaldehyde concentrations measured at an
Oklahoma City site were measured at BROK, each of which was measured between
the end of August and mid-October 2015. For each site, the third quarter average
18-31
-------
concentration of formaldehyde is the highest quarterly average for each year. Though
the statistical significance of this varies by site, higher concentrations were most often
measured during the warmer months of the year. Of the 48 concentrations of
formaldehyde greater than 5 |ig/m3 measured at these sites, the majority (34) were
measured during the third quarter of either year, with another seven measured in June.
Across the Oklahoma City sites, annual average concentrations of acetaldehyde range
from 1.46 ± 0.14 |ig/m3 (BROK, 2016) to 4.06 ± 1.42 |ig/m3 (BROK, 2015). Similar
to formaldehyde, both the highest and lowest annual averages of acetaldehyde were
calculated for BROK, with the annual average for 2015 nearly three times the annual
average for 2016. If the annual averages for BROK are excluded, the annual averages
across the remaining sites vary by about 0.3 |ig/m3. The 12 highest acetaldehyde
concentrations measured at an Oklahoma City site were measured at BROK (those
greater than 4 |ig/m3), each of which was measured between the end of August and
the end of December 2015. These higher measurements are reflected in BROK's third
and fourth quarter averages for 2015.
The trend in BROK's carbonyl compound data continues with propionaldehyde.
BROK's annual average concentration for 2015 is three times greater than the annual
average for 2016. Among the Oklahoma City sites, propionaldehyde concentrations
greater than 1 |ig/m3 were only measured at BROK. Further, the eight highest
propionaldehyde concentrations measured across the program were measured at
BROK.
Among the VOC pollutants of interest, benzene and carbon tetrachloride had the
highest annual average concentrations for each Oklahoma City site. Annual averages
for the remaining VOC pollutants of interest are less than 0.1 |ig/m3.
Across the Oklahoma City sites, annual average concentrations of benzene range
from 0.52 ± 0.05 |ig/m3 (YUOK, 2016) to 0.80 ± 0.10 |ig/m3 (BROK, 2016). The site-
specific annual average concentrations vary little across each year, where two annual
averages could be calculated. Among the Oklahoma City sites, the highest benzene
concentration was measured at OCOK (3.71 |ig/m3), which is reflected in this site's
first quarter average concentration for 2016 (0.76 ± 0.42 |ig/m3). While the quarterly
average itself is only slightly higher than the other quarterly averages, the confidence
interval is four to five times higher than the confidence intervals calculated for the
remaining quarterly averages. Other Oklahoma City sites have quarterly average
concentrations of similar or greater magnitude (NROK, BROK), but the confidence
intervals are considerably less. NROK's third quarter average concentration for 2016
(1.17 ± 0.10 |ig/m3) is the highest quarterly average concentration of benzene among
the Oklahoma City sites.
Across the Oklahoma City sites, annual average concentrations of carbon
tetrachloride vary by about 0.05 |ig/m3, ranging from 0.59 ± 0.04 |ig/m3 (BROK,
2016) to 0.64 ± 0.02 |ig/m3 (OCOK, 2015). The quarterly average concentrations
exhibit somewhat more variability across the sites, ranging from 0.50 ± 0.06 |ig/m3
(NROK, third quarter 2016) to 0.70 ± 0.03 |ig/m3 (BROK, second quarter 2016).
18-32
-------
• Arsenic is a pollutant of interest for OCOK and YUOK. Annual average
concentrations of arsenic for these two sites vary by only 0.05 ng/m3, ranging from
0.51 ± 0.06 ng/m3 (OCOK, 2016) to 0.56 ± 0.07 ng/m3 (OCOK, 2015). The quarterly
average concentrations of arsenic exhibit more variability, ranging from
0.39 ± 0.09 ng/m3 (YUOK, first quarter 2016) to 0.67 ±0.15 ng/m3 (YUOK, third
quarter 2016).
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for the
Oklahoma sites include the following:
• The Oklahoma sites appear in Tables 4-10 through 4-13 a total of 27 times. However,
because they are the only sites sampling TSP metals, each Oklahoma site appears for
arsenic, accounting for 10 of the appearances.
• TOOK is the only Oklahoma site to appear in Table 4-10 for benzene; this site has the
eighth (2015) and tenth (2016) highest annual average concentrations of benzene.
• TMOK is the only Oklahoma site to appear in Table 4-10 for /;-dichlorobenzene; this
site has the seventh (2016) and eighth (2015) highest annual average concentrations
of />dichlorobenzene.
• Four Oklahoma sites appear in Table 4-10 for their annual average concentrations of
1,2-dichloroethane; TMOK, TOOK, TROK, and BROK's 2016 annual averages rank
seventh through tenth, respectively, for 1,2-dichloroethane.
• Each of the Tulsa sites appear in Table 4-10 for ethylbenzene, ranking fifth (TOOK),
seventh (TMOK), and ninth (TROK) for their 2016 annual averages; TOOK's annual
average concentration for 2015 ranks tenth.
• Each of the Tulsa sites appear in Table 4-10 for hexachloro-l,3-butadiene, ranking
second (TMOK), third (TROK), and seventh (TOOK) for their 2016 annual averages.
• BROK is the only Oklahoma City site to appear in Table 4-10 for VOCs.
• BROK is also the only Oklahoma site that appears in Table 4-11 for the carbonyl
compounds. BROK's 2015 annual average concentration of acetaldehyde ranks
highest among NMP sites sampling this pollutant, and BROK's 2015 annual average
concentration of formaldehyde ranks fourth. BROK's 2016 annual average
concentrations of these pollutants do not appear in Table 4-11.
• Each of the Tulsa sites' annual averages of arsenic for 2015 and 2016 rank higher
than the annual averages for OCOK and YUOK, as shown in Table 4-13. There is a
considerable difference in the annual averages shown between the sites from the two
metro areas. TROK has the highest annual average arsenic concentration among the
Oklahoma sites (2016).
18-33
-------
18.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 18-3 for the seven Oklahoma sites. Figures 18-11 through 18-23 overlay these sites'
minimum, annual average, and maximum concentrations onto the program-level minimum, first
quartile, median, average, third quartile, and maximum concentrations, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations is still provided in the figures that follow.
18-34
-------
Figure 18-11. Program vs. Site-Specific Average Acetaldehyde Concentrations
TMOK
TROK
J
¦I
El
9
¦ !
N
1
B
|
•
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
18-35
-------
Figure 18-11 presents the box plots for acetaldehyde for all seven Oklahoma sites and
shows the following:
• The maximum acetaldehyde concentration measured across the program was
measured at BROK in 2015 (17.2 |ig/m3). Three of the four highest concentrations of
acetaldehyde across the program were measured at BROK (17.2 |ig/m3, 17.0 |ig/m3,
16.7 |ig/m3), Acetaldehyde concentrations greater than 4 |ig/m3 were not measured at
the remaining Oklahoma sites.
• The range of acetaldehyde concentrations measured at each Tulsa site is fairly similar
between the two years of sampling. The ranges measured at the Oklahoma City sites
are more variable.
• Both annual averages for all three Tulsa sites are greater than the program-level
average concentration (1.67 |ig/m3) but less than the program-level third quartile
(2.11 |ig/m3). This is also true for OCOK's annual average for 2015, while this site's
annual average for 2016 is just less than the program-level average. This is also true
for both of YUOK's annual averages.
• BROK has the largest difference between its annual averages. The annual average
concentration of acetaldehyde for 2015 is more than twice the annual average for
2016. BROK's annual average for 2016 is similar in magnitude to the annual
averages for 2016 for the other Oklahoma City sites.
18-36
-------
Figure 18-12. Program vs. Site-Specific Average Arsenic (TSP) Concentrations
TMOK
TOOK
TROK
OCOK
YUOK
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
9
o
Figure 18-12 presents the box plots for arsenic for the five Oklahoma sites sampling TSP
metals and shows the following:
• Because the Oklahoma sites are the only sites sampling TSP metals, Figure 18-12
compares each Oklahoma site's arsenic data against the combined Oklahoma data.
• The range of arsenic concentrations measured was smallest for OCOK and largest for
TROK. For both years of sampling, the maximum concentration of arsenic was
measured at TROK. Non-detects of arsenic were not measured at these sites.
• TROK and TOOK have the highest annual average concentrations of arsenic (TSP),
with all four greater than the program-level/TSP average arsenic concentration
(0.70 ng/m3). TMOK's annual averages are just less than the program-level/TSP
18-37
-------
average while OCOK and YUOK's annual averages are just less than program-
level/TSP median concentration (0.61 ng/m3).
Figure 18-13. Program vs. Site-Specific Average Benzene Concentrations
TMOK
-a ¦
Q
¦i
1
m
1 !
m
rt
~
I i i i i i i i
012345678
Concentration (|ig/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
18-38
-------
Figure 18-13 presents the box plots for benzene for all seven Oklahoma sites and shows
the following:
• The maximum concentration of benzene measured at an Oklahoma site was measured
at OCOK (3.71 |ig/m3), although this concentration is roughly half the maximum
concentration measured across the program. If this maximum concentration was
excluded from OCOK's 2016 dataset, the range of benzene concentrations measured
at OCOK in 2016 would more closely resemble the range measured in 2015.
• The range of concentrations measured each year at some sites, such as OCOK and
YUOK, vary considerably, while the range of concentrations measured at others, such
as TOOK, do not. Yet, each site's annual averages vary by 0.1 |ig/m3 or less.
• TOOK is the only site for which annual average concentration(s) of benzene greater
than 1 |ig/m3 were calculated. The annual average concentrations of benzene for
TROK, TMOK, and BROK (2016 only) are less than 1 |ig/m3, but still greater than
the program-level average concentration (0.72 |ig/m3). The annual averages for the
remaining sites are similar to or less than the program-level median concentration
(0.58 |ig/m3).
18-39
-------
Figure 18-14. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
TMOK
Hf- ¦
i
Program Max Concentration
= 3.90}ig/m3 !
1
m
Program Max Concentration = 3.90 (.ig/m3
TOOK
it-
! Program Max Concentration = 3.90 (ig/m3
OCOK
b
| Program Max Concentration = 3.90 (.ig/m3 i
¦
i
NROK
Program Max Concentration = 3.90 (ig/m3 i
r
_ r ,
Program Max Concentration = 3.90 (.ig/m3 j
91
BROK
| Program Max Concentration = 3.90 (.ig/m3 i
0.00
0.25 0.50
0.75 1.00
Concentration (jig/m3)
1.25 1.50
Site:
Program: IstQuartile
2015 Avera
2nd Quartile
2016 Ave a
3rd Quartile
Concentration Range, 2015 &2016
4th Quartile
Avera
18-40
-------
Figure 18-14 presents the box plot for 1,3-butadiene for all seven Oklahoma sites and
shows the following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plots in Figure 18-14 because the scale of the box plots would be
too large to readily observe data points at the lower end of the concentration range.
Thus, the scale of the box plots has been reduced. Note that the program-level
average concentration is just less than the program-level third quartile, such that they
are difficult to discern in Figure 18-14.
• Concentrations of 1,3-butadiene measured at the Oklahoma sites are considerably less
than the maximum concentration measured at the program-level, as all concentrations
measured at these sites are less than 0.50 |ig/m3.
• The range of 1,3-butadiene concentrations measured is larger for the Tulsa sites than
the Oklahoma City sites, with the exception of NROK.
• TMOK is the only Oklahoma site for which an annual average concentration of
1,3-butadine greater than the program-level average (0.09 |ig/m3) was calculated
(2016 only). The annual averages for the remaining sites are similar to or less than the
program-level average. The annual averages for the Oklahoma City sites, where they
would be calculated, are all less than the program-level median concentration.
BROK's annual average for 2016 is less than the program-level first quartile.
18-41
-------
Figure 18-15. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
TMOK
TOOK
TROK
OCOK
NROK
YUOK
BROK
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
18-42
-------
Figure 18-15 presents the box plots for carbon tetrachloride for all seven Oklahoma sites
and shows the following:
• The program-level median and average concentrations are similar and plotted nearly
on top of each other.
• The maximum carbon tetrachloride concentrations measured at each site are fairly
similar to each other, while the minimum concentrations exhibited more variability.
Carbon tetrachloride concentrations greater than 1 |ig/m3 were not measured at the
Oklahoma sites.
• The annual average concentrations for these sites vary 0.05 |ig/m3 across these sites,
all of which are similar to or just less than the program level average concentration of
0.64 |ig/m3.
Figure 18-16. Program vs. Site-Specific Average /7-Dichlorobenzene Concentrations
TOOK
V
51
Program Max Concentration = 2.78 (.ig/m3 !
1
Ill
TMOK
H
ft
r
Program Max Concentration = 2.78 (.ig/m3
1 -
1
NROK
1
r
Program MaxConcentration = 2.78 (.ig/m3
I
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Concentration (]ug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 18-16 presents the box plots for p-dichlorobenzene for TOOK, TMOK, and
NROK, and shows the following:
• Similar to 1,3-butadiene, the program-level maximum p-dichlorobenzene
concentration (2.78 |ig/m3) is not shown directly on the box plots in Figure 18-16
because the scale of the box plots would be too large to readily observe data points at
the lower end of the concentration range. Thus, the scale of the box plots has also
been reduced. Note that the program-level first and second quartiles are both zero and
therefore not visible on the box plots.
18-43
-------
/;-Dichlorobenzene is a pollutant of interest for only three Oklahoma sites.
All />dichlorobenzene concentrations measured at these sites are less than
0.35 |ig/m3, an order of magnitude less than the maximum concentration measured
across the program.
The range of/>dichlorobenzene concentrations measured at each site in 2016 are
similar to each other. The range of concentrations for each site in 2015 are also
similar to each other, although NROK was not sampling in 2015.
The annual average />dichlorobenzene concentrations for TOOK and TMOK are
similar to or greater than the program-level average concentration (0.05 |ig/m3), with
TMOK's 2016 annual average also greater than the program-level third quartile
(0.06 |ig/m3).
The total number of non-detects measured at TMOK (37) is just slightly less than the
number measured at TOOK (41). For TMOK, these non-detects were spread across
the two years of sampling, while the majority were measured in 2016 (30 of the 41)
for TOOK.
18-44
-------
Figure 18-17. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
TMOK
r ,
! Program Max Concentration = 45.8 p.g/m3
TOOK
—'
i ¦]
! Program Max Concentration = 45.8 p.g/m3 i
Br
i 1
| Program MaxConcentration =45.8 (.ig/m3
mr
1 -J
] Program Max Concentration = 45.8 p.g/m3 i
______________ ,
! Program Max Concentration = 45.8 }ig/m3 >
\ Program MaxConcentration =45.8 (.ig/m3
| Program MaxConcentration =45.8 (.ig/m3
0.00
0.25
0.50 0.75 1.00
Concentration (|ug/m3)
1.25
1.50
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
18-45
-------
Figure 18-17 presents the box plots for 1,2-dichloroethane for all seven Oklahoma sites
and shows the following:
• The scale of the box plots in Figure 18-17 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• The program-level average concentration of 1,2-dichloroethane is being driven by the
higher concentrations measured at a few monitoring sites. The entire range of
1,2-dichloroethane concentrations measured at the Oklahoma sites is less than the
average concentration across the program.
• The annual average concentrations of 1,2-dichloroethane for the Oklahoma sites fall
on either side of the program-level median concentration (0.08 |ig/m3), with less than
0.03 |ig/m3 separating these sites' annual averages.
Figure 18-18. Program vs. Site-Specific Average Ethylbenzene Concentrations
t-r-r—
rr^
h1
0 0.5 1 1.5 2 2.5 3
Concentration {[jg/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
18-46
-------
Figure 18-18 presents the box plots for ethylbenzene for TOOK, TMOK, TROK, and
NROK, and shows the following:
• Ethylbenzene is a pollutant of interest for the three Tulsa sites and NROK.
• The range of ethylbenzene concentrations measured across both years is largest for
TMOK and smallest for NROK. For each Tulsa site, the range of ethylbenzene
concentrations measured in 2016 is larger than the range of concentrations measured
in 2015.
• Each of the annual average concentrations for the Tulsa sites are greater than the
program-level average concentration (0.26 |ig/m3) and most are also greater than the
program-level third quartile (0.32 |ig/m3),
• Non-detects of ethylbenzene were not measured at these four sites. The minimum
concentration measured at TOOK in 2015 is equivalent to the program-level first
quartile (0.10 |ig/m3); the minimum concentration measured in 2016 is similar in
magnitude.
18-47
-------
Figure 18-19. Program vs. Site-Specific Average Formaldehyde Concentrations
TMOK
TOOK
TROK
OCOK
NROK
YUOK
BROK
0 3 6 9 12 15 18 21 24 27
Concentration (]ug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
a
2016 Ave rag
o
e Concentration Range, 2015 & 2016
18-48
*
*
¦ "
N-
—o
-------
Figure 18-19 presents the box plots for formaldehyde for all seven Oklahoma sites and
shows the following:
• The maximum formaldehyde concentration measured at BROK is among the highest
formaldehyde concentrations measured across the program. Seven formaldehyde
concentrations measured at BROK are greater than the highest formaldehyde
concentration measured at another Oklahoma site. BROK's annual average
concentration for 2015 is just less than the maximum formaldehyde concentration
measured at this site in 2016.
• If BROK is excluded, the range of formaldehyde concentrations measured at OCOK
exhibit the largest difference between the two years of sampling.
• With the exception of BROK, the annual average concentrations of formaldehyde for
each Oklahoma site fall on either side of the program-level average concentration
(3.05 |ig/m3), BROK is the only site for which an annual average concentration does
not fall between the program-level second (median) and third quartiles.
Figure 18-20. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
(
] Program Max Concentration = 1.02 (ig/m3 ;
i -•
Program Max Concentration = 1.02 (ig/m3 ¦
1
Program Max Concentration = 1.02 (.ig/m3
1
1
Program MaxConcent
ration = 1.02 (.ig/m3
f
0.00 0.05 0.10 0.15 0.20 0.25
Concentration (jig/m3)
Program: IstQuartile 2ndQuartile
3rdQuartile 4thQuartile
Ave ra ge
~
~ ~
Site:
2015Average 2016Aveage Concentration Range, 2015 & 2016
a o
18-49
-------
Figure 18-20 presents the box plots for hexachloro-1,3-butadiene for four of the seven
Oklahoma sites and shows the following:
• The program-level maximum concentration of hexachloro-1,3-butadiene (1.02 |ig/m3)
is not shown directly on the box plot as the scale has been reduced to 0.25 |ig/m3 to
allow for the observations data points at the lower end of the concentration range.
• The program-level first, second, and third quartiles are all zero for this pollutant,
indicating that at least 75 percent of the measurements across the program are non-
detects and thus, are not visible on the box plot.
• Annual average concentrations for hexachl oro-1,3-butadiene were not calculated for
2015 due to the use of a contaminated internal standard at the laboratory for Method
TO-15, which resulted in the invalidation of the results from early March 2015
through mid-December 2015, as described above and in Section 2.4.
• The maximum concentration of hexachloro-l,3-butadiene for each Oklahoma site is
0.15 |ig/m3 or less. For each site, non-detects make up the majority of concentrations
measured at these sites.
• The 2016 annual average concentrations of hexachl oro-1,3-butadiene for TMOK and
TROK are slightly greater than the program-level average concentration, while the
annual averages for OCOK and YUOK are more similar to the program-level
average. Less than 0.015 |ig/m3 separates these annual averages.
Figure 18-21. Program vs. Site-Specific Average Manganese (TSP) Concentrations
E—f
0 10 20 30 40 50 60 70 80 90 100
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
*
o
Figure 18-21 presents the box plot for manganese for TOOK and shows the following:
• TOOK is the only Oklahoma site for which manganese is a pollutant of interest.
Because the Oklahoma sites are the only sites sampling TSP metals, Figure 18-21
compares the manganese concentrations measured at TOOK against the combined
Oklahoma data.
• The maximum manganese concentrations measured each year at TOOK are the
highest manganese concentrations measured among the Oklahoma sites.
18-50
-------
• TOOK's annual average manganese concentration for 2015 is similar to the annual
average for 2016, varying by less than 1 ng/m3.
• TOOK's annual averages are greater than the program-level/TSP manganese
concentration and third quartile (TSP only). Similar observations were made in the
2013 and 2014 NMP reports.
Figure 18-22. Program vs. Site-Specific Average Nickel (TSP) Concentrations
l-h—
1 -»
0 3 6 9 12 15 18 21 24
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
9
o
Figure 18-22 presents the box plots for nickel for TOOK and TROK and shows the
following:
• TOOK and TROK are the two Oklahoma sites for which nickel is a pollutant of
interest. Because the Oklahoma sites are the only sites sampling TSP metals,
Figure 18-22 compares the nickel concentrations measured at TOOK and TROK
against the combined Oklahoma data. Note that the majority of concentrations
measured at the Oklahoma sites fall into a more compressed range for nickel than for
manganese, as indicated by the closeness of the first, second, and third quartiles in the
box plots.
• The maximum nickel concentration measured among the Oklahoma sites was
measured at TOOK (22.1 ng/m3). The next highest concentration measured at TOOK
is half the magnitude (10.3 ng/m3). Nine nickel concentrations measured at TOOK are
greater than the highest nickel concentration measured at TROK.
• Both annual average nickel concentrations for TOOK are greater than the program-
level/TSP average concentration (1.25 ng/m3) and the third quartile (1.49 ng/m3), TSP
only. TROK's annual averages are lower.
18-51
-------
Figure 18-23. Program vs. Site-Specific Average Propionaldehyde Concentrations
H a
E
>
0 1 2 3 4 5 6
Concentration (ng/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 18-23 presents the box plot for propionaldehyde for BROK and shows the
following:
• BROK is the only Oklahoma site with propionaldehyde as a pollutant of interest;
BROK is one of only two NMP sites with propionaldehyde as a pollutant of interest
(BTUT is the other).
• The maximum propionaldehyde concentration (5.83 |ig/m3) measured across the
program was measured at BROK in 2015; further, the eight highest propionaldehyde
concentrations measured across the program were measured at BROK in 2015.
Fourteen propionaldehyde concentrations measured at BROK in 2015 are greater than
the maximum concentration measured at this site in 2016.
• BROK's annual average concentration for 2015 is three times greater than the annual
average for 2016, which is similar in magnitude to the program-level average
concentration (0.32 |ig/m3),
18.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
TOOK has sampled TSP metals, carbonyl compounds, and VOCs under the NMP since 2006 and
TMOK and OCOK have sampled these pollutants since 2009. Thus, Figures 18-24 through
18-52 present the 1-year statistical metrics for each of the pollutants of interest first for TOOK,
followed by TMOK and OCOK. The statistical metrics presented for assessing trends include the
substitution of zeros for non-detects. If sampling began mid-year, a minimum of 6 months of
sampling is required for inclusion in the trends analysis; in these cases, a 1-year average
concentration is not provided, although the range and percentiles are still presented.
18-52
-------
Figure 18-24. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
TOOK
. o
T T"
O.
2006 2007 2008 2009
2010 2011 2012
Year
2013 2014 2015 2016
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-24 for acetaldehyde concentrations measured at TOOK
include the following:
• Although TOOK began sampling carbonyl compounds under the NMP in January
2006, equipment complications at the onset of sampling resulted in a low
completeness for 2006; thus, a 1-year average concentration is not presented for 2006,
although the range of measurements is provided.
• The maximum concentration of acetaldehyde was measured in 2011 (8.95 |ig/m3),
although a similar concentration was also measured in 2012 (8.59 |ig/m3). The 10
highest acetaldehyde concentrations were measured in 2011 or 2012. Of the 35
acetaldehyde concentrations greater than 4 |ig/m3 measured at TOOK, 20 were
measured in either 2011 or 2012, five were measured in 2010, and three or fewer
were measured in each of the other years (including none in 2015 and 2016).
• The statistical metrics exhibit an increasing trend between 2008 and 2011, with little
change shown in the acetaldehyde measurements from 2011 to 2012. The 95th
percentiles for 2011 and 2012 are greater than the maximum concentrations measured
prior to 2011. These are the only two years that the median acetaldehyde
concentration is greater than 2 |ig/m3.
• A significant decrease in acetaldehyde concentrations is shown for 2013, with
relatively little change in the central tendency shown for 2014 through 2016.
18-53
-------
Figure 18-25. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at
TOOK
6.0
1E
1 30
C
m
c
O
u
r
1.0
o
Lr
¦¦
L2-
-L 1
T~
1-2-1 L2-
i
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
Observations from Figure 18-25 for arsenic (TSP) concentrations measured at TOOK
include the following:
• Although TOOK began sampling TSP metals in 2006, sampling did not begin until
October, which does not yield enough samples for the statistical metrics to be
calculated; thus, Figure 18-25 excludes data from 2006 per the criteria specified in
Section 3.4.2.2.
• The two highest concentrations of arsenic were measured at TOOK in September
2007 and are the only two concentrations greater than 4 ng/m3 measured at TOOK.
Eight of the nine concentrations of arsenic greater than 2 ng/m3 were measured in
either 2007 or 2008, with the ninth measured in 2016.
• The 1-year average and median concentrations exhibit a decreasing trend between
2007 and 2010, although the difference is relatively small between 2009 and 2010.
The 1-year average and median concentrations exhibit an increasing trend between
2010 and 2012.
• All of the statistical parameters exhibit decreases from 2012 to 2013. Little change is
shown in the central tendency parameters between 2013 and 2015 despite
increasingly higher concentrations measured each year.
• Although the median concentration changed little from 2015 to 2016, the 1-year
average concentration exhibits an increase. The range of arsenic concentrations
18-54
-------
measured at TOOK expands again in 2016, at both ends of the concentration range.
The number of arsenic concentrations greater than 1.5 ng/m3 increased to six for
2016, including the first arsenic concentration greater than 2 ng/m3 measured since
2008.
Figure 18-26. Yearly Statistical Metrics for Benzene Concentrations Measured at TOOK
o
±
2006 2007 2008
o.
X
s*} i i i
2011
Year
2012 2013 2014
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-26 for benzene concentrations measured at TOOK include
the following:
• Although TOOK began sampling VOCs under the NMP in January 2006, equipment
complications at the onset of sampling resulted in a low completeness for 2006; thus,
a 1-year average concentration is not presented for 2006, although the range of
measurements is provided.
• The maximum concentration of benzene was measured at TOOK in 2011
(23.8 |ig/m3). All four benzene concentrations greater than 10 |ig/m3 were measured
at TOOK in 2011. The 95th percentile for 2011 is greater than the maximum
concentration for each of the other years shown.
• The slight increases in the statistical parameters from 2007 to 2008 are followed by
significant decreases from 2008 to 2009. An increasing trend occurred between 2009
through 2011, when most of the statistical parameters are at a maximum. After 2011,
a significant decreasing trend in benzene concentrations is shown, with little change
shown for the most recent years of sampling. Most of the statistical parameters are at
a minimum for 2014, when the smallest range of benzene concentrations was
18-55
-------
measured, though little difference in shown among the concentration profiles for
2014, 2015, and 2016.
Figure 18-27. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
TOOK
o..
¦-r
o
2007 2008
o
o
£
2011
Year
T V
O
2014 2015 2016
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-27 for 1,3-butadiene concentrations measured at TOOK
include the following:
• The maximum concentration of 1,3-butadiene was measured in 2015 (0.366 |ig/m3),
although concentrations of similar magnitude were also measured in 2011
(0.339 |ig/m3) and 2007 (0.326 |ig/m3).
• The minimum concentration for most years is zero, indicating the presence of non-
detects. For 2006, 2010, 2011, and 2013, both the minimum concentration and 5th
percentile are zero, indicating that more than one non-detect was measured during
those years. The percentage of non-detects has ranged from zero (2007, 2012, 2015,
and 2016) to 14 percent (2006).
• After an initial decrease from 2007 to 2008 and little change for 2009, the 1-year
average concentration of 1,3-butadiene has an increasing trend through 2012. This is
also true for the median concentration. Even though the maximum and 95th percentile
decreased for 2012, both the 1-year average and median concentrations are at a
maximum.
18-56
-------
• With the exception of the maximum concentration, all of the statistical parameters
exhibit decreases for 2013. While some of this decrease is attributable to the non-
detects measured, they are not the sole reason. For example, the number of
1,3-butadiene concentrations greater than 0.1 |ig/m3 decreased from 19 measured at
TOOK in 2012 to 12 in 2013.
• Considerable fluctuations in 1,3-butadiene concentrations are shown for the last
several years of sampling, when both the smallest (2014) and largest (2015) range of
concentrations was measured. Between 2013 and 2016, less than 0.015 |ig/m3
separates the 1-year average concentrations; this is also true for the median
concentrations.
Figure 18-28. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at TOOK
T
iti
r
ft
*
¦ -0
pi
i rh
5
"**
1
T
-
-
20061 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-28 for carbon tetrachloride concentrations measured at
TOOK include the following:
• The maximum concentration of carbon tetrachloride was measured in 2011
(1.64 |ig/m3). Four additional concentrations greater than 1 |ig/m3 have been
measured at TOOK.
• With the exception of 2011, the range of carbon tetrachloride measurements spans
approximately 1 |ig/m3 or less. The range of measurements is at a minimum for 2015,
when the difference between the minimum and maximum concentrations is less than
0.4 |ig/m3.
18-57
-------
• The 1-year average concentration increased from 2007 to 2008, after which little
change is shown through 2011. A slight increase is shown for 2012, even though the
concentrations span the smallest range up to this point. After 2012, the 1-year average
concentration of carbon tetrachloride returned to previous levels. Excluding 2007 and
2012, the 1-year average concentrations vary from 0.61 |ig/m3 to 0.63 |ig/m3. Across
all years of sampling, the 1-year average (and median concentrations) have varied by
only 0.11 |ig/m3.
Figure 18-29. Yearly Statistical Metrics for />-Dichlorobenzene Concentrations Measured at
TOOK
o-
~
r—5—1
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
Observations from Figure 18-29 for /;-dichlorobenzene concentrations measured at
TOOK include the following:
• The maximum p-dichlorobenzene concentration was measured at TOOK on
October 9, 2008 (1.33 |ig/m3) and is twice the next highest concentration
(0.669 |ig/m3, measured in 2009). Four additional p-dichlorobenzene concentrations
greater than 0.5 |ig/m3 have been measured at TOOK.
• The increase in the 1-year average concentration from 2007 to 2008 is not solely a
result of the outlier concentration measured in 2008. The range within which the
majority of concentrations lie expanded, nearly doubling from 2007 to 2008, with
additional concentrations measured at the both ends of the concentration range.
18-58
-------
• Between 2008 and 2011, most of the concentrations measured at TOOK fell into a
similar range and the 1-year average concentration did not vary significantly
(although there is a little more variability in the median concentrations).
• Concentrations of p-dichlorobenzene decreased significantly from 2011 to 2012 and
again in 2013. Concentrations greater than 0.2 |ig/m3 were not measured at TOOK in
2012 or 2013; the maximum concentration measured in 2013 is less than the 1-year
average and median concentrations for several of the previous years.
• The concentration profile for 2014 resembles the concentration profile for 2012.
• Despite slight increases in the maximum concentration over the last few years of
sampling, both the 1-year average and median concentration exhibit slight decreases
for 2015 and 2016.
• There were no non-detects of p-dichlorobenzene measured in 2006 or 2007. The
minimum concentration and 5th percentile are zero for most years after 2007,
indicating the presence of non-detects. Between 2008 and 2012, the number of non-
detects measured each year ranges from two (2009) to six (2010, 2011, and 2012).
The number of non-detects increased four-fold for 2013 (24), decreased to 11 for both
2014 and 2015, then increased to 30 for 2016.
Figure 18-30. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at TOOK
2006 2007
o
2011
Year
O 5th Percentile
Maximum o 95th Percentile Avera
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
18-59
-------
Observations from Figure 18-30 for 1,2-dichloroethane concentrations measured at
TOOK include the following:
• The median concentration for each year through 2011 is zero, indicating that at least
half of the measurements were non-detects. In 2006, there was one measured
detection of 1,2-dichloroethane. In 2007 and 2008, there were none. Between 2009
and 2011, the number of measured detections varied from five to six. The number of
measured detections increased significantly for 2012, up from six in 2011 to 38 in
2012. Slightly fewer measured detections were measured in 2013 (31), after which at
least 50 measured detections were measured each year between 2014 and 2016. The
1-year average concentration increases (and decreases) correspondingly, with the
1-year average reaching 0.10 |ig/m3 for the first time in 2016.
• The 1-year average concentration is less than the corresponding median concentration
for the last five years of sampling, which is a little unusual. The 1-year average
concentration is more susceptible to outliers (on either end of the concentration range)
than the median concentration, which represents the midpoint of a group of
measurements. Here, concentrations on the lower end of the concentration range (the
many zeroes representing non-detects) are pulling the average down (just like a
maximum or outlier concentration can drive the average upward). The difference
between the two statistical parameters decreases each year through 2015.
Figure 18-31. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
TOOK
X
20061 2007 2008 2009 2010
2011
Year
2013 2014
Q.
2016
O 5th Percentile
— Minimum — Media
— Maximum
O 95th Percentile
1 A 1-year average is not presented because issues at the onset of sampling resulted in low completeness.
18-60
-------
Observations from Figure 18-31 for ethylbenzene concentrations measured at TOOK
include the following:
• The two highest concentrations of ethylbenzene were both measured during the
summer of 2008 (5.09 |ig/m3 and 4.57 |ig/m3). No other ethylbenzene concentrations
greater than 3 |ig/m3 have been measured at TOOK since the onset of sampling. The
next five highest concentrations, those between 2.50 |ig/m3 and 3 |ig/m3, were all
measured at TOOK in 2012.
• The maximum, 95th percentile, and 1-year average concentrations exhibit increases
from 2007 to 2008; the median also increased, although slightly. Even if the two
highest concentrations measured in 2008 were excluded from the dataset, the 1-year
average concentration would still exhibit a slight increase. A review of the data shows
that the number of ethylbenzene concentrations greater than 1 |ig/m3 nearly doubled
from 2007 (7) to 2008 (13).
• There were no ethylbenzene concentrations greater than 1 |ig/m3 measured at TOOK
in 2009. Both the 1-year average and median concentrations are at a minimum for
2009, both decreasing by more than half from 2008 to 2009.
• After 2009, concentrations of ethylbenzene measured at TOOK exhibit a significant
increasing trend through 2012. The 95th percentile, 1-year average concentration, and
the median concentration are all at a maximum for 2012. The 95th percentile for 2012
is greater than the maximum concentration for all other years except 2008. The 1-year
average concentration for 2012 is just less than 1 |ig/m3.
• Ethylbenzene concentrations measured in 2013 decreased significantly from 2012,
with all of the statistical parameters exhibiting decreases, including the 1-year
average concentration, which decreased by more than half. Relatively little change is
shown for most of the statistical parameters between 2013 and 2015. Slight increases
are shown for most of the statistical parameters for 2016, with the median
concentration as the exception.
18-61
-------
Figure 18-32. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
TOOK
T
T
T
"
-------
• Each of the statistical parameters exhibits an increase for 2016, with the median
concentration exhibiting the largest change.
Figure 18-33. Yearly Statistical Metrics for Manganese (TSP) Concentrations Measured at
TOOK
5th Percentile
O 95th Percentile
Observations from Figure 18-33 for manganese (TSP) concentrations measured at TOOK
include the following:
• The maximum concentration of manganese was measured in 2012 (273 ng/m3), on
the day of a dust storm (October 18, 2012). Measurements greater than 100 ng/m3
were also measured in 2007 (128 ng/m3) and 2011 (104 ng/m3).
• A decreasing trend in manganese concentrations is shown through 2009, which was
followed by an increasing trend through 2012. Even if the maximum concentration
measured in 2012 was excluded from the calculations, the 1-year average and median
concentrations would still exhibit an increasing trend for 2012. This is because there
were more concentrations at the upper end of the concentration range for 2012 (the
number of manganese concentrations greater than 50 ng/m3 increased from four in
2011 to 12 in 2012) as well as fewer concentrations at the lower end of the
concentration range (the number of manganese concentrations less than 20 ng/m3
decreased from 17 in 2011 to 11 in 2012).
• With the exception of the 95th percentile, all of the statistical parameters exhibit
decreases from 2012 to 2013. Both the 1-year average and median concentrations of
manganese decreased by more than 10 ng/m3 from 2012 to 2013.
18-63
-------
• Despite differences in the range of concentrations measured, the central tendency
parameters exhibited relatively little change between 2013 and 2016. These
parameters vary by less than 2.5 ng/m3 across the last four years of sampling shown.
Figure 18-34. Yearly Statistical Metrics for Nickel (TSP) Concentrations Measured at
TOOK
O 5th Percentile
O 95th Percentile
Observations from Figure 18-34 for nickel (TSP) concentrations measured at TOOK
include the following:
• The maximum concentration of nickel (21.6 ng/m3) was measured at TOOK on
September 27, 2016. Three additional nickel concentrations greater than 10 ng/m3
have been measured at TOOK (including one on October 18, 2012, the day of a dust
storm). In total, 18 nickel concentrations greater than 5 ng/m3 have been measured at
TOOK, with all but two measured in 2012 or later (with the two exceptions measured
in 2007).
• A significant decreasing trend in the nickel concentrations measured at TOOK is
shown through 2009. A slight increase is shown for 2010, which was followed by
significant increases for 2011 and 2012.
• With the exception of the maximum concentration, the concentration profiles shown
for the last five years of sampling more closely resemble the concentration profiles
shown for 2007 than the years in-between.
• Despite some of the highest nickel concentrations measured, the median
concentration exhibits a decrease from 2014 to 2015 and again for 2016. An
18-64
-------
increasing number of nickel concentrations less than 1 ng/m3 were measured at
TOOK during this time, from a minimum of one in 2014 to eight in 2015 and 18 in
2016, the most since 2010.
Figure 18-35. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
TMOK
8.0
5.0
"e
§ 1 0
C
m
c
O
u
3.0
2.0
1.0
...o-
I
_L i, _L
1—2-J
LjJ
i
t
2009 1 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-35 for acetaldehyde concentrations measured at TMOK
include the following:
• Sampling for carbonyl compounds began at TMOK under the NMP in April 2009. A
1-year average concentration is not presented for 2009 because a full year's worth of
data is not available, although the range of measurements is provided.
• The maximum acetaldehyde concentration (7.00 |ig/m3) was measured at TMOK on
August 19, 2011. All seven acetaldehyde concentrations greater than 5 |ig/m3 were
measured in either 2011 or 2012.
• The range of acetaldehyde concentrations measured increased considerably from
2010 to 2011, when the number of acetaldehyde concentrations greater than 3 |ig/m3
measured at TMOK increased three-fold, from five in 2010 to 15 in 2011. After 2011
the range of measurements decreased each year through 2014.
• A decreasing trend is shown in the 1-year average concentrations between 2011 and
2014, with little change shown for 2015 and 2016. The median concentration exhibits
a similar trend, although an increase is shown for 2016.
18-65
-------
Figure 18-36. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at
TMOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-36 for arsenic (TSP) concentrations measured at TMOK
include the following:
• Sampling for TSP metals began at TMOK under the NMP in April 2009. A 1-year
average concentration is not presented for 2009 because a full year's worth of data is
not available, although the range of measurements is provided.
• Five of the six arsenic concentrations greater than 2 ng/m3 were measured at TMOK
in 2009, including the two highest measurements (4.65 ng/m3 and 3.11 ng/m3). The
entire range of concentrations measured in other years is less than the 95th percentile
for 2009 and the median concentration is at a maximum for 2009.
• With the exception of 2012, the 1-year average concentrations vary between
0.6 ng/m3 and 0.7 ng/m3. Most of the statistical parameters exhibit increases for 2012,
as the number of arsenic concentrations greater than 1 ng/m3 measured in 2012 (15) is
greater than the number measured during most other years, with the exception of
2009 (16).
• Excluding 2009, the statistical metrics for arsenic concentrations measured at TMOK
resemble those shown in Figure 18-25 for TOOK.
18-66
-------
Figure 18-37. Yearly Statistical Metrics for Benzene Concentrations Measured at TMOK
rh
o
I
LJ
"O-.
1—2—
Ll-1
—2—
i I
2012 2013
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-37 for benzene concentrations measured at TMOK include
the following:
• Sampling for VOCs began at TMOK under the NMP in April 2009. A 1-year average
concentration is not presented for 2009 because a full year's worth of data is not
available, although the range of measurements is provided.
• The maximum benzene concentration (3.91 |ig/m3) was measured at TMOK on
May 7, 2009, although additional benzene concentrations greater than 3 |ig/m3 were
measured each year between 2009 and 2012.
• The 1-year average benzene concentration has a significant decreasing trend between
2010 and 2014, with the largest year-to-year decrease shown from 2012 to 2013. The
1-year average decreased by half during this time. The median concentration also
decreases between 2010 and 2014 but continues to decrease through 2016 (while the
1-year average exhibits a slight increase during the last two years). The median
concentration has also decreased by half since the onset of sampling.
• The range of benzene concentrations measured is at a minimum for 2014 then
increases slightly for 2015 and again for 2016. Between 2014 and 2016, the 1-year
average concentration of benzene increased only slightly (by 0.1 |ig/m3).
18-67
-------
Figure 18-38. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
TMOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-38 for 1,3-butadiene concentrations measured at TMOK
include the following:
• The range of 1,3-butadiene concentrations measured at TMOK is at a minimum for
2009, with all concentrations measured spanning less than 0.2 |ig/m3. The range of
concentrations measured then increases each year through 2012. After 2012, the
range of measurements decreases for two years then increases for two years, such that
the concentration profiles for 2013 and 2016 resemble each other.
• Despite the differences in the concentrations measured, less than 0.04 |ig/m3 separates
the 1-year average concentrations across the years shown, which range from
0.09 |ig/m3 (2015) to 0.13 |ig/m3 (2012).
• The number of non-detects has varied across the years of sampling, from a few as
none (2009) to as many as nine (2011).
18-68
-------
Figure 18-39. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at TMOK
2012 2013
Year
O 5th Percentile
O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-39 for carbon tetrachloride concentrations measured at
TMOK include the following:
• The maximum carbon tetrachloride concentration was measured on August 17, 2009
and is the only concentration greater than 1 |ig/m3 measured at TMOK.
• All of the statistical parameters exhibit decreases from 2009 to 2010, with little
change in the carbon tetrachloride measurements at TMOK shown from 2010 to
2011.
• All of the statistical parameters exhibit increases for 2012, despite the compressed
range of concentrations measured. The number of carbon tetrachloride concentrations
greater than 0.6 |ig/m3 was at a maximum for 2012, accounting for 51 of the 61
measurements (compared to between 30 and 40 for most of the other years shown).
• All of the statistical parameters exhibit decreases from 2012 to 2013, with several
parameters exhibiting additional decreases for 2014. The minimum concentration
measured in 2014 (0.063 |ig/m3) is considerably less than other carbon tetrachloride
concentrations measured at TMOK.
• With the exception of 2012, the 1-year average concentrations of carbon tetrachloride
fall between 0.60 |ig/m3 and 0.65 |ig/m3. This is also true for the median
18-69
-------
concentrations, excluding 2009. The 1-year average and median concentrations for
2012 are only slightly outside this range, at 0.68 |ig/m3 each.
Figure 18-40. Yearly Statistical Metrics for o-Dichlorobenzene Concentrations Measured at
TMOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-40 for /;-dichlorobenzene concentrations measured at
TMOK include the following:
• The maximum /;-dichlorobenzene concentration was measured on June 30, 2009
(0.747 |ig/m3). Only one additional concentration greater than 0.5 |ig/m3 has been
measured at TMOK (0.663 |ig/m3, measured in 2013).
• A decreasing trend in concentrations of p-dichlorobenzene is shown through 2012.
The median decreases by nearly half between 2009 and 2012 and 1-year average
concentration decreased significantly from 2010 to 2011 with little change shown
from 2011 to 2012.
• The increase in the 1-year average concentration shown for 2013 is not solely
attributable to the maximum concentration measured that year, as the median
concentration, which is less influenced by outliers, exhibits a similar increase. The
number of concentrations greater than 0.1 |ig/m3 nearly doubled from 2012 (16) to
2013 (30).
18-70
-------
• The decreasing trend in p-dichlorobenzene concentrations measured at TMOK shown
prior to 2013 resumes in 2014 and 2015, when both central tendency parameters are
at a minimum.
• Most of the statistical parameters exhibit slight increases for 2016, though confidence
intervals indicate that the difference is not statistically significant.
Figure 18-41. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at TMOK
0.25
r
E
1
c
O
c
m
c
\
pU
-
0.05
0.00
o~
...o"'
O-
o
..-O-
20091 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-41 for 1,2-dichloroethane concentrations measured at
TMOK include the following:
• The minimum, 5th percentile, and median concentrations for 2009, 2010, and 2011
are zero, indicating that at least half of the measurements were non-detects. In 2009,
there were three measured detections of 1,2-dichloroethane. In 2010 and 2011, there
were 10 each year. For 2012, the number of measured detections increased by a factor
of four and the median concentration is greater than zero for the first time. Measured
detections also accounted for more than half of measurements in 2013. By 2014,
measured detections account for 85 percent of measurements, with the percentage
increasing each year through 2016, when the 5th percentile is greater than zero for the
first time. The 1-year average concentrations increase (and decrease)
correspondingly, with the 1-year average reaching 0.10 |ig/m3 for the first time in
2016.
18-71
-------
• The 1-year average concentration is more susceptible to outliers (on either end of the
concentration range) than the median concentration. The 1-year average concentration
for each year between 2012 and 2015 is less than the median, indicating that
concentrations on the lower end of the concentration range (the zeroes representing
non-detects) are pulling the 1-year average downward (just like a maximum or outlier
concentration can drive the average upward). The 1-year average concentration for
2016 is greater than the median concentration, when the fewest non-detects were
measured (three).
• Figure 18-41 for 1,2-dichloroethane concentrations measured at TMOK resembles
Figure 18-30 for 1,2-dichloroethane concentrations measured at TOOK.
Figure 18-42. Yearly Statistical Metrics for Ethylbenzene Concentrations Measured at
TMOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-42 for ethylbenzene concentrations measured at TMOK
include the following:
• The maximum ethylbenzene concentration was measured in 2010 (3.63 |ig/m3) and is
the only measurement greater than 2 |ig/m3 measured at TMOK.
• Despite the decrease in the maximum concentrations shown between 2010 and 2012,
little change is shown for most of the statistical parameters. Less than 0.04 |ig/m3
separates the median concentrations for these years and approximately 0.01 |ig/m3
separates the 1-year average concentrations during this period.
18-72
-------
• A significant decreasing trend in ethylbenzene concentrations is shown after 2012.
Most of the statistical parameters are at a minimum for 2015, the first year
ethylbenzene concentrations greater than 1 |ig/m3 were not measured.
• Several of the statistical parameters exhibit increases for 2016. Eight concentrations
measured in 2016 are higher than the maximum concentration measured in 2015. Yet,
both the minimum concentration and 5th percentile are at a minimum for 2016.
Figure 18-43. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
TMOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
Observations from Figure 18-43 for formaldehyde concentrations measured at TMOK
include the following:
• The maximum formaldehyde concentration was measured on August 19, 2011
(10.8 |ig/m3), the same date that the maximum acetaldehyde concentration was
measured at TMOK. Two additional formaldehyde concentrations greater than
10 |ig/m3 were measured at TMOK in 2012.
• Most of the statistical parameters exhibit increases from 2010 to 2011, when the
number of formaldehyde concentrations greater than 5 |ig/m3 nearly doubled (from 10
measured in 2010 to 19 measured in 2011).
• After 2011, the maximum formaldehyde concentration measured at TMOK decreases
each year through 2015. The 95th percentile decreases across most of these years, as
does the 1-year average and median concentrations (though 2014 is the exception).
18-73
-------
Most of the statistical parameters are at a minimum for 2015, the only year in which
the 1-year average concentration is less than 3 |ig/m3.
• Each of the statistical parameters exhibits an increase for 2016.
Figure 18-44. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at TMOK
0.16
0.14
0.12
0.10
1E
3
c
o
4= 0.08
c
CD
c
o
U
0.06
0.04
0.02
0.00
20091 2010 2011 2012 2013 2014 20152 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until April 2009.
2 A concentration profile is not presented due to a laboratory contamination issue affecting numerous
samples.
Observations from Figure 18-44 for hexachloro-l,3-butadiene concentrations measured at
TMOK include the following:
• The use of a contaminated internal standard at the laboratory for Method TO-15
resulted in the invalidation of the hexachloro-l,3-butadiene measurements from early
March 2015 through mid-December 2015, as described in Section 2.4. As a result, a
concentration profile for 2015 is not presented.
• There were few measured detections of hexachloro-1,3-butadiene in the first few
years of sampling at TMOK. The median concentration is zero for all years of
sampling, indicating that at least half of the measurements were non-detects for each
year. The number of measured detections ranges from zero in 2009 to 18 in 2016,
increasing by a few each year.
• All concentrations of hexachloro-l,3-butadiene measured at TMOK are less than the
MDL.
o
...o-
18-74
-------
Figure 18-45. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
OCOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-45 for acetaldehyde concentrations measured at OCOK
include the following:
• Sampling for carbonyl compounds began at OCOK under the NMP in May 2009. A
1-year average concentration is not presented for 2009 because a full year's worth of
data is not available, although the range of measurements is provided.
• The maximum acetaldehyde concentration was measured on May 9, 2011
(6.68 |ig/m3). One additional acetaldehyde concentration greater than 6 |ig/m3 has
been measured at OCOK (6.16 |ig/m3 in 2012).
• The smallest range of acetaldehyde concentrations was measured in 2009, after which
the range of measurements increased considerably. The 1-year average concentration
increased significantly from 2010 to 2011, with the median concentration exhibiting a
similar increase. Fifteen concentrations measured in 2011 (or one-quarter of the
measurements) are greater than the maximum concentration measured in 2010. Little
change in these parameters is shown from 2011 to 2012.
• A decreasing trend in acetaldehyde concentrations is shown after 2012. Slight
increases are shown for 2015, before additional decreases are exhibited for 2016. The
1-year average and median concentrations for 2016 are at their lowest since the first
full year of sampling at OCOK.
18-75
-------
Figure 18-46. Yearly Statistical Metrics for Arsenic (TSP) Concentrations Measured at
OCOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-46 for arsenic (TSP) concentrations measured at OCOK
include the following:
• Sampling for TSP metals began at OCOK under the NMP in May 2009. A 1-year
average concentration is not presented for 2009 because a full year's worth of data is
not available, although the range of measurements is provided.
• The maximum concentration of arsenic was measured at OCOK in 2009 (3.11 ng/m3).
Despite decreasing maximum concentrations after 2009, both the 1-year average and
median concentrations exhibit increases through 2012. This is due to a higher number
of concentrations at the upper end of the concentration range as well as fewer
concentrations at the lower end of the concentration range.
• Each of the statistical parameters exhibit a decrease for 2013, when the entire range
of arsenic concentrations measured at OCOK spans less than 1 ng/m3.
• Although variations in the 1-year average and median concentrations are shown
between 2013 and 2016, the majority of measurements collected during this period
fall into relatively similar ranges, as indicated by the 5th and 95th percentiles.
18-76
-------
Figure 18-47. Yearly Statistical Metrics for Benzene Concentrations Measured at OCOK
1
~ J a
> rh ,-5-. rn
r S
2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum o 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-47 for benzene concentrations measured at OCOK include
the following:
• Sampling for VOCs began at OCOK under the NMP in May 2009. A 1-year average
concentration is not presented for 2009 because a full year's worth of data is not
available, although the range of measurements is provided.
• The maximum benzene concentration was measured at OCOK on November 6, 2013
(9.38 |ig/m3). The next highest concentration was measured on September 18, 2011
(6.80 |ig/m3). Additional benzene concentrations greater than 4 |ig/m3 have not been
measured at OCOK.
• With the exception of 2013, the 1-year average concentration has a decreasing trend
between 2010 and 2015. If the maximum concentration measured in 2013 was
excluded from the calculation, the 1-year average concentration would have a
continuous decreasing trend through 2013, virtually no change for 2014, and further
decreases for 2015.
• Benzene concentrations measured at OCOK in 2015 exhibit the least amount of
variability among the seven full years of sampling, as this year has the smallest range
of measurements, the majority of concentrations fall into the smallest range, and the
difference between the 1-year average and median concentrations is at a minimum.
18-77
-------
• The slight increases shown for 2016 are primarily attributable to the maximum
concentration measured (3.71 |ig/m3). If this measurement was excluded, the
decreasing trend, however slight, would continue.
Figure 18-48. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
OCOK
Maximum Concentration
for 2011 is 10.0 |jg/m3
L,/'"
o..
6 S a
2012 2013
Year
O 5th Percentile
— Minimum
— Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-48 for 1,3-butadiene concentrations measured at OCOK
include the following:
• The maximum 1,3-butadiene concentration was measured at OCOK on
September 18, 2011 (10.0 |ig/m3), which is the same day the second highest benzene
concentration was measured. The next highest concentration was measured in 2012
(1.09 |ig/m3). No other 1,3-butadiene concentrations greater than 0.35 |ig/m3 have
been measured at OCOK.
• The 1-year average concentration for 2011 is being driven by the outlier, as the 1-year
average is greater than the 95th percentile for 2011. If this measurement was excluded
from the calculation, the 1-year average concentration would decrease from
0.21 |ig/m3 to 0.05 |ig/m3, resulting in a negligible change from 2010 levels.
• Excluding 2011, the 1-year average concentration for 2012 is higher than most other
years of sampling (although difficult to discern in Figure 18-48). The 1-year average
for 2012 is also influenced by a single higher concentration, though to a lesser extent
than 2011. But this is not the only reason for the increase. The number of non-detects
18-78
-------
measured decreased significantly from 2011 (20) to 2012 (5) before increasing again
for 2013 (21).
• The median concentrations shown between 2010 and 2016 have varied by less than
0.01 |ig/m3 over the period, ranging from 0.035 |ig/m3 (2013, 2014) to 0.044 |ig/m3
(2012, 2016). Excluding 2011 and 2012, the 1-year average concentrations vary by
less than 0.01 |ig/m3, ranging from 0.038 |ig/m3 (2014, 2015) to 0.046 |ig/m3 (2016).
Figure 18-49. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at OCOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-49 for carbon tetrachloride concentrations measured at
OCOK include the following:
• The two highest concentrations of carbon tetrachloride were measured at OCOK in
2009, including one greater than 1 |ig/m3 (1.10 |ig/m3). The maximum concentration
measured after 2009 are less than 0.90 |ig/m3, and less than 0.8 |ig/m3 for 2014 and
later.
• The range of carbon tetrachloride concentrations measured at OCOK decreased each
year through 2013, when all carbon tetrachloride concentrations measured span less
than 0.50 |ig/m3.
• The 1-year average concentrations of carbon tetrachloride have varied by less than
0.1 |ig/m3, ranging from 0.58 |ig/m3 (2011) to 0.66 |ig/m3 (2012). The median
18-79
-------
concentrations have a similar pattern, ranging from 0.59 |ig/m3 (2011) to 0.67 |ig/m3
(2012).
• With the exception of 2013, the median concentration is greater than the 1-year
average concentration, which can be attributed to the few concentrations on the lower
end of the concentration range, which can pull an average down in a similar manner
to an outlying concentration driving the average up. In total, six carbon tetrachloride
concentrations less than 0.2 |ig/m3 have been measured at OCOK, one each in 2009,
2010, 2014, and 2016, and two in 2011. This explains why the box and whisker plots
for carbon tetrachloride appear "inverted" for several years, with the minimum
concentration extending farther away from the majority of the measurements than the
maximum concentration, which is more common (see acetaldehyde as an example).
Figure 18-50. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at OCOK
0.40
0.35
0.30
0.25
1E
3
c
o
4= 0.20
c
dj
c
o
u
0.15
0.10
0.05
0.00
2009 1 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-50 for 1,2-dichloroethane concentrations measured at
OCOK include the following:
• The median concentration for 2009, 2010, and 2011 is zero, indicating that at least
half of the measurements were non-detects. In 2009, there were four measured
detections of 1,2-dichloroethane, which increased gradually through 2011. For 2012,
the number of measured detections increased by a factor of four (up to 52). The
number of measured detections is greater than 50 each year between 2013 and 2016.
Max i m u m Co n ce nt rati on
for 2011 is 0.706 |^/m3
¦ O
.
o
18-80
-------
• The increase in measured detections results in an increase in the 1-year average
concentrations shown through 2012. Less than 0.01 |ig/m3 separates the 1-year
average concentrations calculated for each year between 2012 and 2016; 0.01 |ig/m3
separates the median concentrations calculated for these years.
• The range within which most of the concentrations fall, as indicated by the 5th and
95th percentiles, changed little between 2010 and 2016, even with the percentage of
measured detections increasing to more than 90 percent during the last two years of
sampling.
Figure 18-51. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
OCOK
O 5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until May 2009.
Observations from Figure 18-51 for formaldehyde concentrations measured at OCOK
include the following:
• The maximum formaldehyde concentration was measured at OCOK on May 9, 2011
(19.6 |ig/m3), the same day as the maximum acetaldehyde concentration was
measured; the only other concentration greater than 10 |ig/m3 was also measured at
OCOK in 2011 (10.6 |ig/m3). In total, 20 formaldehyde concentrations greater than
7 |ig/m3 were measured at OCOK, with more than half (11) measured in 2011 (and
six in 2012 and three in 2015).
• With the exception of the 5th percentile, all of the statistical parameters exhibit an
increase from 2010 to 2011. This is not just a result of the two highest concentrations
measured in 2011, as concentrations were higher overall. Twelve concentrations
18-81
-------
measured in 2011 were greater than the maximum concentration measured in 2010.
The median concentration increased by more than 1 |ig/m3 and the 1-year average
concentration increased by more than 60 percent for 2011.
• Formaldehyde concentrations measured after 2011 have a decreasing trend through
2014, though there is little difference shown in the concentration profiles for 2013
and 2014.
• Each of the statistical parameters exhibits an increase for 2015. The maximum, 95th
percentile, and 1-year average concentration return to 2014 levels for 2016 while the
5th percentile and minimum concentration change little and the median exhibits
further (albeit slight) increases.
Figure 18-52. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at OCOK
0.7 |
-E 04
3
c
O
c
m
o
u
1 pL J
.
-------
• There were few measured detections of hexachloro-1,3-butadiene in the first few
years of sampling at OCOK. The median concentration is zero for all years of
sampling, indicating that at least half of the measurements were non-detects for each
year. The number of measured detections has varied from none (in 2009) to as many
as 14 (in 2013 and again in 2016).
• One concentration of hexachloro-l,3-butadiene measured at OCOK is greater than the
MDL for this pollutant, the maximum concentration measured in 2014 (0.609 |ig/m3).
This is the only hexachloro-1,3-butadiene measurement greater than 0.15 |ig/m3
measured at OCOK.
18.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at each Oklahoma monitoring site. Refer to Sections 3.2,
3.4.2.3, and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time
frames, and calculations associated with these risk-based screenings.
18.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Oklahoma monitoring sites and where annual
average concentrations could be calculated, risk was examined by calculating cancer risk and
noncancer hazard approximations. These approximations can be used as risk estimates for cancer
and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 18-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
18-83
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Public Works, Tulsa, Oklahoma - TOOK
Acetaldehyde
0.0000022
0.009
60/60
2.02
±0.21
4.44
0.22
61/61
1.97
±0.18
4.34
0.22
Benzene
0.0000078
0.03
59/59
1.09
±0.12
8.50
0.04
61/61
1.08
±0.13
8.42
0.04
1.3 -Butadiene
0.00003
0.002
56/59
0.07
±0.01
2.07
0.03
61/61
0.08
±0.01
2.50
0.04
Carbon Tetrachloride
0.000006
0.1
59/59
0.61
±0.02
3.67
0.01
61/61
0.61
±0.02
3.69
0.01
p-Dichlorobcnzcnc
0.000011
0.8
48/59
0.06
±0.01
0.61
<0.01
31/61
0.05
±0.02
0.51
<0.01
1,2 -Dichloroethane
0.000026
2.4
51/59
0.09
±0.01
2.40
<0.01
55/61
0.10
±0.01
2.63
<0.01
Ethylbenzene
0.0000025
1
59/59
0.40
±0.06
1.01
<0.01
61/61
0.47
±0.08
1.18
<0.01
Formaldehyde
0.000013
0.0098
60/60
2.84
±0.39
36.86
0.29
61/61
3.33
±0.39
43.31
0.34
Arsenic (TSP)a
0.0043
0.000015
60/60
0.78
±0.08
3.37
0.05
61/61
0.89
±0.11
3.83
0.06
Manganese (TSP)a
0.0003
60/60
26.01
±3.92
0.09
61/61
26.94
±4.15
0.09
Nickel (TSP)a
0.00048
0.00009
60/60
1.95
±0.36
0.94
0.02
61/61
2.17
±0.75
1.04
0.02
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Fire Station, Tulsa, Oklahoma - TMOK
Acetaldehyde
0.0000022
0.009
60/60
1.84
±0.19
4.04
0.20
60/60
1.77
±0.16
3.89
0.20
Benzene
0.0000078
0.03
60/60
0.84
±0.10
6.59
0.03
60/60
0.90
±0.12
7.03
0.03
1.3 -Butadiene
0.00003
0.002
56/60
0.09
±0.02
2.59
0.04
59/60
0.11
±0.02
3.30
0.05
Carbon Tetrachloride
0.000006
0.1
60/60
0.62
±0.02
3.70
0.01
60/60
0.63
±0.02
3.79
0.01
p-Dichlorobcnzcnc
0.000011
0.8
42/60
0.06
±0.01
0.64
<0.01
41/60
0.08
±0.02
0.85
<0.01
1,2 -Dichloroethane
0.000026
2.4
54/60
0.08
±0.01
2.20
<0.01
57/60
0.10
±0.01
2.64
<0.01
Ethylbenzene
0.0000025
1
60/60
0.32
±0.04
0.80
<0.01
60/60
0.44
±0.08
1.10
<0.01
Formaldehyde
0.000013
0.0098
60/60
2.71
±0.34
35.26
0.28
60/60
3.16
±0.41
41.11
0.32
Hexachloro-1,3 -butadiene
0.000022
0.09
1/60
NR
NR
NR
18/60
0.03
±0.01
0.63
<0.01
Arsenic (TSP)a
0.0043
0.000015
59/59
0.64
±0.07
2.76
0.04
61/61
0.67
±0.08
2.87
0.04
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Riverside, Tulsa, Oklahoma - TROK
Acetaldehyde
0.0000022
0.009
59/59
1.92
±0.19
4.22
0.21
61/61
2.02
±0.19
4.43
0.22
Benzene
0.0000078
0.03
60/60
0.79
±0.06
6.17
0.03
61/61
0.89
±0.09
6.96
0.03
1.3 -Butadiene
0.00003
0.002
58/60
0.07
±0.01
2.01
0.03
61/61
0.08
±0.01
2.44
0.04
Carbon Tetrachloride
0.000006
0.1
60/60
0.61
±0.02
3.65
0.01
61/61
0.60
±0.03
3.60
0.01
1,2 -Dichloroethane
0.000026
2.4
53/60
0.08
±0.01
2.09
<0.01
57/61
0.10
±0.01
2.52
<0.01
Ethylbenzene
0.0000025
1
60/60
0.31
±0.03
0.78
<0.01
61/61
0.41
±0.07
1.02
<0.01
Formaldehyde
0.000013
0.0098
59/59
2.67
±0.38
34.71
0.27
61/61
2.92
±0.35
37.94
0.30
Hexachloro-1,3 -butadiene
0.000022
0.09
1/60
NR
NR
NR
18/61
0.03
±0.01
0.60
<0.01
Arsenic (TSP)a
0.0043
0.000015
60/60
0.85
±0.14
3.64
0.06
61/61
0.95
±0.20
4.08
0.06
Nickel (TSP)a
0.00048
0.00009
60/60
1.44
±0.19
0.69
0.02
61/61
1.20
±0.18
0.58
0.01
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Oklahoma City, Oklahoma - OCOK
Acetaldehyde
0.0000022
0.009
60/60
1.92
±0.22
4.23
0.21
60/60
1.63
±0.14
3.59
0.18
Benzene
0.0000078
0.03
60/60
0.58
±0.04
4.52
0.02
61/61
0.61
±0.11
4.75
0.02
1.3 -Butadiene
0.00003
0.002
50/60
0.04
±0.01
1.14
0.02
54/61
0.05
±0.01
1.37
0.02
Carbon Tetrachloride
0.000006
0.1
60/60
0.64
±0.02
3.81
0.01
61/61
0.61
±0.03
3.68
0.01
1,2 -Dichloroethane
0.000026
2.4
56/60
0.08
±0.01
1.99
<0.01
55/61
0.07
±0.01
1.89
<0.01
Formaldehyde
0.000013
0.0098
60/60
3.16
±0.50
41.05
0.32
60/60
2.72
±0.37
35.41
0.28
Hexachloro-1,3 -butadiene
0.000022
0.09
5/60
NR
NR
NR
14/61
0.02
±0.01
0.41
<0.01
Arsenic (TSP)a
0.0043
0.000015
60/60
0.56
±0.07
2.39
0.04
60/60
0.51
±0.06
2.21
0.03
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Near-road, Oklahoma City, Oklahoma - NROK
Acetaldehyde
0.0000022
0.009
NS
NS
NS
NS
38/38
NA
NA
NA
Benzene
0.0000078
0.03
NS
NS
NS
NS
38/38
NA
NA
NA
1.3 -Butadiene
0.00003
0.002
NS
NS
NS
NS
38/38
NA
NA
NA
Carbon Tetrachloride
0.000006
0.1
NS
NS
NS
NS
38/38
NA
NA
NA
p-Dichlorobcnzcnc
0.000011
0.8
NS
NS
NS
NS
32/38
NA
NA
NA
1,2 -Dichloroethane
0.000026
2.4
NS
NS
NS
NS
32/38
NA
NA
NA
Ethylbenzene
0.0000025
1
NS
NS
NS
NS
38/38
NA
NA
NA
Formaldehyde
0.000013
0.0098
NS
NS
NS
NS
38/38
NA
NA
NA
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Yukon, Oklahoma - YUOK
Acetaldehyde
0.0000022
0.009
59/59
1.64
±0.16
3.61
0.18
61/61
1.60
±0.13
3.51
0.18
Benzene
0.0000078
0.03
59/59
0.56
±0.07
4.34
0.02
61/61
0.52
±0.05
4.07
0.02
1.3 -Butadiene
0.00003
0.002
51/59
0.04
±0.01
1.16
0.02
57/61
0.04
±0.01
1.21
0.02
Carbon Tetrachloride
0.000006
0.1
59/59
0.62
±0.03
3.71
0.01
61/61
0.61
±0.02
3.64
0.01
1,2 -Dichloroethane
0.000026
2.4
57/59
0.08
±0.01
2.03
<0.01
57/61
0.08
±0.01
1.96
<0.01
Formaldehyde
0.000013
0.0098
59/59
2.71
±0.39
35.20
0.28
61/61
2.64
±0.37
34.32
0.27
Hexachloro-1,3 -butadiene
0.000022
0.09
2/59
NR
NR
NR
13/61
0.02
±0.01
0.37
<0.01
Arsenic (TSP)a
0.0043
0.000015
59/59
0.55
±0.09
2.38
0.04
61/61
0.54
±0.08
2.34
0.04
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Table 18-4. Risk Approximations for the Oklahoma Monitoring Sites (Continued)
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-
million)
Noncancer
(HQ)
Bradley, Oklahoma - BROK
Acetaldehyde
0.0000022
0.009
40/40
4.06
± 1.42
8.93
0.45
60/60
1.46
±0.14
3.22
0.16
Benzene
0.0000078
0.03
39/39
NA
NA
NA
59/59
0.80
±0.10
6.23
0.03
1.3 -Butadiene
0.00003
0.002
29/39
NA
NA
NA
45/59
0.02
± <0.01
0.68
0.01
Carbon Tetrachloride
0.000006
0.1
39/39
NA
NA
NA
59/59
0.59
±0.04
3.55
0.01
1,2 -Dichloroethane
0.000026
2.4
34/39
NA
NA
NA
56/59
0.09
±0.01
2.43
<0.01
Formaldehyde
0.000013
0.0098
40/40
4.95
± 1.59
64.31
0.50
60/60
2.07
±0.30
26.97
0.21
Propionaldehyde
0.008
40/40
1.01
±0.42
0.13
60/60
0.33
±0.03
0.04
— = A Cancer URE or Noncancer RfC is not available.
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NS = Sampling was not conducted during this time.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Observations from Table 18-4 include the following:
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
each site.
• Formaldehyde and benzene have the highest cancer risk approximations among the
pollutants of interest for each Oklahoma monitoring site, with one exception (2015,
when annual averages could not be calculated for the VOC pollutants of interest for
BROK). Cancer risk approximations for formaldehyde range from 26.97 in-a-million
(BROK, 2016) to 64.31 in-a-million (BROK, 2015). BROK's 2015 cancer risk
approximation for formaldehyde ranks sixth highest among all cancer risk
approximations program-wide. Benzene cancer risk approximations for the Oklahoma
monitoring sites range from 4.07 in-a-million (YUOK, 2016) to 8.50 in-a-million
(TOOK, 2015). BROK's cancer risk approximation for acetaldehyde for 2015 is
8.93 in-a-million; remaining cancer risk approximations for acetaldehyde are less
than 5 in-a-million, ranging from 3.22 in-a-million for BROK (2015) to
4.44 in-a-million for TOOK (2015).
• None of the pollutants of interest for the Oklahoma sites have noncancer hazard
approximations greater than 1.0, indicating that no adverse noncancer health effects
are expected from these individual pollutants. The highest noncancer hazard
approximation calculated was based on BROK's annual average concentration of
formaldehyde for 2015 (0.50).
18.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 18-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 18-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 18-5 provides the pollutants with the highest cancer risk approximations (in-a-million) for
each site, as presented in Table 18-4. Cancer risk approximations for 2015 are presented in green
while approximations for 2016 are in white. The emissions, toxicity-weighted emissions, and
cancer risk approximations are shown in descending order in Table 18-5. Table 18-6 presents
similar information, but is limited to those pollutants with noncancer toxicity factors.
18-91
-------
Table 18-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Oklahoma Monitoring Sites
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Public Works, Tulsa, Oklahoma (Tulsa County) - TOOK
Benzene
260.73
Hexavalent Chromium PM
2.47E-03
Formaldehyde
43.31
Formaldehyde
185.06
Formaldehyde
2.41E-03
Formaldehyde
36.86
Ethylbenzene
162.19
Benzene
2.03E-03
Benzene
8.50
Acetaldehyde
101.70
Naphthalene
1.21E-03
Benzene
8.42
1.3 -Butadiene
35.95
1,3-Butadiene
1.08E-03
Acetaldehyde
4.44
Naphthalene
35.51
Ethylbenzene
4.05E-04
Acetaldehyde
4.34
Tetrachloroethylene
14.79
POM, Group 2b
3.62E-04
Arsenic
3.83
Trichloroethylene
10.43
POM, Group 5a
2.84E-04
Carbon Tetrachloride
3.69
POM, Group 2b
4.12
POM, Group 2d
2.74E-04
Carbon Tetrachloride
3.67
POM, Group 2d
3.11
Nickel, PM
2.49E-04
Arsenic
3.37
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Benzene
260.73
Hexavalent Chromium PM
2.47E-03
Formaldehyde
41.11
Formaldehyde
185.06
Formaldehyde
2.41E-03
Formaldehyde
35.26
Ethylbenzene
162.19
Benzene
2.03E-03
Benzene
7.03
Acetaldehyde
101.70
Naphthalene
1.21E-03
Benzene
6.59
1,3-Butadiene
35.95
1,3-Butadiene
1.08E-03
Acetaldehyde
4.04
Naphthalene
35.51
Ethylbenzene
4.05E-04
Acetaldehyde
3.89
Tetrachloroethylene
14.79
POM, Group 2b
3.62E-04
Carbon Tetrachloride
3.79
Trichloroethylene
10.43
POM, Group 5a
2.84E-04
Carbon Tetrachloride
3.70
POM, Group 2b
4.12
POM, Group 2d
2.74E-04
1,3-Butadiene
3.30
POM, Group 2d
3.11
Nickel, PM
2.49E-04
Arsenic
2.87
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Riverside, Tulsa, Oklahoma (Tulsa County) - TROK
Benzene
260.73
Hexavalent Chromium. PM
2.47E-03
Formaldehyde
37.94
Formaldehyde
185.06
Formaldehyde
2.41E-03
Formaldehyde
34.71
Ethylbenzene
162.19
Benzene
2.03E-03
Benzene
6.96
Acetaldehyde
101.70
Naphthalene
1.21E-03
Benzene
6.17
1.3 -Butadiene
35.95
1,3-Butadiene
1.08E-03
Acetaldehyde
4.43
Naphthalene
35.51
Ethylbenzene
4.05E-04
Acetaldehyde
4.22
Tetrachloroethylene
14.79
POM, Group 2b
3.62E-04
Arsenic
4.08
Trichloroethylene
10.43
POM, Group 5a
2.84E-04
Carbon Tetrachloride
3.65
POM, Group 2b
4.12
POM, Group 2d
2.74E-04
Arsenic
3.64
POM, Group 2d
3.11
Nickel, PM
2.49E-04
Carbon Tetrachloride
3.60
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Benzene
303.27
Formaldehyde
3.54E-03
Formaldehyde
41.05
Formaldehyde
272.57
Benzene
2.37E-03
Formaldehyde
35.41
Ethylbenzene
178.49
Naphthalene
1.52E-03
Benzene
4.75
Acetaldehyde
138.18
1,3-Butadiene
1.26E-03
Benzene
4.52
Naphthalene
44.62
Hexavalent Chromium. PM
7.60E-04
Acetaldehyde
4.23
1,3-Butadiene
42.06
POM, Group 2b
5.31E-04
Carbon Tetrachloride
3.81
Dichloromethane
15.53
Ethylbenzene
4.46E-04
Carbon Tetrachloride
3.68
POM, Group 2b
6.03
POM, Group 2d
3.32E-04
Acetaldehyde
3.59
POM, Group 2d
3.78
Acetaldehyde
3.04E-04
Arsenic
2.39
Trichloroethylene
3.12
POM, Group 5a
2.92E-04
Arsenic
2.21
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Near-road, Oklahoma City, Oklahoma (Oklahoma County) - NROK
Benzene
303.27
Formaldehyde
3.54E-03
Formaldehyde
272.57
Benzene
2.37E-03
Ethylbenzene
178.49
Naphthalene
1.52E-03
Acetaldehyde
138.18
1,3-Butadiene
1.26E-03
Naphthalene
44.62
Hexavalent Chromium. PM
7.60E-04
1.3 -Butadiene
42.06
POM, Group 2b
5.31E-04
Dichloromethane
15.53
Ethylbenzene
4.46E-04
POM, Group 2b
6.03
POM, Group 2d
3.32E-04
POM, Group 2d
3.78
Acetaldehyde
3.04E-04
Trichloroethylene
3.12
POM, Group 5a
2.92E-04
Yukon, Oklahoma (Canadian County) - YUOK
Formaldehyde
267.69
Formaldehyde
3.48E-03
Formaldehyde
35.20
Acetaldehyde
120.19
Benzene
6.63E-04
Formaldehyde
34.32
Benzene
84.96
1,3-Butadiene
3.59E-04
Benzene
4.34
Ethylbenzene
24.98
Acetaldehyde
2.64E-04
Benzene
4.07
1,3-Butadiene
11.96
Naphthalene
2.48E-04
Carbon Tetrachloride
3.71
Naphthalene
7.30
POM, Group 2b
7.57E-05
Carbon Tetrachloride
3.64
POM, Group 2b
0.86
POM, Group la
7.12E-05
Acetaldehyde
3.61
POM, Group la
0.81
POM, Group 2d
6.51E-05
Acetaldehyde
3.51
POM, Group 2d
0.74
Ethylbenzene
6.24E-05
Arsenic
2.38
Dichloromethane
0.69
1,2-Dibromoethane
6.05E-05
Arsenic
2.34
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bradley, Oklahoma (Grady County) - BROK
Formaldehyde
273.26
Formaldehyde
3.55E-03
Formaldehyde
64.31
Acetaldehyde
67.50
Benzene
4.32E-04
Formaldehyde
26.97
Benzene
55.36
1,3-Butadiene
2.98E-04
Acetaldehyde
8.93
Ethylbenzene
17.11
Acetaldehyde
1.49E-04
Benzene
6.23
1.3 -Butadiene
9.95
Naphthalene
1.35E-04
Carbon Tetrachloride
3.55
Naphthalene
3.97
POM, Group la
7.01E-05
Acetaldehyde
3.22
Dichloromethane
0.88
1,2-Dibromoethane
6.76E-05
1,2-Dichloroethane
2.43
POM, Group la
0.80
POM, Group 2b
5.25E-05
1,3-Butadiene
0.68
POM, Group 2b
0.60
POM, Group 2d
4.51E-05
POM, Group 2d
0.51
Ethylbenzene
4.28E-05
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Oklahoma Monitoring Sites
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations Based
on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Public Works, Tulsa, Oklahoma (Tulsa County) - TOOK
Toluene
1,035.38
Acrolein
732,286.29
Formaldehyde
0.34
Xylenes
598.49
Formaldehyde
18,883.99
Formaldehyde
0.29
Methanol
318.54
1,3-Butadiene
17,975.67
Acetaldehyde
0.22
Hexane
269.09
Naphthalene
11,836.01
Acetaldehyde
0.22
Benzene
260.73
Acetaldehyde
11,299.58
Manganese
0.09
Formaldehyde
185.06
Benzene
8,691.01
Manganese
0.09
Ethylbenzene
162.19
Xylenes
5,984.91
Arsenic
0.06
Ethylene glycol
107.55
Nickel, PM
5,760.01
Arsenic
0.05
Acetaldehyde
101.70
T richloroethylene
5,215.94
1,3-Butadiene
0.04
Hydrofluoric acid
48.54
Cadmium, PM
4,820.24
Benzene
0.04
Fire Station, Tulsa, Oklahoma (Tulsa County) - TMOK
Toluene
1,035.38
Acrolein
732,286.29
Formaldehyde
0.32
Xylenes
598.49
Formaldehyde
18,883.99
Formaldehyde
0.28
Methanol
318.54
1,3-Butadiene
17,975.67
Acetaldehyde
0.20
Hexane
269.09
Naphthalene
11,836.01
Acetaldehyde
0.20
Benzene
260.73
Acetaldehyde
11,299.58
1,3-Butadiene
0.05
Formaldehyde
185.06
Benzene
8,691.01
Arsenic
0.04
Ethylbenzene
162.19
Xylenes
5,984.91
1,3-Butadiene
0.04
Ethylene glycol
107.55
Nickel, PM
5,760.01
Arsenic
0.04
Acetaldehyde
101.70
T richloroethylene
5,215.94
Benzene
0.03
Hydrofluoric acid
48.54
Cadmium, PM
4,820.24
Benzene
0.03
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Riverside, Tulsa, Oklahoma (Tulsa County) - TROK
Toluene
1,035.38
Acrolein
732,286.29
Formaldehyde
0.30
Xylenes
598.49
Formaldehyde
18,883.99
Formaldehyde
0.27
Methanol
318.54
1,3-Butadiene
17,975.67
Acetaldehyde
0.22
Hexane
269.09
Naphthalene
11,836.01
Acetaldehyde
0.21
Benzene
260.73
Acetaldehyde
11,299.58
Arsenic
0.06
Formaldehyde
185.06
Benzene
8,691.01
Arsenic
0.06
Ethylbenzene
162.19
Xylenes
5,984.91
1,3-Butadiene
0.04
Ethylene glycol
107.55
Nickel, PM
5,760.01
1,3-Butadiene
0.03
Acetaldehyde
101.70
T richloroethylene
5,215.94
Benzene
0.03
Hydrofluoric acid
48.54
Cadmium, PM
4,820.24
Benzene
0.03
Oklahoma City, Oklahoma (Oklahoma County) - OCOK
Toluene
1,155.76
Acrolein
1,388,820.96
Formaldehyde
0.32
Xylenes
678.98
Formaldehyde
27,813.75
Formaldehyde
0.28
Methanol
390.58
1,3-Butadiene
21,030.79
Acetaldehyde
0.21
Hexane
374.93
Acetaldehyde
15,353.68
Acetaldehyde
0.18
Benzene
303.27
Naphthalene
14,874.92
Arsenic
0.04
Formaldehyde
272.57
Benzene
10,109.02
Arsenic
0.03
Ethylbenzene
178.49
Xylenes
6,789.78
1,3-Butadiene
0.02
Ethylene glycol
143.51
Cyanide Compounds, gas
2,746.37
Benzene
0.02
Acetaldehyde
138.18
Nickel, PM
2,204.08
Benzene
0.02
Naphthalene
44.62
Glycol ethers, gas
2,109.68
1,3-Butadiene
0.02
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Near-road, Oklahoma City, Oklahoma (Oklahoma County) - NROK
Toluene
1,155.76
Acrolein
1,388,820.96
Xylenes
678.98
Formaldehyde
27,813.75
Methanol
390.58
1.3 -Butadiene
21,030.79
Hexane
374.93
Acetaldehyde
15,353.68
Benzene
303.27
Naphthalene
14,874.92
Formaldehyde
272.57
Benzene
10,109.02
Ethylbenzene
178.49
Xylenes
6,789.78
Ethylene glycol
143.51
Cyanide Compounds, gas
2,746.37
Acetaldehyde
138.18
Nickel, PM
2,204.08
Naphthalene
44.62
Glycol ethers, gas
2,109.68
Yukon, Oklahoma (Canadian County) - YUOK
Formaldehyde
267.69
Acrolein
3,787,368.55
Formaldehyde
0.28
Xylenes
219.21
Formaldehyde
27,315.35
Formaldehyde
0.27
Toluene
170.13
Acetaldehyde
13,354.33
Acetaldehyde
0.18
Methanol
126.11
1,3-Butadiene
5,982.12
Acetaldehyde
0.18
Acetaldehyde
120.19
Benzene
2,832.09
Arsenic
0.04
Benzene
84.96
Naphthalene
2,434.57
Arsenic
0.04
Acrolein
75.75
Xylenes
2,192.08
1,3-Butadiene
0.02
Hexane
61.45
Cyanide Compounds, gas
1,665.91
1,3-Butadiene
0.02
Ethylbenzene
24.98
Lead, PM
1,020.81
Benzene
0.02
Ethylene glycol
21.49
Cadmium, PM
530.43
Benzene
0.02
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 18-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Oklahoma Monitoring Sites (Continued)
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer
Hazard
Approximation
(HQ)
Bradley, Oklahoma (Grady County) - BROK
Formaldehyde
273.26
Acrolein
1,867,435.12
Formaldehyde
0.50
Xylenes
167.13
Formaldehyde
27,884.14
Acetaldehyde
0.45
Toluene
118.51
Acetaldehyde
7,500.28
Formaldehyde
0.21
Acetaldehyde
67.50
1.3 -Butadiene
4,974.14
Acetaldehyde
0.16
Methanol
56.81
Cobalt, PM
2,066.54
Propionaldehyde
0.13
Benzene
55.36
Cyanide Compounds, gas
1,965.13
Propionaldehyde
0.04
Acrolein
37.35
Benzene
1,845.35
Benzene
0.03
Styrene
35.75
Xylenes
1,671.27
1,3-Butadiene
0.01
Hexane
35.64
Naphthalene
1,324.08
Carbon Tetrachloride
0.01
Ethylbenzene
17.11
Propionaldehyde
283.01
1,2-Dichloroethane
<0.01
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more
in-depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 18.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
Observations from Table 18-5 include the following:
• Benzene is the highest emitted pollutant with a cancer URE in Tulsa and Oklahoma
Counties, followed by formaldehyde and ethylbenzene. The highest emitted
pollutants in Canadian and Grady Counties are formaldehyde, acetaldehyde, and
benzene.
• The pollutant with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for Tulsa County is hexavalent chromium, followed by formaldehyde
and benzene. The pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) for Oklahoma County is formaldehyde, followed by
benzene and naphthalene. The pollutant with the highest toxicity-weighted emissions
(of the pollutants with cancer UREs) for Canadian and Grady Counties is also
formaldehyde, followed by benzene and 1,3-butadiene.
• Seven of the highest emitted pollutants in Tulsa County also have the highest
toxicity-weighted emissions. Eight of the highest emitted pollutants in Oklahoma
County also have the highest toxicity-weighted emissions. Nine of the highest emitted
pollutants in Canadian County also have the highest toxicity-weighted emissions; this
is also true for Grady County.
• With the exception of BROK, formaldehyde and benzene have the highest cancer risk
approximations among the Oklahoma sites' pollutants of interest. Both of these
pollutants appear at or near the top of both emissions-based lists for each county.
Acetaldehyde also appears among the pollutants with the highest cancer risk
approximations for each Oklahoma site; acetaldehyde is one of the pollutants listed in
both emissions-based lists for Oklahoma, Canadian, and Grady Counties, but is not
among the pollutants with the highest toxicity-weighted emissions for Tulsa County
(it ranks 11th, though).
• Arsenic is a pollutant of interest for each of the five sites sampling TSP metals.
Although this pollutant has one of the higher cancer risk approximations for each site,
arsenic is not on either emissions-based list for these four counties.
18-100
-------
• Carbon tetrachloride is another pollutant of interest for each site and has some of the
higher cancer risk approximations for each site but does not appear on either
emissions-based list for any of the Oklahoma counties with an NMP site.
• Naphthalene and several POM Groups appear in Table 18-5 for quantity emitted and
toxicity-weighted emissions. PAHs were not sampled for under the NMP at the
Oklahoma sites.
Observations from Table 18-6 include the following:
• Toluene and xylenes are the highest emitted pollutants with noncancer RfCs in Tulsa
and Oklahoma Counties, while the formaldehyde and xylenes are the highest emitted
pollutants with noncancer RfCs in Canadian and Grady Counties. Emissions were
considerably higher in Tulsa and Oklahoma Counties than Canadian and Grady
Counties.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for all four Oklahoma counties. Acrolein has the
highest toxicity-weighted emissions for almost all counties with NMP sites but
appears among the highest emitted for only five. Canadian and Grady Counties are
two of those counties, with acrolein ranking seventh among those with the highest
emissions for each county. Compared to other counties with NMP sites, Canadian
County's acrolein emissions rank fourth highest (75.75 tpy) and Grady County's rank
seventh highest (37.35 tpy) among counties with NMP sites (acrolein emissions in
Oklahoma and Tulsa Counties are 27.78 tpy and 14.65 tpy, respectively). Acrolein
was sampled for at all seven Oklahoma sites, but this pollutant was excluded from the
pollutants of interest designation, and thus subsequent risk-based screening
evaluations, due to questions about the consistency and reliability of the
measurements, as discussed in Section 3.2.
• Four of the highest emitted pollutants in Tulsa County also have the highest toxicity-
weighted emissions; five of the highest emitted pollutants in Oklahoma, Canadian,
and Grady Counties also have the highest toxicity-weighted emissions. Even though
toluene is one of, if not the highest emitted pollutant in all four counties, this pollutant
does not appear among those with the highest toxicity-weighted emissions.
• Formaldehyde and acetaldehyde have the highest noncancer hazard approximations
among the Oklahoma sites, where they could be calculated. These pollutants appear
on both emissions-based lists for each county. Benzene also appears on all three lists
for each site. 1,3-Butadiene also has some of the highest noncancer hazard
approximations for each site, and has some of the highest toxicity-weighted emissions
for each respective county, but is not one of the 10 highest emitted pollutants in each
county (but is just outside the list at 11th or 12th highest for each county).
• Arsenic appears among the pollutants with the highest noncancer hazard
approximations for TOOK, TMOK, TROK, OCOK, and YUOK, but appears on
neither emissions-based list for the four Oklahoma Counties. At least one metal
appears among the pollutants with the highest toxicity-weighted emissions for each
18-101
-------
county, but no metals are listed among the highest emitted pollutants for any of the
counties. This speaks to the relative toxicity of the speciated metals.
18.5 Summary of the 2015-2016 Monitoring Data for the Oklahoma Monitoring Sites
Results from several of the data analyses described in this section include the following:
~~~ Seventeen pollutants failed at least one screen for TOOK; 17 pollutants failed screens
for TMOK; 17 pollutants failed screens for TROK; 16 pollutants failed screens for
OCOK; 10 pollutants failed screens for NROK; 15 pollutants failed screens for
YUOK; and 11 pollutants failed screens for BROK.
~~~ Formaldehyde and acetaldehyde had the highest annual average concentrations for
each site. BROK's annual average concentration of acetaldehyde for 2015 is the
highest annual average among NMP sites sampling this pollutant, although BROK's
annual average for 2016 is significantly lower.
~~~ After several years of increasing, concentrations of several pollutants, including
acetaldehyde, benzene, ethylbenzene, and manganese, decreased at TOOK after 2012
then have remained fairly static in recent years. Other pollutants exhibit this trend as
well but the difference is less significant. Benzene and acetaldehyde concentrations
have also been decreasing at TMOK and have also leveled out in recent years. In
addition, the detection rates of 1,2-dichloroethane have been increasing at TOOK,
TMOK, and OCOK over the last five years of sampling.
~~~ Formaldehyde has the highest cancer risk approximation among the site-specific
pollutants of interest for each site. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
18-102
-------
19.0 Site in Rhode Island
This section summarizes those data from samples collected at the NATTS site in Rhode
Island and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial 7 ~ ~~
Data generated by sources other
and temporal characteristics of the ambient monitoring than ERG, EPA's contract
laboratory for the NMP, are not
concentrations and reviews them through the context of included in the data analyses
risk. Readers are encouraged to refer to Sections 1 contained in this report,
through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
19.1 Site Characterization
This section characterizes the Rhode Island monitoring site by providing a description of
the nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient measurements.
The PRRI monitoring site is located in south Providence. Figure 19-1 presents a
composite satellite image retrieved from ArcGIS Explorer showing the monitoring site and its
immediate surroundings. Figure 19-2 identifies nearby point source emissions locations by
source category, as reported in the 2014 NEI for point sources, version 1. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 19-2. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring site.
Further, this boundary provides both the proximity of emissions sources to the monitoring site as
well as the quantity of such sources within a given distance of the site. Sources outside the
10-mile boundary are still visible on the map for reference but have been grayed out to
emphasize emissions sources within the boundary. Table 19-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates. Each
figure and table is discussed in detail in the paragraphs that follow.
19-1
-------
Figure 19-1. Providence, Rhode Island (PRR1) Monitoring Site
Point Park
¦Henderson .St-
.JWhrtmarstiiSt,
^Ljar£*/jBucklin Park
Princeton Ave
.RnncetohrAve,
Moore St
Da boll St
-------
Figure 19-2. NET Point Sources Located Within 10 Miles of PRRI
MASSACHUSETTS
Bristol
County
Providence
Harbor
Kent
County
Miles
Legend
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
PRRI NATTS site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
Aerospace/Aircraft Manufacturing Facility (1)
"t" Airport/Airline/Airport Support Operations (9)
jji- Asphalt Production/Hot Mix Asphalt Plant (7)
B Bulk Terminal/Bulk Plant (5)
C Chemical Manufacturing Facility (6)
[X Crematory (1)
G) Dry Cleaning Facility (38)
@ Electrical Equipment Manufacturing Facility (16)
f Electricity Generation Facility (4)
E Electroplating, Plating, Polishing, Anodizing, and Coloring Facility (16)
F Food Processing/Agriculture Facility (3)
Industrial Machinery or Equipment Plant (6)
O Institution (school, hospital, prison, etc.) (14)
¦ Landfill (1)
A Metal Coating, Engraving, and Allied Services to Manufacturers (1)
(•) Metals Processing/Fabrication Facility (14)
A Military Base/National Security Facility (4)
X Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility (50)
[j Paint and Coating Manufacturing Facility (3)
cd Pharmaceutical Manufacturing Facility (2)
R Plastic, Resin, or Rubber Products Plant (14)
¦Is Printing, Coating and Dyeing of Fabrics Facility (1)
P Printing/Publishing/Paper Product Manufacturing Facility (8)
-ii. Ship/Boat Manufacturing or Repair Facility (3)
TT Telecommunications/Radio Facility (4)
^ Testing Laboratory (1)
T Textile, Yarn, or Carpet Plant (7)
Truck/Bus/Transportation Operations (1)
* Wastewater Treatment Facility (1)
W Woodwork, Furniture. Millwork and Wood Preserving Facility (4)
\
\
\ Providence
\ County
\
\
\ ^
\
\
\
\
\
\
\
\
^ ^ ^
" " \
19-3
-------
Table 19-1. Geographical Information for the Rhode Island Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual Average
Daily Traffic1
Intersection
Used for
Traffic Data
PRRI
44-007-0022
Providence
Providence
Providence-
Warwick, RI-MA
41.807776,
-71.415105
Residential
Urban/City
Center
148,000
1-95 near 1-195
1AADT reflects 2015 data (RI DOT, 2016)
BOLD ITALICS = EPA-designated NATTS Site
vo
-------
Figure 19-1 shows that the areas to the west and south of PRRI are primarily residential,
but areas to the north and east are commercial. A hospital lies to the northeast of the site, just
north of Dudley Street. Interstate-95 runs north-south about one-half mile to the east of the site,
then turns northwestward, skirting around downtown Providence. Industrial areas are located
between Providence Harbor and 1-95, as shown on the right-hand side of Figure 19-1.
Figure 19-2 shows that a large number of point sources are located within 10 miles of
PRRI, most of which are within about 5 miles of the site. The source categories with the greatest
number of point sources within 10 miles of PRRI include dry cleaners; electroplating, plating,
polishing, anodizing, and coloring facilities; electrical equipment manufacturing facilities;
plastic, resin, or rubber products facilities; institutions (such as schools, prisons, and hospitals);
and metals processing and fabrication facilities. Sources within one-half mile of PRRI include
several hospitals, a heliport at a hospital, and a facility that falls into the miscellaneous
commercial and industrial source category.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 19-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near PRRI is nearly 150,000 and is the seventh highest compared to traffic
volumes near other NMP monitoring sites. The traffic estimate provided is for 1-95 near the
1-195 interchange.
19.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 19-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 19-2. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. PAHs were sampled for at PRRI under the NMP in 2015 and 2016.
19-5
-------
Table 19-2. 2015-2016 Risk-Based Screening Results for the Rhode Island Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Providence, Rhode Island - PRRI
Naphthalene
0.029
99
120
82.50
94.29
94.29
Fluorene
0.011
3
81
3.70
2.86
97.14
Benzo(a)pyrene
0.00057
2
118
1.69
1.90
99.05
Acenaphthene
0.011
1
91
1.10
0.95
100.00
Total
105
410
25.61
Observations from Table 19-2 include the following:
• Concentrations of four PAHs failed at least one screen for PRRI; nearly 26 percent of
concentrations for these four pollutants were greater than their associated risk
screening value (or failed screens).
• Concentrations of naphthalene account for 99 of the 105 total failed screens.
• Naphthalene and fluorene together contribute to 95 percent of failed screens for PRRI
and therefore were identified as pollutants of interest for this site.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
19.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Rhode Island monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of sampling.
• The range of measurements and annual concentration averages are presented
graphically to illustrate how each site's concentrations compare to the program-level
averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
19-6
-------
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at PRRI are provided
in Appendix N.
19.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for the Rhode Island site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for PRRI are presented
in Table 19-3, where applicable. Note that if a pollutant was not detected in a given calendar
quarter, the quarterly average simply reflects "0" because only zeros substituted for non-detects
were factored into the quarterly average concentration.
19-7
-------
Table 19-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Rhode Island Monitoring Site
vO
00
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Providence, Rhode Island - PRRI
Fluorene
41/41/60
1.78
±0.38
2.71
±0.83
4.50
± 1.05
0.48
±0.53
2.24
±0.51
40/40/60
0
2.61
± 1.20
6.76
±2.26
3.47
± 1.06
3.17
±0.90
Naphthalene
60/60/60
59.77
± 11.60
38.04
±7.04
49.98
± 11.49
60.48
±21.46
52.63
±7.62
60/60/60
50.81
±23.16
44.21
± 11.81
55.05
± 12.77
70.35
± 19.62
55.21
±8.76
-------
Observations for PRRI from Table 19-3 include the following:
• Naphthalene was detected in all of the valid PAH samples collected at PRRI.
• Concentrations of naphthalene measured at PRRI vary in magnitude, ranging from
7.70 ng/m3 to 203 ng/m3. For both years, the second quarter has the lowest quarterly
average concentration and the fourth quarter has the highest, although the confidence
intervals are relatively large, indicating considerable variability in the measurements.
The annual average concentration of naphthalene for 2015 is fairly similar to the
annual average for 2016.
• Fluorene was detected in nearly 70 percent of the PAH samples collected each year.
The "0" shown for the first quarter average concentration for 2016 indicates that this
pollutant was not detected during this calendar quarter. There were no measured
detections of fluorene for a four-month period between December 8, 2015 and April
12, 2016. Eighteen of the 20 non-detects measured in 2016 were from this period.
• Concentrations of fluorene measured at PRRI range from 1.19 ng/m3 to 16.1 ng/m3,
plus the non-detects. Measurements collected in 2016 are more variable than those
collected in 2015, based on the confidence intervals shown. The confidence interval
for the fourth quarter of 2015 is larger than the average itself (0.48 ± 0.53 ng/m3);
there were only three measured detections during this quarter, and 15 non-detects.
19.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for each of the pollutants
listed in Table 19-3 for PRRI. Figures 19-3 and 19-4 overlay PRRI's minimum, annual average,
and maximum concentrations for each year onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations, as described in Section 3.4.2.1,
and are discussed below. If an annual average concentration could not be calculated, the range of
concentrations are still provided in the figures that follow.
19-9
-------
Figure 19-3. Program vs. Site-Specific Average Fluorene Concentrations
1
1
•
( ,
Program Max Concentration = 105 ng/m3
r 1
* 1
0 10 20 30 40 50 60
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 19-3 presents the box plot for fluorene for PRRI and shows the following:
• The program-level maximum fluorene concentration (105 ng/m3) is not shown
directly on the box plot in Figure 19-3 as the scale of the box plot has been reduced to
allow for the observation of data points at the lower end of the concentration range.
Note that the first quartile is zero and therefore not visible on the box plot due to the
number of non-detects.
• The range of fluorene concentrations measured at fluorene in 2016 is more than twice
the range of concentrations measured in 2015. Five fluorene concentrations measured
in 2016 are higher than the maximum concentration measured in 2015. The number
of fluorene concentrations greater than 5 ng/m3 measured in 2016 (12) is more than
twice the number measured in 2015 (5).
• PRRI's annual average fluorene concentration for 2015 is similar to the program-
level median concentration of 2.25 ng/m3. PRRI's annual average for 2016 is slightly
higher, falling between the program-level median and average (4.36 ng/m3)
concentrations.
19-10
-------
Figure 19-4. Program vs. Site-Specific Average Naphthalene Concentrations
*1
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2ndQuartile 3rdQuartile 4thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 19-4 presents the box plot for naphthalene for PRRI and shows the following:
• The maximum naphthalene concentration measured at PRRI was measured in 2016
and is half the magnitude of the maximum concentration measured across the
program.
• Both annual average concentrations for PRRI fall between the program-level median
(48.90 ng/m3) and average concentrations (61.23 ng/m3). PRRI's annual average
concentrations of naphthalene are in the lower half of the range compared to other
NMP sites sampling PAHs (ranking 21st and 23 rd).
19.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
PAHs have been sampled at PRRI under the NMP since 2008. Thus, Figures 19-5 and 19-6
presents the 1-year statistical metrics for the pollutants of interest for PRRI. The statistical
metrics presented for assessing trends include the substitution of zeros for non-detects. If
sampling began mid-year, a minimum of 6 months of sampling is required for inclusion in the
trends analysis; in these cases, a 1-year average concentration is not provided, although the range
and percentiles are still presented.
19-11
-------
Figure 19-5. Yearly Statistical Metrics for Fluorene Concentrations Measured at PRRI
5th Percentile
O 95th Percentile
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
Observations from Figure 19-5 for fluorene concentrations measured at PRRI include the
following:
• PRRI began sampling PAHs under the NMP in July 2008. Because a full year's worth
of data is not available, a 1-year average concentration is not presented for 2008,
although the range of measurements is provided.
• The only two fluorene concentrations greater than 20 ng/m3 were both measured in
August 2008 (29 ng/m3 and 22 ng/m3). This is also the only year that a concentration
less than 1 ng/m3 was not measured. Note, however, that 2008 includes half a year's
worth of data.
• The concentration range decreases considerably for 2009. The concentration profile
changes only slightly between 2009 and 2010.
• After 2010, fluorene concentrations measured at PRRI have a decreasing trend, with
most of the statistical parameters at a minimum for 2015.
• With the exception of the minimum and 5th percentile (which did not change), each
of the statistical parameters exhibits an increase for 2016. The maximum
concentration (16.1 ng/m3) and 95th percentile are both at their highest since 2008.
19-12
-------
Figure 19-6. Yearly Statistical Metrics for Naphthalene Concentrations Measured at PRRI
350
mE
c
O
c
QJ
pt-
o
u
o-
-o—
-o-
Lr-
2
L^J
'-2-
I
2
2008
2009
2010
2011
2012
Year
2013
2014
2015
2016
O
5th Percentile
- M
Inin
nurr
-
Median
- Maxim
um
O
95th Percentile
Ave
rag
1
1 A 1-year average is not presented because sampling under the NMP did not begin until July 2008.
Observations from Figure 19-6 for naphthalene concentrations measured at PRRI include
the following:
• The maximum naphthalene concentration was measured at PRRI in 2011
(301 ng/m3). In total, 11 naphthalene concentrations greater than 200 ng/m3 have been
measured at PRRI, of which seven were measured in November of any given year. Of
the 27 naphthalene concentrations greater than 150 ng/m3 measured at PRRI, more
than half (18) were measured during the fourth quarter of any given year and 24 of
these 27 were measured during the first or fourth quarters (or the colder months of the
year).
• Although the range of concentrations measured has varied between 2009 and 2012,
the 1-year average concentrations of naphthalene exhibit little variability, ranging
from 71.39 ng/m3 (2010) to 77.73 ng/m3 (2009). This is also true for the median
concentration, which, including 2008, ranges from 58.90 ng/m3 (2008) to 64.80 ng/m3
(2009).
• The concentrations of naphthalene measured at PRRI have a decreasing trend
between 2012 and 2014. The 1-year average and median concentrations are both at a
minimum for 2014; the median concentration is less than 50 ng/m3 for the first time
since the onset of sampling.
19-13
-------
• For 2015, some statistical parameters, such as the maximum and minimum
concentrations and the 5th percentile, exhibit additional decreases, while the central
tendency parameters exhibit slight increases, although the median concentration
remains less than 50 ng/m3. Relatively little change is shown in the concentration
profile for 2016, although the maximum concentration measured is greater than
200 ng/m3 for the first time in several years.
19.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the PRRI monitoring site. Refer to Sections 3.2, 3.4.2.3, and
3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
19.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Rhode Island monitoring site, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 19-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations for PRRI from Table 19-4 include the following:
• Naphthalene has both a cancer URE and a noncancer RfC while only a cancer RfC is
available for fluorene.
• The cancer risk approximations for naphthalene are both less than 2 in-a-million. The
cancer risk approximations for fluorene are an order of magnitude less.
• The noncancer hazard approximations for naphthalene are negligible (both 0.02),
indicating that no adverse noncancer health effects are expected from this individual
pollutant.
19-14
-------
Table 19-4. Risk Approximations for the Rhode Island Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Providence, Rhode Island - PRRI
Fluorene
0.000088
41/60
2.24
±0.51
0.20
40/60
3.17
±0.90
0.28
Naphthalene
0.000034
0.003
60/60
52.63
±7.62
1.79
0.02
60/60
55.21
±8.76
1.88
0.02
- = A Cancer URE or Noncancer RfC is not available.
-------
19.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 19-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 19-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 19-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for PRRI, as presented in Table 19-4. Cancer risk approximations for 2015 are presented
in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 19-5.
Table 19-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on the site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 19.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
19-16
-------
Table 19-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer
UREs for the Rhode Island Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Providence, Rhode Island (Providence County) - PRRI
Benzene
174.63
Formaldehyde
2.16E-03
Naphthalene
1.88
Formaldehyde
166.18
Benzene
1.36E-03
Naphthalene
1.79
Acetaldehyde
93.86
1.3 -Butadiene
8.05E-04
Fluorene
0.28
Ethylbenzene
82.84
Naphthalene
7.47E-04
Fluorene
0.20
1.3 -Butadiene
26.83
POM, Group 2b
3.22E-04
Bis(2-ethylhexyl) phthalate, gas
23.54
Arsenic, PM
3.14E-04
Naphthalene
21.96
POM, Group 2d
2.55E-04
Tetrachloroethylene
15.82
POM, Group 5a
2.40E-04
Trichloroethylene
8.17
Ethylbenzene
2.07E-04
POM, Group 2b
3.66
Acetaldehyde
2.06E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 19-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with
Noncancer RfCs for the Rhode Island Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Providence, Rhode Island (Providence County) - PRRI
Toluene
585.59
Acrolein
509,316.97
Naphthalene
0.02
Methanol
312.84
2,4-Toluene diisocyanate
28,964.64
Naphthalene
0.02
Xylenes
309.37
Formaldehyde
16,956.90
Benzene
174.63
1.3 -Butadiene
13,412.76
Formaldehyde
166.18
Acetaldehyde
10,428.55
Hexane
99.45
Naphthalene
7,321.59
Acetaldehyde
93.86
Benzene
5,820.86
Ethylbenzene
82.84
Cadmium, PM
5,291.88
Ethylene glycol
34.11
Arsenic, PM
4,864.13
1.3 -Butadiene
26.83
T richloroethylene
4,085.59
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 19-5 include the following:
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Providence County.
• Formaldehyde is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs), followed by benzene and 1,3-butadiene.
• Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Providence County.
• Naphthalene has the seventh highest emissions and the fourth highest toxicity-
weighted emissions for Providence County. Fluorene is part of POM, Group 2b,
which has the tenth highest emissions and the fifth highest toxicity-weighted
emissions for Providence County.
• Additional POM Groups appear among the pollutants with the highest toxicity-
weighted emissions for Providence County. POM, Groups 2d and 5a rank seventh and
eighth, respectively, for their toxicity-weighted emissions. POM, Group 5a includes
benzo(a)pyrene, which failed a single screen for PRRI; POM, Group 2d does not
include any of the PAHs sampled for at PRRI.
Observations from Table 19-6 include the following:
• Toluene, methanol, and xylenes are the highest emitted pollutants with noncancer
RfCs in Providence County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, 2,4-toluene diisocyanate, and formaldehyde.
• Four of the highest emitted pollutants in Providence County also have the highest
toxicity-weighted emissions.
• Although naphthalene ranks sixth among the pollutants with the highest toxicity-
weighted emissions, it is not one of the highest emitted pollutants (with a noncancer
RfC) in Providence County (it ranks 12th). None of the POM Groups appear in either
emissions-based list in Table 19-6.
19.5 Summary of the 2015-2016 Monitoring Data for PRRI
Results from several of the data analyses described in this section include the following:
~~~ Concentrations offour PAHs failed screens for PRRI. Naphthalene andfluorene were
identified as pollutants of interest for PRRI, with concentrations of naphthalene
accounting for the majority offailed screens.
~~~ Concentrations of naphthalene measured at PRRI do not vary significantly across the
two years of sampling. The range offluorene concentrations measured at PRRI in
2016 is twice the range measured in 2015.
19-19
-------
~~~ Concentrations of naphthalene measured at PRR1 have a decreasing trend that has
leveled out over the last two years. Fluorene concentrations also exhibited a
decreasing trend that lasted several years, before increasing for 2016.
19-20
-------
20.0 Site in Utah
This section summarizes those data from samples collected at the NATTS site in Utah
and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial
and temporal characteristics of the ambient monitoring
concentrations and reviews them through the context of
risk. Readers are encouraged to refer to Sections 1 through
4 for detailed discussions and definitions regarding the
various data analyses presented below.
20.1 Site Characterization
This section characterizes the Utah monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient monitoring measurements.
The BTUT monitoring site is located in Bountiful, in northern Utah. Figure 20-1 presents
a composite satellite image retrieved from ArcGIS Explorer showing the monitoring site and its
immediate surroundings. Figure 20-2 identifies nearby point source emissions locations by
source category, as reported in the 2014 NEI for point sources, version 1. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 20-2. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring site.
Further, this boundary provides both the proximity of emissions sources to the monitoring site as
well as the quantity of such sources within a given distance of the site. Sources outside the
10-mile boundary are still visible on the map for reference but have been grayed out to
emphasize emissions sources within the boundary. Table 20-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates. Each
figure and table is discussed in detail in the paragraphs that follow.
Data generated by sources other
than ERG, EPA's contract
laboratory for the NMP, are not
included in the data analyses
contained in this report.
. r
20-1
-------
Figure 20-1. Bountiful, Utah (BTUT) Monitoring Site
¦County
Jiapiffr
yj.RageS'RIaceOc-
¦ElBert Av«
Jlle'Commbrts-Way.
o Village Square Rd
W Pages Ln
.£.1400 N
£ 1300 N.
-W.1225-N.
;&.1250
O
K)
-------
Figure 20-2. NEI Point Sources Located Within 10 Miles of BTUT
112°10'0"W 112°5'0"W 112C0'0"W 111°55'0"W 111C50'0"W 111o45'0"W
I I
Morgan
County
Davis '
I County I
Great Salt l
Lake
I I
Salt Lake
County
i I
iwswv nzww nrss'o-'w nrso'cw mwcrvv m,-40'o,,w
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
~ BTUT NATTS site
O 10 mile radius
County boundary
Source Category Group (No. of Facilities)
"I" Airport/Airline/Airport Support Operations (8)
'i Asphalt Production/Hot Mix Asphalt Plant (2)
3 Brick, Structural Clay, or Clay Ceramics Plant (1)
B Bulk Terminal/Bulk Plant (2)
e Electrical Equipment Manufacturing Facility (2)
f Electricity Generation Facility (2)
+ Industrial Machinery or Equipment Plant (2)
o Institution (school, hospital, prison, etc.) (1)
• Landfill (1)
® Metals Processing/Fabrication Facility (2)
? Miscellaneous Commercial/Industrial Facility (1)
~ Paint and Coating Manufacturing Facility (1)
Petroleum Refinery (5)
X Rail Yard/Rail Line Operations (2)
20-3
-------
Table 20-1. Geographical Information for the Utah Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
BTUT
49-011-0004
Bountiful
Davis
Ogden-Clearfield,
UT
40.902967,
-111.884467
Residential
Suburban
133,965
1-15, north of Hwy 89 junction
lAADT reflects 2014 data (UT DOT, 2014)
BOLD ITALICS = EPA-designated NATTS Site
to
o
-------
Bountiful is north of Salt Lake City and is situated in a valley between the Great Salt
Lake to the west and the Wasatch Mountains to the east. Figure 20-1 shows that BTUT is located
on the property of Viewmont High School, in a primarily residential area. The site is located
about one-third of a mile from 1-15, which runs north-south through most of the surrounding
urban area including Salt Lake City, Clearfield, and Ogden.
Figure 20-2 shows that most of the point sources near BTUT are located to the south of
the site and run parallel to 1-15. The facilities near BTUT are involved in a variety of industries,
although the source categories with the greatest number of point sources surrounding BTUT are
the airport and airport support operations category and the petroleum refinery source category.
The airport source category includes airports and related operations as well as small runways and
heliports, such as those associated with hospitals or television stations. Point sources within 2
miles of BTUT include a metals processing/fabrication facility, a facility generating electricity
via combustion, a petroleum refinery, a paint and coatings manufacturer, and a landfill.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 20-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near BTUT is nearly 134,000 and is among the higher traffic volumes
compared to those for other NMP sites. The traffic estimate provided is for 1-15, north of the
Highway 89 junction, just west of the site.
20.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 20-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 20-2. It is important to note which pollutants were sampled for at each site when reviewing
the results of this analysis. VOCs, carbonyl compounds, SNMOCs, PAHs, and metals (PMio)
20-5
-------
were sampled for at BTUT. BTUT is one of only two NMP sites sampling both SNMOC and
VOCs (NBIL is the other).
Table 20-2. 2015-2016 Risk-Based Screening Results for the Utah Monitoring Site
Pollutant
Screening
Value
frig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Bountiful, Utah - BTUT
Acetaldehyde
0.45
113
113
100.00
12.38
12.38
Formaldehyde
0.077
113
113
100.00
12.38
24.75
Benzene
0.13
109
109
100.00
11.94
36.69
Carbon Tetrachloride
0.17
107
109
98.17
11.72
48.41
1,2 -Dichloroethane
0.038
100
100
100.00
10.95
59.36
Arsenic (PMio)
0.00023
95
113
84.07
10.41
69.77
1.3 -Butadiene
0.03
95
105
90.48
10.41
80.18
Naphthalene
0.029
91
117
77.78
9.97
90.14
Hexachloro-1,3 -butadiene
0.045
17
17
100.00
1.86
92.00
Nickel (PMio)
0.0021
15
115
13.04
1.64
93.65
Propionaldehyde
0.8
15
113
13.27
1.64
95.29
Ethylbenzene
0.4
12
109
11.01
1.31
96.60
Dichloro methane
60
7
109
6.42
0.77
97.37
p-Dichlorobcnzcnc
0.091
5
29
17.24
0.55
97.92
Fluorene
0.011
5
64
7.81
0.55
98.47
Acenaphthene
0.011
4
94
4.26
0.44
98.90
Cadmium (PMio)
0.00056
4
115
3.48
0.44
99.34
1,2 -Dibromoethane
0.0017
2
2
100.00
0.22
99.56
Benzo(a)pyrene
0.00057
1
57
1.75
0.11
99.67
Lead (PMio)
0.015
1
115
0.87
0.11
99.78
Manganese (PMio)
0.03
1
115
0.87
0.11
99.89
1,1,2-Trichloroethane
0.0625
1
1
100.00
0.11
100.00
Total
913
1,934
47.21
Observations from Table 20-2 include the following:
• Concentrations of 22 pollutants failed at least one screen for BTUT; approximately
47 percent of concentrations for these 22 pollutants were greater than their associated
risk screening value (or failed screens). BTUT has the second highest number of
individual pollutants failing screens.
• Concentrations of 11 pollutants contributed to 95 percent of failed screens for BTUT
and therefore were identified as pollutants of interest for this site. These 11 include
three carbonyl compounds, five VOCs, two PMio metals, and one PAH.
20-6
-------
• BTUT is one of only two NMP sites for which propionaldehyde is a pollutant of
interest (BROK is the other).
• Acetaldehyde, formaldehyde, and benzene were detected in every valid carbonyl
compound and VOC sample collected at BTUT and failed 100 percent of screens.
Other pollutants also failed 100 percent of screens but were detected less frequently.
• As described in Section 3.2, if a pollutant was measured by both the TO-15 and
SNMOC methods at the same site, the TO-15 results were used for the risk-based
screening process. As BTUT sampled both VOCs (TO-15) and SNMOCs, the TO-15
results were used for the 12 pollutants these methods have in common.
20.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Utah monitoring site. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of sampling.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at BTUT are provided
in Appendices J, K, M, N, and O.
20-7
-------
20.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for BTUT, as described in Section 3.1. The quarterly average concentration
of a particular pollutant is simply the average concentration of the preprocessed daily
measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the Utah monitoring site are presented in
Table 20-3, where applicable. Note that concentrations of the PAHs and PMio metals are
presented in ng/m3 in Table 20-3 for ease of viewing. Also note that if a pollutant was not
detected in a given calendar quarter, the quarterly average concentration simply reflects "0"
because only zeros substituted for non-detects were factored into the quarterly average
concentration.
20-8
-------
Table 20-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Utah Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
Oig/m3)
Annual
Average
Oig/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
frig/m3)
Q2
Avg
frig/m3)
Q3
Avg
frig/m3)
Q4
Avg
frig/m3)
Annual
Average
frig/m3)
Bountiful, Utah - BTUT
Acetaldehyde
54/54/54
3.90
±0.86
3.35
±0.79
NA
4.42
± 1.24
3.64
±0.48
59/59/59
3.28
±0.70
1.59
±0.43
2.75
±0.46
2.96
±0.54
2.62
±0.31
Benzene
50/50/50
1.26
±0.34
0.53
±0.08
0.58
±0.18
0.91
±0.40
NA
59/59/59
1.03
±0.29
0.46
±0.10
0.69
±0.18
0.72
±0.12
0.72
±0.10
1.3 -Butadiene
47/42/50
0.13
±0.06
0.04
±0.01
0.03
±0.01
0.09
±0.04
NA
58/30/59
0.10
±0.04
0.05
±0.01
0.06
±0.02
0.08
±0.02
0.07
±0.01
Carbon Tetrachloride
50/49/50
0.55
±0.11
0.60
±0.02
0.64
±0.05
0.58
±0.03
NA
59/59/59
0.63
±0.05
0.68
±0.04
0.40
±0.11
0.48
±0.06
0.56
±0.04
1,2 -Dichloroethane
47/44/50
0.10
±0.02
0.08
±0.02
0.05
±0.01
0.09
±0.01
NA
53/53/59
0.11
±0.01
0.10
±0.01
0.05
±0.02
0.07
±0.02
0.08
±0.01
Formaldehyde
54/54/54
8.64
± 1.97
9.43
±3.47
NA
9.54
±3.09
8.42
± 1.37
59/59/59
7.34
± 1.26
3.10
± 1.19
6.10
± 1.19
6.42
± 1.34
5.68
±0.72
Hexachloro-1,3 -butadiene
3/0/50
NR
NR
NR
NR
NR
14/1/59
0.08
±0.14
0.04
±0.03
0.02
±0.02
0.03
±0.02
0.04
±0.04
Propionaldehyde
54/54/54
0.70
±0.15
0.67
±0.16
NA
0.76
±0.20
0.67
±0.08
59/59/59
0.58
±0.11
0.34
±0.09
0.57
±0.10
0.54
±0.09
0.50
±0.05
Arsenic (PMi0)a
58/56/60
1.40
±0.64
0.45
±0.17
0.56
±0.12
0.39
±0.16
0.70
±0.19
55/55/55
0.86
±0.72
0.43
±0.14
0.54
±0.11
1.18
±0.57
0.77
±0.23
Naphthalene1
56/56/56
60.53
± 20.83
46.01
± 17.22
45.09
± 12.29
56.37
± 16.42
52.25
±8.19
61/61/61
44.79
±9.93
30.03
±7.95
44.45
±9.83
46.26
±9.35
41.44
±4.69
Nickel (PMi,;,)a
60/59/60
2.05
±0.58
1.26
±0.27
1.55
±0.35
1.25
±0.34
1.53
±0.21
55/55/55
1.08
±0.30
1.08
±0.26
1.32
±0.32
1.64
±0.30
1.30
±0.15
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating a quarterly and/or annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
-------
Observations for BTUT from Table 20-3 include the following:
• The pollutants of interest with the highest annual average concentrations for both
2015 and 2016 are formaldehyde and acetaldehyde. For both pollutants, the annual
average concentrations for 2015 are significantly higher than the annual averages for
2016. This is also true for propionaldehyde, the third carbonyl compound that is a
pollutant of interest for BTUT. Each of the available quarterly average concentrations
of these pollutants for 2015 are greater than the quarterly average concentrations for
2016.
• A review of the formaldehyde data shows that concentrations measured at BTUT
range from 0.843 |ig/m3 to 25.5 |ig/m3, with the nine highest formaldehyde
concentrations measured in 2015. All but two of the 19 formaldehyde concentrations
greater than 10 |ig/m3 measured at BTUT were measured in 2015, including the two
highest concentrations measured across the program. Similarly, 20 of the 25 highest
acetaldehyde concentrations (those greater than 4 |ig/m3) were measured at BTUT in
2015. The five propionaldehyde concentrations greater than 1 |ig/m3 measured at
BTUT were also measured in 2015.
• A number of invalid canister samples scattered throughout the year resulted in a VOC
completeness less than 85 percent for BTUT in 2015. As a result, annual average
concentrations could not be calculated for 2015. For 2016, the VOCs with the highest
annual average concentrations are benzene and carbon tetrachloride; the annual
averages for the remaining VOCs are all less than 0.1 |ig/m3.
• For both years, the first quarter average concentration of benzene is greater than
1 |ig/m3. Nine of the 12 benzene concentrations greater than 1 |ig/m3 measured at
BTUT in 2015 were measured during the first quarter of the year (with none
measured during the second quarter, one during the third, and two during the fourth).
Similarly, seven of the 11 benzene concentrations greater than 1 |ig/m3 measured at
BTUT in 2016 were measured during the first quarter of the year (with none
measured during the second quarter, two during the third, and two during the fourth).
• The third and fourth quarter average concentrations of carbon tetrachloride for 2016
are the lowest quarterly average concentrations of this pollutant among all NMP sites
sampling VOCs. BTUT has the highest number of carbon tetrachloride measurements
less than 0.4 |ig/m3 among all NMP sites (15), 13 of which were measured during the
second half of 2016. The other two were both measured during the first quarter of
2015, including the minimum concentration of carbon tetrachloride measured across
the program (0.0378 |ig/m3).
• Quarterly and annual averages for hexachloro-1,3-butadiene for 2015 are not
presented in Table 20-3 due to the use of a contaminated internal standard at the
laboratory for Method TO-15, which resulted in the invalidation of the results from
early March 2015 through mid-December 2015, as described in Section 2.4. For
2016, the first quarterly average concentration for 2016 (0.08 ± 0.14 |ig/m3) is twice
the next highest quarterly average and has an associated confidence interval greater
than the average itself. The maximum concentration of this pollutant among NMP
sites sampling VOCs was measured at BTUT (1.02 |ig/m3, with all other
20-10
-------
measurements program-wide 0.15 |ig/m3 or less). Concentrations of hexachloro-1,3-
butadiene measured at BTUT in 2016 range from 0.0748 |ig/m3 to 1.02 |ig/m3, with
45 non-detects.
• The first quarter average concentration of arsenic for 2015 is two to three times
higher than the other quarterly average concentrations for that year. Six of the seven
arsenic concentrations greater than 1 ng/m3 that were measured in 2015 were
measured in January or February. The first quarter also has the fewest arsenic
concentrations (3) less than 0.5 ng/m3 among the quarterly average concentrations for
2015 (with between six and nine measured during each of the remaining calendar
quarters). For 2016, the nine arsenic concentrations greater than 1 ng/m3 were
measured at BTUT during the first and fourth quarters of the year, including two
measurements greater than 4 ng/m3.
• The first quarter average concentration of nickel for 2015 (2.05 ± 0.58 ng/m3) is the
highest of the quarterly average concentrations for BTUT, both for 2015 and 2016. Of
the 11 nickel concentrations greater than 2 ng/m3 measured at BTUT in 2015, seven
were measured during the first quarter, while the remaining four were measured in
August. The six nickel concentrations greater than 2 ng/m3 measured at BTUT in
2016 were measured between the end of September and the end of the year. This is
reflected in the quarterly average concentrations of nickel for 2016.
• Of the 15 naphthalene concentrations greater than 75 ng/m3 measured at BTUT, 11
were measured in 2015, including all five concentrations greater than 100 ng/m3. The
annual average concentration of naphthalene for 2015 is higher than the annual
average for 2016, though the difference is not statistically significant. Note that the
confidence intervals calculated for the quarterly and annual average concentrations
for 2015 are larger than each of those calculated for 2016, indicating that naphthalene
measurements collected in 2015 exhibit more variability.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for BTUT from
those tables include the following:
• BTUT has the highest annual average concentration for hexachloro-1,3-butadiene
among NMP sites sampling this pollutant, as shown in Table 4-10. Note that the
confidence interval associated with BTUT's 2016 annual average concentration of
this pollutant is the same as the annual average itself, indicating the presence of
outliers and/or a high-level of variability. The maximum concentration measured at
BTUT is an order of magnitude higher than any other measurement of hexachloro-
1,3-butadiene measured across the program.
• For the fifth year in a row, BTUT has the highest annual average concentration of
formaldehyde among NMP sites sampling carbonyl compounds, as shown in
Table 4-11. BTUT's annual averages of formaldehyde rank highest (2015) and third
highest (2016), despite the statistically significant difference shown between the two.
BTUT's annual average concentrations of acetaldehyde both rank among the five
highest annual averages.
20-11
-------
• BTUT does not appear in Table 4-12 for PAHs or Table 4-13 for metals.
20.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 20-3 for BTUT. Figures 20-3 through 20-13 overlay the site's minimum, annual average,
and maximum concentrations for each year onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations, as described in Section 3.4.2.1,
and are discussed below. If an annual average concentration could not be calculated, the range of
concentrations are still provided in the figures that follow.
Figure 20-3. Program vs. Site-Specific Average Acetaldehyde Concentrations
g%
I
¦5
0 2 4 6 8 10 12 14 16 18
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
a
o
Figure 20-3 presents the box plot for acetaldehyde for BTUT and shows the following:
• The range of acetaldehyde concentrations measured BTUT in 2015 is larger than the
range of concentrations measured in 2016.
• Despite the differences in the magnitude of measurements across the two years of
sampling, both annual average acetaldehyde concentrations for BTUT are greater
than the program-level average concentration (1.67 |ig/m3), as well as the program-
level third quartile. The annual average concentrations of acetaldehyde for BTUT are
the second (2015) and fifth (2016) highest among NMP sites sampling carbonyl
compounds.
20-12
-------
Figure 20-4. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
012345678
Concentration (ng/m3)
Program: IstQuartile
2ndQuartile 3rdQuartile 4thQuartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
9
o —
Figure 20-4 presents the box plot for arsenic (PMio) for BTUT and shows the following:
• Although BTUT's maximum arsenic concentration is not the maximum arsenic
concentration measured at the program-level, two of the eight arsenic concentrations
greater than 4 ng/m3 measured across the program were measured at BTUT (both in
2016).
• Two of the five non-detects of arsenic measured across the program were measured at
BTUT, both in 2015.
• BTUT's annual average concentration of arsenic for 2015 is similar to the program-
level average concentration, while the annual average for 2016 is just slightly higher.
Figure 20-5. Program vs. Site-Specific Average Benzene Concentrations
m
012345678
Concentration (ng/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 20-5 presents the box plot for benzene for BTUT and shows the following:
• An annual average concentration for 2015 could not be calculated, although the range
of measurements is provided in Figure 20-5.
• Benzene concentrations greater than 3 |ig/m3 were not measured at BTUT in 2015 or
2016.
20-13
-------
• The range of benzene concentrations measured BTUT in 2015 is larger than the range
of concentrations measured at this site in 2016.
• The minimum benzene concentration measured at BTUT in 2015 (0.144 |ig/m3) is not
the minimum concentration measured across the program (0.109 |ig/m3), although it
is among the lowest.
• BTUT's annual average concentration of benzene for 2016 is similar to the program-
level average concentration (0.72 |ig/m3).
Figure 20-6. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
1
m
Program MaxConcentration = 3.90 (.ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 20-6 presents the box plot for 1,3-butadiene for BTUT and shows the following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plot because the scale of the box plot would be too large to readily
observe data points at the lower end of the concentration range. Thus, the scale of the
box plot has been reduced.
• An annual average concentration for 2015 could not be calculated, although the range
of measurements is provided in Figure 20-6.
• The range of 1,3-butadiene concentrations measured at BTUT in 2015 is twice the
range of concentrations measured in 2016. However, if the maximum concentration
measured in 2015 was excluded, the range of concentrations measured each year
would be similar.
• BTUT's annual average concentration of 1,3-butadiene for 2016 falls between the
program-level median (0.06 |ig/m3) and program-level average (0.09 |ig/m3)
concentrations.
• Four non-detects were measured at BTUT, three in 2015 and one in 2016.
20-14
-------
Figure 20-7. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
BTUT
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 20-7 presents the box plot for carbon tetrachloride for BTUT and shows the
following:
• As with other VOCs, an annual average concentration for 2015 could not be
calculated, although the range of measurements is provided in Figure 20-7.
• Carbon tetrachloride concentrations measured at BTUT are all less than 0.9 |ig/m3.
• The minimum carbon tetrachloride concentration measured across the program
(0.0378 |ig/m3) was measured at BTUT in 2015. BTUT is the only site for which a
concentration less than 0.1 |ig/m3 was measured.
• BTUT's annual average concentration of carbon tetrachloride for 2016 is less than the
program-level first quartile and is the lowest annual average concentration among
NMP sites sampling this pollutant, although the range of annual averages is relatively
small for most of the sites.
Figure 20-8. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1
"
1 "J
! Program Max Concentration = 45.8 }ig/m3 ¦
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (jug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
o
20-15
-------
Figure 20-8 presents the box plot for 1,2-dichloroethane for BTUT and shows the
following:
• The scale of the box plot in Figure 20-8 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• As with other VOCs, an annual average concentration for 2015 could not be
calculated, although the range of measurements is provided in Figure 20-8.
• All of the concentrations of 1,2-dichloroethane measured at BTUT are less than the
program-level average concentration of 0.30 |ig/m3, which is being driven by the
measurements at the upper end of the concentration range.
• BTUT's annual average concentration for 2016 is similar to the program-level
median concentration.
• Nine non-detects of 1,2-dichloroethane were measured at BTUT, three in 2015 and
six in 2016.
Figure 20-9. Program vs. Site-Specific Average Formaldehyde Concentrations
BTUT I |
& * . . _ _ . .
0 3 6 9 12 15 18 21 24 27
Concentration (jug/m3)
Program: IstQuartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
0
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 20-9 presents the box plot for formaldehyde for BTUT and shows the following:
• The two highest formaldehyde concentrations measured at BTUT (25.5 |ig/m3 and
25.4 |ig/m3) are maximum concentrations of formaldehyde measured across the
program. These two concentrations were measured in 2015; the maximum
concentration measured at BTUT in 2016 is less than half the magnitude of those
measured in 2015.
• Both of BTUT's annual average concentrations of formaldehyde are greater than the
program-level average concentration (3.05 |ig/m3), even though the annual average
for 2016 is one-third less than the annual average for 2015. As discussed in the
previous section, BTUT's annual average concentrations of formaldehyde are the
highest and third-highest annual averages of formaldehyde among NMP sites
sampling carbonyl compounds.
20-16
-------
Figure 20-10. Program vs. Site-Specific Average Hexachloro-l,3-butadiene Concentrations
Program MaxConcentration = 1.02 (.ig/m3
'
! BTUT 2016 Max Concentration = 1.02 (.ig/m3 ]
0.00 0.05 0.10 0.15 0.20 0.25
Concentration (jug/m3)
Program: IstQuartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 20-10 presents the box plot for hexachloro-1,3-butadiene for BTUT and shows the
following:
• The program-level maximum concentration of hexachloro-1,3-butadiene (1.02 |ig/m3)
is not shown directly on the box plot as the scale has been reduced to 0.25 |ig/m3 to
allow for the observations data points at the lower end of the concentration range.
• The program-level first, second, and third quartiles are all zero for this pollutant,
indicating that at least 75 percent of the measurements across the program are non-
detects and thus, are not visible on the box plot.
• Annual average concentrations for hexachloro-1,3-butadiene were not calculated for
2015 due to the use of a contaminated internal standard at the laboratory for Method
TO-15, which resulted in the invalidation of the results from early March 2015
through mid-December 2015, as described above and in Section 2.4.
• The maximum concentration of hexachloro-l,3-butadiene measured across the
program was measured at BTUT and is the only measurement of this pollutant across
the program greater than the MDL. All other concentrations of this pollutant
measured at BTUT are 0.15 |ig/m3 or less.
• Fifty-six non-detects of hexachloro-1,3-butadiene were measured at BTUT.
• BTUT's annual average concentration of hexachloro-1,3-butadiene for 2016 is nearly
two and half times greater than the program-level average concentration
(0.017 |ig/m3) and is the highest annual average concentration among NMP sites
sampling this pollutant. This is mostly attributable to the maximum concentration
measured; if the maximum concentration was excluded from the dataset, BTUT's
annual average for 2016 would decrease by almost half, but would still be greater
than the program-level average concentration.
20-17
-------
Figure 20-11. Program vs. Site-Specific Average Naphthalene Concentrations
50
100
150 200 250 300
Concentration (ng/m3)
350
400
450
Program: IstQuartile 2nd Quartile 3rd Quartile 4thQuartile Ave rag
¦ ~ ~ ~
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o
Figure 20-11 presents the box plot for naphthalene for BTUT and shows the following:
• The maximum concentration of naphthalene measured at BTUT (136 ng/m3) is
considerably less than the maximum concentration measured across the program
(403 ng/m3).
• The annual average concentrations of naphthalene for BTUT fall on either side of the
program4evel median concentration (48.90 ng/m3).
Figure 20-12. Program vs. Site-Specific Average Nickel (PMio) Concentrations
1
h?—•
Program Max Concentration
69.5 ng/m3
9 12 15
Concentration (ng/m3)
Program: IstQuartile
¦
Site: 2015 Average
2ndQuartile 3rdQuartile 4thQuartile Averag
~ ~ ~ I
2016 Average Concentration Range, 2015 & 2016
o —
Figure 20-12 presents the box plot for nickel (PMio) for BTUT and shows the following:
• The program-level maximum concentration of nickel (69.5 ng/m3) is not shown
directly on the box plot as the scale has also been reduced to allow for the
observations data points at the lower end of the concentration range.
• The three highest nickel concentrations measured at BTUT were measured in 2015.
• Both annual average concentrations of nickel for BTUT are greater than the program-
level average concentration (1.09 ng/m3); the annual average for 2015 is also greater
than the program-level third quartile (1.30 ng/m3) while the annual average for 2016
is similar to the third quartile.
20-18
-------
Figure 20-13. Program vs. Site-Specific Average Propionaldehyde Concentrations
BTUT
0 1 2 3 4 5 6
Concentration (ng/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 20-13 presents the box plot for propionaldehyde for BTUT and shows the
following:
• Although the maximum concentration of propionaldehyde was not measured at
BTUT, concentrations greater than the maximum concentration measured at BTUT
were measured at only two NMP sites (BROK, where the maximum concentration
was measured, and SPIL).
• Both annual average concentrations of propionaldehyde for BTUT are greater than
the program-level average concentration (0.32 |ig/m3) and the third quartile, with
BTUT's annual average for 2015 more than twice the program-level average.
20.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
BTUT has sampled carbonyl compounds, VOCs, metals, and SNMOCs under the NMP since
2003 and PAHs since 2008. Thus, Figures 20-14 through 20-24 present the 1-year statistical
metrics for each of the pollutants of interest for BTUT. The statistical metrics presented for
assessing trends include the substitution of zeros for non-detects. If sampling began mid-year, a
minimum of 6 months of sampling is required for inclusion in the trends analysis; in these cases,
a 1-year average concentration is not provided, although the range and percentiles are still
presented.
20-19
-------
Figure 20-14. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
BTUT
35.0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Observations from Figure 20-14 for acetaldehyde concentrations measured at BTUT
include the following:
• Sampling for carbonyl compounds under the NMP began at BTUT in late July 2003.
Because this represents less than half of the sampling year, Figure 20-14 excludes
data from 2003.
• The maximum acetaldehyde concentration was measured in 2004 (32.7 |ig/m3).
Concentrations of acetaldehyde greater than 10 |ig/m3 were also measured at BTUT
in 2008 (20.0 |ig/m3), 2007 (15.3 |ig/m3), another in 2004 (10.8 |ig/m3), and one in
2015 (10.1 |ig/m3).
• After 2005, the 1-year average concentration exhibits a steady decreasing trend
through 2009, when the 1-year average concentration reaches a minimum
(1.97 |ig/m3), although the most significant changes occurred between 2005 and
2007. Between 2007 and 2011, the 1-year average concentration varied by less than
0.3 |ig/m3, ranging from 1.97 |ig/m3 (2009) to 2.25 |ig/m3 (2010).
• Although the range of concentrations measured in 2012 is smaller than the range
measured in 2011, a slight increase is shown in both the 1-year average and median
concentrations for 2012. The slight increase for 2012 is followed by a significant
increase for 2013, with both the 1-year average and median concentrations at a
maximum for the period of sampling. The number of acetaldehyde concentrations
greater than 4 |ig/m3 increased from seven in 2011 to 11 in 2012 and 32 in 2013.
20-20
-------
Additionally, 11 concentrations measured in 2012 are less than the minimum
concentration measured in 2013.
• The significant increase shown by the central tendency statistics for 2013 is followed
by a significant decrease in these same parameters for 2014. The increases shown for
2015 are followed by significant decreases for 2016, with the 1-year average
concentration nearly returning to 2012 levels.
Figure 20-15. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
BTUT
Maximum Concentration
for 2004 is 32.99 ng/m3
12.0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Observations from Figure 20-15 for arsenic concentrations measured at BTUT include
the following:
• Sampling for PMio metals under the NMP began at BTUT in late July 2003. Because
this represents less than half of the sampling year, Figure 20-15 excludes data from
2003.
• The maximum arsenic concentration was measured at BTUT in 2004 (32.99 ng/m3)
and is nearly twice the next highest concentration (16.84 ng/m3), also measured in
2004. Eight of the 15 highest concentrations of arsenic (those greater than 5 ng/m3)
were measured at BTUT in 2004.
• Of the 44 arsenic concentrations greater than 3 ng/m3 measured at BTUT, 40 were
measured during the colder months of the year, with 21 measured during the first
quarter of the calendar year and 19 measured during the fourth quarter of the calendar
year, suggesting a seasonality in the measurements.
20-21
-------
• The average concentration of arsenic decreased significantly from 2004 to 2005, with
the 1-year average decreasing from 2.79 ng/m3 to 0.96 ng/m3. Between 2006 and
2010, there is an undulating pattern in the 1-year average concentrations, with years
with higher concentrations followed by years with lower concentrations. During this
period, the 1-year average arsenic concentration fluctuated between 0.61 ng/m3
(2010) and 1.13 ng/m3 (2009). However, the 1-year average concentrations for 2007
and 2009 are being driven primarily by a single "high" measurement. If the maximum
concentrations measured in 2007 and 2009 were removed from the data sets, the
1-year average concentrations for this period would all be less than 1 ng/m3.
• Little change in the arsenic concentrations is shown between 2010 and 2011. The
smallest range of arsenic concentrations was measured at BTUT in 2012, when all
arsenic concentrations measured at BTUT were less than 2 ng/m3. The 1-year average
concentration, along with the 95th percentile and maximum concentration, are at a
minimum for 2012.
• Concentrations of arsenic measured at BTUT increased significantly for 2013, as
indicated by the increase shown in all of the statistical parameters. Although the
1-year average concentration doubled from 2012 to 2013, the increase in the median
concentration is less dramatic.
• With the exception of the median concentration, each of the statistical parameters
shown for 2014 exhibits a decrease from the previous year. Little change is shown in
most of the statistical parameters after 2014.
Figure 20-16. Yearly Statistical Metrics for Benzene Concentrations Measured at BTUT
rh
i
©¦
&
T
_o
•o-
L2-
&
£
i
LjJ
A
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 20151 2016
Year
1 A 1-year average is not presented due to low method completeness in 2015.
20-22
-------
Observations from Figure 20-16 for benzene concentrations measured at BTUT include
the following:
• Sampling for VOCs under the NMP also began at BTUT in late July 2003. Because
this represents less than half of the sampling year, Figure 20-16 excludes data from
2003.
• A 1-year average concentration is not presented for 2015 due to low method
completeness, as described in the previous sections.
• The maximum concentration of benzene shown was measured in 2009 (8.16 |ig/m3).
The next highest concentration (6.56 |ig/m3) was also measured in 2009, although
concentrations greater than 6 |ig/m3 were also measured in 2005 and 2007. Benzene
concentrations greater than 4 |ig/m3 have not been measured at BTUT in recent years.
• Concentrations of benzene appear to be higher during the colder months of the year,
as 52 of the 56 highest concentrations (those greater than 2.50 |ig/m3) were measured
during the first (29) or fourth (23) quarters of the calendar year.
• The 1-year average and median benzene concentrations have a decreasing trend
through 2007. An increasing trend in the 1-year average concentration is then shown
through 2009, after which another decreasing trend follows. The 1-year average
benzene concentration is at a minimum for 2016 (0.72 |ig/m3), although the 1-year
average for 2014 is similar. The median concentration is also at a minimum for 2016
(0.62 |ig/m3), though it does not always follow the same pattern as the 1-year average
concentration.
20-23
-------
Figure 20-17. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
BTUT
O
T
T
i-j-i
T
2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Pe r ce nti le Ave r z
1 A 1-year average is not presented due to low method completeness in 2015.
Observations from Figure 20-17 for 1,3-butadiene concentrations measured at BTUT
include the following:
• The maximum concentration of 1,3-butadiene shown was measured in 2005
(0.75 |ig/m3). The second highest concentration was also measured in 2005
(0.53 |ig/m3), although a similar concentration was also measured in 2006. These are
the only concentrations of 1,3-butadiene greater than 0.5 |ig/m3 measured at BTUT.
• The minimum, 5th percentile, and median concentrations are all zero for 2004,
indicating that at least half of the measurements were non-detects. The detection rate
of 1,3-butadiene increased after 2004, as indicated by the increase in the median
concentrations for 2005 and 2006, the 5th percentile for 2007, and the minimum
concentration for 2008 and 2009. The percentage of non-detects decreased from
75 percent for 2004 to 0 percent for 2008 and 2009. The percentage of non-detects
increased to 7 percent for 2010 and 18 percent for 2011, explaining why the 5th
percentile returned to zero. Five or fewer non-detects of this pollutant were measured
in each year between 2012 and 2016.
• The 1-year average concentration increased from 0.061 |ig/m3 for 2004 to
0.104 |ig/m3 for 2005. This increase is likely due to the decrease in non-detects (and
thus zeros substituted for them) as well as the higher concentrations measured in
2005, as discussed above. Between 2005 and 2012, the 1-year average concentration
hovers around 0.1 |ig/m3, ranging from 0.099 |ig/m3 (2011) to 0.117 |ig/m3 (2012).
20-24
-------
The median concentration varies more, ranging from 0.044 |ig/m3 (2005) to
0.089 |ig/m3 (2006), although the median concentration varies less after these two
years.
• A decreasing trend in the 1,3-butadiene concentrations measured at BTUT is shown
between 2012 and 2014, the first year that concentrations greater than 0.3 |ig/m3 were
not measured. Although a few "higher" concentrations were measured in 2015, the
median concentration (0.050 |ig/m3) is at its lowest since 2005. Concentrations
greater than 0.3 |ig/m3 were again not measured in 2016, and the 1-year average
concentration, the 95th percentile, and the maximum concentration are at a minimum.
Figure 20-18. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at BTUT
o
X
i
r-5-,
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
— Maximum O 95th Percentile Averc
1 A 1-year average is not presented due to low method completeness in 2015.
Observations from Figure 20-18 for carbon tetrachloride concentrations measured at
BTUT include the following:
• Non-detects of carbon tetrachloride were measured only in 2004 (nine) and 2005
(five). Concentrations of carbon tetrachloride greater than 1 |ig/m3 were measured in
2006 (two), 2008 (three), and 2011 (one).
• A significant increasing trend is shown in the 1-year average concentrations between
2004 and 2008, with the exception of 2007. The range and magnitude of
concentrations measured decreased substantially for 2007, which is reflected in the
dip in the 1-year average concentration. Concentrations greater than 0.8 |ig/m3 were
20-25
-------
not measured at BTUT in 2007, although concentrations of this magnitude account
for more than 20 percent of measurements for 2006 and 2008.
• After decreasing between 2008 and 2010, an increasing trend in the carbon
tetrachloride measurements is shown through 2012. Several of the statistical
parameters, including the 1-year average and median concentrations, are at a
maximum in 2012.
• A significant decrease in the 1-year average concentration, and the other statistical
parameters, is shown for 2013. This year has the lowest maximum concentration
since 2007 and the lowest minimum concentration since 2006.
• Each of the statistical parameters exhibits at least a slight increase for 2014, with the
exception of the 5th percentile, which did not change. The median concentration
changed little for 2015, despite variations at the upper and lower ends of the
concentration range for 2015.
• The 5th percentile for 2016 is at its lowest since 2005 (when the 5th percentile was
zero due to non-detects). A review of the data shows that the number of carbon
tetrachloride concentrations less than 0.4 |ig/m3 measured in 2016 (13) is at its
highest in more than 10 years, accounting for 22 percent of measurements in 2016
(with concentrations of this magnitude accounting for fewer than 10 percent of
measurements for most years since 2006).
Figure 20-19. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at BTUT
I
o
1
T
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
1 A 1-year average is not presented due to low method completeness in 2015.
20-26
-------
Observations from Figure 20-19 for 1,2-dichloroethane concentrations measured at
BTUT include the following:
• For the first several years of sampling, all of the statistical parameters shown are zero,
indicating that 1,2-dichloroethane was not detected. Between 2004 and 2008, there
was a single measured detection of 1,2-dichloroethane, which was measured in 2007.
Beginning with 2009, the number of measured detections began to increase; there
were two measured detections in 2009, seven in 2010, 15 in 2011, and 47 in 2012, the
first year with a median concentration greater than zero (indicating that there were
more measured detections than non-detects for the first time since the onset of
sampling). Between 2012 and 2016, the percentage of measured detections ranges
from 71 percent (2013) to 98 percent (2014).
• The range of concentrations measured in 2013 is considerably larger than the range of
concentrations measured in previous years, as the 1-year average concentration for
2013 is similar to the 95th percentile shown for previous years. All seven
1,2-dichloroethane concentrations greater than 0.25 |ig/m3 measured at BTUT were
measured in 2013. Concentrations measured in 2013 account for more than
40 percent of the 29 1,2-dichloroethane concentrations greater than 0.15 |ig/m3
measured at BTUT since the onset of sampling.
• Little change is shown in the central tendency parameters between 2013 and 2014,
despite the differences shown in the concentration profiles. 2014 is also the first year
the 5th percentile is greater than zero, indicating that non-detects accounted for less
than 5 percent of the measurements; a single non-detect was measured in 2014.
• Most of the statistical parameters exhibit decreases for 2015. Additional decreases are
shown for some of the parameters for 2016, as the 5th percentile returns to zero.
Although a 1-year average concentration is not available for 2015, median
concentrations are available for both years. The median concentrations for both years
are identical.
20-27
-------
Figure 20-20. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
BTUT
50.0
45.0
40.0
35.0
sr 30.0
E
o
U 20.0
15.0
10.0
5.0
0.0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
Observations from Figure 20-20 for formaldehyde concentrations measured at BTUT
include the following:
• The maximum formaldehyde concentration (45.4 |ig/m3) was measured on
August 31, 2004, on the same day that the highest acetaldehyde concentration was
measured. This measurement is nearly twice the next highest concentration
(25.5 |ig/m3), measured in 2015. Concentrations greater than 15 |ig/m3 were
measured several times each year between 2004 and 2007, as well as in 2011 and
2015.
• Although the maximum concentration decreased significantly from 2004 to 2005, the
other statistical metrics exhibit increases for 2005. The median increased by nearly
2 |ig/m3 from 2004 to 2005, indicating that concentrations were higher in 2005 than
2004 (as opposed to being driven by an outlier, as in 2004). To illustrate, the number
of concentrations greater than 5 |ig/m3 increased from 11 measured in 2004 to 31
measured in 2005.
• After 2005, concentrations of formaldehyde measured at BTUT have a decreasing
trend, with the 1-year average concentration decreasing from 6.21 |ig/m3 for 2005 to
2.44 |ig/m3 for 2008. In 2008, 95 percent of the concentrations measured were less
than 4 |ig/m3, which is less than the 1-year average and/or median concentrations for
several of the previous years. After 2008, a steady increasing trend is shown in the
1-year average formaldehyde concentrations, as well as most other statistical
parameters, through 2011.
T
X
-
T
X
rh
T
A
©¦
i&
O-
%/
'-2-
"""2
s
3- E
ki"'
o
~
^-2-
LJ
L?
'-2r
20-28
-------
• Although little change is shown in the 1-year average concentration between 2011
and 2012 and the range of concentrations measured is smaller for 2012, the median
concentration exhibits an increase. The decrease in the concentrations at the upper
end of the range from 2011 to 2012 is balanced by a higher number of measurements
at the mid-to-upper part of the range. The number of measurements greater than
10 |ig/m3 decreased from nine to one from 2011 to 2012 while the number of
measurements between 5 |ig/m3 and 10 |ig/m3 increased from six to 14 during the
same period. Also, six concentrations measured in 2011 are less than the minimum
concentration measured in 2012.
• Significant increases are shown for the central tendency parameters for 2013, with all
of the statistical parameters exhibiting increases. The 1-year average concentration
nearly doubled and the median concentration increased by 160 percent from 2012.
The number of formaldehyde concentrations greater than 10 |ig/m3 increased from
one in 2012 to 16 in 2013, with concentrations greater than 5 |ig/m3 accounting for
more than 75 percent of the measurements in 2013. This is also the only year for
which a formaldehyde concentration less than 2 |ig/m3 was not measured.
• Significant decreases are shown for 2014, which is followed by significant increases
again for 2015, when the second highest formaldehyde concentration was measured
and when the 1-year average concentration is at a maximum (8.42 |ig/m3). The
undulating pattern continues into 2016. The central tendency parameters calculated
for the years between 2013 and 2016, though variable, are greater than most other
years of sampling.
Figure 20-21. Yearly Statistical Metrics for Hexachloro-l,3-butadiene Concentrations
Measured at BTUT
I
rh
O
O 5th Percentile
O 95th Pe r ce nti le Ave r z
1 A 1-year average is not presented due to a laboratory contamination issue affecting numerous samples.
20-29
-------
Observations from Figure 20-21 for hexachloro-l,3-butadiene concentrations measured at
BTUT include the following:
• The use of a contaminated internal standard at the laboratory for Method TO-15
resulted in the invalidation of the hexachloro-l,3-butadiene measurements from early
March 2015 through mid-December 2015, as described in Section 2.4.
• The maximum hexachloro-1,3-butadiene concentration was measured at BTUT in
2016 (1.02 |ig/m3). Two additional concentrations greater than 0.25 |ig/m3 have been
measured at BTUT since the onset of sampling (0.36 |ig/m3 in 2010 and 0.27 |ig/m3
in 2005).
• The median concentration of hexachloro-l,3-butadiene is zero for all years of
sampling except 2014, indicating that at least half of the measurements were non-
detects each year. There were no measured detections of this pollutant in 2004, 2007,
and 2008 and there were five or fewer measured detections in 2006, each year
between 2009 and 2013, and 2015 (as shown in Table 20-3). Hexachloro-1,3-
butadiene was detected 12 times in 2005, 29 times in 2014, and 14 times in 2016.
• The magnitude of the maximum hexachloro-1,3-butadiene concentration measured in
2016 is balanced by the higher number of non-detects (or zeros) measured that year,
such that the 1-year average concentrations for 2014 (0.045 |ig/m3) and 2016
(0.040 |ig/m3) are fairly similar to each other.
Figure 20-22. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
BTUT
5th Percentile
O 95th Percentile
20-30
-------
Observations from Figure 20-22 for naphthalene concentrations measured at BTUT
include the following:
• Although PAH sampling began at BTUT in April 2008, complications with the
collection system lead to a 6-month lapse in sampling until mid-October. Thus,
Figure 20-22 begins with 2009.
• The maximum naphthalene concentration (421 ng/m3) was measured in 2009. The
second highest naphthalene concentration (242 ng/m3), measured in 2013, is the only
other naphthalene measurement greater than 200 ng/m3 measured at BTUT since the
onset of PAH sampling.
• A decreasing trend in naphthalene concentrations measured at BTUT is shown
through 2011. Although little change in the range of measurements or the 1-year
average concentration shown for 2012, the median concentration exhibits an increase.
The biggest change in concentrations between the two years occurs in the middle of
the concentration range. The number of naphthalene concentrations measured at
BTUT between 50 ng/m3 and 75 ng/m3 increased from 11 to 20 from 2011 to 2012.
• After 2012, years when the majority of naphthalene concentrations fell into a wider
range alternate with years when most naphthalene concentrations fell into a tighter
range.
• Most of the statistical parameters are at a minimum for 2016, including the 1-year
average concentration (41.44 ng/m3). The maximum concentration is less than
100 ng/m3 for the first time in 2016.
• Concentrations of naphthalene exhibit seasonality. Of the 51 naphthalene
concentrations greater than 100 ng/m3 measured at BTUT since 2009, all but four
were measured during the first or fourth quarters of any given year, or the colder
months of the year, with the majority measured in January (17), November (11), or
December (16). Only one naphthalene concentration greater than 100 ng/m3 has been
measured at BTUT between April and August.
20-31
-------
Figure 20-23. Yearly Statistical Metrics for Nickel (PMio) Concentrations Measured at
BTUT
O 5th Percentile
O 95th Percentile
Observations from Figure 20-23 for nickel concentrations measured at BTUT include the
following:
• The maximum nickel concentration was measured in 2005 (29.6 ng/m3), although a
similar concentration was also measured in 2007. Two additional nickel
concentrations greater than 20 ng/m3 were measured in 2008. Additional nickel
concentrations greater than 10 ng/m3 have not been measured at BTUT.
• All 24 non-detects of nickel were measured in 2009 and, with one exception, were
measured on consecutive sample days between June and October.
• The range of nickel concentrations measured each year is highly variable, particularly
through 2010. Concentrations measured over a given year have spanned a little as
2.57 ng/m3 (2010) or to nearly 30 ng/m3 (2005). This variability is reflected in the
undulating pattern shown in the central tendency statistics, particularly in the years
between 2004 and 2011. During this time period, the 1-year average concentrations
ranged from 0.75 ng/m3 (2009) to 4.05 ng/m3 (2005).
• The concentrations measured between 2012 and 2016 exhibit less variability then the
preceding years. The 1-year average concentrations calculated for each year during
this period vary by less than 0.25 ng/m3.
20-32
-------
Figure 20-24. Yearly Statistical Metrics for Propionaldehyde Concentrations Measured at
BTUT
4.0
3.0
2.5
E
3
c
~ 2.0
C
m
c
O
u
1.5
X
1
0.5
o
T
4
T
•A
rZ,
0_
I
LjJ
LjJ
n
o
i
O
V
£
LjJ
£
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
Observations from Figure 20-24 for propionaldehyde concentrations measured at BTUT
include the following:
• The maximum propionaldehyde concentration (3.38 |ig/m3) was measured in 2004,
although a similar concentration (3.36 |ig/m3) was measured in 2007. The maximum
propionaldehyde concentration was measured on August 31, 2004, the same day the
highest acetaldehyde and formaldehyde concentrations were measured. An additional
propionaldehyde concentration greater than 2 |ig/m3 was measured each year between
2004 and 2006.
• Although the range of propionaldehyde concentrations decreased somewhat from
2004 to 2005, both central tendency parameters exhibit an increase for 2005; the
median concentration more than doubled during this time. The number of
propionaldehyde concentrations greater than 0.5 |ig/m3 increased considerably from
2004 to 2005, increasing from nine measured in 2004 to 33 (accounting for more than
half of the measurements collected in in 2005). A few additional propionaldehyde
concentrations greater than 0.5 |ig/m3 were measured in 2006 (35).
• A significant decreasing trend in propionaldehyde concentrations is shown after 2006
through 2009, when the smallest range of concentrations was measured, although
most of change occurred between 2006 and 2007.
20-33
-------
• The range of concentrations measured nearly doubled from 2009 to 2010, with both
central tendency parameters increasing significantly. Eight concentrations measured
in 2010 were greater than the maximum propionaldehyde concentration measured in
2009; further, fewer concentrations less than 0.5 |ig/m3 were measured in 2009 (42)
compared to 2010 (30). Although the range of concentrations measured changed little
between 2010 and 2011, the number of concentrations less than 0.5 |ig/m3 returned to
2009 levels (43).
• A significant increasing trend in propionaldehyde concentrations is shown between
2011 and 2013. Both central tendency parameters are at a maximum for 2013. The
number of propionaldehyde concentrations greater than 1 |ig/m3 is at its highest for
2013 (24); with the exceptions of 2005 and 2006, during which 13 concentrations
greater than 1 |ig/m3 were measured, five or fewer were measured during each of the
other years of sampling. Each of the carbonyl compounds exhibits an increase for
2013, as shown in Figures 20-14 and 20-20. A return to 2012 levels is shown for
2014.
• Years with "higher" concentrations continue to alternate with years with "lower"
concentrations through 2016, though a return to 2013 levels has not occurred.
20.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the BTUT monitoring site. Refer to Sections 3.2, 3.4.2.3, and
3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
20.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for BTUT, risk was examined by calculating cancer risk and
noncancer hazard approximations for each year annual average concentrations could be
calculated. These approximations can be used as risk estimates for cancer and noncancer effects
attributable to the pollutants of interest. Although the use of these approximations is limited, they
may help identify where policy-makers want to shift their air monitoring priorities. Refer to
Section 3.4.2.3 for an explanation of how cancer risk and noncancer hazard approximations are
calculated and what limitations are associated with them. Annual averages, cancer UREs and/or
noncancer RfCs, and cancer risk and noncancer hazard approximations are presented in
Table 20-4, where applicable. Cancer risk approximations are presented as probabilities while
the noncancer hazard approximations are ratios and thus, unitless values.
20-34
-------
Table 20-4. Risk Approximations for the Utah Monitoring Site
2015
2016
# of
Measured
Risk Approximations
# of
Measured
Risk Approximations
Cancer
Noncancer
Detections
Annual
Detections
Annual
URE
RfC
vs. # of
Average
Cancer
Noncancer
vs. # of
Average
Cancer
Noncancer
Pollutant
(jig/m3)1
(mg/m3)
Samples
Oig/m3)
(in-a-million)
(HQ)
Samples
frig/m3)
(in-a-million)
(HQ)
Bountiful, Utah - BTUT
3.64
2.62
Acetaldehyde
0.0000022
0.009
54/54
±0.48
8.01
0.40
59/59
±0.31
5.76
0.29
0.72
Benzene
0.0000078
0.03
50/50
NA
NA
NA
59/59
±0.10
5.64
0.02
0.07
1.3 -Butadiene
0.00003
0.002
47/50
NA
NA
NA
58/59
±0.01
2.13
0.04
0.56
Carbon Tetrachloride
0.000006
0.1
50/50
NA
NA
NA
59/59
±0.04
3.33
0.01
0.08
1,2 -Dichloroethane
0.000026
2.4
47/50
NA
NA
NA
53/59
±0.01
2.19
<0.01
8.42
5.68
Formaldehyde
0.000013
0.0098
54/54
± 1.37
109.51
0.86
59/59
±0.72
73.79
0.58
0.04
Hexachloro-1,3 -butadiene
0.000022
0.09
3/50
NR
NR
NR
14/59
±0.04
0.89
<0.01
0.67
0.50
Propionaldehyde
--
0.008
54/54
±0.08
--
0.08
59/59
±0.05
—
0.06
0.70
0.77
Arsenic (PMi0)a
0.0043
0.000015
58/60
±0.19
3.01
0.05
55/55
±0.23
3.30
0.05
52.25
41.44
Naphthalene1
0.000034
0.003
56/56
±8.19
1.78
0.02
61/61
±4.69
1.41
0.01
1.53
1.30
Nickel (PMi,;,)a
0.00048
0.00009
60/60
±0.21
0.73
0.02
55/55
±0.15
0.62
0.01
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
NA = Not available because the criteria for calculating an annual average were not met.
NR = Not reportable due to invalidation related to a contaminated internal standard.
— = A Cancer URE or Noncancer RfC is not available.
-------
Observations for BTUT from Table 20-4 include the following:
• Formaldehyde is the pollutant with the highest annual average concentrations for both
years of sampling.
• Formaldehyde is also the pollutant with the highest cancer risk approximations for
BTUT. Formaldehyde's cancer risk approximation for 2015 is 109.51 in-a-million,
which is the highest cancer risk approximation across the program, and the only
cancer risk approximation greater than 100 in-a-million. The remaining cancer risk
approximations calculated for BTUT are all less than 10 in-a-million and most are
less than 5 in-a-million.
• There were no pollutants of interest with noncancer hazard approximations greater
than 1.0, indicating that no adverse noncancer health effects are expected from these
individual pollutants. The highest noncancer hazard approximation was calculated for
formaldehyde (0.86, 2015), which is also the highest noncancer hazard approximation
calculated among the site-specific pollutants of interest with noncancer toxicity
factors.
As an extension of this analysis, pollution roses were created for each of the site-specific
pollutants of interest that have a cancer risk approximation greater than 75 in-a-million and/or a
noncancer hazard approximation greater than 1.0, where applicable. Thus, a pollution rose was
created for BTUT's formaldehyde measurements. A pollution rose is a plot of the ambient
concentration versus the wind direction; the magnitude of the concentration is indicated using
different colored dots and are shown in relation to the average wind direction oriented about a
16-point compass. Thus, high concentrations may be shown in relation to the direction of
potential emissions sources. Hourly wind observations collected at the NWS station at the Salt
Lake City International Airport and obtained from NOAA are used in this analysis and were
averaged (using vector averaging techniques) to compute daily wind direction averages for
comparison to the 24-hour concentration data. This analysis is intended to help identify the
geographical area where emissions sources of these pollutants may have originated. Additional
information regarding this analysis is also presented in Section 3.4.2.3. Figure 20-25 presents the
pollution rose for all 113 formaldehyde concentrations measured at BTUT over the two-year
sampling period.
20-36
-------
Figure 20-25. Pollution Rose for Formaldehyde Concentrations Measured at BTUT
360/0
O < 5 |ig/m3 0 5-10 |ig/m3 © > 10 |ig/m3
Observations from Figure 20-25 include the following:
• Formaldehyde concentrations of varying magnitude are shown in relation to varying
average wind directions.
• The majority of the formaldehyde concentrations are shown in relation to sample days
with an average wind direction from the southeast (51 of the 113 concentrations) and
northwest (42) quadrants. Relatively few measurements were measured on sample
days with an average wind direction from the northeast (11) and southwest (9)
quadrants.
• For each concentration range shown on the pollution rose, the largest number of
concentrations were associated with average winds from either the southeast or
northwest quadrant. For example, among the 19 formaldehyde concentrations
measured at BTUT greater than 10 |ig/m3 (indicated by the blue dots), 11 were
measured on a sample day with an average wind direction within the southeast
quadrant, more specifically, between 135° and 180°).
• The facility map in Figure 20-2 shows that most of the point sources within 10 miles
of BTUT are located to the south and southwest of the site, along the 1-15 corridor
and towards Salt Lake City.
20-37
-------
• If the formaldehyde concentrations are grouped by average compass direction using
an 8-point compass, the direction with the most concentrations is southeast followed
by northwest. If the formaldehyde concentrations are averaged by compass direction
using an 8-point compass, the highest average concentration is calculated for north.
The northerly direction includes one formaldehyde concentration less than 5 |ig/m3,
two formaldehyde concentrations greater than 10 |ig/m3 (including one greater than
25 |ig/m3), and 13 in between.
• The wind data for many of the sample days reflect a lake breeze/valley breeze system,
one in which the wind direction in the morning is different from the
afternoon/evening, switching directions with regularity due to daytime heating and
geographic features such as the Great Salt Lake and the mountains on either side of
the Salt Lake Valley (NHMU, 2018).
20.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 20-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 20-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 20-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for BTUT, as presented in Table 20-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 20-5.
Table 20-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 20.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
20-38
-------
Table 20-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Utah Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Bountiful, Utah (Davis County) - BTUT
Benzene
80.80
Formaldehyde
9.84E-04
Formaldehyde
109.51
Formaldehyde
75.69
Hexavalent Chromium. PM
7.51E-04
Formaldehyde
73.79
Ethylbenzene
47.83
Benzene
6.30E-04
Acetaldehyde
8.01
Acetaldehyde
39.64
Naphthalene
3.69E-04
Acetaldehyde
5.76
Dichloro methane
18.27
1,3-Butadiene
3.48E-04
Benzene
5.64
T etrachloroethylene
12.20
POM, Group 2b
1.60E-04
Carbon Tetrachloride
3.33
1.3 -Butadiene
11.61
Nickel, PM
1.24E-04
Arsenic
3.30
Naphthalene
10.86
Ethylbenzene
1.20E-04
Arsenic
3.01
POM, Group 2b
1.82
POM, Group 2d
1.13E-04
1,2-Dichloroethane
2.19
POM, Group 2d
1.29
Acetaldehyde
8.72E-05
1,3-Butadiene
2.13
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 20-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Utah Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Bountiful, Utah (Davis County) - BTUT
Toluene
305.13
Acrolein
300,503.52
Formaldehyde
0.86
Xylenes
178.94
Chlorine
19,043.33
Formaldehyde
0.58
Methanol
147.88
Formaldehyde
7,723.68
Acetaldehyde
0.40
Benzene
80.80
1.3 -Butadiene
5,803.27
Acetaldehyde
0.29
Hexane
77.67
Acetaldehyde
4,403.90
Propionaldehyde
0.08
Formaldehyde
75.69
Naphthalene
3,618.63
Propionaldehyde
0.06
Ethylene glycol
57.74
Nickel, PM
2,862.10
Arsenic
0.05
Ethylbenzene
47.83
Benzene
2,693.47
Arsenic
0.05
Acetaldehyde
39.64
Xylenes
1,789.40
1,3-Butadiene
0.04
Dichloromethane
18.27
Lead, PM
1,009.88
Benzene
0.02
'Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 20-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Davis County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, hexavalent chromium, and benzene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Davis County.
• Formaldehyde, which has the highest cancer risk approximations for BTUT, ranks
second for quantity emitted and first for its toxicity-weighted emissions.
Acetaldehyde, benzene, and 1,3-butadiene also appear on all three lists in Table 20-5.
Arsenic, carbon tetrachloride, and 1,2-dichloroethane, the remaining pollutants of
interest listed for BTUT, appear on neither emissions-based list.
• POM, Groups 2b and 2d appear on both emissions-based lists in Table 20-5. POM,
Group 2b includes several PAHs sampled for at BTUT, including acenaphthene and
fluorene, which failed a few screens but were not identified as pollutants of interest
for BTUT. POM, Group 2d does not include any of the PAHs sampled for at BTUT.
Observations from Table 20-6 include the following:
• Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in Davis County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, chlorine, and formaldehyde. Although acrolein was
sampled for at BTUT, this pollutant was excluded from the pollutants of interest
designation, and thus subsequent risk-based screening evaluations, due to questions
about the consistency and reliability of the measurements, as discussed in Section 3.2.
• Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions in Davis County.
• The carbonyl compound pollutants of interest (formaldehyde, followed by
acetaldehyde, and propionaldehyde) have the highest noncancer hazard
approximations for BTUT (although all are less than 1.0). Formaldehyde and
acetaldehyde both appear on both emissions-based lists, while propionaldehyde does
not.
• Benzene, 1,3-butadiene, and arsenic are also among the pollutants of interest with the
highest noncancer hazard approximations for BTUT. Benzene also appears on both
emissions-based lists while 1,3-butadiene ranks among the pollutants with the highest
toxicity-weighted emissions in Davis County but does not appear among those with
the highest total emissions. Arsenic does not appear on either emissions-based list.
20-41
-------
20.5 Summary of the 2015-2016 Monitoring Data for BTUT
Results from several of the data analyses described in this section include the following:
~~~ Twenty-two pollutants failed at least one screen for BTUT.
~~~ Formaldehyde had the highest annual average concentrations among the pollutants
of interest for BTUT, followed by acetaldehyde.
~~~ For the fifth year in a row, BTUT has the highest annual average formaldehyde
concentration (2015) among NMP sites sampling this pollutant.
~~~ Concentrations of benzene have an overall decreasing trend at BTUT; the 1-year
average concentration for 2016 is the lowest 1-year average concentration of
benzene calculated since the onset of sampling at BTUT. Concentrations of
1,3-butadiene and naphthalene have also decreased in recent years.
~~~ The cancer risk approximation calculatedfor formaldehyde, based on the annual
average concentration for 2015, is the highest cancer risk approximation across the
program. None of the pollutants of interest for BTUT have noncancer hazard
approximations greater than an HQ of 1.0.
20-42
-------
21.0 Site in Vermont
This section summarizes those data from samples collected at the NATTS site in
Vermont and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and
2016 monitoring efforts. This section also examines the
context of risk. Readers are encouraged to refer to Sections 1 (.Una anal\ scs coniamcd in ihis
report
through 4 for detailed discussions and definitions regarding
the various data analyses presented below.
21.1 Site Characterization
This section characterizes the Vermont monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient measurements.
The Vermont NATTS site (UNVT) is located in the town of Underhill, in northwest
Vermont, in the Burlington-South Burlington, VT CBSA. Figure 21-1 presents a composite
satellite image retrieved from ArcGIS Explorer showing the Underhill monitoring site and its
immediate surroundings. Figure 21-2 identifies nearby point source emissions locations by
source category, as reported in the 2014 NEI for point sources, version 1. Note that only sources
within 10 miles of the site are included in the facility counts provided in Figure 21-2. A 10-mile
boundary was chosen to give the reader an indication of which emissions sources and emissions
source categories could potentially have a direct effect on the air quality at the monitoring site.
Further, this boundary provides both the proximity of emissions sources to the monitoring site as
well as the quantity of such sources within a given distance of the site. Sources outside the
10-mile boundary are still visible on the map for reference but have been grayed out to
emphasize emissions sources within the boundary. Table 21-1 provides supplemental
geographical information such as land use, location setting, and locational coordinates for the
site. Each figure and table is discussed in detail in the paragraphs that follow.
IXlUl L!CllcralCl.l li\ SHUIVCS
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
oilier lhail I !R( i. LIJ.\'s
contract lahoralorx lor ilic
\\ll\ arc nol includcil in the
21-1
-------
Figure 21-1. Underbill, Vermont (UNYT) Monitoring Site
-------
Figure 21-2. NEI Point Sources Located Within 10 Miles of UNVT
ra-s-crw 73 0 0 W 72'55'0"W 72 50 0 W 72 45'0"W 72"40,0"W 72"350"W
Franklin
County
\ Lamoille ^
County
Chittenden
County
Lake
Champlain
\
<2>\
#<•> \
\ Washington ^
^ County
\
I
73°15'0"W 73°10'0"W 73°5'0"W 73°0'0"W 72'J55'0"W 72°50,0"W 72°45'0"W
Note: Due to facility density and collocation, the total facilities
Legend displayed may not represent all facilities within the area of interest.
~ UNVT NATTS site O 10 mile radius | County boundary
Source Category Group (No. of Facilities)
*1" Airport/Airline/Airport Support Operations (2)
P Printing/Publishing/Paper Product Manufacturing Facility (2)
21-3
-------
Table 21-1. Geographical Information for the Vermont Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for Traffic Data
UNVT
50-007-0007
Underhill
Chittenden
Burlington-South
Burlington, VT
44.528390,
-72.868840
Forest
Rural
970
Pleasant Valley Rd, North of
Harvey Rd
1AADT reflects 2014 data forUNVT (Vtrans, 2015)
BOLD ITALICS = EPA-designated NATTS Site
-------
The UNVT monitoring site is located on the Proctor Maple Research Center in Underhill,
Vermont, which is east of the Burlington area. This research station is part of the University of
Vermont, with research focused on the sugar maple tree and sap collection methods (UVM,
2015). Figure 21-1 shows that the area surrounding the site is rural in nature and heavily
forested. Mount Mansfield, the highest peak in Vermont, lies to the east in Underhill State Park,
less than 3 miles away. This site is intended to serve as a regional background site for trends
assessment, standards compliance, and long-range transport assessment.
Most of the emissions sources near UNVT are located to the west and southwest of the
monitoring site, primarily closer to the Burlington area. The closest sources to UNVT are both in
the airport source category, which includes airports and related operations as well as small
runways and heliports, such as those associated with hospitals or television stations. These two
sources are private airports. Two sources in the printing and publishing source category are also
located within 10 miles of UNVT.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 21-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume near UNVT is relatively light, with less than 1,000 vehicles passing near UNVT on a
daily basis. The traffic estimate near UNVT is the third lowest compared to other NMP sites. The
traffic estimate for UNVT is provided for Pleasant Valley Road, north of Harvey Road.
21.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site in
order to identify site-specific "pollutants of interest," which allows analysts and readers to focus
on a subset of pollutants through the context of risk. Each pollutant's preprocessed daily
measurement was compared to its associated risk screening value. If the concentration was
greater than the risk screening value, then the concentration "failed the screen." The site-specific
results of this risk-based screening process are presented in Table 21-2 and incorporate
measurements from both 2015 and 2016. Pollutants of interest are those for which the individual
pollutant's total failed screens contribute to the top 95 percent of the site's total failed screens
and are shaded in gray in Table 21-2. It is important to note which pollutants were sampled for at
each site when reviewing the results of this analysis. PAHs were sampled for at UNVT under the
NMP in 2015 and 2016.
21-5
-------
Table 21-2. 2015-2016 Risk-Based Screening Results for the Vermont Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Underhill, Vermont - UNVT
Benzo(a)pyrene
0.00057
1
41
2.44
50.00
50.00
Naphthalene
0.029
1
121
0.83
50.00
100.00
Total
2
162
1.23
Observations from Table 21-2 include the following:
• Two individual concentrations failed screens for UNVT.
• Of the 41 measured detections of benzo(a)pyrene, one concentration failed a screen,
representing a 2 percent failure rate. Of the 121 measured detections of naphthalene,
one concentration failed a screen, representing less than a 1 percent failure rate.
• Because these two pollutants contributed equally to the total number of failed screens
(two), they are both considered pollutants of interest for UNVT. UNVT is one of only
two NMP sites with benzo(a)pyrene as a pollutant of interest.
• UNVT has the fewest failed screens of any NMP site (by a considerable margin), as
shown in Table 4-9.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
21.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Vermont monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of sampling.
• The range of measurements and annual average concentrations are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
21-6
-------
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at UNVT are provided
in Appendix N.
21.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for the UNVT site, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the UNVT
monitoring site are presented in Table 21-3, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
21-7
-------
Table 21-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Vermont Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
Underhill, Vermont - UNVT
Benzo(a)pyrene
19/0/60
0.04
±0.02
0.01
±0.01
0.01
±0.01
0.01
±0.01
0.02
±0.01
22/3/61
0.03
±0.02
0.06
±0.12
0
0.01
±0.01
0.03
±0.03
Naphthalene
60/60/60
17.56
±3.99
5.97
± 1.85
6.58
±2.95
9.39
±2.68
9.87
± 1.83
61/61/61
12.12
±2.09
5.86
± 1.24
4.87
±0.81
10.44
±2.66
8.39
± 1.17
-------
Observations from Table 21-3 include the following:
• Benzo(a)pyrene was detected in 34 percent of samples collected over the two-year
period but was detected over the MDL only three times (all of which were measured
in 2016).
• Concentrations of benzo(a)pyrene measured at UNVT range from 0.0074 ng/m3 to
0.860 ng/m3, with 80 non-detects. A total of five benzo(a)pyrene concentrations
greater than 0.1 ng/m3 were measured at UNVT; two concentrations between
0.10 ng/m3 and 0.15 ng/m3 were measured during the first quarter of each year, plus
the maximum concentration, which was measured in June 2016.
• Measured detections of benzo(a)pyrene were measured during every calendar quarter
except during the third quarter of 2016. In 2015, 13 of the 19 measured detections
were measured during the first quarter of the year, with two each measured during
each of the other three calendar quarters. In 2016, most of the measured detections
were measured during the first (nine) and fourth (10) quarters of the year, with three
measured during the second and none measured during the third. The maximum
benzo(a)pyrene concentration measured at UNVT was measured on June 26, 2016
(0.860 ng/m3) and is nearly eight time greater than the next highest concentration
measured at UNVT; the magnitude of this measurement is reflected in the second
quarter average concentration for 2016 (0.06 ± 0.12 ng/m3).
• Naphthalene was detected in all 121 valid PAH samples collected in 2015 and 2016.
Concentrations of naphthalene measured at UNVT range from 2.64 ng/m3 to
34.5 ng/m3. The annual average for 2015 (9.87 ± 1.83 ng/m3) is higher than the
annual average for 2016 (8.39 ±1.17 ng/m3), though the difference is not statistically
significant.
• For 2015, the first quarter average concentration of naphthalene is significantly higher
than the other quarterly average concentrations. Nine of the 12 naphthalene
concentrations greater than 15 ng/m3 measured in 2015 were measured during the
first calendar quarter (with none measured during the second quarter, one during the
third, and two during the fourth). Additionally, none of the 20 naphthalene
concentrations less than 5 ng/m3 were measured during the first quarter of 2015.
• For 2016, the first and fourth quarter average concentrations of naphthalene are
higher than the quarterly averages for the second and third quarters of the year.
Nineteen of the 21 naphthalene concentrations greater than 10 ng/m3 were measured
during the first or fourth calendar quarters (with the other two measured in April).
Additionally, none of the 18 naphthalene concentrations less than 5 ng/m3 were
measured during the first quarter of 2016 and only one was measured during the
fourth quarter.
• Among NMP sites sampling PAHs, UNVT has the lowest annual average
concentrations of naphthalene. This is also true for benzo(a)pyrene.
21-9
-------
21.3.2 Concentration Comparison
In order to better illustrate how a site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 21-3 for UNVT. Figure 21-3 overlays the site's minimum, annual average, and maximum
concentrations for each year onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.2.1, and are discussed
below. If an annual average concentration could not be calculated, the range of concentrations
are still provided in the figures that follow.
Figure 21-3. Program vs. Site-Specific Average Benzo(a)pyrene Concentrations
J
•H
|_i ! Program Max Concentration = 5.82 ng/m3 j
1
1
I
0 0.2 0.4 0.6 0.8 1 1.2
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
» o
Figure 21-3 presents the box plot for benzo(a)pyrene for UNVT and shows the following:
• The program-level maximum benzo(a)pyrene concentration (5.82 ng/m3) is not
shown directly on the box plot in Figure 21-3 because the scale of the box plot would
be too large to readily observe data points at the lower end of the concentration range.
Thus, the scale of the box plot has been reduced to 1.2 |ig/m3.
• Although the maximum benzo(a)pyrene concentration measured at UNVT is
considerably less than the maximum concentration measured across the program, it is
the 15th highest concentration measured across the program. If this concentration is
excluded, the range of concentrations measured each year at UNVT would be similar
to each other.
• Both annual average concentrations for UNVT are greater than the program-level first
quartile and less than the program-level median concentration.
21-10
-------
Figure 21-4. Program vs. Site-Specific Average Naphthalene Concentrations
H-
50
100
150
200 250
Concentration (ng/m3)
300
350
400
450
Program: IstQuartile 2nd Quartile 3rd Quartile 4-thQuartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
o o —
Figure 21-4 presents the box plot for naphthalene for UNVT and shows the following:
• The maximum naphthalene concentration measured at UNVT (34.5 ng/m3) is an order
of magnitude less than the program-level maximum naphthalene concentration
(403 ng/m3).
• The entire range of naphthalene concentrations measured at UNVT in 2016 is less
than the program-level first quartile (28.3 ng/m3) and only one concentration
measured in 2015 is greater than the first quartile.
• UNVT is the only NMP site for which an annual average concentration of
naphthalene less than 10 ng/m3 was calculated.
21.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
UNVT has sampled PAHs under the NMP since 2008. Thus, Figures 21-5 and 21-6 present the
1-year statistical metrics for the pollutants of interest for UNVT. The statistical metrics presented
for assessing trends include the substitution of zeros for non-detects. If sampling began mid-year,
a minimum of 6 months of sampling is required for inclusion in the trends analysis; in these
cases, a 1-year average concentration is not provided, although the range and percentiles are still
presented.
21-11
-------
Figure 21-5. Yearly Statistical Metrics for Benzo(a)pyrene Concentrations Measured at
UNVT
T
. ~ P
h
2012
Year
O 5th Percentile
O 95th Pe r ce nti le Ave r z
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 21-5 for benzo(a)pyrene concentrations measured at UNVT
include the following:
• UNVT began sampling PAHs under the NMP in June 2008. Because a full year's
worth of data is not available, a 1-year average for 2008 is not presented, although the
range of measurements is provided.
• The median benzo(a)pyrene concentration for each year of sampling is zero,
indicating that non-detects account for at least half of the measurements each year at
UNVT.
• The range of benzo(a)pyrene concentrations measured each year is highly variable,
with a year with a "higher" concentration or two alternating with a year with "lower"
concentrations. In total, six benzo(a)pyrene concentrations greater than 0.3 ng/m3
have been measured at UNVT; five of these are visible as the maximum
concentrations for 2008, 2010, 2012, 2014, and 2016, with the sixth also measured in
2010.
• The 1-year average concentration has ranged from 0.014 ng/m3 (2011, 2014) to
0.037 ng/m3 (2010) over the years of sampling.
21-12
-------
Figure 21-6. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
UNVT
200
180
160
140
o
4= 100
c
CD
u 80
60
40
20
0
O 5th Percentile — Minimum — Median — Maximum O 95th Percentile Average
1 A 1-year average is not presented because sampling under the NMP did not begin until June 2008.
Observations from Figure 21-6 for naphthalene concentrations measured at UNVT
include the following:
• A single naphthalene concentration greater than 100 ng/m3 has been measured at
UNVT (186 ng/m3). In total, only six concentrations greater than 50 ng/m3 have been
measured at this site, three in 2011, two in 2009, and one in 2008. Naphthalene
concentrations greater than 35 ng/m3 have not been measured since 2013.
• A significant decreasing trend in naphthalene concentrations measured at UNVT is
shown in Figure 21-6. The 1-year average concentration, the 95th percentile, and
maximum concentration are all at a minimum for 2016.
• The 1-year average concentration has decreased by nearly half since 2009, from
16.38 ng/m3 in 2009 to 8.39 ng/m3 in 2016.
• The median concentration decreased each year between 2009 and 2013. Although the
median has increased slightly each year since 2013, they are still less than those
calculated for the first full-years of sampling at UNVT.
21-13
-------
21.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the Vermont monitoring site. Refer to Sections 3.2, 3.4.2.3,
and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames,
and calculations associated with these risk-based screenings.
21.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Vermont monitoring site, risk was examined by
calculating cancer risk and noncancer hazard approximations for each year annual average
concentrations could be calculated. These approximations can be used as risk estimates for
cancer and noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers want to shift their air
monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 21-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
Observations from Table 21-4 include the following:
• The annual average concentrations of naphthalene are greater than the annual average
concentrations of benzo(a)pyrene.
• Although the cancer risk approximations for naphthalene are an order of magnitude
greater than the cancer risk approximations for benzo(a)pyrene, all of these are less
than 1 in-a-million.
• The noncancer hazard approximations for naphthalene are both less than 0.01,
indicating that no adverse noncancer health effects are expected from this individual
pollutant. A noncancer RfC is not available for benzo(a)pyrene.
21-14
-------
Table 21-4. Risk Approximations for the Vermont Monitoring Site
2015
2016
# of
Risk Approximations
# of
Risk Approximations
Measured
Measured
Cancer
Noncancer
Detections
Annual
Detections
Annual
URE
RfC
vs. # of
Average
Cancer
Noncancer
vs. # of
Average
Cancer
Noncancer
Pollutant
(jig/m3)1
(mg/m3)
Samples
(ng/m3)
(in-a-million)
(HQ)
Samples
(ng/m3)
(in-a-million)
(HQ)
Underhill, Vermont - UNVT
0.02
0.03
Benzo(a)pyrene
0.00176
—
19/60
±0.01
0.03
—
22/61
±0.03
0.04
—
9.87
8.39
Naphthalene
0.000034
0.003
60/60
± 1.83
0.34
<0.01
61/61
± 1.17
0.29
<0.01
- = A Cancer URE or Noncancer RfC is not available.
-------
21.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screenings discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 21-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 21-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 21-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for UNVT, as presented in Table 21-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 21-5.
Table 21-6 presents similar information but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 21.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
21-16
-------
Table 21-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Vermont Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Underhill, Vermont (Chittenden County) - UNVT
Benzene
101.44
POM, Group 3
1.19E-03
Naphthalene
0.34
Formaldehyde
81.09
Formaldehyde
1.05E-03
Naphthalene
0.29
Acetaldehyde
43.22
Benzene
7.91E-04
Benzo(a)pyrene
0.04
Ethylbenzene
25.84
1,3-Butadiene
3.83E-04
Benzo(a)pyrene
0.03
1.3 -Butadiene
12.76
Naphthalene
3.31E-04
Naphthalene
9.73
Arsenic, PM
2.94E-04
Bis(2-ethylhexyl) phthalate, gas
5.96
POM, Group 2b
2.01E-04
Dichloromethane
2.87
POM, Group 5a
1.42E-04
POM, Group 2b
2.28
POM, Group 2d
1.27E-04
POM, Group 2d
1.45
Hexavalent Chromium PM
1.09E-04
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 21-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Vermont Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Underhill, Vermont (Chittenden County) - UNVT
Toluene
173.71
Acrolein
689,481.49
Naphthalene
<0.01
Benzene
101.44
Chlorine
15,172.33
Naphthalene
<0.01
Xylenes
94.73
Manganese, PM
12,474.61
Methanol
81.93
Formaldehyde
8,274.96
Formaldehyde
81.09
2,4-Toluene diisocyanate
7,296.86
Hydrochloric acid
44.59
1.3 -Butadiene
6,378.93
Acetaldehyde
43.22
Acetaldehyde
4,802.45
Hexane
30.90
Arsenic, PM
4,550.92
Ethylbenzene
25.84
Benzene
3,381.22
Acrolein
13.79
Naphthalene
3,243.73
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 21-5 include the following:
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in Chittenden County.
• POM, Group 3 is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with cancer UREs) in Chittenden County, followed by formaldehyde and
benzene.
• Six of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Chittenden County.
• Naphthalene has the fifth highest toxicity-weighted emissions and is the sixth highest
emitted pollutant with a cancer URE in Chittenden County.
• Benzo(a)pyrene is part of POM, Group 5a. POM, Group 5a has the eighth highest
toxicity-weighted emissions but is not one of the highest emitted in Chittenden
County (it ranks 19th).
• POM, Groups 2b and 2d also appear on both-emissions based lists in Table 21-5.
POM, Groups 2b includes several PAHs sampled for with Method TO-13A, although
none of these failed screens for UNVT. POM, Group 3, which has the highest
toxicity-weighted emissions (of the pollutants with cancer UREs) in Chittenden
County, does not include any of the PAHs sampled for at UNVT.
Observations from Table 21-6 include the following:
• Toluene, benzene, and xylenes are the highest emitted pollutants with noncancer
RfCs in Chittenden County.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for Chittenden County, followed by chlorine and
manganese (PM).
• Four of the highest emitted pollutants for Chittenden County also have the highest
toxicity-weighted emissions.
• Naphthalene has the tenth highest toxicity-weighted emissions but is not one of the
highest emitted in Chittenden County (it ranks 13th). As discussed in the previous
section, benzo(a)pyrene, which is part of POM, Group 5a, does not have a noncancer
RfC.
21.5 Summary of the 2015-2016 Monitoring Data for the Vermont Monitoring Site
Results from several of the data analyses described in this section include the following:
~~~ PAHs were sampledfor at UNVT in 2015 and 2016.
~~~ Naphthalene and benzo(a)pyrene were identified as pollutants of interest for UNVT.
21-19
-------
~~~ UNVT's annual average concentrations of naphthalene are the lowest annual
averages of this pollutant among NMP sites sampling PAHs. This is also true for
benzo(a)pyrene.
~~~ Concentrations of naphthalene have a decreasing trend at UNVT.
21-20
-------
22.0 Site in Virginia
This section summarizes those data from samples collected at the NATTS site in Virginia
and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and 2016
monitoring efforts. This section also examines the spatial
through 4 for detailed discussions and definitions
regarding the various data analyses presented below.
22.1 Site Characterization
This section characterizes the Virginia monitoring site by providing a description of the
nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient monitoring measurements.
The RIVA monitoring site is located just outside the Richmond, Virginia city limits in
East Highland Park. Figure 22-1 presents a composite satellite image retrieved from ArcGIS
Explorer showing the monitoring site and its immediate surroundings. Figure 22-2 identifies
nearby point source emissions locations by source category, as reported in the 2014 NEI for
point sources, version 1. Note that only sources within 10 miles of the site are included in the
facility counts provided in Figure 22-2. A 10-mile boundary was chosen to give the reader an
indication of which emissions sources and emissions source categories could potentially have a
direct effect on the air quality at the monitoring site. Further, this boundary provides both the
proximity of emissions sources to the monitoring site as well as the quantity of such sources
within a given distance of the site. Sources outside the 10-mile boundary are still visible on the
map for reference but have been grayed out to emphasize emissions sources within the boundary.
Table 22-1 provides supplemental geographical information such as land use, location setting,
and locational coordinates. Each figure and table is discussed in detail in the paragraphs that
follow.
and temporal characteristics of the ambient monitoring
concentrations and reviews them through the context of
risk. Readers are encouraged to refer to Sections 1
IXiUi jjcncmlct.1 h> sources oilier
limn LR(i. UW's conliucl
kiliomioiA lor iIk- \\ll\ ;iic noi
included m I lie d^ila ;uuil\ses
coiiLiinei.1 in ilns report
22-1
-------
Figure 22-1. East Highland Park, Virginia (RIVA) Monitoring Site
. Source: USGS
Source: N^SA, NGA, USGS
v^200 S*M iCrbSOft Corp.
Cre*9wor''
-------
Figure 22-2. NEI Point Sources Located Within 10 Miles of RIVA
J King William '
i / County \
Hanover
County
Henrico
County
Richmond
City
James
River,
]j J King William '
i / County \
77°30,0"W 77°25'0"W 77°20,0"W
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
\ _ _ - "*
n" ^
\ i
\ Chesterfield ^
\ County I
V \
\ ^
Legend
\
\
\
V
Goochland
- County
~
RIVA NATTS site
10 mile radius
County boundary
Source Category Group (No. of Facilities)
r
Airport/Airline/Airport Support Operations (10)
B
Bulk Terminal/Bulk Plant (2)
D
C
Chemical Manufacturing Facility (2)
R
f
Electricity Generation Facility (5)
P
F
Food Processing/Agriculture Facility (1)
X
o
Institution (school, hospital, prison, etc.) (1)
M
A
Landfill (1)
Miscellaneous Commercial/Industrial Facility (2)
Paint and Coating Manufacturing Facility (1)
Plastic, Resin, or Rubber Products Plant (2)
Printing/Publishing/Paper Product Manufacturing Facility (2)
Rail Yard/Rail Line Operations (4)
Tobacco Manufacturing Facility (1)
22-3
-------
Table 22-1. Geographical Information for the Virginia Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude and
Longitude
Land Use
Location
Setting
Annual Average
Daily Traffic1
Intersection
Used for
Traffic Data
RIVA
51-087-0014
East Highland
Park
Henrico
Richmond, VA
37.556520,
-77.400270
Residential
Suburban
80,000
1-64 at Mechanicsville
Turnpike/US-360
1AADT reflects 2016 data (VA DOT, 2016)
BOLD ITALICS = EPA-designated NATTS Site
to
to
-------
The RIVA monitoring site is located just northeast of the capital city of Richmond, in
east-central Virginia. The site is located at the MathScience Innovation Center in a residential
area about one-quarter mile from 1-64. The 1-64 interchange with Mechanicsville Turnpike
(US-360) is one-half mile west of the site, as shown in Figure 22-1. Beyond the residential areas
surrounding the school property are a golf course to the southeast, a high school to the south (on
the south side of 1-64), and commercial areas to the west.
As Figure 22-2 shows, RIVA is located near several point sources, most of which are
located on the southern half of the 10-mile boundary and within the city of Richmond. The
source categories with the greatest number of emissions sources within 10 miles of RIVA are the
airport source category, which includes airports and related operations as well as small runways
and heliports, such as those associated with hospitals or television stations; facilities generating
electricity via combustion; and rail yard and rail line operations. The source closest to RIVA is a
heliport at the Medical College of Virginia.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 22-1 also contains traffic volume information for the site as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near RIVA is 80,000, which is in the top third of the range compared to
other NMP monitoring sites, ranking 15th. The traffic volume provided is for 1-64 at US-360
(Mechanicsville Turnpike).
22.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each
monitoring site to identify site-specific "pollutants of interest," which allows analysts and
readers to focus on a subset of pollutants through the context of risk. Each pollutant's
preprocessed daily measurement was compared to its associated risk screening value. If the
concentration was greater than the risk screening value, then the concentration "failed the
screen." The site-specific results of this risk-based screening process are presented in Table 22-2
and incorporate measurements from both 2015 and 2016. Pollutants of interest are those for
which the individual pollutant's total failed screens contribute to the top 95 percent of the site's
total failed screens and are shaded in gray in Table 22-2. It is important to note which pollutants
were sampled for at each site when reviewing the results of this analysis. PAHs and hexavalent
22-5
-------
chromium were sampled for at RIVA, although hexavalent chromium sampling was discontinued
at the end of June 2016.
Table 22-2. 2015-2016 Risk-Based Screening Results for the Virginia Monitoring Site
Pollutant
Screening
Value
(Ug/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
East Highland Park, Virginia - RIVA
Naphthalene
0.029
105
116
90.52
99.06
99.06
Hexavalent Chromium
0.000083
1
59
1.69
0.94
100.00
Total
106
175
60.57
100.00
Observations from Table 22-2 include the following:
• Concentrations naphthalene and hexavalent chromium failed at least one screen for
RIVA.
• Nearly 61 percent of concentrations of these two pollutants failed screens, although
concentrations of naphthalene account for all but one of the 106 failed screens for
RIVA. Thus, naphthalene is RIVA's only pollutant of interest.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
22.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Virginia monitoring site. Where applicable, the following calculations
and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of sampling.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
22-6
-------
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at RIVA are provided
in Appendices N and P.
22.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for RIVA, as described in Section 3.1. The quarterly average concentration
of a particular pollutant is simply the average concentration of the preprocessed daily
measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutant of interest for the Virginia
monitoring site are presented in Table 22-3, where applicable. Note that if a pollutant was not
detected in a given calendar quarter, the quarterly average simply reflects "0" because only zeros
substituted for non-detects were factored into the quarterly average concentration.
Observations for RIVA from Table 22-3 include the following:
• Naphthalene was detected in every valid PAH sample collected at RIVA in 2015 and
2016. Concentrations of naphthalene measured at RIVA range from 11.0 ng/m3 to
183 ng/m3.
• The annual average concentration of naphthalene for 2015 is similar to the annual
average concentration for 2016.
• The quarterly average concentrations range from 56.64 ± 18.04 ng/m3 (second quarter
2016) to 86.02 ± 27.24 ng/m3 (fourth quarter 2015). At least one naphthalene
concentration greater than 100 ng/m3 was measured during each calendar quarter,
with the most measured during the fourth quarter of each year (between one and three
were measured during the first, second, or third quarters each year, compared to five
measured during the fourth quarter of 2015 and six during the fourth quarter of 2016).
22-7
-------
Table 22-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Virginia Monitoring Site
Pollutant
2015
2016
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Qi
Avg
(ng/m3)
Annual
Average
(ng/m3)
# of
Detects/
# >MDL/
# Samples
Qi
Avg
(ng/m3)
Q2
Avg
(ng/m3)
Q3
Avg
(ng/m3)
Q4
Avg
(ng/m3)
Annual
Average
(ng/m3)
East Highland Park, Virginia - RIVA
Naphthalene
56/56/56
64.21
± 12.74
70.04
± 15.26
64.28
± 17.02
86.02
± 27.24
71.24
±9.12
60/60/60
67.57
±22.93
56.64
± 18.04
63.71
± 15.81
81.61
± 24.97
67.57
± 10.11
to
to
00
-------
22.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, a box plot was created for the pollutant listed in
Table 22-3 for RIVA. Figure 22-3 overlays the site's minimum, annual average, and maximum
concentrations for each year onto the program-level minimum, first quartile, median, average,
third quartile, and maximum concentrations, as described in Section 3.4.2.1, and are discussed
below. If an annual average concentration could not be calculated, the range of concentrations
are still provided in the figures that follow.
Figure 22-3. Program vs. Site-Specific Average Naphthalene Concentrations
RIVA |
:
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
» o
Figure 22-3 presents the box plot for naphthalene for RIVA and shows the following:
• The range of naphthalene concentrations measured each year at RIVA are similar to
each other.
• The maximum naphthalene concentrations measured at RIVA each year (183 ng/m3
for 2015 and 182 ng/m3 for 2016) are considerably less than the program-level
maximum concentration (403 ng/m3).
• There were no non-detects of naphthalene measured at RIVA, or across the program
(although difficult to discern in Figure 22-3).
• The annual average concentrations of naphthalene for RIVA are both just greater than
the program-level average concentration (61.23 ng/m3).
22-9
-------
22.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
RIVA began sampling PAHs under the NMP in October 2008. Thus, Figure 22-4 presents the
1-year statistical metrics for the pollutant of interest for RIVA. The statistical metrics presented
for assessing trends include the substitution of zeros for non-detects. If sampling began mid-year,
a minimum of 6 months of sampling is required for inclusion in the trends analysis; in these
cases, a 1-year average concentration is not provided, although the range and percentiles are still
presented.
Figure 22-4. Yearly Statistical Metrics for Naphthalene Concentrations Measured at RIVA
5th Percentile
95th Percentile
Observations from Figure 22-4 for naphthalene concentrations measured at RIVA include
the following:
• RIVA began sampling PAHs under the NMP in October 2008. Because less than
6 months of data are available for 2008, Figure 22-4 begins with 2009.
• Three naphthalene concentrations greater than 400 ng/m3 were measured at RIVA
during the fall of 2009. The next highest concentration was measured in 2013
(354 ng/m3) and is the only other concentration greater than 300 ng/m3 measured at
RIVA.
22-10
-------
• Most of the statistical parameters exhibit a decreasing trend through 2011, with the
most significant change occurring between 2010 and 2011. All of the statistical
parameters exhibit an increase for 2012 before decreasing slightly for 2013 (with the
exception of the maximum concentration) and again for 2014. The 1-year average
concentration is at a minimum for 2014.
• Relatively little change is shown in the range of concentrations measured between
2014 and 2016. During this 3-year period, the 1-year average concentration varied by
less than 10 ng/m3 and the median concentration varied by less than 7 ng/m3.
22.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the RIVA monitoring site. Refer to Sections 3.2, 3.4.2.3, and
3.4.2.4 for definitions and explanations regarding the various toxicity factors, time frames, and
calculations associated with these risk-based screenings.
22.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for RIVA, risk was examined by calculating cancer risk and
noncancer hazard approximations for each year annual average concentrations could be
calculated. These approximations can be used as risk estimates for cancer and noncancer effects
attributable to the pollutants of interest. Although the use of these approximations is limited, they
may help identify where policy-makers want to shift their air monitoring priorities. Refer to
Section 3.4.2.3 for an explanation of how cancer risk and noncancer hazard approximations are
calculated and what limitations are associated with them. Annual averages, cancer UREs and/or
noncancer RfCs, and cancer risk and noncancer hazard approximations are presented in
Table 22-4, where applicable. Cancer risk approximations are presented as probabilities while
the noncancer hazard approximations are ratios and thus, unitless values.
22-11
-------
Table 22-4. Risk Approximations for the Virginia Monitoring Site
Pollutant
Cancer
URE
frig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
(ng/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
East Highland Park, Virginia - RIVA
Naphthalene
0.000034
0.003
56/56
71.24
±9.12
2.42
0.02
60/60
67.57
± 10.11
2.30
0.02
-------
Observations for RIVA from Table 22-4 include the following:
• The annual average concentrations of naphthalene are 71.24 ± 9.12 ng/m3 for 2015
and 67.57 ± 10.11 ng/m3 for 2016.
• The cancer risk approximations for naphthalene, based on RIVA's annual average
concentrations, are 2.42 in-a-million for 2015 and 2.30 in-a-million for 2016.
• The noncancer hazard approximations for naphthalene are both considerably less than
1.0 (0.02 for both years), indicating that no adverse noncancer health effects are
expected from this individual pollutant.
22.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 22-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 22-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 22-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for RIVA, as presented in Table 22-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 22-5.
Table 22-6 presents similar information, but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual average concentrations to be
calculated. A more in-depth discussion of this analysis is provided in Section 3.4.2.4. Similar to
the cancer risk and noncancer hazard approximations provided in Section 22.4.1, this analysis
may help policy-makers prioritize their air monitoring activities.
22-13
-------
Table 22-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Virginia Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based
on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
East Highland Park, Virginia (Henrico County) - RIVA
Benzene
90.56
Formaldehyde
1.13E-03
Naphthalene
2.42
Formaldehyde
86.58
Benzene
7.06E-04
Naphthalene
2.30
Ethylbenzene
51.26
1,3-Butadiene
4.48E-04
Acetaldehyde
47.24
Naphthalene
2.61E-04
1.3 -Butadiene
14.92
POM, Group 2b
1.55E-04
Naphthalene
7.69
Ethylbenzene
1.28E-04
POM, Group 2b
1.76
POM, Group 2d
1.11E-04
POM, Group 2d
1.26
Acetaldehyde
1.04E-04
Trichloroethylene
0.49
POM, Group 5a
9.23E-05
Dichloromethane
0.44
Arsenic, PM
4.75E-05
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 22-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Virginia Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tl»)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
East Highland Park, Virginia (Henrico County) - RIVA
Ethylene glycol
383.19
Acrolein
332,228.50
Naphthalene
0.02
Toluene
303.05
Formaldehyde
8,834.97
Naphthalene
0.02
Xylenes
178.93
1.3 -Butadiene
7,459.46
Methanol
146.28
Acetaldehyde
5,248.37
Hexane
114.66
Benzene
3,018.68
Benzene
90.56
Naphthalene
2,562.65
Formaldehyde
86.58
Xylenes
1,789.35
Ethylbenzene
51.26
Ethylene glycol
957.97
Acetaldehyde
47.24
Glycol ethers, gas
798.69
Glycol ethers, gas
15.98
Cyanide Compounds, gas
784.97
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 22-5 include the following:
• Benzene, formaldehyde, and ethylbenzene are the highest emitted pollutants with
cancer UREs in Henrico County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) are formaldehyde, benzene, and 1,3-butadiene.
• Eight of the highest emitted pollutants also have the highest toxicity-weighted
emissions for Henrico County.
• Naphthalene, the only pollutant of interest for RIVA, has the sixth highest emissions
and the fourth highest toxicity-weighted emissions for Henrico County.
• Several POM Groups appear in Table 22-5. POM, Group 2b is the seventh highest
emitted "pollutant" in Henrico County and ranks fifth for toxicity-weighted
emissions. POM, Group 2b includes several PAHs sampled for at RIVA, including
fluorene, perylene, and acenaphthene. None of the PAHs sampled for at RIVA
included in POM, Group 2b failed screens. POM, Group 2d also appears on both
emissions-based lists for Henrico County but does not include any PAHs sampled for
at RIVA. POM, Group 5a includes benzo(a)pyrene and ranks ninth for toxicity-
weighted emissions but is not among the highest emitted. Concentrations of
benzo(a)pyrene measured at RIVA did not fail screens.
Observations from Table 22-6 include the following:
• Ethylene glycol, toluene, and xylenes are the highest emitted pollutants with
noncancer RfCs in Henrico County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
noncancer RfCs) are acrolein, formaldehyde, and 1,3-butadiene.
• Six of the highest emitted pollutants in Henrico County also have the highest toxicity-
weighted emissions.
• Naphthalene has the sixth highest toxicity-weighted emissions for Henrico County
but is not among the highest emitted pollutants with a noncancer toxicity factor in
Henrico County (it ranks 12th).
22.5 Summary of the 2015-2016 Monitoring Data for RIVA
Results from several of the data analyses described in this section include the following:
~~~ Concentrations of naphthalene and hexavalent chromium failed at least one screen,
although naphthalene was the only pollutant identified as a pollutant of interest for
RIVA.
~~~ The range of naphthalene concentrations measured at RIVA varied little across the
two years of sampling.
22-16
-------
~~~ Concentrations of naphthalene have decreased at RIVA since the onset of PAH
sampling at this site, although concentrations have leveled off over the last few years.
22-17
-------
23.0 Site in Washington
This section summarizes those data from samples collected at the NATTS site in
Washington and generated by ERG, EPA's contract laboratory for the NMP, over the 2015 and
2016 monitoring efforts. This section also examines the
Sections 1 through 4 for detailed discussions and
definitions regarding the various data analyses presented below.
23.1 Site Characterization
This section characterizes the Washington monitoring site by providing a description of
the nearby area surrounding the monitoring site; plotting emissions sources surrounding the
monitoring site; and presenting traffic data and other characterizing information for the site. This
information is provided to give the reader insight regarding factors that may influence the air
quality near the site and assist in the interpretation of the ambient measurements.
The NATTS site in Washington is located in Seattle. Figure 23-1 presents a composite
satellite image retrieved from ArcGIS Explorer showing the monitoring site and its immediate
surroundings. Figure 23-2 identifies nearby point source emissions locations by source category,
as reported in the 2014 NEI for point sources, version 1. Note that only sources within 10 miles
of the site are included in the facility counts provided in Figure 23-2. A 10-mile boundary was
chosen to give the reader an indication of which emissions sources and emissions source
categories could potentially have a direct effect on the air quality at the monitoring site. Further,
this boundary provides both the proximity of emissions sources to the monitoring site as well as
the quantity of such sources within a given distance of the site. Sources outside the 10-mile
boundary are still visible on the map for reference but have been grayed out to emphasize
emissions sources within the boundary. Table 23-1 provides supplemental geographical
information such as land use, location setting, and locational coordinates. Each figure and table
is discussed in detail in the paragraphs that follow.
spatial and temporal characteristics of the ambient
monitoring concentrations and reviews them through the
context of risk. Readers are encouraged to refer to
IXiUi jjcncr;ilct.l h> sources oilier
lhail I!R(i. UW's coniriicl
kiboixiloi'N lor I lie \\ll\ ;ire nol
included in I lie dala anal> ses
conlained in llns report
23-1
-------
Figure 23-1. Seattle, Washington (SEWA) Monitoring Site
'Norton, St.
•S«H|nds-St,
li 13PT 'I |
S Dakota St
S Nevada St
S Oregon St
-------
Figure 23-2. NEI Point Sources Located Within 10 Miles of SEW A
122C40'0"W 122°35'0"W 122530'0"W 122°25,0"W 122°26"0"W 122°15,0"W 122°10'0"W
Sound
%¦
1 F Lakl
Washington
County
' Lake
^L Sammamish
I ^ I
County 1 USB
County
122"30'0"W 122"25'0"W 122"20'0"W 122"15'0"W 122n10'0"W 122"5'0"W 122WW
Note: Due to facility density and collocation, the total facilities
displayed may not represent all facilities within the area of interest.
Legend
~ SEWA NATTS site
0 10 mile radius
County boundary
Source Category Group (No. of Facilities)
i®i Aerospace/Aircraft Manufacturing Facility (2)
T Airport/Airline/Airport Support Operations (27)
H Automobile/Truck Manufacturing Facility (4)
b Bulk Terminal/Bulk Plant (2)
e Electrical Equipment Manufacturing Facility (2)
F Food Processing/Agriculture Facility (2)
Glass Plant (1)
o Institution (school, hospital, prison, etc.) (1)
A Metal Coating, Engraving, and Allied Services
to Manufacturers (1)
© Metals Processing/Fabrication Facility (4)
x Mine/Quarry/Mineral Processing Facility (1)
? Miscellaneous Commercial/Industrial Facility
' (1)
Paint and Coating Manufacturing Facility (1)
7 Portland Cement Manufacturing Facility (1)
X Rail Yard/Rail Line Operations (1)
A Ship/Boat Manufacturing or Repair Facility (1)
V Steel Mill (1)
i Wastewater Treatment Facility (1)
23-3
-------
Table 23-1. Geographical Information for the Washington Monitoring Site
Site
Code
AQS Code
Location
County
Micro- or
Metropolitan
Statistical Area
Latitude
and
Longitude
Land Use
Location
Setting
Annual
Average Daily
Traffic1
Intersection
Used for
Traffic Data
SEWA
53-033-0080
Seattle
King
Seattle-Tacoma-
Bellevue, WA
47.568236,
-122.308628
Residential
Urban/City
Center
186,000
1-5 S at Spokane St Viaduct
lAADT reflects 2015 data (WS DOT, 2015)
BOLD ITALICS = EPA-designated NATTS Site
to
-U
-------
The SEWA monitoring site is located in Seattle, at the southeast corner of the Beacon
Hill Reservoir. The reservoir is covered and the entire area is part of Jefferson Park (Seattle,
2018). The reservoir and park are separated from the Jefferson Park Golf Course to the east by
Beacon Avenue, as shown in Figure 23-1. A middle school and a hospital can be seen to the
south of the site in the bottom-center portion of Figure 23-1. The site is surrounded by residential
neighborhoods to the west, north, and east. Interstate-5, which runs north-south through Seattle,
is just over one-half mile to the west of SEWA and intersects with 1-90 a couple of miles farther
north of the site. The area to the west of 1-5 is highly industrial while the area to the east is
primarily residential. Although the emissions sources within 10 miles of the site are involved in a
variety of industries, the airport source category, which includes airports and related operations
as well as small runways and heliports, such as those associated with hospitals or television
stations, has the greatest number of sources. The closest point sources to SEWA are a metals
processing and fabrication facility, a food processing facility, and a rail yard/rail line operation,
as shown in Figure 23-2.
In addition to providing city, county, CBS A, and land use/location setting information,
Table 23-1 also contains traffic volume information for SEWA as well as the location for which
the traffic volume was obtained. This information is provided because emissions from motor
vehicles can significantly affect concentrations measured at a given monitoring site. The traffic
volume experienced near SEWA is 186,000, which is the fourth highest compared to traffic
volumes near other NMP monitoring sites. The traffic estimate provided is for 1-5 at the Spokane
Street Viaduct.
23.2 Pollutants of Interest
The risk-based screening process described in Section 3.2 was performed for each site to
identify site-specific "pollutants of interest," which allows analysts and readers to focus on a
subset of pollutants through the context of risk. Each pollutant's preprocessed daily measurement
was compared to its associated risk screening value. If the concentration was greater than the risk
screening value, then the concentration "failed the screen." The site-specific results of this risk-
based screening process are presented in Table 23-2 and incorporate measurements from both
2015 and 2016. Pollutants of interest are those for which the individual pollutant's total failed
screens contribute to the top 95 percent of the site's total failed screens and are shaded in gray in
Table 23-2. It is important to note which pollutants were sampled for at the site when reviewing
23-5
-------
the results of this analysis. PMio metals, VOCs, PAHs, and carbonyl compounds were sampled
for at SEWA.
Table 23-2. 2015-2016 Risk-Based Screening Results for the Washington Monitoring Site
Pollutant
Screening
Value
Oig/m3)
# of
Failed
Screens
# of
Measured
Detections
%of
Screens
Failed
%of
Total
Failures
Cumulative
%
Contribution
Seattle, Washington - SEWA
Formaldehyde
0.077
119
120
99.17
14.15
14.15
Benzene
0.13
118
118
100.00
14.03
28.18
Carbon Tetrachloride
0.17
118
118
100.00
14.03
42.21
1,2-Dichloroethane
0.038
108
110
98.18
12.84
55.05
Arsenic (PMio)
0.00023
101
116
87.07
12.01
67.06
1.3 -Butadiene
0.03
99
113
87.61
11.77
78.83
Naphthalene
0.029
79
118
66.95
9.39
88.23
Acetaldehyde
0.45
77
120
64.17
9.16
97.38
Nickel (PMio)
0.0021
8
116
6.90
0.95
98.34
Ethylbenzene
0.4
7
118
5.93
0.83
99.17
Cadmium (PMio)
0.00056
3
116
2.59
0.36
99.52
Acenaphthene
0.011
1
110
0.91
0.12
99.64
p-Dichlorobenzene
0.091
1
7
14.29
0.12
99.76
Fluorene
0.011
1
92
1.09
0.12
99.88
Manganese (PMio)
0.03
1
116
0.86
0.12
100.00
Total
841
1,608
52.30
Observations from Table 23-2 for SEWA include the following:
• Concentrations of 15 pollutants failed at least one screen for SEWA; 52 percent of
concentrations for these 15 pollutants were greater than their associated risk screening
value (or failed screens).
• Concentrations of eight pollutants contributed to 95 percent of failed screens for
SEWA and therefore were identified as pollutants of interest for the site. These eight
include two carbonyl compounds, four VOCs, one PMio metal, and one PAH.
• Benzene and carbon tetrachloride were detected in every valid VOC sample collected
at SEWA and failed 100 percent of screens.
For each of the data analyses described in the remaining sections, the focus is on the site-
specific pollutants of interest identified via the risk-based screening process, as described in
Section 3.4.2.
23-6
-------
23.3 Concentrations
This section presents various concentration averages used to characterize air toxics
concentration levels at the Washington monitoring site. Where applicable, the following
calculations and data analyses were performed for each of the site-specific pollutants of interest:
• Time period-based concentration averages (quarterly and annual) are provided for
each monitoring site for each year of sampling.
• The range of measurements and annual concentration averages are presented
graphically for each site to illustrate how the site's concentrations compare to the
program-level averages, as presented in Section 4.1.
• Concentration averages and other statistical metrics for 2015, 2016, and from
previous years of monitoring are presented in order to characterize concentration
trends at each site.
Each data analysis is performed where the applicable criteria are met (as specified in the
appropriate sections discussed below) and is limited to the site-specific pollutants of interest.
However, site-specific statistical summaries for all pollutants sampled for at SEWA are provided
in Appendices J, M, N, and O.
23.3.1 2015 and 2016 Concentration Averages
Quarterly and annual concentration averages for 2015 and 2016 were calculated for the
pollutants of interest for SEWA, as described in Section 3.1. The quarterly average
concentration of a particular pollutant is simply the average concentration of the preprocessed
daily measurements over a given calendar quarter. Quarterly average concentrations include the
substitution of zeros for all non-detects. A site must have a minimum of 75 percent valid samples
compared to the total number of samples possible within a given calendar quarter for a quarterly
average to be calculated. An annual average concentration includes all measured detections and
substituted zeros for non-detects for an entire year of sampling. Annual averages were calculated
for pollutants where three valid quarterly averages could be calculated for a given year and
where method completeness was greater than or equal to 85 percent, as presented in Section 2.4.
Quarterly and annual average concentrations for the pollutants of interest for the Washington
monitoring site are presented in Table 23-3, where applicable. Note that concentrations of the
PAHs and PMio metals are presented in ng/m3 for ease of viewing. Also, note that if a pollutant
was not detected in a given calendar quarter, the quarterly average simply reflects "0" because
only zeros substituted for non-detects were factored into the quarterly average concentration.
23-7
-------
Table 23-3. Quarterly and Annual Average Concentrations of the Pollutants of Interest for the Washington Monitoring Site
2015
2016
# of
# of
Detects/
Qi
Q2
Q3
Q4
Annual
Detects/
Qi
Q2
Q3
Q4
Annual
Pollutant
# >MDL/
# Samples
Avg
frig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Avg
Oig/m3)
Average
frig/m3)
# >MDL/
# Samples
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Avg
frig/m3)
Average
frig/m3)
Seattle, Washington - SEWA
0.78
0.65
0.85
0.54
0.71
0.48
0.54
0.83
0.69
0.64
Acetaldehyde
59/59/59
±0.22
±0.20
±0.21
±0.17
±0.10
61/61/61
±0.22
±0.11
±0.19
±0.25
±0.10
0.75
0.36
0.37
0.60
0.52
0.70
0.32
0.27
0.58
0.47
Benzene
57/57/57
±0.22
±0.06
±0.09
±0.11
±0.07
61/61/61
±0.17
±0.04
±0.05
±0.13
±0.07
0.10
0.04
0.04
0.07
0.06
0.09
0.04
0.03
0.06
0.06
1.3 -Butadiene
55/46/57
±0.04
±0.01
±0.01
±0.02
±0.01
58/21/61
±0.03
±0.01
±0.01
±0.03
±0.01
0.67
0.66
0.70
0.66
0.67
0.68
0.74
0.67
0.70
0.70
Carbon Tetrachloride
57/57/57
±0.03
±0.03
±0.04
±0.03
±0.02
61/61/61
±0.04
±0.03
±0.05
±0.04
±0.02
0.08
0.07
0.04
0.06
0.06
0.07
0.07
0.03
0.06
0.06
1,2 -Dichloroethane
56/45/57
±0.01
± <0.01
±0.01
± <0.01
±<0.01
54/44/61
± <0.01
±0.01
±0.01
±<0.01
±0.01
0.68
0.51
0.72
0.48
0.60
0.37
0.48
0.86
0.90
0.66
Formaldehyde
59/59/59
±0.20
±0.13
±0.23
±0.18
±0.09
61/61/61
±0.15
±0.10
±0.19
±0.25
±0.10
1.27
0.52
0.65
0.67
0.77
0.74
0.64
0.45
0.55
0.60
Arsenic (PMi0)a
58/57/58
±0.44
±0.14
±0.36
±0.23
±0.16
58/58/58
±0.34
±0.13
±0.16
±0.21
±0.12
47.03
37.71
43.42
44.87
43.30
50.64
30.70
42.72
44.96
42.59
Naphthalene1
57/57/57
± 13.61
±9.36
± 11.11
± 11.97
±5.54
61/61/61
± 14.14
±5.43
± 10.59
± 15.34
±6.03
a Average concentrations provided for the pollutants below the blue line are presented in ng/m3 for ease of viewing.
-------
Observations from Table 23-3 include the following:
• Similar to previous years, the annual average concentrations for all of SEWA's
pollutants of interest are less than 1.0 |ig/m3. The pollutants with the highest annual
average concentrations are acetaldehyde, carbon tetrachloride, and formaldehyde.
• Even though acetaldehyde and formaldehyde have some of the highest annual
average concentrations among SEWA's pollutants of interest, the annual averages for
these pollutants are among the lowest for NMP sites sampling carbonyl compounds.
For formaldehyde, only one site (BRCO) has an annual average concentration less
than SEWA's annual averages. For acetaldehyde, the only sites with annual average
concentrations less than SEWA's are located in Garfield County, Colorado. Few
NMP sites have annual average concentrations of these two pollutants less than
1 |ig/m3; SEWA (and BRCO) are the only sites for which both year's annual averages
for acetaldehyde and formaldehyde are less than 1 |ig/m3. Similar observations were
made in previous NMP reports.
• The first and fourth quarter average benzene concentrations for both years are roughly
twice the magnitude of the remaining quarterly average concentrations, indicating
that concentrations of benzene tended to be higher during the colder months of the
year at SEWA. A review of the data shows that benzene concentrations measured at
SEWA span an order of magnitude, ranging from 0.170 |ig/m3 to 1.76 |ig/m3. Of the
39 benzene concentrations greater than 0.5 |ig/m3 measured at SEWA, all but three
were measured during the first (20) or fourth (16) quarters of the years. Conversely,
only one of the 28 concentrations of benzene less than 0.30 |ig/m3 were measured at
SEWA during the first or fourth quarters of 2014. Concentrations of 1,3-butadiene
exhibit a similar pattern, though the differences are not statistically significant.
• Concentrations of 1,2-dichloroethane measured during the third quarter of both years
are lower than those measured during the rest of the year, based on the quarterly
average concentrations. A review of the data shows that all eight non-detects of this
pollutant were measured in August or September, with seven of these measured in
2016. Further, 10 of the 12 lowest measured detections were measured in July,
August, or September of either year, with the two exceptions measured in early
October. Similar observations were made in the 2013 and 2014 NMP reports.
• The first quarter average concentration of arsenic for 2015 is roughly twice the other
quarterly average concentrations calculated for 2015. This is true for most of the
quarterly averages for 2016 as well. Three of the five arsenic concentrations greater
than 2 ng/m3 measured at SEWA were measured during the first quarter of 2015.
Further, one-third of the 24 arsenic concentrations greater than 1 ng/m3 measured
over the two-year period were measured during this calendar quarter.
• For both years, the quarterly average concentration of naphthalene was highest for the
first calendar quarter and lowest for the second quarter, though the difference
between the two is larger for 2016. Concentrations of naphthalene measured at
SEWA range from 13.3 ng/m3 to 129 ng/m3. Confidence intervals associated with the
quarterly averages of naphthalene indicate that there is considerable variability in the
measurements. Naphthalene concentrations greater than 50 ng/m3 were measured
23-9
-------
during seven of the eight calendar quarters, with the exception being the second
quarter of 2016.
• Several of the highest concentrations of the pollutants of interest were measured on
the same day. For example, the highest concentrations of naphthalene, acetaldehyde,
and formaldehyde were all measured at SEWA on November 11, 2016. Higher
concentrations of benzene and 1,3-butadiene were also measured on this date. The
highest concentrations of arsenic, 1,3-butadiene, and 1,2-dichloroethane were all
measured on February 11, 2015, with the second highest benzene concentration also
measured on this date.
Tables 4-10 through 4-13 present the NMP sites with the 10 highest annual average
concentrations for each of the program-level pollutants of interest. Observations for SEWA from
those tables include the following:
• SEWA only appears in Table 4-10 for VOCs for one pollutant; SEWA has the fourth
(2016) and seventh (2015) highest annual average concentrations of carbon
tetrachloride among sites sampling VOCs. SEWA is the first NMP site outside of
Calvert City, Kentucky to appear in Table 4-10 for carbon tetrachloride. A similar
observation was made in the last several NMP reports. However, most of the site-
specific annual average concentrations for carbon tetrachloride span less than
0.1 |ig/m3.
• SEWA does not appear in Table 4-11 for carbonyl compounds. As indicated
previously, SEWA has some of the lowest annual average concentrations of
acetaldehyde and, in particular, formaldehyde among NMP sites sampling these
pollutants.
• SEWA does not appear in Table 4-12 for the PAHs pollutants of interest.
• As shown in Table 4-13, SEWA's annual average concentration of arsenic for 2015
ranks tenth highest among NMP sites sampling metals (PMio), with SEWA's annual
average for 2016 ranking 11th (though not shown in Table 4-13).
23.3.2 Concentration Comparison
In order to better illustrate how each site's annual average concentrations compare to the
program-level averages, a site-specific box plot was created for each of the site-specific
pollutants of interest, where applicable. Thus, box plots were created for the pollutants listed in
Table 23-3 for SEWA. Figures 23-3 through 23-10 overlay the site's minimum, annual average,
and maximum concentrations for each year onto the program-level minimum, first quartile,
median, average, third quartile, and maximum concentrations for each pollutant, as described in
Section 3.4.2.1, and are discussed below. If an annual average concentration could not be
calculated, the range of concentrations are still provided in the figures that follow.
23-10
-------
Figure 23-3. Program vs. Site-Specific Average Acetaldehyde Concentrations
Hi
0 2 4 6 8 10 12 14 16 18
Concentration (]ug/m3)
Program: 1st Quartile
¦
2nd Quartile
~
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
~ ~ 1
Site: 2015 Average
e
2016 Ave rag
o
e Concentration Range, 2015 & 2016
Figure 23-3 presents the box plot for acetaldehyde for SEWA and shows the following:
• The maximum acetaldehyde concentration measured at SEWA is equivalent to the
program-level third quartile (2.11 |ig/m3); few acetaldehyde concentrations measured
at SEWA are greater than the program-level average concentration (1.67 |ig/m3).
• Both of SEWA's annual average acetaldehyde concentrations are less than the
program-level first quartile (0.96 |ig/m3).
• The minimum acetaldehyde concentration measured at SEWA in 2016
(0.0307 |ig/m3) is the second lowest acetaldehyde concentration measured among
NMP sites sampling this pollutant.
Figure 23-4. Program vs. Site-Specific Average Arsenic (PMio) Concentrations
¦ Ii
E H
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile 3rd Quartile 4th Quartile Average
¦
~
~
~
Site: 2015 Average
2016 Average Concentration Range, 2015 & 2016
e
o —
Figure 23-4 presents the box plot for arsenic (PMio) for SEWA and shows the following:
• The range of arsenic concentrations measured at SEWA in 2015 is larger than the
range of concentrations measured in 2016.
• SEWA's annual average concentrations of arsenic fall between the program-level
median concentration and third quartile, with the program-level average concentration
(0.70 ng/m3) falling in between SEWA's two annual averages.
23-11
-------
• There were no non-detects of arsenic measured at SEWA.
Figure 23-5. Program vs. Site-Specific Average Benzene Concentrations
1 i
012345678
Concentration (ng/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
0
o
Figure 23-5 presents the box plot for benzene for SEWA and shows the following:
• All benzene concentrations measured at SEWA in 2015 and 2016 are less than
2 |ig/m3.
• The annual average benzene concentrations for SEWA fall between the program-level
first quartile (0.42 |ig/m3) and the program-level median concentration (0.58 |ig/m3).
SEWA is one of only five NMP sites with an annual average concentration of
benzene less than 0.5 |ig/m3.
Figure 23-6. Program vs. Site-Specific Average 1,3-Butadiene Concentrations
1
SEWA
| Program MaxConcentration = 3.90 (.ig/m3
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75
Concentration (|ug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
23-12
-------
Figure 23-6 presents the box plot for 1,3-butadiene for SEWA and shows the following:
• The program-level maximum 1,3-butadiene concentration (3.90 |ig/m3) is not shown
directly on the box plot in Figure 23-6 because the scale of the box plot would be too
large to readily observe data points at the lower end of the concentration range. Thus,
the scale of the box plot has been reduced.
• The maximum 1,3-butadiene concentration measured at SEWA is an order of
magnitude less than the maximum concentration measured at the program-level.
• Both annual average concentrations of 1,3-butadiene for SEWA are just greater than
the program-level median concentration and less than the program-level average
concentration.
• Five non-detects of 1,3-butadiene were measured at SEWA, two in 2015 and three in
2016.
Figure 23-7. Program vs. Site-Specific Average Carbon Tetrachloride Concentrations
Wr
0 0.5 1 1.5 2 2.5 3 3.5 4
Concentration (jjg/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
e
o
Figure 23-7 presents the box plot for carbon tetrachloride for SEWA and shows the
following:
• The entire range of carbon tetrachloride concentrations measured at SEWA spans less
than 0.5 |ig/m3. The range of concentrations measured in 2015 is smaller than the
range measured in 2016.
• Both annual average concentrations of carbon tetrachloride for SEWA are greater
than the program-level average concentration (0.64 |ig/m3). SEWA's 2016 annual
average is also just greater than the program-level third quartile. However, less than
0.06 |ig/m3 separates SEWA's annual averages and these program-level summary
statistics.
23-13
-------
Figure 23-8. Program vs. Site-Specific Average 1,2-Dichloroethane Concentrations
1 1
9r
i ¦]
Program Max Concentration = 45.8 fig/m3 ¦
&
0.00 0.25 0.50 0.75 1.00 1.25 1.50
Concentration (|ug/m3)
Program: 1st Quartile
2nd Quartile
3rd Quartile 4th Quartile Average
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave age
Concentration Range, 2015 &2016
9
o
Figure 23-8 presents the box plot for 1,2-dichloroethane for SEWA and shows the
following:
• The scale of the box plot in Figure 23-8 has also been reduced to allow for the
observation of data points at the lower end of the concentration range, as the
program-level maximum 1,2-dichloroethane concentration (45.8 |ig/m3) is
considerably greater than the majority of measurements.
• The maximum 1,2-dichloroethane concentration measured at SEWA is equivalent to
the program-level third quartile (0.101 |ig/m3); the remaining 1,2-dichloroethane
concentrations measured at SEWA are less than 0.1 |ig/m3.
• SEWA's annual average concentrations of 1,2-dichloroethane are similar to the
program-level first quartile. Note that the program-level average concentration of
0.30 |ig/m3 is being driven by measurements at the upper end of the concentration
range.
• Eight non-detects of 1,2-dichloroethane were measured at SEWA, one in 2015 and
seven in 2016.
Figure 23-9. Program vs. Site-Specific Average Formaldehyde Concentrations
Et
0 3 6 9 12 15 18 21 24 27
Concentration (]ug/m3)
Program: 1st Quartile
2nd Quartile
3rd Qu a rti 1 e 4th Qu a rti 1 e Ave ra ge
¦
~
~ ~ 1
Site: 2015 Average
2016 Ave rag
e Concentration Range, 2015 & 2016
O
o
23-14
-------
Figure 23-9 presents the box plot for formaldehyde for SEWA and shows the following:
• The entire range of formaldehyde concentrations measured at SEWA is less than the
program-level median concentration of 2.47 |ig/m3. A single formaldehyde
concentration measured at SEWA is greater than 2 |ig/m3.
• Both annual average concentrations for SEWA are less than the program-level first
quartile (1.54 |ig/m3); SEWA's annual averages are less than half the program-level
first quartile. One only NMP site (BRCO) has an annual average concentration less
than SEWA's annual averages.
Figure 23-10. Program vs. Site-Specific Average Naphthalene Concentrations
^1—
0 50 100 150 200 250 300 350 400 450
Concentration (ng/m3)
Program: IstQuartile 2nd Quartile 3rd Quartile 4th Quartile Average
¦ ~ ~ ~ i
Site: 2015Average 2016Avereage Concentration Range, 2015 & 2016
e o
Figure 23-10 presents the box plot for naphthalene for SEWA and shows the following:
• The two naphthalene concentrations greater than 100 ng/m3 measured at SEWA were
both measured in 2016. These measurements are considerably less than the maximum
naphthalene concentration measured across the program.
• SEWA's annual average concentrations of naphthalene for 2015 and 2016 are similar
to each other, both of which are just less than the program-level median concentration
of 48.90 ng/m3.
23.3.3 Concentration Trends
A site-specific trends evaluation was completed for sites that have sampled one or more
of the pollutants of interest for 5 consecutive years or longer, as described in Section 3.4.2.2.
Sampling for PMio metals, VOCs, and carbonyl compounds under the NMP began at SEWA in
2007 and sampling for PAHs began in 2008. Thus, Figures 23-11 through 23-18 present the
1-year statistical metrics for each of the pollutants of interest for SEWA. If sampling began mid-
year, a minimum of 6 months of sampling is required for inclusion in the trends analysis; in these
cases, a 1-year average concentration is not provided, although the range and percentiles are still
presented.
23-15
-------
Figure 23-11. Yearly Statistical Metrics for Acetaldehyde Concentrations Measured at
SEWA
2011 2012
Year
O 5th Percentile
O 95th Percentile
Observations from Figure 23-11 for acetaldehyde concentrations measured at SEWA
include the following:
• The maximum acetaldehyde concentration was measured at SEWA on July 17, 2007
(9.73 |ig/m3). The next highest concentration is considerably less (3.38 |ig/m3,
measured in September 2009). One additional acetaldehyde concentration greater
than 3 |ig/m3 has been measured at SEWA (September 2012).
• The 1-year average concentration has a slight undulating pattern through 2012, with
years with slightly lower concentrations alternating with years with slightly higher
concentrations. Through 2012, the 1-year average concentrations ranged from
0.74 |ig/m3 (2012) to 0.98 |ig/m3 (2009). The median concentration exhibits a steady
increasing trend for the first 5 years of sampling, ranging from 0.61 |ig/m3 (2007) to
0.85 |ig/m3 (2011), before decreasing considerably for 2012.
• Although the 1-year average concentration has varied by only 0.1 |ig/m3 over the last
five years of sampling, the 1-year average for 2016 is at a minimum compared to
other years of sampling (0.64 |ig/m3). In fact, several of the statistical parameters are
at a minimum for 2016, including the 5th and 95th percentiles and the 1-year average
and median concentrations.
23-16
-------
Figure 23-12. Yearly Statistical Metrics for Arsenic (PMio) Concentrations Measured at
SEWA
o
•O"
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
O 95th Pe r ce nti le Ave r z
Observations from Figure 23-12 for arsenic (PMio) concentrations measured at SEWA
include the following:
• The maximum arsenic concentration was measured at SEWA on February 11, 2015
(2.73 ng/m3), although a similar concentration was also measured in 2009
(2.69 ng/m3). In total, 16 arsenic concentrations greater than 2 ng/m3 have been
measured at SEWA, at least one in each year of sampling except 2014.
• There have been no non-detects of arsenic measured at SEWA since the onset of
sampling, including 2008 and 2014, when the minimum concentration is less than
other years. For these two years, the minimum concentration of arsenic is around
0.01 ng/m3.
• Despite the fluctuations shown, the 1-year average concentration of arsenic for
SEWA has only varied by about 0.2 ng/m3, ranging from 0.58 ng/m3 (2010) to
0.79 ng/m3 (2013). Confidence intervals indicate that the changes are not statistically
significant. The median concentration has varied less, from 0.47 ng/m3 (2016) to
0.63 ng/m3 (2013); 2016 is the first time that the median concentration is less than
0.5 ng/m3.
23-17
-------
Figure 23-13. Yearly Statistical Metrics for Benzene Concentrations Measured at SEWA
2008 2009
2010 2011 2012 2013
Year
2014 2015
O 5th Percentile — Minimum
Median — Maximum o 95th Percentile Averc
Observations from Figure 23-13 for benzene concentrations measured at SEWA include
the following:
• The maximum benzene concentration was measured at SEWA on January 19, 2009
(5.38 |ig/m3), which is the same day the maximum arsenic concentration was
measured. The next highest concentration is roughly half as high (2.55 |ig/m3,
measured in January 2011). In total, five benzene concentrations greater than 2 |ig/m3
have been measured at SEWA.
• Overall, benzene concentrations have a slight decreasing trend at SEWA. Despite
differences in the magnitude of concentrations measured, little change in the 1-year
average concentration is shown for the first three years of sampling. The year with the
largest range of benzene concentrations measured (2009) is followed by the year with
the small range of concentrations measured (2010). The range of measurements
expands again for 2011; between 2012 and 2016, the range of concentrations
measured each year changes little. The 1-year average concentration of benzene is
highest for 2009 (0.81 |ig/m3) and lowest for 2016, when the 1-year average is less
than 0.50 |ig/m3 for the first time (0.47 |ig/m3).
• Concentrations of benzene appear to have a seasonal trend at SEWA. Of the
72 benzene concentrations greater than 1 |ig/m3, 61 have been measured during the
colder months of the year, either during the first quarter (27) or fourth quarter (34) of
any given year. Conversely, few of the lower benzene concentrations have been
measured during the warmer months of the year. Of the 71 benzene concentrations
23-18
-------
less than 0.3 |ig/m3 measured at SEW A, 66 have been measured during the warmer
months of the year, during the second (23) and third (43) quarters of the year.
Figure 23-14. Yearly Statistical Metrics for 1,3-Butadiene Concentrations Measured at
SEWA
¦¦
I
I
O"
o
rh
L~2—
L£"
'-s
1"
^
«
Lr
\a\
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
Year
Observations from Figure 23-14 for 1,3-butadiene concentrations measured at SEWA
include the following:
• The maximum 1,3-butadiene concentration (0.89 |ig/m3) was measured at SEWA on
the same day as the maximum arsenic and benzene concentrations were measured,
January 19, 2009. The next highest concentration was approximately half as high
(0.47 |ig/m3) and was measured on the same day in January 2011 as the second
highest benzene concentration.
• At least one non-detect of 1,3-butadiene has been measured at SEWA each year since
the onset of sampling, with the exception of 2007, as indicated by the minimum
concentration. For 2010, 2011, 2013, and 2014, both the minimum and 5th percentile
are zero, indicating that at least 5 percent of the measurements were non-detects. For
each of these years, the percentage of non-detects ranged from 10 percent to
15 percent.
• The 1-year average concentration has an undulating pattern over the first several
years of sampling, when a year with a higher 1-year average concentration is
followed by a year with a lower 1-year average. Between 2007 and 2011, the 1-year
average concentration ranged from 0.06 |ig/m3 (2008) to 0.09 |ig/m3 (2011).
23-19
-------
• After 2011, the 1-year average concentration has a steady decreasing trend, and is at a
minimum for 2016. The median concentration exhibits a similar trend.
Figure 23-15. Yearly Statistical Metrics for Carbon Tetrachloride Concentrations
Measured at SEWA
~
I
-
rn
"O"
2011 2012
Year
O 5th Percentile
— Minimum
— Maximum
O 95th Percentile
Observations from Figure 23-15 for carbon tetrachloride concentrations measured at
SEWA include the following:
• The maximum carbon tetrachloride concentration (1.22 |ig/m3) has been measured
twice at SEWA, once in 2008 and once in 2010. Nineteen concentrations of carbon
tetrachloride greater than 1.0 |ig/m3 have been measured at SEWA since the onset of
sampling in 2007. All but two of these were measured in 2008 and 2009, with one
each in 2010 and 2013.
• All of the statistical metrics increased from 2007 to 2008. Eleven concentrations
measured in 2008 were greater than the maximum concentration measured in 2007. In
addition, the number of carbon tetrachloride concentrations greater than 0.75 |ig/m3
increased from 12 in 2007 to 43 for 2008.
• A steady decreasing trend in the concentrations is shown between 2008 and 2011,
when the 1-year average concentration is at a minimum (0.65 |ig/m3).
• The range of measurements tightened considerably for 2012. Yet, both the 1-year
average and median concentrations exhibit significant increases. As the number of
concentrations falling into the 0.65 |ig/m3 to 0.85 |ig/m3 range increased, from 29 for
23-20
-------
2011 to 42 in 2012, the number of concentrations less than 0.6 |ig/m3 fell from 20 to
seven during this time frame.
• After 2012, the 1-year average concentration exhibits slight decreases each year, until
returning to 2012 levels for 2016.
Figure 23-16. Yearly Statistical Metrics for 1,2-Dichloroethane Concentrations Measured
at SEW A
o-
2007 2008
o
2009 2010
o
2011 2012
Year
o*
o
2013 2014 2015 2016
O 5th Percentile
Maximum O 95th Percentile Avera
Observations from Figure 23-16 for 1,2-dichloroethane concentrations measured at
SEWA include the following:
• The minimum, 5th percentile, and median concentrations are zero for 2007 through
2011. This indicates that at least half of the measurements were non-detects. In 2008,
there were no measured detections of 1,2-dichloroethane. The percentage of measured
detections in 2007 and 2009 was around 10 percent, after which the percentage
increased. For 2012, the percentage of measured detections is 93 percent, a
considerable increase from 26 percent in 2011. This percentage leveled off somewhat
for 2013 and 2014 (at 88 percent each) before reaching a maximum of 98 percent for
2015 (note the 5th percentile is greater than zero for the first time). The percentage of
measured detections decreased to 88 percent for 2016.
• As the number of measured detections increased, particularly for 2012 and the years
that follow, the median and 1-year average concentrations increased correspondingly.
• The median concentration is greater than the 1-year average concentration for each of
the last five years of sampling. This is because there were still several non-detects (or
23-21
-------
zeros) factoring into the 1-year average concentration for these years, which can pull
an average down in a similar manner that an outlier can drive an average upward,
while the range of measured detections is rather small.
Figure 23-17. Yearly Statistical Metrics for Formaldehyde Concentrations Measured at
SEWA
Maximum
Concentration for
2009 is 16.6 (Jg/m3
—J—
1
X
2009 2010 2011 2012 2013 2014 2015 2016
Year
O 5th Percentile
— Minimum — Median — Maximum o 95th Percentile Avera
Observations from Figure 23-17 for formaldehyde concentrations measured at SEWA
include the following:
• The maximum formaldehyde concentration measured at SEWA was measured on
January 13, 2009 (16.6 |ig/m3). The next highest concentration (9.44 |ig/m3) was
measured on the same day in 2007 as the maximum acetaldehyde concentration. Only
one other formaldehyde concentration greater than 3 |ig/m3 has been measured at
SEWA and was also measured in 2009. In total, 10 concentrations greater than
2 |ig/m3 have been measured at SEWA since the onset of carbonyl compound
sampling, and all but two of these were measured prior to 2010.
• The 1-year average concentration has an undulating pattern through 2012, with a
"down" year followed by an "up" year. Between 2007 and 2012, the 1-year average
concentration varied considerably, from 0.53 |ig/m3 (2012) to 1.04 |ig/m3 (2009). The
median concentration is static during the first four years of sampling, with each year's
median just less than 0.6 |ig/m3, after which it too exhibits the up/down pattern.
• The 1-year average formaldehyde concentration exhibits a very subtle increase
between 2012 and 2016, although the changes are not statistically significant.
23-22
-------
Figure 23-18. Yearly Statistical Metrics for Naphthalene Concentrations Measured at
SEWA
o
O:
2012
Year
-2
2014
O 5th Percentile
Maximum O 95th Percentile Avera
1 A 1-year average is not presented because sampling under the NMP did not begin until March 2008.
Observations from Figure 23-18 for naphthalene concentrations measured at SEWA
include the following:
• SEWA began sampling PAHs under the NMP in March 2008. Because a full year's
worth of data is not available, a 1-year average concentration is not presented for
2008, although the range of measurements is provided.
• The maximum naphthalene concentration measured at SEWA was measured in 2011
(317 ng/m3). This is the only naphthalene concentration greater than 250 ng/m3
measured at this site. Eight additional concentrations greater than 200 ng/m3 have
been measured at SEWA, with each measured between 2009 and 2013.
• Each of the statistical parameters shown exhibits an increase from 2008 to 2009.
Although the range of concentrations measured in 2009 is similar to those measured
in 2010, the 95th percentile decreased by almost half from one year to the next. The
number of naphthalene concentrations greater than 100 ng/m3 decreased by nearly
two-thirds, decreasing from 19 in 2009 to seven in 2010.
• With the exception of the median concentration, each of the statistical parameters
exhibits an increase for 2011, with the 1-year average concentration nearly returning
to 2009 levels. This is partially driven by the maximum concentration measured in
2011. Another factor is the minimum concentration measured in 2011, which is
considerably higher than the minimum concentration measured in the previous two
23-23
-------
years. The median concentration decreased each year between 2009 and 2011, as
naphthalene concentrations less than 50 ng/m3 account for an increasing number of
measurements during this time.
• Little change in the 1-year average concentration is shown between 2011 and 2013,
after which a significant decrease is shown for 2014, when the 1-year average
concentration is less than 50 ng/m3 for the first time (as is the median concentration).
Only one naphthalene concentration greater than 100 ng/m3 was measured in 2014,
with most of the previous years having 10 or more. Additional decreases are shown
for 2015 (when concentrations greater than 100 ng/m3 were not measured) and 2016
(when the 1-year average concentration is at a minimum). With the exception of the
minimum concentration, each of the statistical parameters is at a minimum in 2014
(5th percentile), 2015 (median, maximum, and 95th percentile), or 2016 (1-year
average).
23.4 Additional Risk-Based Screening Evaluations
The following risk-based screening evaluations were conducted to characterize risk
related to the air toxics measured at the Washington monitoring site. Refer to Sections 3.2,
3.4.2.3, and 3.4.2.4 for definitions and explanations regarding the various toxicity factors, time
frames, and calculations associated with these risk-based screenings.
23.4.1 Cancer Risk and Noncancer Hazard Approximations
For the pollutants of interest for the Washington site, risk was examined by calculating
cancer risk and noncancer hazard approximations for each year annual average concentrations
could be calculated. These approximations can be used as risk estimates for cancer and
noncancer effects attributable to the pollutants of interest. Although the use of these
approximations is limited, they may help identify where policy-makers may want to shift their
air monitoring priorities. Refer to Section 3.4.2.3 for an explanation of how cancer risk and
noncancer hazard approximations are calculated and what limitations are associated with them.
Annual averages, cancer UREs and/or noncancer RfCs, and cancer risk and noncancer hazard
approximations are presented in Table 23-4, where applicable. Cancer risk approximations are
presented as probabilities while the noncancer hazard approximations are ratios and thus, unitless
values.
23-24
-------
Table 23-4. Risk Approximations for the Washington Monitoring Site
Pollutant
Cancer
URE
(jig/m3)1
Noncancer
RfC
(mg/m3)
2015
2016
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
# of
Measured
Detections
vs. # of
Samples
Annual
Average
frig/m3)
Risk Approximations
Cancer
(in-a-million)
Noncancer
(HQ)
Cancer
(in-a-million)
Noncancer
(HQ)
Seattle, Washington - SEWA
Acetaldehyde
0.0000022
0.009
59/59
0.71
±0.10
1.55
0.08
61/61
0.64
±0.10
1.40
0.07
Benzene
0.0000078
0.03
57/57
0.52
±0.07
4.03
0.02
61/61
0.47
±0.07
3.66
0.02
1.3 -Butadiene
0.00003
0.002
55/57
0.06
±0.01
1.90
0.03
58/61
0.06
±0.01
1.72
0.03
Carbon Tetrachloride
0.000006
0.1
57/57
0.67
±0.02
4.03
0.01
61/61
0.70
±0.02
4.17
0.01
1,2 -Dichloroethane
0.000026
2.4
56/57
0.06
±<0.01
1.67
<0.01
54/61
0.06
±0.01
1.51
<0.01
Formaldehyde
0.000013
0.0098
59/59
0.60
±0.09
7.75
0.06
61/61
0.66
±0.10
8.53
0.07
Arsenic (PMi0)a
0.0043
0.000015
58/58
0.77
±0.16
3.32
0.05
58/58
0.60
±0.12
2.56
0.04
Naphthalene1
0.000034
0.003
57/57
43.30
±5.54
1.47
0.01
61/61
42.59
±6.03
1.45
0.01
a Average concentrations provided below the blue line for this pollutant are presented in ng/m3 for ease of viewing.
-------
Observations from Table 23-4 for SEWA include the following:
• The pollutants with the highest annual average concentrations for SEWA are
acetaldehyde, carbon tetrachloride, and formaldehyde, although the order varies by
year.
• Despite its comparatively low annual averages, the pollutant with the highest cancer
risk approximation for each year is formaldehyde. SEWA's cancer risk
approximations for formaldehyde are among the lowest site-specific cancer risk
approximations for this pollutant; SEWA is one of only two sites for which the cancer
risk approximations for formaldehyde are less than 10 in-a-million.
• The noncancer hazard approximations for SEWA are all considerably less than 1.0,
with the highest calculated for acetaldehyde (0.08) for 2015. This indicates that no
adverse noncancer health effects are expected from these individual pollutants.
23.4.2 Risk-Based Emissions Assessment
In addition to the risk-based screening discussed above, this section presents an
evaluation of county-level emissions based on cancer and noncancer toxicity, respectively.
Table 23-5 presents the 10 pollutants with the highest emissions from the 2014 NEI (version 1)
that have cancer toxicity factors. Table 23-5 also presents the 10 pollutants with the highest
toxicity-weighted emissions, based on the weighting schema described in Section 3.4.2.4. Lastly,
Table 23-5 provides the pollutants of interest with the highest cancer risk approximations (in-a-
million) for SEWA, as presented in Table 23-4. Cancer risk approximations for 2015 are
presented in green while approximations for 2016 are in white. The emissions, toxicity-weighted
emissions, and cancer risk approximations are shown in descending order in Table 23-5.
Table 23-6 presents similar information but is limited to those pollutants with noncancer toxicity
factors.
Because not all pollutants have both cancer and noncancer toxicity factors, the highest
emitted pollutants in the cancer table may be different from the noncancer table, although the
actual quantity of emissions is the same. The cancer risk and noncancer hazard approximations
based on each site's annual averages are limited to the pollutants of interest identified for each
site. In addition, the cancer risk and noncancer hazard approximations are limited to those
pollutants with enough data to meet the criteria for annual averages to be calculated. A more in-
depth discussion of this analysis is provided in Section 3.4.2.4. Similar to the cancer risk and
noncancer hazard approximations provided in Section 23.4.1, this analysis may help policy-
makers prioritize their air monitoring activities.
23-26
-------
Table 23-5. Top 10 Emissions, Toxicity-Weighted Emissions, and Cancer Risk Approximations for Pollutants with Cancer UREs for
the Washington Monitoring Site
Top 10 Total Emissions for Pollutants with
Cancer UREs
(County-Level)
Top 10 Cancer Toxicity-Weighted
Emissions
(County-Level)
Top 10 Cancer Risk Approximations Based on
Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Cancer
Toxicity
Weight
Pollutant
Cancer Risk
Approximation
(in-a-million)
Seattle, Washington (King County) - SEWA
Benzene
814.74
Formaldehyde
1.01E-02
Formaldehyde
8.53
Formaldehyde
780.67
Benzene
6.35E-03
Formaldehyde
7.75
Acetaldehyde
463.17
POM, Group 3
5.39E-03
Carbon Tetrachloride
4.17
Ethylbenzene
376.38
Naphthalene
4.82E-03
Carbon Tetrachloride
4.03
Naphthalene
141.81
1,3-Butadiene
3.71E-03
Benzene
4.03
1.3 -Butadiene
123.76
Hexavalent Chromium PM
2.51E-03
Benzene
3.66
Bis(2-ethylhexyl) phthalate, gas
95.13
POM, Group 2b
1.73E-03
Arsenic
3.32
POM, Group 2b
19.63
POM, Group 5a
1.24E-03
Arsenic
2.56
POM, Group 2d
13.43
POM, Group 2d
1.18E-03
1,3-Butadiene
1.90
2,4-Toluene diisocyanate
8.14
Acetaldehyde
1.02E-03
1,3-Butadiene
1.72
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Table 23-6. Top 10 Emissions, Toxicity-Weighted Emissions, and Noncancer Hazard Approximations for Pollutants with Noncancer
RfCs for the Washington Monitoring Site
Top 10 Total Emissions for Pollutants with
Noncancer RfCs
(County-Level)
Top 10 Noncancer Toxicity-Weighted Emissions
(County-Level)
Top 10 Noncancer Hazard Approximations
Based on Annual Average Concentrations
(Site-Specific)1
Pollutant
Emissions
(tpy)
Pollutant
Noncancer
Toxicity
Weight
Pollutant
Noncancer Hazard
Approximation
(HQ)
Seattle, Washington (King County) - SEWA
Toluene
2,384.23
Acrolein
2,917,859.78
Acetaldehyde
0.08
Xylenes
1,360.91
2,4-Toluene diisocyanate
116,325.71
Acetaldehyde
0.07
Methanol
965.12
Formaldehyde
79,660.05
Formaldehyde
0.07
Benzene
814.74
1,3-Butadiene
61,882.43
Formaldehyde
0.06
Formaldehyde
780.67
Acetaldehyde
51,463.20
Arsenic
0.05
Acetaldehyde
463.17
Cyanide Compounds, gas
51,064.88
Arsenic
0.04
Hexane
437.73
Naphthalene
47,269.92
1,3-Butadiene
0.03
Ethylbenzene
376.38
Benzene
27,158.03
1,3-Butadiene
0.03
Naphthalene
141.81
Lead, PM
15,660.71
Benzene
0.02
Ethylene glycol
132.33
Nickel, PM
14,362.01
Benzene
0.02
1 Green shading represents a risk approximation based on a 2015 annual average concentration and no shading represents a 2016 risk approximation.
-------
Observations from Table 23-5 for SEWA include the following:
• Benzene, formaldehyde, and acetaldehyde are the highest emitted pollutants with
cancer UREs in King County.
• The pollutants with the highest toxicity-weighted emissions (of the pollutants with
cancer UREs) for King County are formaldehyde, benzene, and POM, Group 3.
• Seven of the highest emitted pollutants also have the highest toxicity-weighted
emissions for King County.
• Formaldehyde has the highest cancer risk approximations for SEWA, the highest
toxicity-weighted emissions for King County, and ranks second for quantity emitted.
Benzene and 1,3-butadiene also appear on all three lists.
• Carbon tetrachloride and arsenic, which also have some of the highest cancer risk
approximations for SEWA, do not appear on either emissions-based list.
• Several POM Groups appear on the emissions-based lists for King County, including
POM, Group 2b. POM, Group 2b includes several PAHs sampled for at SEWA
including acenaphthene and fluorene, both of which failed a single a screen (and thus,
were not identified as pollutants of interest for SEWA).
Observations from Table 23-6 for SEWA include the following:
• Toluene, xylenes, and methanol are the highest emitted pollutants with noncancer
RfCs in King County.
• Acrolein is the pollutant with the highest toxicity-weighted emissions (of the
pollutants with noncancer RfCs) for King County, followed by 2,4-toluene
diisocyante and formaldehyde. Although acrolein was sampled for at SEWA, this
pollutant was excluded from the pollutants of interest designation, and thus
subsequent risk-based screening evaluations, due to questions about the consistency
and reliability of the measurements, as discussed in Section 3.2.
• Four of the highest emitted pollutants also have the highest toxicity-weighted
emissions for King County.
• Acetaldehyde, formaldehyde, and benzene appear on all three lists in Table 23-6.
• 1,3-Butadiene is also a pollutant of interest for SEWA that also appears among those
with the highest toxicity-weighted emissions but does not appear among the highest
emitted (of those with a noncancer RfC).
• Arsenic is a pollutant of interest for SEWA that appears on neither emissions-based
list.
23-29
-------
23.5 Summary of the 2015-2016 Monitoring Data for SEWA
Results from several of the data analyses described in this section include the following:
~~~ Fifteen pollutants failed at least one screen for SEWA.
~~~ All of the pollutants of interest for SEWA have annual average concentrations less
than 1 ng/m3.
~~~ The annual average concentrations of several of SEWA's pollutants of interest are
among the lowest compared to other NMP sites sampling these pollutants.
~~~ Concentrations of benzene exhibit an overall decreasing trend over the period
sampling period at SEWA. Concentrations of 1,3-butadiene and naphthalene have a
decreasing trend in recent years.
23-30
-------
24.0 Data Quality
This section discusses the data quality of the ambient air measurements that constitute the
2015-2016 NMP dataset. Each monitoring program under the NMP has its own specific Data
Quality Objectives (DQOs) which have been established and approved by EPA, consistent with
the specific data use needs of the individual monitoring programs. Because the DQOs are
program-specific and the ERG laboratory is contracted to perform services for a subset of the
overall program participants, attainment of the individual program DQO(s) is not assessed in this
report. This section establishes data quality through the assessment of Data Quality Indicators
(DQI) in the form of MQOs specific to the program elements conducted by the ERG laboratory.
MQOs are established to control and evaluate the various phases of the measurement process
(sample collection, preparation, and analysis) to ensure that the total measurement quality meets
the overall program data quality objectives. In accordance with ERG's EPA-approved QAPP
(ERG, 2015 and ERG, 2016), the following MQOs were assessed: completeness, precision, and
accuracy (also called bias).
The quality assessments presented in this section show that the 2015-2016 monitoring
data are of a known and high quality, consistent with the intended data use. The overall method-
specific completeness was greater than 85 percent for each method. The method precision for
collocated and duplicate analyses met the precision MQO of 15 percent Coefficient of Variation
(CV) for most methods, with the exceptions of TO-13A for PAHs (which is just outside the
15 percent MQO). The analytical precision for replicate analyses for all methods met the
precision MQO of 15 percent CV, with all methods less than 7 percent. Audit samples show that
ERG is meeting the accuracy requirements of the NATTS TAD (EPA, 2009a). These data
quality indicators are discussed in further detail in the following sections.
24.1 Completeness
Completeness refers to the number of valid samples successfully collected and analyzed
compared to the number of total samples scheduled to be collected and analyzed. The MQO for
completeness based on the EPA-approved QAPP specifies that at least 85 percent of samples
collected at a given monitoring site must be analyzed successfully to be considered sufficient for
data trends analysis (ERG, 2015 and ERG, 2016). The MQO of 85 percent completeness was
met by 204 of the 215 site-method-year datasets while 11 datasets (seven from 2015 and four
from 2016) did not. Completeness statistics are presented and discussed more thoroughly in
Section 2.4.
24-1
-------
24.2 Method Precision
Precision defines the level of agreement between independent measurements performed
according to identical protocols and procedures. Method precision, which includes sampling and
analytical precision, quantifies random errors associated with collecting ambient air samples in
the field, transporting them to the laboratory, and analyzing the samples in the laboratory.
Method precision is evaluated by comparing concentrations measured in duplicate or collocated
samples. A duplicate sample is a sample collected simultaneously with a primary sample through
a common inlet probe such that the same air parcel is being sampled. This simultaneous
collection is typically achieved by teeing the line from the collection system to two canisters (or
other sampling media) and doubling the flow rate applied to achieve integration over the 24-hour
collection period. Collocated samples are samples collected simultaneously through separate
inlet probes, regardless of collection system set-up (i.e., either two separate sampling systems or
a single sampling system with multiple inlets). Because the samples are not collected using a
common inlet, the system is sampling potentially different air parcels. The overarching
difference between the two sample types is whether or not the potential for non-homogeneity of
the air parcel is being considered as part of the precision calculation. Duplicate samples provide
an indication of "intra-system" variability while collocated samples provide an indication of
"inter-system" variability, of which the non-homogeneity of the air parcels sampled factors into
the level of precision measured.
During the 2015 and 2016 monitoring efforts, duplicate or collocated samples were
collected on at least 10 percent of the scheduled sample days, where possible, as outlined in the
EPA-approved QAPP. This provides a minimum of six pairs of either duplicate or collocated
samples per site and method per year. For the VOC, SNMOC, and carbonyl compound methods,
samples may be duplicate or collocated. For PAHs, metals, and hexavalent chromium, only
collocated samples may be collected due to limitations of the sampling media/instrumentation.
For each method, these duplicate or collocated samples were then analyzed in replicate at the
laboratory (nested approach). Replicate measurements are repeated analyses performed on a
duplicate or collocated pair of samples and are discussed in greater detail in Section 24.3. Where
duplicate or collocated events were not possible at a given monitoring site, replicate analyses
were run on individual samples to provide an indication of analytical precision, and are discussed
further in Section 24.3.
24-2
-------
Method precision is calculated by comparing the concentrations of the
duplicates/collocates for each pollutant. The CV for duplicate or collocated samples was
calculated for each pollutant and each site. The following approach was employed to estimate
how closely the collected and analyzed samples agree with one another:
Coefficient of Variation (CV) provides a relative measure of variability. CV is often
expressed as a ratio of the standard deviation and the mean, and is used for a single
variable. The CV listed below is ideal when comparing paired values, such as a primary
concentration and a duplicate concentration, and has been used to evaluate NATTS data
for years (EPA, 2017g). A coefficient of variation of 1 percent would indicate that the
analytical results could vary slightly due to sampling error, while a variation of
50 percent means that the results are more imprecise.
CVs were calculated for every pair of duplicate or collocated samples where both
measurements were at or above the MDL. Thus, the number of pairs included in the calculations
varies significantly from pollutant to pollutant. To make an overall estimate of method precision,
program-level average CVs were calculated as follows:
• A site-specific CV was calculated for each pollutant, per the equation above.
• A pollutant-specific average CV was calculated for each method.
• A method-specific average CV was calculated and compared to the precision MQO.
Table 24-1 presents the 2015-2016 NMP method precision for VOCs, SNMOCs,
methane, carbonyl compounds, PAHs, metals, and hexavalent chromium, presented as the
average CV and expressed as a percentage. CVs exceeding the 15 percent MQO are bolded in
the table. Six of the seven analytical methods met the program MQO of 15 percent CV for
precision. TO-13A/PAH results did not meet the MQO of 15 percent, although they are just
outside the criteria (and are discussed further in the individual method sections). This table also
includes the number of pairs that were included in the calculation of the method precision. The
total number of pairs including those with concentrations less than the MDL (and with two
numerical results) is also included in Table 24-1 for each method to provide an indication of the
Where:
p = the primary result from a duplicate or collocated pair;
r = the secondary result from a duplicate or collocated pair;
n = the number of valid data pairs (the 2 adjusts for the fact that there are two
values with error).
24-3
-------
effect that excluding those with concentrations less than the MDL has on the population of pairs
in the dataset. For some methods, such as TO-11A for carbonyl compounds, the difference is
small; for others, such as TO-15 for VOCs, the difference is relatively large.
Table 24-1. Method Precision by Analytical Method
Method/Pollutant
Group
Average
Coefficient of
Variation
(%)
Number of
Pairs Included
in the
Calculation
Total Number
of Pairs Without
the > MDL
Exclusion
VOC
(TO-15)
9.27
5,596
8,073
SNMOC
12.23
1,694
1,931
Methane
3.70
13
13
Carbonyl Compounds
(TO-11 A)
6.85
3,418
3,428
PAHs
(TO-13A)
15.27
583
796
Metals Analysis
(Method IO-3.5/FEM)
12.85
3,486
4,101
Hexavalent Chromium
(ASTM D7614)
13.48
5
8
MQO
15.00 percent CV
Bold = CV greater than or equal to 15 percent
Tables 24-2 through 24-8 present method precision for VOCs, SNMOCs, methane,
carbonyl compounds, PAHs, metals, and hexavalent chromium, respectively, as the CV per
pollutant per site and the average CV per site, per pollutant, and per method. Also included in
these tables is the number of duplicate and/or collocated pairs included in the CV calculations.
For methods where duplicate or collocated samples are both possible, the type of sample
collected at each site is identified and the average CV based on sample type is also provided.
CVs exceeding the 15 percent MQO are bolded in each table. The CVs that exceed the program
MQO for precision are often driven by relatively low concentrations, even though they are
greater than the MDL.
24.2.1 VOC Method Precision
Table 24-2 presents the method precision for all duplicate and collocated VOC samples
as the CV per pollutant per site, the average CV per site, the average CV per pollutant, and the
overall average CV across the VOCs listed. The duplicate and collocated sample precision
results exhibit low- to high-level variability, where the CV ranges from 0 percent (a few
pollutants for several sites) to 88.50 percent (dichloromethane for GLKY). The CV for
24-4
-------
dichloromethane for GLKY is based on 11 pairs of samples greater than the MDL. For eight of
these 11 pairs, the duplicate sample was considerably greater than the primary sample. The
number of sites for which a given pollutant has a CV greater than or equal to 15 percent varies,
from none (28 pollutants) to 19 (methyl isobutyl ketone). Propylene and chloroethane are the
only other pollutants besides methyl isobutyl ketone with a CV greater than or equal to
15 percent for at least 10 sites.
The pollutant-specific average CV, as shown in orange in Table 24-2, ranges from
0 percent (chlorobenzene and 1,1,1 -trichloroethane) to 24.19 percent (methyl isobutyl ketone).
For both chlorobenzene and 1,1,1-trichloroethane, the precision is based on a single pair of
measurements greater than the MDL. The site-specific average CV, as shown in green in
Table 24-2, ranges from 6.31 percent (NBNJ) to 16.49 percent (GLKY). GLKY is the only site
with a site-specific average CV greater than or equal tol5 percent. Note that TVKY collected
collocated samples more frequently than the 10 percent requirement, as indicated by the number
of collocated sample pairs collected at this site (55). The overall average method precision for
VOCs is 9.27 percent, meeting the MQO of 15 percent. Note that the results for acrolein,
acetonitrile, acrylonitrile, and carbon disulfide were excluded from the precision calculations due
to the issues described in Section 3.2.
Sites at which duplicate samples were collected are highlighted in blue in Table 24-2
while sites at which collocated samples were collected are highlighted in purple. Collocated
VOC samples were collected at two of the sites shown in Table 24-2 (PXSS and TVKY);
duplicate VOC samples were collected at the remainder of sites. The average CV for sites that
collected duplicate samples was calculated and is shown at the end of Table 24-2 in blue while
the average CV for sites collecting collocated samples is shown in purple. The average CV for
both precision types is less than 15 percent, with less than 1 percent separating them; the
variability associated with collocated samples (10.60 percent) is slightly more than the variability
associated with duplicate samples (9.91 percent).
24-5
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
BROK
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Acetylene
6.56
7.59
27.32
3.94
3.97
4.71
4.49
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
Benzene
7.09
12.85
6.60
3.39
15.30
3.75
28.64
Bromochloromethane
—
—
6.43
—
18.86
—
0.00
Bromodichloro methane
—
—
—
—
—
—
—
Bromofonn
—
—
—
—
—
—
—
Bromomethane
10.65
3.87
7.54
7.65
10.91
6.43
8.71
1.3 -Butadiene
—
20.74
8.33
6.46
9.69
7.07
—
Carbon Tetrachloride
11.82
18.55
4.73
33.76
5.77
17.56
5.88
Chlorobenzene
—
—
—
—
—
—
—
Chloroethane
18.55
15.61
20.29
14.46
7.82
16.29
12.10
Chloroform
7.33
5.67
4.85
4.37
21.57
4.65
12.80
Chloromethane
6.58
6.16
5.36
5.80
4.21
5.29
6.45
Chloroprene
—
—
—
—
—
—
—
Dibromochloromethane
—
—
—
—
—
—
—
1,2 -Dibromoethane
—
—
—
—
—
—
—
w-Dichlorobenzene
—
—
—
—
—
—
—
o-Dichlorobenzene
—
—
—
—
—
—
—
p-Dichlorobcnzcnc
—
—
—
—
—
4.29
—
Dichlorodifluoromethane
7.22
8.79
3.10
3.42
3.24
4.58
4.65
1,1 -Dichloroethane
—
—
—
—
—
—
—
1,2 -Dichloroethane
8.28
9.48
7.68
4.84
4.59
7.57
7.97
1,1 -Dichloroethene
—
—
—
—
—
—
—
67.Y- 1.2-Dichlorocthvlcnc
—
—
—
—
—
—
—
trans-1,2 -Dichloroethy lene
36.94
—
—
—
—
0.00
—
Dichloro methane
15.43
19.61
6.15
11.27
6.73
12.42
88.50
1,2 -Dichloropropane
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
5.47
3.34
2.67
5.46
3.67
6.13
4.30
Ethyl Acrylate
—
—
—
—
—
—
—
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-6
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
BROK
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Ethyl ter/-Butyl Ether
—
—
3.52
4.56
—
15.71
—
Ethylbenzene
8.42
26.24
6.80
4.98
6.76
4.08
15.32
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
9.80
11.40
19.71
16.21
16.66
26.90
40.14
Methyl Methacrylate
—
—
—
—
—
2.18
—
Methyl ter/-Butyl Ether
—
—
~
~
—
1.96
—
//-Octane
4.70
20.72
14.43
6.37
7.52
5.92
14.28
Propylene
25.44
18.04
18.80
15.79
8.01
10.94
26.51
Styrene
13.63
32.89
4.71
19.63
12.74
9.27
49.19
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
T etrachloroethylene
—
0.00
1.67
5.36
3.93
4.51
—
Toluene
12.37
23.47
4.37
8.82
6.69
2.89
13.29
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
—
~
~
—
~
—
1,1,2-Trichloroethane
—
—
~
~
—
~
—
T richloroethylene
—
0.00
—
2.65
—
0.00
—
T richlorofluoro methane
6.91
7.48
3.49
3.10
3.33
4.18
3.71
T richlorotrifluoroethane
6.24
5.92
3.97
2.90
3.47
3.91
4.58
1,2,4-Trimethylbenzene
4.95
7.33
5.99
11.30
7.69
4.62
6.41
1.3.5 -T rimethy lbenzene
—
—
—
8.61
10.13
4.07
0.00
Vinyl chloride
—
—
7.44
4.28
—
8.32
—
m,p-Xylene
8.92
28.11
8.12
3.33
7.22
5.00
19.39
o-Xylene
7.73
26.40
10.34
6.27
6.97
5.33
18.43
Average CV by Site
10.92
13.61
8.31
8.05
8.36
6.89
16.49
# of pairs by site
10
9
11
10
12
11
11
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-7
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GPCO
NBIL
NBNJ
NRNJ
NROK
OCOK
PXSS
Acetylene
4.70
6.41
3.66
6.04
5.07
16.15
4.98
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
Benzene
6.32
6.05
5.19
7.78
6.63
7.41
6.60
Bromochloromethane
—
—
—
—
—
—
—
Bromodichloro methane
—
23.32
—
—
—
—
—
Bromofonn
—
7.69
—
—
—
—
—
Bromomethane
11.14
7.19
4.49
—
—
11.69
11.42
1.3 -Butadiene
10.30
7.44
5.22
4.20
5.88
16.43
9.57
Carbon Tetrachloride
4.27
6.51
5.65
4.40
5.93
7.27
5.45
Chlorobenzene
—
—
—
—
—
—
—
Chloroethane
47.54
32.31
2.01
25.86
—
10.96
19.47
Chloroform
6.91
19.30
4.22
8.87
4.76
10.33
5.72
Chloromethane
4.77
8.54
4.80
5.01
6.57
5.31
4.96
Chloroprene
—
—
—
—
—
—
—
Dibromochloromethane
—
22.16
—
—
—
—
0.00
1,2 -Dibromoethane
—
—
—
—
—
—
—
w-Dichlorobenzene
—
—
—
—
—
—
—
o-Dichlorobenzene
—
—
—
—
—
—
—
p-Dichlorobcnzcnc
—
22.23
—
—
9.12
—
10.58
Dichlorodifluoromethane
4.40
5.77
4.93
4.67
5.56
4.54
4.41
1,1 -Dichloroethane
—
—
—
—
—
—
—
1,2 -Dichloroethane
9.30
12.90
6.12
8.96
7.99
11.01
8.34
1,1 -Dichloroethene
—
—
—
—
—
—
—
67.Y- 1.2-Dichlorocthvlcnc
—
—
—
—
—
—
—
trans-1,2 -Dichloroethy lene
4.12
—
—
—
12.85
4.88
0.00
Dichloro methane
10.90
35.42
10.30
5.74
6.69
12.27
35.30
1,2 -Dichloropropane
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
6.87
4.44
6.23
1.94
—
10.80
3.81
Ethyl Acrylate
—
—
—
—
—
—
—
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-8
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GPCO
NBIL
NBNJ
NRNJ
NROK
OCOK
PXSS
Ethyl ter/-Butyl Ether
8.40
4.06
—
9.21
—
—
—
Ethylbenzene
6.54
6.97
4.87
11.79
5.51
4.14
7.71
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
30.01
15.54
22.41
30.06
23.95
27.64
37.66
Methyl Methacrylate
—
—
—
—
—
—
—
Methyl ter/-Butyl Ether
—
10.88
—
7.19
—
—
—
//-Octane
8.43
14.29
7.68
9.83
9.77
12.79
18.72
Propylene
14.00
19.60
7.92
16.89
11.75
26.50
27.80
Styrene
40.65
5.46
10.79
12.64
6.45
9.50
22.53
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
T etrachloroethylene
10.50
7.20
2.89
4.92
9.43
4.04
8.23
Toluene
4.92
9.04
7.72
5.96
4.72
5.76
13.49
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
—
—
—
—
—
—
1,1,2-Trichloroethane
—
—
—
—
—
—
—
T richloroethylene
—
13.35
—
—
—
9.23
—
T richlorofluoro methane
4.31
5.74
4.29
4.70
7.29
3.68
4.70
T richlorotrifluoroethane
4.49
7.55
4.82
5.47
6.78
4.75
4.55
1,2,4-Trimethylbenzene
11.02
9.14
4.34
11.67
6.38
7.95
15.94
1.3.5 -T rimethy lbenzene
6.43
6.01
—
1.69
7.69
4.16
14.74
Vinyl chloride
—
12.29
—
—
—
—
—
m,p-Xylene
12.58
7.43
5.81
10.61
5.30
4.79
7.58
o-Xylene
10.66
7.90
5.10
10.63
5.89
9.16
6.08
Average CV by Site
11.28
11.82
6.31
9.08
7.83
9.75
11.44
# of pairs by site
13
13
5
6
3
12
13
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-9
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
ROIL
S4MO
SEWA
SPIL
TMOK
TOOK
TROK
Acetylene
5.53
7.27
10.74
7.06
14.39
4.90
6.67
tert-Amyl Methyl Ether
—
—
—
—
—
—
—
Benzene
14.74
9.50
5.09
12.91
7.05
6.19
6.03
Bromochloromethane
—
—
—
—
9.43
—
—
Bromodichloro methane
—
—
—
—
—
—
—
Bromofonn
—
—
—
—
—
—
—
Bromomethane
3.89
13.16
24.40
7.38
14.91
6.69
4.56
1.3 -Butadiene
3.07
12.16
12.56
9.46
12.29
6.74
14.39
Carbon Tetrachloride
8.72
7.80
6.46
20.72
5.78
27.08
27.86
Chlorobenzene
—
—
—
—
—
—
—
Chloroethane
16.39
20.34
33.42
13.72
12.61
16.86
27.10
Chloroform
6.64
24.32
6.79
7.92
7.35
4.28
6.54
Chloromethane
4.92
6.73
6.33
6.26
7.12
4.29
6.88
Chloroprene
—
—
—
—
—
—
—
Dibromochloromethane
—
—
—
—
—
—
—
1,2 -Dibromoethane
—
—
—
—
—
—
—
w-Dichlorobenzene
—
—
—
—
—
—
—
o-Dichlorobenzene
—
—
—
—
—
—
—
p-Dichlorobcnzcnc
—
10.12
—
—
2.77
—
—
Dichlorodifluoromethane
5.10
6.01
4.49
5.88
5.77
4.30
6.69
1,1 -Dichloroethane
—
—
—
—
—
—
—
1,2 -Dichloroethane
7.75
10.75
7.04
8.54
9.24
5.70
7.43
1,1 -Dichloroethene
—
2.67
—
0.00
—
—
—
67.Y- 1.2-Dichlorocthvlcnc
—
2.18
—
—
—
—
—
trans-1,2 -Dichloroethy lene
—
—
—
—
—
—
2.18
Dichloro methane
7.58
36.79
42.68
8.48
13.58
16.83
16.19
1,2 -Dichloropropane
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
6.07
12.29
7.40
6.36
7.20
5.37
6.56
Ethyl Acrylate
—
—
—
—
—
—
—
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-10
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
ROIL
S4MO
SEWA
SPII
TMOK
TOOK
TROK
Ethyl ter/-Butyl Ether
—
—
—
7.62
—
—
—
Ethylbenzene
4.75
13.63
4.38
12.00
9.55
8.39
6.01
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
12.92
21.11
37.26
21.79
26.17
9.76
30.08
Methyl Methacrylate
—
—
—
—
—
—
—
Methyl ter/-Butyl Ether
—
—
—
11.79
—
—
26.19
//-Octane
7.57
14.15
6.64
17.90
10.30
5.21
9.52
Propylene
5.27
17.47
25.82
16.55
12.21
13.13
19.36
Styrene
—
16.05
10.76
14.67
9.49
9.90
10.92
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
T etrachloroethylene
10.15
6.30
8.60
4.12
2.51
11.40
Toluene
3.85
7.71
5.46
11.10
7.53
4.84
6.13
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
—
—
~
—
—
—
1,1,2-Trichloroethane
—
—
—
~
—
—
—
T richloroethylene
—
9.68
—
13.07
14.01
—
—
T richlorofluoro methane
4.64
6.24
5.08
5.76
6.01
4.22
5.33
T richlorotrifluoroethane
4.36
6.78
12.09
6.65
6.07
5.08
5.99
1,2,4-Trimethylbenzene
8.26
15.90
5.38
12.18
11.63
8.17
9.12
1.3.5 -T rimethy lbenzene
—
8.59
—
2.77
—
—
—
Vinyl chloride
—
7.31
—
6.73
—
—
—
m,p-Xylene
6.79
18.52
5.22
10.67
9.15
7.13
6.86
o-Xylene
5.31
17.13
4.31
10.72
8.80
7.05
8.21
Average CV by Site
7.01
12.42
12.34
10.18
9.90
7.91
11.20
# of pairs by site
4
12
12
10
11
12
12
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-11
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
TVKY
YUOK
# of pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetylene
7.83
20.63
279
8.29
8.47
6.40
tert-Amyl Methyl Ether
—
—
0
—
—
—
Benzene
7.61
3.92
279
8.55
8.69
7.11
Bromochloromethane
18.88
—
7
10.32
8.61
18.88
Bromodichloro methane
—
—
9
23.32
23.32
—
Bromofonn
—
—
4
7.69
7.69
~
Bromomethane
16.26
10.82
127
9.70
9.27
13.84
1.3 -Butadiene
7.96
4.65
157
9.27
9.32
8.76
Carbon Tetrachloride
7.03
5.82
278
11.08
11.54
6.24
Chlorobenzene
0.00
—
1
0.00
—
0.00
Chloroethane
21.85
47.63
135
20.60
20.59
20.66
Chloroform
11.87
6.57
261
8.85
8.86
8.79
Chloromethane
5.96
16.89
279
6.31
6.39
5.46
Chloroprene
8.84
—
1
8.84
—
8.84
Dibromochloromethane
—
—
10
11.08
22.16
0.00
1,2 -Dibromoethane
—
—
0
—
—
—
w-Dichlorobenzene
—
—
0
—
—
~
o-Dichlorobenzene
—
—
0
—
—
—
p-Dichlorobcnzcnc
—
—
18
9.85
9.71
10.58
Dichlorodifluoro methane
4.55
3.26
279
5.02
5.07
4.48
1,1 -Dichloroethane
11.71
—
18
11.71
—
11.71
1,2 -Dichloroethane
20.68
8.59
238
8.73
8.18
14.51
1,1 -Dichloroethene
9.68
—
4
4.12
1.33
9.68
67.Y- 1.2-Dichlorocthvlcnc
—
—
1
2.18
2.18
—
trans-1,2 -Dichloroethy lene
3.82
—
12
8.10
10.16
1.91
Dichloro methane
13.93
13.61
279
19.41
18.91
24.61
1,2 -Dichloropropane
6.43
—
1
6.43
—
6.43
cis-1,3 -Dichloropropene
—
—
0
—
—
—
trans-1,3 -Dichloropropene
—
—
0
—
—
—
Dichlorotetrafluoroethane
9.99
5.24
150
5.98
5.89
6.90
Ethyl Acrylate
—
—
0
—
—
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-12
-------
Table 24-2. VOC Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
TVKY
YUOK
# of pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Ethyl fcrt-Butvl Ether
—
—
22
7.58
7.58
—
Ethylbenzene
10.98
11.60
214
8.76
8.70
9.34
Hexachloro-1,3 -butadiene
—
—
0
—
—
—
Methyl Isobutyl Ketone
31.82
37.43
199
24.19
23.19
34.74
Methyl Methacrylate
—
—
3
3.32
3.32
—
Methyl ter/-Butyl Ether
—
—
8
11.60
11.60
—
//-Octane
11.96
11.25
205
10.87
10.44
15.34
Propylene
24.63
28.86
279
17.88
17.09
26.21
Styrene
14.09
2.48
115
15.38
15.09
18.31
1.1,2,2-Tetrachloroethane
—
—
0
—
—
—
T etrachloroethylene
7.91
4.04
110
5.88
5.64
8.07
Toluene
23.41
9.02
278
8.81
7.89
18.45
1,2,4-Trichlorobenzene
—
—
0
—
—
—
1,1,1 -T richloroethane
0.00
—
1
0.00
—
0.00
1,1,2-Trichloroethane
7.32
—
10
7.32
—
7.32
T richloroethylene
4.35
—
19
7.37
7.75
4.35
T richlorofluoro methane
9.15
3.32
279
5.07
4.90
6.93
T richlorotrifluoroethane
5.01
2.62
279
5.39
5.45
4.78
1,2,4-Trimethylbenzene
9.93
12.80
182
9.05
8.68
12.94
1.3.5 -T rimethy lbenzene
5.34
4.88
67
6.41
5.92
10.04
Vinyl chloride
12.76
—
42
8.45
7.73
12.76
m,p-Xylene
10.17
9.82
230
9.50
9.56
8.87
o-Xylene
11.24
14.34
227
9.74
9.84
8.66
Average CV by Site
10.97
12.00
5,596
9.27
9.91
10.60
# of pairs by site
55
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-13
-------
24.2.2 SNMOC Method Precision
The SNMOC method precision for duplicate and collocated samples is presented in
Table 24-3 as the CV per pollutant per site, the average CV per site, the average CV per
pollutant, and the overall average CV across the SNMOCs listed. The precision results from
duplicate and collocated samples exhibit low- to high-level variability among the pollutants and
sites, ranging from a CV of 0 percent (//-butane and isobutane for PACO) to 66.84 percent (sum
of unknowns for GSCO). Among the 35 pollutants listed in Table 24-3 for which a CV could be
calculated for each of the seven sites, 15 pollutants have CVs for all seven sites that are less than
15 percent (e.g., acetylene); conversely, there are two pollutants listed for which all seven CVs
are greater than or equal to 15 percent: 1-nonene and sum of unknowns.
The pollutant-specific average CV, as shown in orange in Table 24-3, ranges from
1.98 percent (//-butane) to 37.77 percent (//-tridecane). The site-specific average CV, as shown in
green in Table 24-3, ranges from 7.76 percent (NROK) to 18.08 percent (PACO). The overall
average method precision for SNMOCs is 12.23 percent, meeting the MQO of 15 percent. Note
that the results for TNMOC were not included in the precision calculations.
Sites at which duplicate samples were collected are highlighted in blue in Table 24-3
while sites at which collocated samples were collected are highlighted in purple. Collocated
SNMOC samples were collected at three of the sites shown in Table 24-3 (GSCO, PACO, and
RICO); however, relatively few collocated sample pairs were collected at these sites in 2015 and
2016. Duplicate SNMOC samples were collected at the remaining four sites. The average CV for
sites that collected duplicate samples was calculated and is shown at the end of Table 24-3 in
blue while the average CV for sites collecting collocated samples is shown in purple. The
average CV for both precision types is less than 15 percent, with the variability associated with
collocated samples (13.62 percent) is slightly more than the variability associated with duplicate
samples (11.30 percent).
24-14
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant
Pollutant
BROK
BTUT
GSCO
NBIL
NROK
Acetylene
3.55
3.25
0.75
2.68
0.66
Benzene
5.79
6.71
15.30
5.53
6.83
1.3 -Butadiene
—
3.97
—
—
3.83
//-Butane
0.67
1.16
4.49
5.02
0.75
1-Butene
—
—
—
—
—
67.Y-2-Bute nc
—
11.70
13.93
5.40
2.77
;ra«.v-2-Butcne
19.85
11.41
—
—
8.25
Cyclohexane
2.65
4.59
5.19
8.62
1.52
Cyclopentane
1.19
21.29
3.41
7.90
1.23
Cyclopentene
—
—
—
—
—
w-Decane
12.46
15.67
—
16.04
11.07
1-Decene
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
p-Diethylbenzene
—
—
—
12.83
—
2,2-Dimethylbutane
13.23
7.08
13.80
4.46
2.80
2,3 -Dimethy lbutane
1.99
2.61
9.32
4.42
1.08
2,3 -Dimethy lpentane
6.34
2.52
7.81
7.47
1.61
2,4-Dimethylpentane
13.52
4.94
8.81
9.49
8.09
n-Dodecane
19.98
9.76
—
14.95
—
1-Dodecene
—
—
—
—
—
Ethane
0.94
1.73
4.63
11.27
0.49
2-Ethyl-l-butene
—
—
—
—
—
Ethylbenzene
13.23
6.08
5.48
8.93
6.57
Ethylene
17.78
5.88
24.60
12.24
9.52
/w-Ethyltoluene
17.64
7.83
30.87
14.58
4.30
o-Ethyltoluene
20.60
1.69
—
4.65
12.07
p-Ethyltoluene
17.13
4.39
33.43
12.98
3.31
//-Heptane
7.74
7.24
5.82
16.41
1.61
1-Heptene
—
—
—
—
—
n-Hexane
1.56
4.34
12.39
7.49
5.07
1-Hexene
23.75
32.33
13.55
34.63
20.73
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites
collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-15
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BROK
BTUT
GSCO
NBIL
NROK
c/.v-2-Hc.\cnc
—
-Pincnc
—
—
—
—
—
Propane
0.70
2.11
4.84
7.69
0.99
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites
collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-16
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant (Continued)
Pollutant
BROK
BTUT
GSCO
NBIL
NROK
w-Propylbenzene
44.11
5.49
—
13.08
34.25
Propylene
26.22
8.39
27.85
15.73
9.14
Propyne
—
—
—
—
5.66
Styrene
—
—
—
—
—
Toluene
17.97
5.47
6.13
8.42
3.35
«-T ridccanc
49.83
—
—
25.70
—
1-Tridecene
—
--
--
—
—
1,2,3 -Trimethylbenzene
—
—
—
19.05
32.57
1,2,4 -T rimethylbenzene
20.27
13.33
9.77
23.95
27.99
1,3,5 -T rimethylbenzene
19.95
8.72
—
7.21
3.98
2,2,3 -Trimethylpentane
—
—
—
7.68
—
2,2,4 -T rimethy lpentane
5.21
4.59
9.17
9.99
5.54
2,3,4 -T rimethy lpentane
20.88
9.56
33.77
29.84
14.56
/7-Undecane
18.91
18.29
—
8.91
—
1-Undecene
—
—
—
—
—
w-Xylene/p-Xylene
9.06
13.03
4.35
10.39
1.35
o-Xylene
11.81
12.93
12.76
8.53
1.11
SNMOC (Sum of Knowns)
2.88
4.96
4.90
10.00
1.11
Sum of Unknowns
34.03
25.84
66.84
22.48
18.51
Average CV by Site
15.11
9.32
12.88
12.58
7.76
# of pairs by site
10
8
1
13
3
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this
method is calculated from the pollutant-specific averages and is provided in the final column of
the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites
collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-17
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant (Continued)
Pollutant
PACO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetylene
4.33
1.47
39
2.38
2.54
2.18
Benzene
2.05
3.71
38
6.56
6.21
7.02
1.3 -Butadiene
—
4.30
5
4.03
3.90
4.30
//-Butane
0.00
1.76
39
1.98
1.90
2.08
1-Butene
—
—
—
—
—
—
67.Y-2-Bute nc
—
16.35
15
10.03
6.62
15.14
;ra«.v-2-Butcne
—
24.28
9
15.95
13.17
24.28
Cyclohexane
6.68
2.61
38
4.55
4.34
4.83
Cyclopentane
—
5.95
31
6.83
7.90
4.68
Cyclopentene
—
—
—
—
—
—
w-Decane
29.39
11.89
26
16.09
13.81
20.64
1-Decene
—
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
—
p-Diethylbenzene
—
—
1
12.83
12.83
—
2,2-Dimethylbutane
16.75
11.78
31
9.99
6.89
14.11
2,3 -Dimethy lbutane
3.72
2.53
39
3.67
2.53
5.19
2,3 -Dimethy lpentane
11.25
4.15
39
5.88
4.48
7.74
2,4-Dimethylpentane
14.19
5.57
35
9.23
9.01
9.53
n-Dodecane
—
—
11
14.90
14.90
—
1-Dodecene
—
—
—
—
—
—
Ethane
0.49
2.44
39
3.14
3.61
2.52
2-Ethyl-l-butene
—
—
—
—
—
—
Ethylbenzene
29.71
10.51
35
11.50
8.70
15.23
Ethylene
18.41
13.45
39
14.56
11.36
18.82
/w-Ethyltoluene
—
21.53
28
16.13
11.09
26.20
o-Ethyltoluene
—
16.36
12
11.07
9.75
16.36
p-Ethyltoluene
—
29.08
24
16.72
9.45
31.25
//-Heptane
5.55
3.47
39
6.83
8.25
4.94
1-Heptene
—
—
—
—
—
—
n-Hexane
0.44
4.56
39
5.12
4.62
5.80
1-Hexene
51.48
47.04
18
31.93
27.86
37.36
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-18
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant (Continued)
Pollutant
PACO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
c/.v-2-Hc.\cnc
—
—
—
—
—
—
;ra«.v-2-Hc\cnc
—
—
4
4.82
4.82
—
Isobutane
0.00
2.43
39
2.95
3.50
2.22
Isobutylene
—
—
—
—
—
—
Isopentane
—
0.69
5
4.17
5.91
0.69
Isoprene
—
3.68
28
10.45
11.71
7.94
Isopropylbenzene
—
—
2
18.59
18.59
—
2-Methy 1-1 -butene
—
12.78
12
9.33
8.17
12.78
3 -Methyl-1 -butene
—
—
—
—
—
—
2-Methyl-l-pentene
—
—
—
—
—
—
4-Methyl-l-pentene
—
—
—
—
—
—
2-Methyl-2-butene
—
17.22
20
14.55
11.77
20.11
Methylcyclohexane
5.40
3.58
30
5.40
5.86
4.49
Methylcyclopentane
0.68
4.55
38
3.96
4.60
3.11
2-Methy lheptane
7.90
4.82
27
9.47
12.71
5.15
3-Methylheptane
9.34
2.54
29
9.11
10.66
7.04
2-Methy lhexane
43.66
14.08
38
23.64
20.13
28.32
3-Methylhexane
—
7.70
15
8.25
8.38
7.70
2-Methy lpentane
—
14.50
37
10.49
5.89
19.69
3-Methylpentane
5.66
3.97
39
4.38
4.03
4.86
//-Nona ne
26.62
8.80
33
15.13
13.84
17.71
1-Nonene
49.75
31.20
21
27.51
24.10
32.05
//-Octane
18.61
6.62
37
12.77
13.75
11.47
1-Octene
50.68
36.94
24
32.05
32.37
31.62
n-Pentane
1.28
1.58
39
8.05
12.38
2.27
1-Pentene
36.31
16.57
32
18.51
18.10
19.05
c/.v-2-Pcntcnc
—
9.87
11
6.97
4.59
10.55
;ra/?.v-2-Pcntene
46.00
13.19
28
17.95
13.31
24.13
fl-Pincnc
—
—
15
12.50
13.46
8.69
/>-Pincnc
—
—
—
—
—
—
Propane
0.38
2.47
39
2.74
2.87
2.56
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24-19
-------
Table 24-3. SNMOC Method Precision: Coefficient of Variation
Based on Collocated and Duplicate Samples by Site and Pollutant (Continued)
Pollutant
PACO
RICO
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
w-Propylbenzene
—
22.27
11
23.84
24.23
22.27
Propylene
38.45
18.53
39
20.62
14.87
28.28
Propyne
—
—
1
5.66
5.66
—
Styrene
—
—
—
—
—
—
Toluene
8.84
3.38
39
7.65
8.80
6.12
«-T ridccanc
—
—
4
37.77
37.77
—
1-Tridecene
—
—
—
—
—
—
1,2,3 -Trimethylbenzene
—
16.44
9
22.69
25.81
16.44
1,2,4 -T rimethylbenzene
44.05
22.03
31
23.06
21.38
25.28
1,3,5 -T rimethylbenzene
—
20.54
18
12.08
9.97
20.54
2,2,3 -Trimethylpentane
—
1.79
2
4.73
7.68
1.79
2,2,4 -T rimethy lpentane
—
5.65
29
6.69
6.33
7.41
2,3,4 -T rimethy lpentane
45.80
17.29
34
24.53
18.71
32.29
/7-Undecane
—
—
12
15.37
15.37
—
1-Undecene
—
—
—
—
—
—
w-Xylene/p-Xylene
2.44
5.82
38
6.63
8.46
4.20
o-Xylene
13.80
6.88
38
9.69
8.59
11.15
SNMOC (Sum of Knowns)
4.27
3.48
39
4.52
4.74
4.22
Sum of Unknowns
32.83
36.85
39
33.91
25.22
45.51
Average CV by Site
18.08
11.33
1,694
12.23
11.30
13.62
# of pairs by site
1
3
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24.2.3 Methane Method Precision
Table 24-4 presents the method precision for all duplicate methane samples as the CV per
site and the overall average CV for the method. All samples evaluated in this section are
duplicate samples. Only two NMP sites sampled methane during the 2015 and/or 2016
monitoring efforts, BROK and NROK. The site-specific CV ranges from 2.99 percent for BROK
to 4.41 percent for NROK; note that these CVs are based on 10 duplicate sample pairs for BROK
and three pairs for NROK. The overall average method precision for methane is 3.70 percent, as
24-20
-------
shown in orange in Table 24-4, which is considerably less than the MQO of 15 percent CV for
method precision.
Table 24-4. Methane Method Precision: Coefficient of Variation
Based on Duplicate Samples by Site and Pollutant
Pollutant
BROK
NROK
# of
pairs
Average by
Pollutant
Methane
2.99
4.41
13
3.70
# of pairs
10
3
Bold = CV greater than or equal to 15 percent
Orange shading indicates the pollutant-specific average CV.
BOLD ITALICS = EPA-designated NATTS Site
24.2.4 Carbonyl Compound Method Precision
Table 24-5 presents the method precision for duplicate and collocated carbonyl
compound samples as the CV per pollutant per site, the average CV per site, the average CV per
pollutant, and the overall average CV across the carbonyl compounds listed. The duplicate and
collocated sample results exhibit low- to mid-level variability, ranging from a CV of 0.45 percent
(butyraldehyde for NBNJ) to 41.86 percent (2-butanone for SYFL). SYFL's 2-butanone CV was
also the highest CV among the sites sampling carbonyl compounds for the 2014 NMP report.
The number of sites for which a given pollutant has a CV greater than or equal to 15 percent
varies from none (propionaldehyde) to four (2-butanone and valeraldehyde).
The pollutant-specific average CV, as shown in orange in Table 24-5, ranges from
3.98 percent (formaldehyde) to 8.75 percent (2-butanone). The site-specific average CV, as
shown in green in Table 24-5, ranges from 2.29 percent (NROK) to 21.19 percent (NRNJ). Three
sites collecting duplicate or collocated carbonyl compound samples have site-specific average
CVs greater than or equal to 15 percent (DEMI, NRNJ, and SYFL). The NBNJ site was moved
at the end of 2015 and relocated to NRNJ at the beginning of 2016. The relatively high CVs
shown for NRNJ result primarily from one particularly imprecise pair of samples collected on
July 17, 2016, compared to the higher CVs shown for DEMI and SYFL, where disagreement
among the sample pairs was more common. The overall average method precision is
6.85 percent for carbonyl compounds.
Sites at which duplicate samples were collected are highlighted in blue in Table 24-5
while sites at which collocated samples were collected are highlighted in purple. Collocated
carbonyl compound samples were collected at three sites shown in Table 24-5 (DEMI, INDEM,
and PXSS); duplicate samples were collected at the remaining sites. The average CV for sites
24-21
-------
that collected duplicate samples was calculated and is shown at the end of Table 24-5 in blue
while the average CV for sites collecting collocated samples is shown in purple. The average CV
for both precision types is less than 15 percent, with the variability associated with collocated
samples (10.72 percent) greater than the variability associated with duplicate samples
(6.38 percent).
Table 24-5. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant
Pollutant
A/.II
BROK
BTUT
CHNJ
CSNJ
DEMI
ELNJ
GLKY
Acetaldehyde
3.67
18.05
5.32
4.34
2.20
16.59
1.98
1.23
Acetone
5.99
23.81
6.49
6.95
6.72
8.20
5.16
4.46
Benzaldehyde
7.84
6.14
7.42
6.39
5.15
17.05
5.25
6.99
2-Butanone
13.42
7.68
4.54
6.49
5.47
8.17
5.20
5.32
Butyraldehyde
7.27
3.71
5.85
4.33
2.41
26.07
2.50
4.06
Crotonaldehyde
5.56
8.02
6.58
4.91
3.32
14.85
4.66
2.79
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
7.68
0.91
5.12
4.73
2.35
11.94
2.06
1.30
Hexaldehyde
8.38
4.74
6.27
6.11
6.58
26.46
2.57
5.01
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
5.72
8.46
5.14
2.71
3.53
7.32
2.67
1.25
Tolualdehydes
9.03
5.33
6.66
7.44
7.87
16.44
7.60
4.92
Valeraldehyde
5.98
6.28
5.82
5.81
5.38
16.66
4.11
8.01
Average CV by Site
7.32
8.47
5.93
5.47
4.64
15.43
3.98
4.12
# of pairs by site
13
10
11
8
11
12
12
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-22
-------
Table 24-5. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
GPCO
IN I)IM
NBIL
NBNJ
NRNJ
NROK
OCOK
OR II
Acetaldehyde
3.95
5.86
2.90
1.18
2.90
0.84
1.51
4.50
Acetone
10.77
10.91
6.70
2.55
21.62
0.67
2.81
9.46
Benzaldehyde
19.54
12.21
6.17
5.05
31.41
3.48
7.86
7.89
2-Butanone
8.20
8.43
7.37
4.22
36.82
2.15
3.08
16.86
Butyraldehyde
3.82
6.99
3.61
0.45
28.23
1.66
2.70
7.63
Crotonaldehyde
7.72
8.24
4.74
2.78
31.69
1.49
1.93
16.41
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
3.39
5.49
3.37
1.39
2.92
1.30
1.93
3.36
Hexaldehyde
3.87
5.72
7.28
5.99
30.89
2.13
3.48
5.83
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
5.24
10.57
4.36
3.34
3.40
3.09
3.50
6.57
Tolualdehydes
9.60
6.89
9.74
7.63
17.70
5.89
5.31
13.77
Valeraldehyde
2.79
12.19
5.72
5.91
25.55
2.52
4.26
8.03
Average CV by Site
7.17
8.50
5.63
3.68
21.19
2.29
3.49
9.12
# of pairs by site
10
18
16
5
6
4
12
11
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-23
-------
Table 24-5. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
PXSS
ROIL
S4MO
SEWA
SKFL
S P11
SYFL
TMOK
Acetaldehyde
4.18
4.20
1.92
6.39
2.85
3.15
7.24
0.98
Acetone
6.86
3.84
4.27
3.77
11.07
2.35
14.05
3.35
Benzaldehyde
5.53
4.40
6.37
11.82
6.48
6.77
11.59
6.41
2-Butanone
2.90
5.94
3.32
2.31
23.66
3.34
41.86
4.16
Butyraldehyde
6.98
6.83
3.07
3.98
6.76
3.65
21.86
2.77
Crotonaldehyde
4.92
5.41
4.26
6.88
11.87
3.20
33.07
2.78
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
12.05
3.61
1.83
5.05
3.05
4.50
12.55
1.16
Hexaldehyde
10.44
5.02
3.06
6.18
5.39
7.67
15.82
5.92
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
5.14
4.79
3.70
5.96
5.67
5.44
14.53
1.74
Tolualdehydes
11.60
4.37
7.40
7.05
8.10
8.27
16.08
4.42
Valeraldehyde
19.85
7.07
6.35
7.26
11.47
6.92
17.58
4.87
Average CV by Site
8.22
5.04
4.14
6.06
8.76
5.03
18.75
3.50
# of pairs by site
15
5
12
12
11
10
11
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-24
-------
Table 24-5. Carbonyl Compound Method Precision: Coefficient of Variation
Based on Duplicate and Collocated Samples by Site and Pollutant (Continued)
Pollutant
TOOK
TROK
WIMN
YUOK
# of
pairs
Average
by
Pollutant
Average
for
Duplicate
Pairs
Average
for
Collocated
Pairs
Acetaldehyde
0.95
1.62
3.90
0.86
315
4.12
3.55
8.87
Acetone
3.31
1.97
3.82
1.70
315
6.91
6.71
8.66
Benzaldehyde
6.43
6.30
5.38
5.21
307
8.52
8.15
11.60
2-Butanone
1.93
2.87
6.75
2.59
313
8.75
9.02
6.50
Butyraldehyde
1.78
1.96
6.33
4.15
315
6.48
5.66
13.35
Crotonaldehyde
1.67
3.36
5.87
3.01
313
7.57
7.36
9.34
2,5 -Dimethy lbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
1.53
1.41
4.42
1.06
315
3.98
3.28
9.83
Hexaldehyde
4.16
3.88
6.23
3.73
312
7.46
6.65
14.21
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
3.55
3.70
4.40
1.56
312
4.90
4.56
7.68
Tolualdehydes
7.13
6.87
9.73
6.87
289
8.56
8.19
11.64
Valeraldehyde
4.13
4.97
5.97
4.42
312
8.07
7.09
16.23
Average CV by Site
3.32
3.54
5.71
3.20
3,418
6.85
6.38
10.72
# of pairs by site
12
12
20
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated from
the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24.2.5 PAH Method Precision
The method precision results for collocated PAH samples are shown in Table 24-6 as the
CV per pollutant per site, the average CV per site, the average CV per pollutant, and the overall
average CV across the PAHs listed. All samples evaluated in this section are collocated samples.
Ownership, maintenance, and use of collocated systems were the responsibility of the
participating agency for sites sampling PAHs. Thus, collocated samples were not collected at
most PAH sites because few had collocated collection systems. The method precision presented
for PAHs is based on data from three sites (DEMI, RUCA, and SEW A) and a total of 42 sample
pairs. The results from collocated samples exhibit low- to mid-level variability, ranging from a
CV of 0.57 percent (benzo(a)pyrene for SEW A) to 45.08 percent (anthracene for RUCA). The
overall average method precision is 15.27 percent, which is just greater than the MQO of
15 percent CV.
24-25
-------
The pollutant-specific average CV, as shown in orange in Table 24-6, ranges from
6.41 percent (benzo(a)anthracene) to 27.15 percent (anthracene). The site-specific average CVs,
as shown in green in Table 24-6, range from 4.93 percent for SEWA to 18.19 percent for DEMI
to 21.54 percent forRUCA. None of the CVs for SEWA are greater than or equal to 15 percent
while many PAHs are greater than or equal to 15 percent for DEMI (11 pollutants) and RUCA
(16 pollutants).
Table 24-6. PAH Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
DEMI
RUCA
SEWA
# of
pairs
Average
by
Pollutant
Acenaphthene
5.03
36.54
4.15
38
15.24
Acenaphthylene
6.97
25.24
7.11
19
13.11
Anthracene
29.93
45.08
6.42
25
27.15
Benzo(a)anthracene
8.91
9.26
1.07
17
6.41
Benzo(a)pyrene
9.06
17.38
0.57
14
9.01
Benzo(b)fluoranthene
36.60
34.38
4.31
23
25.10
Benzo(e)pyrene
34.61
32.07
1.32
22
22.67
Benzo(g,h,i)perylene
30.52
23.73
3.59
23
19.28
Benzo(k)fluoranthene
15.06
15.06
—
9
15.06
Chrysene
37.81
29.37
5.91
28
24.36
Coronene
23.81
12.61
9.16
31
15.19
Cyclopenta(c,d)pyrene
4.31
11.53
8.63
14
8.16
Dibenz(a,h)anthracene
4.29
15.15
—
9
9.72
Fluoranthene
30.96
16.80
4.81
42
17.52
Fluorene
6.33
19.98
4.48
35
10.27
9-Fluorenone
9.08
11.48
3.98
42
8.18
Indeno( 1,2,3 -c,d)pyrene
34.09
33.37
1.73
21
23.06
Naphthalene
4.55
13.19
5.23
42
7.65
Perylene
14.07
16.77
—
10
15.42
Phenanthrene
16.41
13.62
4.44
42
11.49
Pyrene
31.56
18.51
5.49
42
18.52
Retene
6.27
22.81
11.21
35
13.43
Average CV by Site
18.19
21.54
4.93
583
15.27
# of pairs by site
10
21
11
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
24-26
-------
24.2.6 Metals Method Precision
The method precision for all collocated metals samples are presented in Table 24-7 as the
C V per pollutant per site, the average CV per site, the average C V per pollutant, and the overall
average CV across the metals listed. All samples evaluated in this section are collocated samples.
The results from collocated samples exhibit low- to high-level variability, ranging from a CV of
2.75 percent (arsenic for ASKY-M) to 56.57 percent (mercury for BOMA). The number of sites
for which a given pollutant has a CV greater than or equal to 15 percent varies from none
(antimony and manganese) to four (cadmium). Note that BOMA, GLKY, S4MO, and TOOK
collected collocated samples more frequently than the 10 percent requirement.
The pollutant-specific average CV, as shown in orange in Table 24-7, ranges from
7.74 percent (antimony) to 24.22 percent (mercury); four of the 11 metals have an average CV
greater than 15 percent. The site-specific average CV, as shown in green in Table 24-7, ranges
from 8.41 percent (TOOK) to 20.32 percent (BOMA). Two sites (BOMA and BTUT) have site-
specific average CVs greater than or equal to 15 percent. The overall average method precision
for metals is 12.85 percent.
Table 24-7. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant
Pollutant
ASKY-M
BOMA
BTUT
GLKY
GPCO
Antimony
3.10
9.39
8.95
13.92
8.79
Arsenic
2.75
20.12
9.71
13.10
4.25
Beryllium
12.01
22.04
17.83
24.54
14.74
Cadmium
27.14
18.19
8.99
13.10
15.40
Chromium
9.58
2.82
15.78
—
—
Cobalt
6.16
12.67
11.43
9.75
43.84
Lead
2.79
11.23
26.46
11.60
4.06
Manganese
4.67
11.84
14.47
7.95
5.72
Mercury
23.68
56.57
34.71
11.77
14.99
Nickel
11.17
42.69
9.21
23.65
10.86
Selenium
6.38
15.95
17.24
7.49
7.90
Average CV by Site
9.95
20.32
15.89
13.69
13.05
# of pairs by site
11
58
12
46
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
24-27
-------
Table 24-7. Metals Method Precision: Coefficient of Variation
Based on Collocated Samples by Site and Pollutant (Continued)
Pollutant
S4MO
TOOK
# of pairs
Average
by
Pollutant
Antimony
4.11
5.92
373
7.74
Arsenic
7.02
4.12
369
8.73
Beryllium
14.26
8.40
331
16.26
Cadmium
5.99
21.54
373
15.76
Chromium
13.32
6.83
147
9.67
Cobalt
7.69
7.48
301
14.14
Lead
4.71
6.14
374
9.57
Manganese
4.42
5.63
374
7.81
Mercury
13.81
14.01
152
24.22
Nickel
18.57
7.82
342
17.71
Selenium
8.32
4.62
350
9.70
Average CV by Site
9.29
8.41
3,486
12.85
# of pairs by site
114
121
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method
is calculated from the pollutant-specific averages and is provided in the final column of the table.
BOLD ITALICS = EPA-designated NATTS Site
24.2.7 Hexavalent Chromium Method Precision
Table 24-8 presents the method precision results from collocated hexavalent chromium
samples as the C V per site and the overall average C V for the method. All samples evaluated in
this section are collocated samples. Hexavalent chromium was sampled at RIVA throughout
2015 and through the end of June in 2016. The overall average method precision for hexavalent
chromium is 13.48 percent, as shown in orange in Table 24-8, which is less than the MQO of
15 percent CV for method precision. Note that the precision calculations are based on five
collocated pairs.
Table 24-8. Hexavalent Chromium Method Precision: Coefficient of Variation
Based on Collocated Samples by Site
Pollutant
RIVA
# of
pairs
Average by
Pollutant
Hexavalent Chromium
13.48
5
13.48
Bold = CV greater than or equal to 15 percent
Orange shading indicates the average CV for this method.
BOLD ITALICS = EPA-designated NATTS Site
24-28
-------
24.3 Analytical Precision
Analytical precision is a measurement of random errors associated with the process of
analyzing environmental samples. These errors may result from various factors, including
random "noise" inherent to analytical instruments. Laboratories can evaluate the analytical
precision of ambient air samples by comparing concentrations measured during multiple
analyses of a single sample (i.e., replicate samples). Replicate analyses were run on duplicate or
collocated samples collected during the program years. CVs were calculated for every replicate
analysis run on duplicate or collocated samples collected during the program year. In addition,
replicate analyses were also run on select individual samples, which can provide an indication of
analytical precision for monitoring sites unable to collect duplicate or collocated samples
(i.e., collection systems "unequipped" to collect duplicate or collocated samples). Individual
samples with replicate analyses were also factored into the CV calculations for analytical
precision. Only results at or above the MDL were used in these calculations, similar to the
calculation of method precision discussed in Section 24.2.
Table 24-9 presents the 2015-2016 NMP analytical precision for VOCs, SNMOCs,
methane, carbonyl compounds, PAHs, metals, and hexavalent chromium, presented as average
CV and expressed as a percentage. The average CV for each method met the program MQO of
15 percent for precision. The analytical precision for all methods is less than 7 percent. This table
also includes the number of pairs that were included in the calculation of the analytical precision.
The total number of pairs including those with concentrations less than the MDL (and two
numerical results) is also included in Table 24-9 to provide an indication of the effect that
excluding those with concentrations less than the MDL has on the population of pairs in the
dataset.
24-29
-------
Table 24-9. Analytical Precision by Analytical Method
Method/Pollutant
Group
Average
Coefficient of
Variation
(%)
Number of
Pairs Included
in the
Calculation
Total Number of
Pairs Without
the > MDL
Exclusion
VOCs
(TO-15)
5.64
12,345
18,350
SNMOCs
4.78
6,098
6,682
Methane
2.91
40
40
Carbonyl Compounds
(TO-11A)
2.48
7,244
7,263
PAHs
(TO-13A)
1.79
4,097
5,727
Metals Analysis
(Method IO-3.5/FEM)
4.79
8,477
9,827
Hexavalent Chromium
(ASTM D7641)
6.53
13
14
MQO
15.00 percent CV
Bold = CV greater than or equal to 15 percent
Tables 24-10 through 24-16 present analytical precision for VOCs, SNMOCs, methane,
carbonyl compounds, PAHs, metals, and hexavalent chromium, respectively, as the CV per
pollutant per site and the average CV per pollutant, per site, and per method. Pollutants
exceeding the 15 percent MQO for CV are bolded in each table. In Tables 24-10 through 24-16,
the number of pairs in comparison to the respective tables listed for duplicate or collocated
analyses in Tables 24-2 through 24-8 is higher, the reason for which is two-fold. One reason is
because each primary and duplicate (or collocated) sample produces a replicate analysis. The
second reason is due to replicate analyses run on individual samples. This is also the reason the
number of sites provided in Tables 24-10 through 24-16 is higher than Tables 24-2 through 24-8
(with the exception of methane and hexavalent chromium). The replicate analyses of duplicate,
collocated, and individual samples indicate that the analytical precision level is within the
program MQOs.
24.3.1 VOC Analytical Precision
Table 24-10 presents analytical precision results from replicate analyses of duplicate,
collocated, and select individual VOC samples as the CV per pollutant per site, the average CV
per site, the average CV per pollutant, and the overall average CV across the VOCs listed. Sites
at which duplicate samples were collected are highlighted in blue in Table 24-10, sites at which
collocated samples were collected are highlighted in purple, and sites for which analytical
replicates were run only on individual field samples are highlighted in brown. Collocated VOC
24-30
-------
samples were collected at two of the sites shown in Table 24-10 (PXSS, and TVKY); replicates
were run on only individual VOC samples for six sites, and duplicate VOC samples were
collected at the remaining of sites. However, analytical replicates could be run on any of the sites
collecting VOC samples (e.g., replicates could be run on individual VOC samples as well as
collocated samples for additional precision information for PXSS).
The analytical precision results from replicate analyses show that, for most of the
pollutants, the VOC analytical precision is within 15 percent. The CV ranged from 0 percent
(several pollutants and several sites) to 20.20 percent (vinyl chloride for YUOK). The number of
sites for which a given pollutant has a CV greater than or equal to 15 percent varies from none
(48 pollutants) to two (vinyl chloride).
The pollutant-specific average CV, as shown in orange in Table 24-10, ranges from
2.01 percent (chlorobenzene) to 8.83 percent (ethyl tert-butyl ether). The site-specific average
CV, as shown in green in Table 24-10, ranges from 3.44 percent (RFCO) to 8.70 percent
(YUOK). The overall average analytical precision is 5.65 percent. Note that the results for
acrolein, acetonitrile, acrylonitrile, and carbon disulfide were excluded from the precision
calculations due to the issues described in Section 3.2.
24-31
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ASKY
ATKY
BLKY
BROK
BTUT
CHNJ
CSNJ
DEMI
Acetylene
4.47
5.75
4.82
5.53
7.35
4.41
3.60
3.88
fcrt-Amvl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
7.52
3.87
6.47
5.23
6.39
4.47
2.53
7.56
B ro mochloro methane
3.82
4.04
3.82
—
—
7.35
4.29
7.26
Bromodichloromethane
—
—
—
—
—
—
—
—
Bromoform
—
—
—
—
—
—
—
—
Bro mo methane
9.40
7.64
8.32
6.96
5.75
4.49
6.75
5.44
1.3 -Butadiene
6.07
9.92
14.51
0.00
7.49
13.36
5.31
6.30
Carbon Tetrachloride
5.02
4.04
5.22
6.99
6.80
4.57
2.87
5.04
Chlorobenzene
..
..
..
..
..
—
..
..
Chloroethane
15.34
6.69
5.76
10.11
4.83
6.48
4.45
6.29
Chloroform
6.72
5.64
6.46
9.93
7.01
5.85
3.94
6.92
Chloro methane
4.13
4.45
4.16
5.54
6.95
3.91
2.73
4.36
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloro methane
..
..
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
w-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
--
..
--
--
--
--
--
--
p-Dichlorobenzene
—
—
—
—
—
—
—
—
Dichlorodifluoro methane
3.46
4.28
4.24
5.98
7.12
3.99
2.99
4.05
1,1 -Dichloroethane
—
0.00
-
-
-
-
-
--
1,2-Dichloroethane
9.50
7.64
5.79
7.11
7.79
8.50
5.73
5.83
1,1 -Dichloroethene
—
—
—
—
—
—
—
—
cis-1,2 -Dichloroethy lene
—
—
—
—
—
—
—
—
/ra«.v- 1.2-Dichlorocthvlcnc
—
—
—
7.37
—
—
5.66
—
Dichloromethane
3.78
4.57
2.94
6.44
6.70
3.50
4.69
5.36
1,2-Dichloropropane
—
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
6.01
7.49
5.02
5.79
6.48
4.37
7.22
3.88
Ethyl Acrylate
..
..
..
..
..
..
..
..
Ethyl tert-Butyl Ether
—
—
—
—
—
8.61
6.25
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-32
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ASKY
ATKY
BLKY
BROK
BTUT
CHNJ
CSNJ
DEMI
Ethylbenzene
13.05
8.68
8.74
4.65
7.24
8.60
3.54
5.88
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
10.04
6.06
10.87
6.65
9.65
8.47
5.31
6.44
Methyl Methacrylate
—
3.45
—
—
—
—
7.02
—
Methyl tert-Butyl Ether
1.89
—
—
—
—
0.00
—
—
//-Octane
15.40
3.87
7.04
4.69
7.21
9.48
3.62
8.04
Propylene
6.01
4.20
5.08
5.05
7.53
4.15
3.39
3.82
Styrene
9.41
3.79
3.84
4.79
8.27
7.17
4.58
7.25
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
Tetrachloroethylene
12.18
—
—
—
6.48
6.08
4.76
4.49
Toluene
6.55
4.13
8.03
3.39
5.49
4.87
2.80
6.31
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
0.00
—
—
—
—
—
—
1.1.2 -T richloroethane
—
1.17
—
—
—
—
—
—
Trichloroethylene
—
12.30
—
—
9.83
—
2.38
—
Trichlorofluoromethane
3.22
4.14
4.00
5.20
8.36
3.70
2.81
3.62
Trichlorotrifluoroethane
3.89
4.39
4.13
5.56
7.50
4.28
3.85
4.10
1,2,4 -T rimethylbenzene
11.58
4.87
11.21
4.91
7.89
5.08
5.20
6.37
1,3,5 -T rimethylbenzene
9.66
5.92
2.77
—
—
4.88
6.52
6.76
Vinyl chloride
—
4.21
5.61
15.79
—
0.00
7.18
—
»/,p-Xylene
9.38
7.71
7.54
4.34
6.97
9.58
4.15
6.24
o-Xylene
10.93
7.46
6.75
5.10
6.44
9.99
4.35
5.97
Average CV by Site
7.72
5.24
6.28
6.12
7.18
5.87
4.53
5.67
# of pairs by site
13
11
12
20
16
23
20
24
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-33
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ELNJ
GLKY
GPCO
LEKY
NBIL
NBNJ
NRNJ
NROK
Acetylene
3.60
4.16
5.02
3.53
6.07
5.52
3.42
2.02
fcrt-Amvl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
5.76
4.97
5.36
6.41
3.88
4.29
4.87
3.85
B ro mochloro methane
—
11.11
—
—
—
—
8.99
—
Bromodichloromethane
—
—
—
—
5.56
—
—
—
Bromoform
—
—
—
—
4.53
—
—
—
Bro mo methane
4.25
10.50
7.19
8.70
7.62
7.52
0.00
..
1.3 -Butadiene
5.11
10.00
9.26
7.95
7.64
7.91
4.78
2.43
Carbon Tetrachloride
4.80
3.70
4.68
3.97
5.16
5.07
5.06
3.43
Chlorobenzene
..
..
..
..
..
..
..
..
Chloroethane
7.04
8.25
7.86
6.65
5.47
7.04
6.36
0.00
Chloroform
7.25
8.92
8.86
5.59
4.40
5.44
6.67
4.45
Chloro methane
4.75
4.22
4.16
3.93
5.05
4.72
3.82
3.02
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloro methane
..
..
..
..
5.09
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
w-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
..
--
--
..
..
--
--
..
p-Dichlorobenzene
3.25
—
—
5.24
5.91
—
—
6.45
Dichlorodifluoro methane
4.49
3.79
3.99
3.24
4.85
4.44
4.12
2.83
1,1 -Dichloroethane
—
—
—
—
—
—
—
—
1,2-Dichloroethane
12.28
10.97
8.88
7.91
7.98
5.40
8.48
5.04
1,1 -Dichloroethene
—
—
—
—
—
—
—
—
cis-1,2 -Dichloroethy lene
—
—
—
—
—
—
—
—
/ra«.v- 1.2-Dichlorocthvlcnc
3.23
—
2.18
0.00
—
—
—
4.05
Dichloromethane
6.12
7.09
5.29
4.75
4.66
5.70
4.39
5.42
1,2-Dichloropropane
—
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
7.34
9.03
5.85
2.86
7.14
6.05
1.37
—
Ethyl Acrylate
..
..
..
..
..
..
..
..
Ethyl tert-Butyl Ether
11.61
—
7.97
—
3.46
—
8.61
—
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-34
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ELNJ
GLKY
GPCO
LEKY
NBIL
NBNJ
NRNJ
NROK
Ethylbenzene
4.50
11.54
6.46
8.94
4.73
4.59
7.13
7.00
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
8.65
6.75
8.46
7.50
8.23
7.37
8.87
4.15
Methyl Methacrylate
1.54
—
—
—
—
—
—
—
Methyl tert-Butyl Ether
5.23
—
—
4.33
4.00
—
7.49
—
//-Octane
6.81
6.92
8.12
11.19
8.87
4.01
9.34
5.53
Propylene
3.90
4.15
5.82
7.56
5.38
4.20
3.46
2.59
Styrene
6.42
12.39
4.47
7.44
8.06
9.01
10.48
8.28
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
Tetrachloroethylene
5.59
—
8.53
3.14
3.94
3.68
6.78
5.44
Toluene
3.78
5.49
4.09
6.28
3.52
3.18
5.10
6.07
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
—
—
—
—
—
—
—
1.1.2 -T richloroethane
—
—
—
—
—
—
—
—
Trichloroethylene
0.00
—
—
—
9.82
—
—
—
Trichlorofluoromethane
4.27
3.50
4.15
2.40
4.42
4.29
4.05
2.40
Trichlorotrifluoroethane
5.27
5.38
5.08
3.61
5.82
3.88
5.39
3.51
1,2,4 -T rimethylbenzene
5.81
10.02
6.39
6.57
5.23
3.41
11.55
6.92
1,3,5 -T rimethylbenzene
6.59
4.16
7.62
2.48
5.41
—
5.28
8.74
Vinyl chloride
4.80
—
—
7.44
9.99
—
—
—
»/,p-Xylene
4.16
13.74
5.51
7.01
4.36
4.09
7.13
6.79
o-Xylene
4.11
13.91
6.00
7.08
5.15
2.79
7.36
6.73
Average CV by Site
5.38
7.79
6.19
5.65
5.80
5.15
6.08
4.69
# of pairs by site
22
22
26
9
27
10
12
6
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-35
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
OCOK
PXSS
RFCO
ROIL
S4MO
SEWA
SPAZ
SPIL
Acetylene
6.28
5.16
1.13
5.85
6.45
4.65
4.26
5.70
fcrt-Amvl Methyl Ether
—
—
—
—
—
—
—
—
Benzene
4.18
5.63
2.78
6.01
7.38
5.73
6.33
7.00
B ro mochloro methane
—
—
—
—
—
—
7.89
—
Bromodichloromethane
—
—
—
—
—
—
—
—
Bromoform
—
—
—
—
—
—
—
—
Bro mo methane
13.98
12.27
6.97
6.82
12.93
6.08
4.18
8.31
1.3 -Butadiene
7.74
11.27
2.78
8.42
9.86
10.95
5.63
6.63
Carbon Tetrachloride
4.53
4.30
2.61
5.72
8.16
5.15
4.49
7.24
Chlorobenzene
..
..
0.00
..
..
..
—
..
Chloroethane
8.62
5.53
2.05
6.87
8.16
8.21
4.32
6.26
Chloroform
7.11
4.81
6.63
8.78
6.75
5.68
5.56
8.74
Chloro methane
5.67
4.59
2.06
5.33
5.69
4.17
4.37
5.54
Chloroprene
..
..
..
..
..
..
..
..
Dibromochloro methane
..
6.67
..
..
..
..
..
..
1,2-Dibromoethane
..
..
..
..
..
..
..
..
w-Dichlorobenzene
..
..
..
..
..
..
..
..
o-Dichlorobenzene
--
..
--
--
..
--
..
--
p-Dichlorobenzene
—
5.45
—
—
6.15
—
6.36
—
Dichlorodifluoro methane
5.40
4.90
2.00
5.24
5.62
4.31
4.24
5.56
1,1 -Dichloroethane
—
—
—
—
—
—
—
—
1,2-Dichloroethane
6.07
8.49
6.16
7.IS
9.58
6.58
6.70
12.95
1,1 -Dichloroethene
0.00
—
—
—
3.96
—
—
5.26
cis-1,2 -Dichloroethy lene
—
—
—
—
4.22
—
—
—
/ra«.v- 1.2-Dichlorocthvlcnc
7.14
4.72
—
—
—
8.32
—
—
Dichloromethane
5.07
5.70
4.24
6.06
4.56
6.93
4.04
6.26
1,2-Dichloropropane
—
—
—
—
—
—
—
—
cis-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
trans-1,3 -Dichloropropene
—
—
—
—
—
—
—
—
Dichlorotetrafluoroethane
9.11
6.97
3.36
8.51
10.35
4.50
2.67
7.37
Ethyl Acrylate
..
..
..
..
..
..
..
..
Ethyl tert-Butyl Ether
—
—
—
—
—
—
—
9.47
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-36
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
OCOK
PXSS
RICO
ROIL
S4MO
SEWA
SPAZ
SPIL
Ethylbenzene
5.13
7.29
2.43
7.18
9.74
5.40
6.39
11.18
Hexachloro-1,3 -butadiene
—
—
—
—
—
—
—
—
Methyl Isobutyl Ketone
6.69
6.04
8.80
10.00
8.52
6.45
9.70
10.15
Methyl Methacrylate
—
—
—
—
—
—
—
—
Methyl tert-Butyl Ether
—
—
—
—
—
—
—
4.35
//-Octane
5.98
10.47
6.41
6.25
12.69
3.84
5.47
12.67
Propylene
6.25
4.80
1.77
5.26
5.10
4.51
8.75
5.93
Styrene
8.05
9.67
1.64
7.86
13.97
4.08
8.05
14.47
1.1,2,2-Tetrachloroethane
—
—
—
—
—
—
—
—
Tetrachloroethylene
4.04
5.49
4.04
—
9.02
3.69
4.47
6.66
Toluene
4.65
5.02
2.19
4.80
4.95
4.93
6.19
6.28
1,2,4-Trichlorobenzene
—
—
—
—
—
—
—
—
1,1,1 -T richloroethane
—
—
—
—
0.00
—
—
—
1.1.2 -T richloroethane
—
—
—
—
—
—
—
—
Trichloroethylene
6.66
—
—
—
6.32
—
3.34
9.24
Trichlorofluoromethane
5.18
5.02
2.94
5.12
5.58
4.44
4.28
5.51
Trichlorotrifluoroethane
5.04
4.51
3.01
4.79
5.90
6.35
4.87
6.45
1,2,4 -T rimethylbenzene
6.23
7.66
2.10
9.25
12.12
5.42
8.15
10.85
1,3,5 -T rimethylbenzene
1.54
11.66
—
—
7.87
0.00
5.00
12.81
Vinyl chloride
—
—
6.33
—
7.21
—
—
6.73
»/,p-Xylene
5.83
6.71
1.59
6.57
7.70
5.46
6.49
9.65
o-Xylene
6.00
7.56
3.45
5.97
9.10
5.32
6.41
10.41
Average CV by Site
6.01
6.73
3.44
6.69
7.60
5.43
5.66
8.19
# of pairs by site
24
27
3
9
24
24
15
21
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-37
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TMOK
TOOK
TROK
TVKY
YUOK
# of
pairs
Average
by
Pollutant
Acetylene
4.90
5.70
3.91
5.70
5.89
623
4.78
tert-Amyl Methyl Ether
—
—
—
—
—
0
—
Benzene
5.14
4.52
2.88
5.13
6.39
623
5.26
Bromochloromethane
7.51
—
—
7.59
—
21
6.70
Bromodichloro methane
—
—
—
—
—
19
5.56
Bromofonn
—
—
—
—
—
9
4.53
Bromomethane
5.46
6.59
8.82
8.78
6.98
293
7.45
1.3 -Butadiene
8.74
6.48
9.11
6.09
9.44
374
7.63
Carbon Tetrachloride
4.81
4.09
3.15
5.99
5.63
622
4.91
Chlorobenzene
—
—
—
4.02
—
4
2.01
Chloroethane
6.41
6.72
7.22
6.16
6.97
333
6.62
Chloroform
6.21
6.10
5.33
7.50
13.35
590
6.78
Chloromethane
5.52
4.49
3.84
3.39
5.63
623
4.49
Chloroprene
—
—
—
4.31
—
2
4.31
Dibromochloromethane
—
—
—
—
—
21
5.88
1,2 -Dibromoethane
—
—
—
—
—
0
—
w-Dichlorobenzene
—
—
—
—
—
0
—
o-Dichlorobenzene
—
—
—
—
—
0
—
p-Dichlorobcnzcnc
5.80
—
2.57
—
—
51
5.24
Dichlorodifluoromethane
4.04
4.32
5.32
3.26
5.63
623
4.40
1,1 -Dichloroethane
—
—
—
5.09
—
38
2.55
1,2 -Dichloroethane
7.80
6.00
4.81
6.22
13.76
538
7.83
1,1 -Dichloroethene
—
—
—
6.70
15.71
15
6.33
67.Y- 1.2-Dichlorocthvlcnc
—
—
—
—
—
2
4.22
trans-1,2 -Dichloroethy lene
—
—
0.00
2.70
15.22
34
5.05
Dichloro methane
5.01
5.83
4.22
7.50
7.29
623
5.31
1,2 -Dichloropropane
—
—
—
3.45
—
1
3.45
cis-1,3 -Dichloropropene
—
—
—
—
—
0
—
trans-1,3 -Dichloropropene
—
—
—
—
—
0
—
Dichlorotetrafluoroethane
6.75
7.46
6.86
7.90
6.55
335
6.22
Ethyl Acrylate
—
—
—
—
—
0
—
Ethyl ter/-Butyl Ether
—
12.86
4.88
—
14.63
49
8.83
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-38
-------
Table 24-10. VOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TMOK
TOOK
TROK
TVKY
YUOK
# of
pairs
Average
by
Pollutant
Ethylbenzene
7.60
7.26
4.31
10.97
8.78
503
7.21
Hexachloro-1,3 -butadiene
—
—
—
—
—
0
—
Methyl Isobutyl Ketone
8.21
4.96
4.83
8.96
10.09
493
7.82
Methyl Methacrylate
4.42
—
—
—
8.32
10
4.95
Methyl ter/-Butyl Ether
—
—
5.30
0.00
—
22
3.62
//-Octane
6.58
4.61
3.62
10.56
7.57
476
7.48
Propylene
4.71
4.66
3.28
6.93
6.17
623
4.98
Styrene
5.30
9.52
8.50
8.87
5.21
288
7.62
1.1,2,2-Tetrachloroethane
—
—
—
—
—
0
—
T etrachloroethylene
3.64
4.52
4.89
6.67
9.46
254
5.74
Toluene
4.22
3.46
3.31
5.42
6.70
622
4.87
1,2,4-Trichlorobenzene
—
—
—
—
—
0
—
1,1,1 -T richloroethane
—
—
—
7.03
—
4
2.34
1,1,2-Trichloroethane
—
—
—
8.03
—
21
4.60
T richloroethylene
16.81
—
—
6.16
—
43
7.53
T richlorofluoro methane
4.37
4.09
3.17
3.36
5.92
623
4.26
T richlorotrifluoroethane
4.34
4.77
3.36
3.96
6.51
623
4.81
1,2,4-Trimethylbenzene
7.57
6.70
6.00
7.36
8.31
431
7.27
1.3.5 -T rimethy lbenzene
8.36
6.96
7.19
3.16
2.40
163
5.99
Vinyl chloride
—
—
—
4.65
20.20
104
7.70
»/,p-Xylene
6.06
5.79
4.44
9.41
8.03
529
6.64
o-Xylene
6.81
6.09
5.04
10.03
8.38
519
6.92
Average CV by Site
6.31
5.94
4.83
6.19
8.70
12,817
5.65
# of pairs by site
22
24
24
109
24
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-39
-------
24.3.2 SNMOC Analytical Precision
Table 24-11 presents analytical precision results from replicate analyses of duplicate,
collocated, and select individual samples as the CV per pollutant per site, the average CV per
site, the average CV per pollutant, and the overall average CV across the SNMOCs listed. Sites
at which duplicate samples were collected (4) are highlighted in blue in Table 24-11, sites at
which collocated samples were collected (2) are highlighted in purple, and sites for which
analytical replicates were run only on individual field samples (4) are highlighted in brown.
However, analytical replicates could be run on any of the sites collecting SNMOC samples
(e.g., replicates could be run on individual SNMOC samples as well as collocated samples for
additional precision information for PACO).
The CVs range from 0 percent (1,2,3-trimethylbenzene for BMCO) to 23.06 percent
(//-tridecane for NBIL). Three SNMOCs have a site-specific CV greater than or equal to
15 percent. The pollutant-specific average CV, as shown in orange in Table 24-11, ranges from
0.44 percent (1-dodecene) to 13.28 percent (//-tridecane). Note that the average CV for
1-dodecene is based on a single individual sample and its replicate analysis. None of the
SNMOCs shown in Table 24-11 have an average CV greater than or equal to 15 percent. The
site-specific average CV, as shown in green in Table 24-11, varies by less than 2 percent,
ranging from 3.76 percent (NROK) to 5.57 percent (RFCO). The overall average analytical
precision is 4.78 percent. Note that the results for TNMOC were not included in the precision
calculations.
24-40
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BMCO
BRCO
BROK
BTUT
GSCO
NBIL
Acetylene
1.58
3.42
3.42
2.48
2.35
2.49
Benzene
6.80
2.97
3.40
3.98
3.52
4.77
1.3 -Butadiene
—
—
—
4.89
7.36
—
//-Butane
0.45
0.99
0.92
1.38
1.28
1.85
1-Butene
—
—
—
—
—
—
67.Y-2-Bute nc
9.53
15.36
—
6.27
3.73
2.21
;ra«.v-2-Butcne
3.66
2.66
6.44
6.59
3.45
Cyclohexane
2.49
1.62
1.54
3.08
3.11
6.45
Cyclopentane
1.78
2.97
1.55
2.18
2.80
5.54
Cyclopentene
—
5.20
—
—
—
—
w-Decane
7.45
3.84
5.41
4.85
3.92
6.03
1-Decene
—
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
—
p-Dicthvlbcnzcnc
—
—
—
—
—
6.78
2,2-Dimethylbutane
2.45
4.02
3.32
9.42
7.09
5.83
2,3 -Dimethy lbutane
1.15
2.42
2.49
2.53
4.68
4.56
2,3 -Dimethy lpentane
4.72
5.61
3.20
2.70
6.82
6.55
2,4-Dimethylpentane
6.16
7.21
4.63
4.33
6.73
5.91
n-Dodecane
—
1.72
6.30
11.60
10.54
8.68
1-Dodecene
—
—
—
—
—
—
Ethane
0.57
0.51
1.00
0.68
0.75
0.97
2-Ethyl-l-butene
—
—
—
—
—
—
Ethylbenzene
7.06
6.29
6.90
5.37
6.77
7.86
Ethylene
1.65
1.18
1.50
1.02
1.04
1.08
/w-Ethyltoluene
7.79
7.68
3.96
5.00
4.52
4.45
o-Ethyltoluene
—
—
7.90
7.86
13.69
7.09
p-Ethyltoluene
11.72
7.33
7.40
8.78
9.33
7.26
//-Heptane
4.86
2.83
2.16
3.90
3.66
4.15
1-Heptene
—
8.14
—
—
—
—
n-Hexane
3.49
2.25
2.45
2.36
2.07
3.25
1-Hexene
8.09
8.24
7.73
7.35
3.12
10.30
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-41
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BROK
BTUT
GSCO
NBIL
c/.v-2-Hc.\cnc
—
—
—
—
—
—
-Pincnc
—
—
—
—
—
—
Propane
0.58
0.61
0.87
0.83
0.70
0.92
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-42
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
BMCO
BRCO
BROK
BTUT
GSCO
NBIL
n-Propylbenzene
—
3.02
2.08
6.57
12.22
3.35
Propylene
1.33
2.41
2.21
1.49
2.77
2.97
Propyne
—
—
—
—
—
—
Styrene
—
0.66
—
—
—
—
Toluene
5.78
3.79
3.12
2.59
2.86
3.46
«-T ridccanc
—
—
9.59
—
—
23.06
1-Tridecene
—
—
—
—
—
—
1,2,3 -Trimethylbenzene
0.00
6.63
7.91
5.12
7.62
7.66
1,2,4 -T rimethylbenzene
6.82
6.15
4.82
5.38
6.67
4.23
1,3,5 -T rimethylbenzene
10.38
5.79
4.71
8.08
0.63
6.50
2,2,3 -Trimethylpentane
—
7.04
—
—
—
6.53
2,2,4 -T rimethy lpentane
—
—
8.94
4.72
5.62
4.93
2,3,4 -T rimethy lpentane
4.07
11.47
12.07
10.84
6.40
5.63
/7-Undecane
—
—
6.13
6.83
0.23
4.49
1-Undecene
—
—
—
—
—
—
w-Xylene/p-Xylene
5.51
5.67
3.51
3.45
5.61
4.94
o-Xylene
6.74
9.06
5.10
5.26
5.54
5.51
SNMOC (Sum of Knowns)
1.35
0.58
0.75
1.43
0.73
1.81
Sum of Unknowns
2.31
2.34
3.86
3.22
3.42
4.55
Average CV by Site
4.28
4.69
4.33
4.64
4.47
5.01
# of pairs by site
5
14
20
16
10
26
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-43
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
NROK
PACO
RICO
RICO
# of
pairs
Average
by
Pollutant
Acetylene
1.66
2.85
3.19
1.99
139
2.54
Benzene
3.95
3.85
4.23
5.67
136
4.32
1.3 -Butadiene
2.39
—
13.62
3.84
17
6.42
//-Butane
0.93
1.39
2.07
1.58
139
1.28
1-Butene
—
—
—
—
—
—
67.Y-2-Bute nc
4.77
4.64
8.34
5.75
66
6.73
;ra«.v-2-Butcne
4.40
6.45
5.93
4.88
48
4.94
Cyclohexane
3.22
1.71
5.72
1.89
137
3.08
Cyclopentane
1.95
2.82
3.74
2.44
113
2.78
Cyclopentene
0.36
—
—
4.72
4
3.43
w-Decane
7.08
4.30
8.11
6.64
81
5.76
1-Decene
—
—
—
—
—
—
/w-Diethylbenzene
—
—
—
—
—
—
p-Dicthvlbcnzcnc
—
—
—
2.33
3
4.55
2,2-Dimethylbutane
1.35
3.57
3.66
4.69
110
4.54
2,3 -Dimethy lbutane
1.06
1.56
4.97
2.60
139
2.80
2,3 -Dimethy lpentane
1.55
4.72
6.32
3.39
136
4.56
2,4-Dimethylpentane
4.45
6.97
6.50
5.14
124
5.80
n-Dodecane
—
6.06
10.32
6.24
35
7.68
1-Dodecene
—
—
0.44
—
1
0.44
Ethane
0.73
0.78
0.72
1.17
139
0.79
2-Ethyl-l-butene
—
—
—
—
—
—
Ethylbenzene
4.18
7.80
6.55
8.22
122
6.70
Ethylene
2.00
1.04
1.12
1.70
139
1.33
/w-Ethyltoluene
4.73
6.53
5.13
6.05
102
5.58
o-Ethyltoluene
8.05
3.87
6.55
5.30
41
7.54
p-Ethyltoluene
7.51
7.47
9.93
4.25
84
8.10
//-Heptane
2.90
2.78
5.00
4.19
139
3.64
1-Heptene
—
—
5.76
—
4
6.95
n-Hexane
3.68
3.44
5.04
3.93
139
3.20
1-Hexene
7.36
7.43
6.63
6.68
78
7.29
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-44
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
NROK
PACO
RICO
RICO
# of
pairs
Average
by
Pollutant
c/.v-2-Hc.\cnc
—
—
12.37
—
1
12.37
-Pincnc
—
—
—
—
—
—
Propane
0.47
0.77
1.00
1.30
139
0.81
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-45
-------
Table 24-11. SNMOC Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
NROK
PACO
RICO
RICO
# of
pairs
Average
by
Pollutant
n-Propylbenzene
6.81
3.08
4.67
7.14
44
5.44
Propylene
0.59
1.57
1.48
1.70
139
1.85
Propyne
4.55
—
—
—
2
4.55
Styrene
—
3.06
0.97
9.27
10
3.49
Toluene
3.30
2.56
3.49
4.77
139
3.57
«-T ridccanc
—
—
7.20
—
9
13.28
1-Tridecene
—
—
—
—
—
—
1,2,3 -Trimethylbenzene
4.53
8.73
6.99
10.76
34
6.59
1,2,4 -T rimethylbenzene
2.96
6.82
5.10
4.48
111
5.34
1,3,5 -T rimethylbenzene
5.46
6.35
11.92
9.35
71
6.92
2,2,3 -Trimethylpentane
—
2.98
7.36
6.19
10
6.02
2,2,4 -T rimethy lpentane
3.06
2.80
6.46
4.52
89
5.13
2,3,4 -T rimethy lpentane
6.86
17.78
8.68
6.67
106
9.05
/7-Undecane
—
3.95
8.56
9.79
34
5.71
1-Undecene
—
—
4.37
—
2
4.37
w-Xylene/p-Xylene
3.36
2.82
5.44
5.52
138
4.58
o-Xylene
2.42
6.29
7.07
4.49
137
5.75
SNMOC (Sum of Knowns)
1.08
11.34
1.13
1.98
139
2.22
Sum of Unknowns
3.18
3.53
3.46
4.38
139
3.43
Average CV by Site
3.76
4.41
5.57
4.72
6,098
4.78
# of pairs by site
6
14
12
16
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicate analyses were run on only
individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-46
-------
24.3.3 Methane Analytical Precision
Table 24-12 presents the analytical precision results from replicate analysis of duplicate
and select individual methane samples as the C V per site and the overall average CV for the
method. As discussed in Section 24.2.3, only BROK and NROK sampled methane during the
2015 and/or 2016 monitoring efforts. The site-specific CV ranges from 2.74 percent for BROK
to 3.09 percent for NROK. The overall average method precision for methane is 2.91 percent, as
shown in orange in Table 24-12, which is considerably less than the MQO of 15 percent CV.
Table 24-12. Methane Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BROK
NROK
# of
pairs
Average by
Pollutant
Methane
2.74
3.09
40
2.91
# of pairs
28
12
Bold = CV greater than or equal to 15 percent
Orange shading indicates the pollutant-specific average CV.
BOLD ITALICS = EPA-designated NATTS Site
24.3.4 Carbonyl Compound Analytical Precision
Table 24-13 presents the analytical precision results from replicate analyses of duplicate,
collocated, and select individual carbonyl compound samples as the CV per pollutant per site, the
average CV per site, the average CV per pollutant, and the overall average CV for the carbonyl
compounds listed. Sites at which duplicate samples were collected are highlighted in blue in
Table 24-13, sites at which collocated samples were collected are highlighted in purple, and sites
for which replicates were run on only individual samples are highlighted in brown. Collocated
carbonyl compound samples were collected at three of the sites shown in Table 24-13 (DEMI,
INDEM, and PXSS); replicates were run on only individual field samples for five sites, and
duplicate samples were collected at the remaining sites. Analytical replicates were typically run
on duplicate or collocated sample pairs, although replicates could be analyzed for any sample
type.
The overall average CV is 2.48 percent, which is well within the program MQO of
15 percent CV. The analytical precision results from replicate analyses range from 0 percent
(several pollutants at different sites) to 9.07 percent (valeraldehyde for GSCO), indicating that
every pollutant-site combination has a CV less than 15 percent.
24-47
-------
The pollutant-specific average CV, as shown in orange in Table 24-13, ranges from
0.75 percent (acetone) to 4.36 percent (tolualdehydes), indicating that all of the pollutant-specific
average CVs are less than 5 percent. The site-specific average CV, as shown in green in
Table 24-13, ranges from 1.83 percent (NROK) to 3.53 percent (RICO), indicating that all of the
site-specific average CVs are also less than 5 percent.
Table 24-13. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
A/.II
BMCO
BRCO
BROK
BTUT
CHNJ
CSNJ
DEMI
Acetaldehyde
1.61
1.19
2.32
0.46
0.59
0.36
0.47
0.97
Acetone
1.50
0.31
0.38
1.56
0.46
0.55
0.73
0.55
Benzaldehyde
4.29
3.14
5.33
3.70
3.30
4.50
2.46
4.08
2-Butanone
3.24
1.55
2.95
1.55
2.15
1.92
1.79
2.45
Butyraldehyde
3.94
4.93
3.38
2.28
2.26
2.86
1.56
2.33
Crotonaldehyde
2.54
3.80
4.26
1.68
3.15
0.87
2.22
2.32
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
—
Formaldehyde
0.83
0.51
1.51
1.01
0.54
0.55
0.76
0.86
Hexaldehyde
3.78
5.44
5.25
3.63
2.14
3.49
2.90
3.11
Isovaleraldehyde
—
—
—
—
—
—
—
—
Propionaldehyde
2.69
4.62
3.31
2.60
1.48
2.34
2.01
1.58
Tolualdehydes
4.19
5.66
3.89
4.39
3.69
4.71
3.66
4.19
Valeraldehyde
3.88
4.30
0.00
3.02
3.91
4.58
2.65
2.84
Average CV by Site
2.95
3.22
2.96
2.35
2.15
2.43
1.93
2.30
# of pairs by site
26
5
9
20
22
16
22
21
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-48
-------
Table 24-13. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ELNJ
GLKY
GPCO
GSCO
IN I)IM
NBIL
NBNJ
Acetaldehyde
0.44
1.08
0.66
0.96
1.43
0.68
0.55
Acetone
0.51
0.87
0.59
0.31
0.70
0.51
0.62
Benzaldehyde
3.10
4.83
3.66
2.44
4.52
4.65
3.15
2-Butanone
1.28
2.06
1.29
1.28
2.75
1.68
1.94
Butyraldehyde
1.34
2.80
2.27
0.00
3.35
2.27
2.22
Crotonaldehyde
1.53
2.04
1.75
3.07
3.27
3.66
2.16
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
Formaldehyde
0.57
0.73
0.83
0.99
0.79
0.97
1.11
Hexaldehyde
2.43
5.02
3.77
0.00
3.66
4.51
4.38
Isovaleraldehyde
—
—
—
—
—
—
—
Propionaldehyde
1.89
1.55
2.13
2.12
2.94
2.10
3.45
Tolualdehydes
4.62
4.82
4.66
7.54
3.79
3.84
3.73
Valeraldehyde
3.65
6.00
3.64
9.07
4.52
4.00
3.30
Average CV by Site
1.94
2.89
2.29
2.53
2.88
2.62
2.42
# of pairs by site
24
24
21
4
36
32
10
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-49
-------
Table 24-13. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
NRNJ
NROK
OCOK
OR II
PACO
PXSS
RICO
Acetaldehyde
0.72
0.33
0.40
1.05
0.31
0.44
1.54
Acetone
0.60
0.35
1.41
1.43
1.62
0.57
1.08
Benzaldehyde
3.36
2.24
2.64
3.31
4.35
2.57
4.04
2-Butanone
1.56
1.15
1.32
3.62
1.95
1.80
2.92
Butyraldehyde
1.70
1.33
2.74
3.43
3.83
2.39
4.00
Crotonaldehyde
3.16
0.43
1.95
2.20
2.73
2.12
5.70
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
Formaldehyde
0.62
0.87
0.59
1.02
0.52
0.76
1.63
Hexaldehyde
2.73
3.27
3.69
3.64
0.00
3.05
5.10
Isovaleraldehyde
—
—
—
—
—
—
—
Propionaldehyde
2.24
1.98
2.43
2.94
2.10
2.16
2.99
Tolualdehydes
4.95
4.80
3.69
4.83
4.21
3.54
5.01
Valeraldehyde
2.33
3.34
3.00
3.39
5.44
3.12
4.84
Average CV by Site
2.18
1.83
2.17
2.80
2.46
2.05
3.53
# of pairs by site
12
8
24
22
7
43
9
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-50
-------
Table 24-13. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
ROIL
S4MO
SEWA
SKFL
S P11
SYFL
TMOK
Acetaldehyde
0.70
0.83
2.14
0.48
0.48
1.05
0.35
Acetone
0.71
0.79
0.72
0.56
0.32
0.93
1.41
Benzaldehyde
2.24
3.76
3.87
3.67
3.16
4.28
3.48
2-Butanone
2.74
1.66
1.52
2.68
0.97
3.41
1.76
Butyraldehyde
3.82
2.53
3.47
3.53
1.62
3.15
2.70
Crotonaldehyde
2.07
2.36
4.18
2.16
2.44
1.98
2.19
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
—
Formaldehyde
0.81
0.84
1.88
0.86
0.61
0.59
0.46
Hexaldehyde
4.30
3.54
4.44
3.79
3.60
3.51
4.17
Isovaleraldehyde
—
—
—
—
—
—
—
Propionaldehyde
3.66
2.60
3.06
2.72
1.70
3.08
1.56
Tolualdehydes
3.89
3.81
4.40
4.10
4.37
4.60
3.33
Valeraldehyde
4.13
3.61
4.36
4.65
2.97
4.50
3.43
Average CV by Site
2.64
2.40
3.09
2.65
2.02
2.83
2.26
# of pairs by site
10
24
24
22
20
21
24
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting collocated
samples; and brown shading identifies sites for which replicate analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-51
-------
Table 24-13. Carbonyl Compound Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
TOOK
TROK
WIMN
YUOK
# of
pairs
Average
by
Pollutant
Acetaldehyde
0.57
0.40
0.95
0.33
671
0.81
Acetone
0.60
0.50
0.54
0.29
671
0.75
Benzaldehyde
3.44
3.90
3.70
3.32
656
3.59
2-Butanone
1.19
0.77
2.81
1.01
667
1.96
Butyraldehyde
1.58
2.22
3.03
3.11
668
2.67
Crotonaldehyde
2.05
2.10
3.10
2.00
666
2.52
2,5 -Dimethylbenzaldehyde
—
—
—
—
—
—
Formaldehyde
0.64
0.78
0.68
0.85
672
0.84
Hexaldehyde
3.55
3.54
3.27
3.73
655
3.53
Isovaleraldehyde
—
—
—
—
—
—
Propionaldehyde
1.83
1.76
2.39
2.57
660
2.44
Tolualdehydes
3.83
3.95
4.24
4.95
605
4.36
Valeraldehyde
3.53
3.16
3.97
4.05
653
3.85
Average CV by Site
2.07
2.10
2.61
2.38
7,244
2.48
# of pairs by site
24
24
38
24
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting duplicate samples; purple shading identifies sites collecting
collocated samples; and brown shading identifies sites for which replicates were run on only individual
samples.
BOLD ITALICS = EPA-designated NATTS Site
24-52
-------
24.3.5 PAH Analytical Precision
Table 24-14 presents analytical precision results from replicate analyses of collocated and
select individual samples as the CV per pollutant per site, the average CV per site, the average
CV per pollutant, and the overall average CV across the PAHs listed. Sites at which collocated
PAH samples were collected are highlighted in blue in Table 24-14 while sites for which
replicates were run on individual field samples are highlighted in brown. Collocated PAH
samples were collected at DEMI, RUCA, and SEW A; replicate analyses were run on only
individual PAH samples for the remaining sites. However, analytical replicates could be run on
samples collected at any of sites (e.g., replicates could be run on individual PAH samples as well
as collocated samples for additional precision information for RUCA).
The CVs range from 0 percent (benzo(a)pyrene for BTUT and benzo(k)fluoranthene for
WADC) to 10.96 percent (cyclopenta(c,d)pyrene for BXNY), indicating that every pollutant-site
combination has a CV less than 15 percent. The pollutant-specific average CV, as shown in
orange in Table 24-14, ranges from 0.70 percent (phenanthrene) to 3.62 percent
(benzo(k)fluoranthene), indicating that all of the pollutant-specific average CVs are less than
4 percent. The site-specific average CV, as shown in green in Table 24-14, ranges from
1.37 percent (GPCO) to 2.27 percent (UNVT). The overall average analytical precision is
1.79 percent CV.
24-53
-------
Table 24-14. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
BOMA
BTUT
BXNY
CELA
DEMI
GLKY
GPCO
Acenaphthene
1.00
2.57
1.33
1.53
1.25
2.12
0.93
Acenaphthylene
1.09
2.90
1.26
1.73
1.74
0.74
1.75
Anthracene
5.99
2.34
2.35
1.94
3.31
2.90
1.70
Benzo(a)anthracene
1.34
1.43
0.81
0.46
0.63
0.48
0.63
Benzo(a)pyrene
1.66
0.00
1.88
1.65
1.86
1.66
1.25
Benzo(b)fluoranthene
6.35
5.43
1.32
2.45
1.65
9.19
2.04
Benzo(e)pyrene
1.21
0.40
0.92
0.72
0.85
0.63
0.70
Benzo(g,h,i)perylene
1.36
0.46
0.88
1.66
0.70
0.99
0.69
Benzo(k)fluoranthene
4.91
—
6.24
—
10.03
—
0.61
Chrysene
0.73
2.79
0.45
0.67
0.50
1.00
0.93
Coronene
6.56
1.53
2.43
2.26
2.44
1.17
1.67
Cyclopenta(c,d)pyrene
1.82
1.22
10.96
2.86
1.81
1.12
2.09
Dibenz(a,h)anthracene
1.37
—
1.89
—
1.23
0.86
2.53
Fluoranthene
1.24
1.84
0.95
0.94
1.32
0.83
1.61
Fluorene
0.83
0.99
0.90
0.63
1.55
1.93
1.01
9-Fluorenone
1.14
1.98
0.88
1.28
1.03
1.29
1.79
Indeno( 1,2,3 -c,d)pyrene
1.71
0.51
1.00
1.32
1.31
0.64
0.89
Naphthalene
1.11
2.23
0.94
2.53
1.38
3.38
1.87
Perylene
1.65
—
1.19
2.53
9.10
1.02
1.12
Phenanthrene
0.83
1.66
0.77
0.55
0.62
0.49
0.50
Pyrene
1.04
1.90
0.81
1.05
1.12
0.99
1.83
Retene
2.08
2.72
0.97
1.55
1.59
1.42
1.92
Average CV by Site
2.14
1.84
1.87
1.52
2.14
1.66
1.37
# of pairs by site
14
15
17
14
21
16
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-54
-------
Table 24-14. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
NBIL
PRRI
PXSS
RIVA
ROCH
RUCA
S4MO
Acenaphthene
1.12
0.75
0.74
1.34
5.95
1.65
0.81
Acenaphthylene
3.35
0.94
1.73
3.61
2.28
1.68
1.70
Anthracene
1.37
4.93
2.05
1.16
2.02
3.05
1.95
Benzo(a)anthracene
0.70
1.11
0.61
0.36
0.83
0.68
1.42
Benzo(a)pyrene
1.46
2.48
0.31
—
1.40
1.52
0.63
Benzo(b)fluoranthene
6.49
3.96
0.73
1.22
2.94
0.75
2.47
Benzo(e)pyrene
0.83
0.75
0.74
0.37
1.05
1.14
0.69
Benzo(g,h,i)perylene
0.68
0.84
0.75
0.65
1.11
1.67
1.27
Benzo(k)fluoranthene
7.27
3.55
1.27
—
3.43
2.16
0.16
Chrysene
0.61
0.70
0.92
0.76
0.53
0.82
0.56
Coronene
3.62
4.53
1.53
5.38
3.42
3.19
3.07
Cyclopenta(c,d)pyrene
2.92
2.63
2.04
1.74
2.45
1.80
1.13
Dibenz(a,h)anthracene
3.65
4.83
1.58
—
3.95
2.26
1.32
Fluoranthene
0.94
1.00
1.41
0.95
1.56
1.40
1.08
Fluorene
0.96
0.81
1.22
2.07
0.84
1.77
1.53
9-Fluorenone
0.99
1.18
2.49
1.21
1.75
1.74
1.30
Indeno( 1,2,3 -c,d)pyrene
0.65
2.69
0.65
0.57
1.14
1.35
1.09
Naphthalene
1.73
1.36
2.75
2.20
1.54
2.09
2.05
Perylene
1.57
2.61
2.97
1.61
1.91
2.45
2.06
Phenanthrene
0.53
0.51
0.75
0.66
0.66
0.69
0.76
Pyrene
0.93
0.86
1.48
1.15
1.75
1.74
1.05
Retene
1.32
1.14
1.75
1.50
2.46
2.14
2.59
Average CV by Site
1.99
2.01
1.39
1.50
2.04
1.72
1.40
# of pairs by site
15
13
14
14
13
47
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is
calculated from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-55
-------
Table 24-14. PAH Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SEWA
SJJCA
SKFL
UNVT
WADC
# of
Pairs
Average
by
Pollutant
Acenaphthene
1.22
0.80
2.17
2.68
1.29
259
1.65
Acenaphthylene
1.75
2.49
0.53
1.09
1.97
155
1.81
Anthracene
1.74
4.89
3.99
1.29
5.99
205
2.89
Benzo(a)anthracene
1.12
1.21
—
—
0.32
106
0.83
Benzo(a)pyrene
1.61
2.65
—
—
0.39
85
1.40
Benzo(b)fluoranthene
2.46
1.56
1.26
1.03
1.57
165
2.89
Benzo(e)pyrene
1.55
1.44
0.89
—
0.98
148
0.88
Benzo(g,h,i)perylene
1.03
0.99
1.28
—
1.18
179
1.01
Benzo(k)fluoranthene
—
3.84
—
—
0.00
50
3.62
Chrysene
0.88
1.01
1.03
1.38
0.61
228
0.89
Coronene
6.02
2.56
2.07
10.66
2.47
177
3.50
Cyclopenta(c,d)pyrene
5.62
2.38
2.87
4.73
3.05
79
2.91
Dibenz(a,h)anthracene
—
—
—
—
0.61
48
2.17
Fluoranthene
1.74
1.16
1.08
1.77
1.18
309
1.26
Fluorene
2.42
1.81
1.63
1.28
0.92
224
1.32
9-Fluorenone
1.55
1.41
0.96
1.41
1.47
311
1.41
Indeno( 1,2,3 -c,d)pyrene
1.00
0.71
0.85
—
0.72
144
1.04
Naphthalene
2.29
2.10
1.99
1.04
1.14
314
1.88
Perylene
—
—
—
—
1.43
63
2.37
Phenanthrene
0.73
0.61
0.54
0.80
0.57
313
0.70
Pyrene
1.83
1.35
1.15
1.53
1.23
303
1.31
Retene
1.70
1.38
1.88
1.06
1.81
232
1.74
Average CV by Site
2.01
1.82
1.54
2.27
1.40
4,097
1.79
# of pairs by site
23
15
13
13
13
- = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-56
-------
24.3.6 Metals Analytical Precision
Table 24-15 presents analytical precision results from replicate analyses of collocated and
select individual samples as the CV per pollutant per site, the average CV per site, the average
CV per pollutant, and the overall average CV across the metals listed. Sites at which collocated
metals samples were collected are highlighted in blue in Table 24-15 while sites for which
replicates were run on only individual field samples are highlighted in brown. Collocated metals
samples were collected at six sites; replicate analyses were run on individual metals samples for
the remaining sites. However, analytical replicates could be run on any of the sites collecting
metals samples (e.g., replicates could be run on individual metals samples as well as collocated
samples for additional precision information for ASKY-M).
The CVs exhibit low- to mid-level variability, ranging from 0 percent (for several sites
and pollutants) to 34.57 percent (mercury for BLKY). The pollutant-specific average CV, as
shown in orange in Table 24-15, ranges from 1.33 percent (lead and manganese) to 14.73 percent
(mercury), indicating that none of the pollutant-specific average CVs are greater than or equal to
15 percent. The site-specific average CV, as shown in green in Table 24-15, ranges from
2.66 percent (TOOK) to 6.85 percent (PAFL); all 19 sites sampling metals have site-specific
average CVs less than 15 percent. The overall average analytical precision is 4.79 percent CV.
24-57
-------
Table 24-15. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant
Pollutant
ASKY-M
BAKY
BLKY
BOMA
BTUT
GLKY
GPCO
Antimony
0.93
1.23
1.52
0.89
1.05
1.16
1.36
Arsenic
4.14
6.61
3.15
9.87
7.33
9.23
3.83
Beryllium
8.18
7.27
15.28
15.31
8.60
19.03
27.06
Cadmium
1.97
3.19
5.39
4.05
4.51
4.57
5.39
Chromium
0.25
0.36
1.17
0.92
0.58
1.20
—
Cobalt
1.17
1.36
2.23
2.22
1.84
3.70
1.48
Lead
0.91
1.74
0.59
0.73
1.62
0.77
0.64
Manganese
0.74
2.03
0.53
1.05
0.81
1.16
0.70
Mercury
12.95
24.54
34.57
11.26
13.86
13.25
6.75
Nickel
0.93
3.27
4.63
2.43
3.84
3.16
1.70
Selenium
4.89
2.15
5.97
13.00
11.29
6.39
7.74
Average CV by Site
3.37
4.89
6.82
5.61
5.03
5.78
5.66
# of pairs by site
24
10
8
116
26
93
26
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-58
-------
Table 24-15. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
LEKY
NBIL
OCOK
PAFL
PXSS
S4MO
SEWA
Antimony
0.94
1.66
1.73
1.97
0.83
1.33
0.92
Arsenic
4.13
1.85
2.66
5.94
4.10
6.55
9.62
Beryllium
6.89
17.85
6.76
17.89
5.89
13.60
0.00
Cadmium
2.91
4.53
4.88
12.92
4.27
2.68
3.41
Chromium
0.14
4.92
1.73
0.00
—
2.54
—
Cobalt
4.64
4.92
3.71
0.00
0.74
2.32
4.30
Lead
0.90
2.21
2.24
0.92
0.54
0.82
0.62
Manganese
1.12
1.93
2.65
2.04
0.83
0.77
0.65
Mercury
18.52
7.02
7.17
26.13
10.48
12.92
21.60
Nickel
2.19
12.74
3.54
0.99
0.96
1.88
1.65
Selenium
9.10
3.01
3.52
6.60
6.27
7.15
10.54
Average CV by Site
4.68
5.70
3.69
6.85
3.49
4.78
5.33
# of pairs by site
10
11
16
7
11
233
12
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24-59
-------
Table 24-15. Metals Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site and Pollutant (Continued)
Pollutant
SJJCA
TMOK
TOOK
TROK
YUOK
# of Pairs
Average
by
Pollutant
Antimony
1.46
1.50
1.10
1.33
3.25
897
1.38
Arsenic
5.75
2.53
1.44
2.28
1.86
889
4.89
Beryllium
18.24
4.35
4.63
5.98
6.29
794
11.01
Cadmium
4.09
8.29
3.74
2.52
7.78
897
4.79
Chromium
0.00
2.24
1.70
2.85
1.73
386
1.40
Cobalt
2.61
5.50
1.72
4.73
2.55
730
2.72
Lead
0.54
3.00
1.31
3.04
2.05
897
1.33
Manganese
0.58
1.98
1.25
2.33
2.18
897
1.33
Mercury
26.75
6.54
4.82
15.19
5.60
414
14.73
Nickel
1.26
5.72
5.24
3.22
1.29
832
3.19
Selenium
6.22
2.88
2.30
2.36
1.84
844
5.96
Average CV by Site
6.14
4.05
2.66
4.17
3.31
8,477
4.79
# of pairs by site
9
16
242
14
13
— = No pairs with concentrations greater than or equal to the MDL.
Bold = CV greater than or equal to 15 percent
Green shading indicates the site-specific average CV for this method.
Orange shading indicates the pollutant-specific average CV; the overall average CV for this method is calculated
from the pollutant-specific averages and is provided at the bottom of the orange column.
Blue shading identifies sites collecting collocated samples and brown shading identifies sites for which replicate
analyses were run on only individual samples.
BOLD ITALICS = EPA-designated NATTS Site
24.3.7 Hexavalent Chromium Analytical Precision
Table 24-16 presents analytical precision results for replicate analyses of collocated
samples as the CV per site and the overall average CV for hexavalent chromium. RIVA is the
only site at which hexavalent chromium sampling was conducted in 2015 and 2016 (with
sampling discontinued in June 2016). The analytical precision for hexavalent chromium is
6.53 percent, as shown in orange in Table 24-16, which is considerably less than the MQO of
15 percent CV for method precision.
24-60
-------
Table 24-16. Hexavalent Chromium Analytical Precision: Coefficient of Variation
Based on Replicate Analyses by Site
Pollutant
RIVA
# of pairs
Average
by
Pollutant
Hexavalent Chromium
6.53
13
6.53
Bold = CV greater than or equal to 15 percent
Orange shading indicates the overall average CV for this method.
Blue shading identifies sites collecting collocated samples.
BOLD ITALICS = EPA-designated NATTS Site
24.4 Accuracy
Laboratories typically evaluate their accuracy by analyzing audit samples that are
prepared by an external source. The pollutants and the respective concentrations of the audit
samples are unknown to the laboratory. The laboratory analyzes the samples and the external
source evaluates the measured concentrations against a comparison concentration of those audit
samples and calculates a percent difference. Accuracy, or bias, indicates the extent to which
experimental measurements represent their corresponding "true" or "actual" values.
Laboratories participating in the NATTS program are provided with proficiency test (PT)
audit samples for VOCs, carbonyl compounds, PAHs, and metals, which are used to
quantitatively measure analytical accuracy. PT samples for hexavalent chromium were
discontinued after monitoring for this pollutant was no longer a mandatory part of the NATTS
program (2014). Tables 24-17 through 24-20 present ERG's results for PT audit samples
analyzed in 2015 and 2016. Results for PT audit samples are presented as a percent difference.
Percent difference audit results are calculated as follows:
Percent Difference = ———— x 100
y
true
Where:
Xhb is the analytical result from the laboratory;
Xtruc is the true concentration of the audit sample.
Percent differences of ± 25 percent are acceptable. Note that the "true" value used in the
calculations above can be based on the mean value of the confirmation laboratory's results or the
mean result of all participating NATTS laboratories and is also indicated in the tables that
follow.
24-61
-------
The results of the 2015 and 2016 PT audit samples show that few of the pollutants for
which PT audit samples were analyzed exceed the MQO for accuracy. Of the 143 results
provided in Tables 24-17 through 24-20, only three exceed the MQO for accuracy (two VOCs
and one PAH). However, none failed multiple audits in 2015 and 2016.
Table 24-17. TO-15 NATTS PT Audit Samples1
Pollutant
May
2015
August
2016
February
2016 "
June
2016
Acrolein
12.5
NS
7.1
-9.0
Benzene
0.1
0.8
11.5
11.2
1,3-Butadiene
5.9
-25.4
4.8
-12.9
Carbon Tetrachloride
28.6
8.6
NS
21.3
Chloroform
24.8
3.7
17.9
13.1
1,2-Dibromoethane
-1.8
-11
1.2
-3.3
1,2-Dichloroethane
15.7
-1.4
7.3
8.5
Dichloromethane
NS
7.9
15.4
13.2
1,2-Dichloropropane
3.1
-6.7
1.0
8.6
cis-1,3 -Dichloropropene
8.4
-12
12.1
6.8
trans-1,3 -Dichloropropene
1.0
-22.6
0.2
-5.3
1,1,2,2-Tetrachloroethane
-4.3
-12.4
-0.3
14.5
Tetrachloroethylene
0.8
-2.4
0.4
0.5
Trichloroethylene
6.8
-0.3
6.0
-3.7
Vinyl chloride
7.4
-4.5
11.4
1.0
1 The true value is based on the mean of participating NATTS laboratories.
Bold = Greater than ± 25 percent MQO
NS = Not spiked onto PT audit sample provided to the laboratory
Table 24-18. TO-11A NATTS PT Audit Samples1
March
February
June
Pollutant
2015
2016
2016
Acetaldehyde
-13.2
-0.9
-2.1
Benzaldehyde
3.7
1.3
-0.4
Formaldehyde
-12.6
-6.7
-7.9
Propionaldehyde
-10.5
-8.3
-8.9
NS = Not spiked onto PT audit sample provided to the laboratory
1 The true value is based on the mean of participating NATTS laboratories.
Bold = Greater than ± 25 percent MQO
24-62
-------
Table 24-19. TO-13A NATTS PT Audit Samples1
Pollutant
June
2015
November
2015
November
2016
Acenaphthene
8.8
NS
7.1
Anthracene
NS
1.1
9.3
Benzo(a)pyrene
10.1
1.9
14.5
Fluoranthene
10.2
6.2
12.7
Fluorene
7.3
11.4
18
Naphthalene
35.5
18.5
24.8
Phenanthrene
-4.7
0.8
10.6
Pyre nc
-2.6
-1.9
15.1
NS = Not spiked onto PT audit sample provided to the laboratory.
Bold = Greater than ± 25 percent MQO
1 Audit result based on percent difference from mean of participating NATTS laboratories.
Table 24-20. Metals NATTS PT Audit Samples1
Pollutant
June
20152
November
20152
June 2016
November 2016
Teflon'
Quartz
Teflon'
Quartz
Antimony
-11.3
-11.1
-4.6
-15.7
8.3
-24.2
Arsenic
6.3
2.3
-3.8
0.3
0.2
0.8
Beryllium
11.7
-1.0
-6.3
4.1
-3.8
0.2
Cadmium
10.4
6.5
-4.8
6.9
0
3.8
Cobalt
6.3
7.4
-4.4
7.2
-2.8
3.0
Lead
7.5
1.0
-5.0
-1.0
-0.6
1.2
Manganese
6.5
-0.4
10.1
6.9
0.3
4.6
Nickel
21.4
NS
-18.3
5.2
-0.5
1.5
Selenium
NS
13.3
0.9
3.6
-2.0
-2.4
NS = Not spiked onto PT audit sample provided to the laboratory.
Bold = Greater than ± 25 percent MQO
1 Audit result based on percent difference from mean of participating NATTS laboratories.
2 Teflon® audit samples not available prior to 2016.
In 2012, ERG was approved for the sampling and analysis of lead for adherence to the
National Ambient Air Quality Standards (NAAQS) using ICP-MS analysis (EPA, 2012). This
approval requires additional quality assurance steps, including the analysis of quarterly audit
strips. Tables 24-21 and 24-22 provide the results of the quarterly NAAQS audit results for lead
for ERG for 2015 and 2016, respectively. Audit results are presented for Teflon® filters, the only
filter type for which ERG receives under the NMP. More than 80 percent of the audit results are
within the percent recovery target of ± 10 percent.
24-63
-------
Table 24-21. 2015 Lead NAAQS Quarterly Audit Samples1
Analysis
#
Q1 2015
Q2 2015
Q3 2015
Q4 2015
Pollutant
Level
Teflon'
1
-9.2
2.5
-8.1
-15.4
Lead
Low
2
-4.3
-2.9
-4.9
-17.1
3
-7.9
-4.3
-4.9
-14.5
1
-2.2
0.3
-6.3
-11.2
Lead
High
2
-5.7
5.6
-4.7
-6.0
3
-2.2
-2.5
-9.3
-3.4
1 Audit result based on percent difference from mean of participating NATTS laboratories.
Bold = Greater than ±10 percent difference target
Table 24-22. 2016 Lead NAAQS Quarterly Audit Samples1
Analysis
#
Q1 2016
Q2 2016
Q3 2016
Q4 2016
Pollutant
Level
Teflon'
1
-1.8
-6.6
3.9
-3.8
Lead
Low
2
-1.7
-17.8
5.8
-9.3
3
-2.0
-17.1
5.8
-1.6
1
-5.8
-5.1
-5.6
-9.1
Lead
High
2
-6.1
-11.6
-6.8
-13.6
3
-6.1
-12.4
-11.4
-12.0
1 Audit result based on percent difference from mean of participating NATTS laboratories.
Bold = Greater than ±10 percent difference target
The accuracy of the 2015 and 2016 NMP monitoring data can also be assessed
qualitatively by reviewing the accuracy of the monitoring methods and how they were
implemented:
• The sampling and analytical methods used during the 2015 and 2016 monitoring
effort have been approved by EPA for accurately measuring ambient levels of various
pollutants - an approval that is based on many years of research into the development
of ambient air monitoring methodologies.
• When collecting and analyzing ambient air samples, field sampling staff and
laboratory analysts are required to strictly adhere to quality control and quality
assurance guidelines detailed in the respective monitoring methods. This strict
adherence to the well-documented sampling and analytical methods suggests that the
2015 and 2016 monitoring data accurately represent ambient air quality.
24-64
-------
25.0 Results, Conclusions, and Recommendations
This section summarizes the results of the data analyses contained in this report, renders
conclusions based on those results, and presents recommendations applicable to future air toxics
monitoring efforts. As demonstrated by the results of the data analyses discussed throughout this
report, NMP data offer a wealth of information for assessing air quality by evaluating trends,
patterns, correlations, and the potential for health risk. NMP data should ultimately assist a wide
range of audiences in understanding the complex nature of ambient air pollution.
25.1 Summary of Results
Analyses of the 2015 and 2016 monitoring data identified the following notable results,
observations, trends, and patterns in the program-level and state- and site-specific air monitoring
data.
25.1.1 Program-level Results Summary
• Number of participating sites. Twenty of the 53 monitoring sites are EPA-designated
NATTS sites. An additional 33 UATMP sites participated in the NMP in 2015 and
2016.
• Total number of samples collected and analyzed. Over 15,300 valid samples were
collected at participating program sites and analyzed at the ERG laboratory, yielding
more than 445,000 valid measurements of air toxics, including primary, duplicate,
collocated, and replicate results.
• Detects. Of the 199 pollutants for which statistical summaries are provided in
Tables 4-1 through 4-6, all but four were detected at least once over the course of the
two-year monitoring effort. The detection of a given pollutant is subject to the
sensitivity limitations associated with the analytical methods used and the limitations
of the instruments. Simply stated, an MDL is the lowest concentration of a target
pollutant that can be measured and reported with 99 percent confidence that the
pollutant concentration is greater than zero. Approximately 54 percent of the reported
measurements were greater than the associated MDLs. At the method level, this
percentage varies considerably, from 40 percent for VOCs to 100 percent for
methane. Quantification less than the MDL is possible and an acceptable analytical
result; therefore, these results are incorporated into the data analyses. These
measurements account for 12 percent of concentrations. Non-detects account for the
remaining 34 percent of results.
• Program-level Pollutants of Interest. The pollutants of interest at the program-level
are based on the total number of concentrations greater than the associated risk
screening value, or those "failing the screen". Concentrations of 39 pollutants failed
at least one scree; of those pollutants, 13 were identified as program-level pollutants
of interest, seven VOCs (benzene, 1,3-butadiene, carbon tetrachloride,
/;-dichlorobenzene, 1,2-dichloroethane, ethylbenzene, and hexachloro-l,3-butadiene),
25-1
-------
two carbonyl compounds (acetaldehyde and formaldehyde), three PAHs
(acenaphthene, fluorene, and naphthalene), and one metal (arsenic).
• Seasonal Trends. Fewer pollutants exhibited identifiable seasonal trends in the
concentrations measured during the 2015 and 2016 program years (at least from a
program-level perspective). Formaldehyde concentrations tended to be highest during
the warmer months of the year, similar to past years. Acetaldehyde concentrations
exhibit a similar pattern, but to a lesser degree. Fluorene concentrations also exhibited
this seasonal trend. Conversely, benzene and 1,3-butadiene concentrations tended to
be higher during the colder months of the year; this is also true for naphthalene,
particularly during the fourth quarter.
25.1.2 State-level Results Summary
Arizona.
• The Arizona monitoring sites are located in Phoenix. PXSS is a NATTS site; SPAZ is
a UATMP site.
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at PXSS.
VOCs were sampled for at SPAZ.
• Twenty pollutants failed screens for PXSS, 10 of which contributed to 95 percent of
failed screens. PXSS failed the second highest number of screens among all NMP
sites. Eight pollutants failed screens for SPAZ, six of which contributed to 95 percent
of failed screens.
• Formaldehyde has the highest annual average concentration for 2016 among the
pollutants of interest for PXSS, followed by acetaldehyde and benzene. Annual
average concentrations for the carbonyl compounds could not be calculated for 2015
due to contamination issues with the collection system. Among the remaining
pollutants of interest, benzene has the highest annual average concentration in 2015.
• Benzene has the highest annual average concentration for SPAZ for both years and is
the only pollutant of interest with an annual average concentration greater than 1
|ig/m3.
• SPAZ and PXSS have the highest annual average concentrations of ethylbenzene
among NMP sites sampling this pollutant for both years. SPAZ also has the highest
annual average concentrations of benzene and />dichlorobenzene in 2015 and 2016.
• Sampling for the site-specific pollutants of interest has occurred at PXSS and SPAZ
for at least 5 consecutive years; thus, a trends analysis was conducted for each of the
site-specific pollutants of interest. The detection rate and measured concentrations of
1,2-dichloroethane at PXSS has been steadily increasing in the recent years. The
maximum concentration of 1,3-butadiene measured since the onset of sampling was
measured at SPAZ in 2016.
25-2
-------
Formaldehyde has the highest cancer risk approximation among the pollutants of
interest for PXSS; benzene has the highest cancer risk approximation among the
pollutants of interest for SPAZ. None of the pollutants of interest for either site have
noncancer hazard approximations greater than an HQ of 1.0.
Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
Maricopa County, while toluene is the highest emitted pollutant with a noncancer
toxicity factor. Formaldehyde has the highest cancer toxicity-weighted emissions,
while acrolein has the highest noncancer toxicity-weighted emissions for Maricopa
County.
California.
The three California monitoring sites are located in Los Angeles (CELA), Rubidoux
(RUCA), and San Jose (SJJCA). All three are NATTS sites.
PAHs were sampled for at each of the three sites. In addition, PMio metals were also
sampled for at SJJCA.
Naphthalene failed the majority screens for CELA and RUCA, and thus, was
identified as the sole pollutant of interest for these two sites. In addition to
naphthalene, arsenic and nickel were also identified as pollutants of interest for
SJJCA.
Among the three California sites, CELA has the highest annual average
concentrations of naphthalene for both 2015 and 2016.
Sampling for the site-specific pollutants of interest has occurred at CELA, RUCA,
and SJJCA for at least 5 consecutive years; thus, a trends analysis was conducted for
each site for the site-specific pollutants of interest. Naphthalene concentrations
exhibit a decreasing trend at CELA in recent years. This is also true for RUCA
through 2015, but concentrations exhibit an increase for 2016.
None of the pollutants of interest for the California sites have cancer risk
approximations greater than 3 in-a-million; none of the pollutants of interest for the
California sites have noncancer hazard approximations greater than an HQ of 1.0.
Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Los
Angeles, Riverside, and Santa Clara Counties. Formaldehyde has the highest cancer
toxicity-weighted emissions for Los Angeles County, while hexavalent chromium has
the highest cancer toxicity-weighted emissions for Riverside and Santa Clara
Counties.
Toluene is the highest emitted pollutant with a noncancer toxicity factor in Los
Angeles, Riverside, and Santa Clara Counties, while acrolein has the highest
noncancer toxicity-weighted emissions for all three counties.
25-3
-------
Colorado.
• The NATTS site in Colorado is located in Grand Junction (GPCO). There are also six
UATMP sites located northeast of Grand Junction in Garfield County. The sites are
located in the towns of Battlement Mesa (BMCO), Silt (BRCO), Glenwood Springs
(GSCO), Parachute (PACO), Carbondale (RFCO), and Rifle (RICO).
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at GPCO.
Carbonyl compounds and SNMOCs were sampled for at each of the Garfield County
sites except RFCO. Between January and September 2015, canister samples collected
at RFCO were analyzed for both VOC and SNMOCs, after which only SNMOCs
were analyzed for the rest of 2015, as well as throughout 2016.
• Twenty pollutants failed at least one screen for GPCO, 12 of which contributed to
95 percent of failed screens. Three pollutants failed screens for BMCO, four
pollutants failed screens for BRCO, GSCO, and RFCO, and five pollutants failed
screens for PACO and RICO. Benzene is a pollutant of interest for all seven Colorado
sites.
• Of the pollutants of interest for GPCO, formaldehyde had the highest annual
concentration for 2015, although annual average concentrations could not be
calculated for VOCs in 2015. Dichloromethane has the highest annual average
concentration for GPCO for 2016, followed by formaldehyde.
• Where they could be calculated, benzene and formaldehyde had the highest annual
average concentrations among the pollutants of interest for the Garfield County sites.
RICO and PACO were the only Garfield County sites with annual average
concentrations of these pollutants greater than 1 |ig/m3.
• PACO has the second (2015) and third (2016) highest annual average concentrations
of benzene among all NMP sites sampling this pollutant.
• Sampling for the site-specific pollutants of interest has occurred at GPCO, BRCO,
PACO, RICO, and RFCO for at least 5 consecutive years; thus, a trends analysis was
conducted for the site-specific pollutants of interest. Notable trends include: The
1-year average concentration for several pollutants for GPCO are at a minimum for
2016, including acenaphthene, benzene, 1,3-butadiene, ethylbenzene, and
naphthalene. The significant increase in the detection rate of 1,2-dichloroethane
beginning at GPCO in 2012 continues through 2016. The highest fluoranthene
concentrations measured since the onset of sampling at GPCO were measured in 2015
and 2016. Concentrations of acetaldehyde and formaldehyde decreased considerably
at BRCO in 2016. Concentrations of acetaldehyde appear to have a decreasing trend
at PACO. The detection rate of 1,3-butadiene has decreased considerably at PACO
and RFCO, particularly in 2016.
• Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for GPCO. Benzene and formaldehyde have the highest cancer risk
approximations for the Garfield County sites, depending on the year and whether
annual average concentrations could be calculated. None of the pollutants of interest
25-4
-------
for the Colorado monitoring sites have noncancer hazard approximations greater than
an HQ of 1.0 (where they could be calculated).
• Benzene is the highest emitted pollutant with a cancer toxicity factor in both Mesa
and Garfield Counties, while formaldehyde has the highest cancer toxicity-weighted
emissions for both counties.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor for Mesa
County, while xylenes is the highest emitted pollutant with a noncancer toxicity
factor for Garfield County. Acrolein has the highest noncancer toxicity-emissions for
both counties.
District of Columbia.
• The Washington, D.C. monitoring site (WADC) is a NATTS site.
• PAHs were sampled for at WADC.
• Naphthalene and benzo(a)pyrene both failed screens for WADC, although
naphthalene accounted for 98 percent of the total failed screens and therefore, was the
only pollutant identified as a pollutant of interest.
• Naphthalene was detected in every valid PAH sample collected at WADC. The
annual average concentration of naphthalene for 2015 is similar to the annual average
concentration for 2016.
• Sampling for the site-specific pollutants of interest has occurred at WADC for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of naphthalene have a decreasing trend at
WADC.
• The cancer risk approximations for naphthalene are 2.06 in-a-million and 2.22 in-a-
million for 2015 and 2016, respectively. The noncancer hazard approximations for
naphthalene are both considerably less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in the District of
Columbia, while toluene is the highest emitted pollutant with a noncancer toxicity
factor. Formaldehyde has the highest cancer toxicity-weighted emissions, while
acrolein has the highest noncancer toxicity-weighted emissions in the District.
Florida,
• Three of the Florida monitoring sites are located in the Tampa-St. Petersburg-
Clearwater CBS A (SYFL, AZFL, and SKFL) and two are located in the Orlando-
Kissimmee-Sanford CBSA (ORFL and PAFL). SKFL and SYFL are NATTS sites
while the other three are UATMP sites.
25-5
-------
• Carbonyl compounds were sampled for at AZFL, ORFL, and SYFL. PAHs were
sampled for at SKFL in addition to carbonyl compounds. Metals (PMio) were
sampled for at PAFL. Sampling at ORFL and PAFL was discontinued at the end of
September 2016.
• Acetaldehyde and formaldehyde failed screens for all four Florida sites sampling
carbonyl compounds. Naphthalene, fluorene, and acenaphthene also failed screens for
SKFL. Arsenic and nickel are the speciated metals that failed screens for PAFL.
• Formaldehyde has the highest annual average concentration for all four sites sampling
carbonyl compounds; annual average concentrations of acetaldehyde were nearly half
the magnitude of the annual average of formaldehyde for each site.
• The 25 highest formaldehyde concentrations measured across the program in 2016
were measured at AZFL and SKFL. These sites have the second and fifth highest
annual averages concentration of formaldehyde, respectively.
• The annual average concentrations of nickel were more than twice the annual average
concentrations of arsenic for PAFL.
• Sampling for the site-specific pollutants of interest has occurred at the Florida sites
for at least 5 consecutive years; thus, a trends analysis was conducted for the site-
specific pollutants of interest. The following notable observations regarding trends
include: Concentrations of naphthalene have a decreasing trend at SKFL through
2015 but increased somewhat for 2016. Concentrations of arsenic at PAFL have a
decreasing trend.
• Formaldehyde has the highest cancer risk approximation for all four sites sampling
carbonyl compounds, ranging from roughly 23 in-a-million to 95 in-a-million.
Arsenic has the highest cancer risk approximation for PAFL. All noncancer hazard
approximations for the pollutants of interest for the Florida sites are less than an HQ
of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Pinellas,
Hillsborough, and Orange Counties. Formaldehyde has the highest cancer toxicity-
weighted emissions for Pinellas and Hillsborough Counties, while hexavalent
chromium has the highest cancer toxicity-weighted emissions for Orange County.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in Pinellas
and Orange Counties, while hydrochloric acid this the highest emitted pollutant with
a noncancer toxicity factor in Hillsborough County. Acrolein has the highest
noncancer toxicity-weighted emissions for all three counties.
Illinois.
• Two Illinois monitoring sites are located near Chicago. NBIL is a NATTS site located
in Northbrook and SPIL is a UATMP site located in Schiller Park. A third site, ROIL,
is located in Roxana, on the Illinois border near St. Louis. Sampling at ROIL was
discontinued at the end of July 2015, ending a three-year monitoring effort at this
location.
25-6
-------
• VOCs and carbonyl compounds were sampled for at all three Illinois sites. SNMOCs,
PAHs, and metals (PMio) were also sampled for at NBIL. NBIL is one of only two
NMP sites sampling both VOCs and SNMOCs across both years of sampling.
• Twenty-three pollutants failed screens for NBIL; 14 pollutants failed screens for
SPIL; and 10 pollutants failed screens for ROIL. Among the site-specific pollutants
of interest, the three Illinois sites have six pollutants in common: two carbonyl
compounds (acetaldehyde and formaldehyde) and four VOCs (benzene,
1,3-butadiene, carbon tetrachloride, and 1,2-dichloroethane).
• Formaldehyde and acetaldehyde have the highest and second-highest annual average
concentrations, respectively, for NBIL and SPIL. Annual average concentrations for
ROIL could not be calculated.
• NBIL has the highest annual average concentrations of acenaphthene and fluorene
among NMP sites sampling PAHs.
• Sampling for the site-specific pollutants of interest has occurred at NBIL and SPIL
for at least 5 consecutive years; thus, a trends analysis was conducted for the site-
specific pollutants of interest. After several years of increasing, concentrations of
acetaldehyde decreased significantly in 2015 and 2016 at NBIL while the opposite is
true of formaldehyde concentrations measured at this site. Concentrations of
fluoranthene have an increasing trend at NBIL while concentrations of naphthalene
have a decreasing trend. Concentrations of benzene measured at NBIL are at a
minimum in 2016; this is also true at SPIL.
• Formaldehyde has the highest cancer risk approximations for NBIL and SPIL, each of
which is a magnitude higher than other cancer risk approximations for both sites. All
noncancer hazard approximations for the pollutants of interest for the Illinois sites are
less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Cook County,
while formaldehyde has the highest cancer toxicity-weighted emissions.
Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
Madison County, while coke oven emissions (PM) have the highest cancer toxicity
emissions.
• Ethylene glycol is the highest emitted pollutant with a noncancer toxicity factor in
Cook County, while toluene is the highest emitted pollutant with a noncancer toxicity
factor Madison County. Acrolein has the highest noncancer toxicity-weighted
emissions for both counties.
Indiana.
• There are two Indiana monitoring sites sampling under the NMP, one located in
Indianapolis (WPIN) and a second located in Gary, near Chicago (INDEM). Both are
UATMP sites.
• Carbonyl compounds were sampled for at WPIN and INDEM.
25-7
-------
• Formaldehyde and acetaldehyde failed screens for both INDEM and WPIN. All
measured detections of formaldehyde failed screens for both sites. All measured
detections of acetaldehyde failed screens for INDEM while all but one acetaldehyde
concentration measured at WPIN failed screens.
• Formaldehyde has the highest annual average concentrations for both sites.
• Sampling for the site-specific pollutants of interest has occurred at WPIN and
INDEM for at least 5 consecutive years; thus, a trends analysis was conducted for the
site-specific pollutants of interest. Concentrations of formaldehyde have a slight
increasing trend at INDEM in recent years, with concentrations of acetaldehyde
exhibiting a similar trend through 2015. Acetaldehyde concentrations have been
decreasing at WPIN since 2010, although the rate of decrease slowed considerably in
recent years. After a few years of decreasing, formaldehyde concentrations have an
increasing trend at WPIN in 2015 and 2016.
• The cancer risk approximations for formaldehyde are an order of magnitude greater
than the cancer risk approximations for acetaldehyde for both sites. The noncancer
hazard approximations for the pollutants of interest for the Indiana sites are
considerably less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in both Marion
and Lake Counties while formaldehyde has the highest cancer toxicity-weighted
emissions for both counties.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in both Lake
and Marion Counties while acrolein has the highest noncancer toxicity-weighted
emissions for both counties.
Kentucky.
• Three Kentucky monitoring sites are located in northeast Kentucky, two in Ashland
(ASKY and ASKY-M) and one near Grayson Lake (GLKY). The Grayson Lake
monitoring site is aNATTS site. One monitoring site is located south of Evansville,
Indiana (BAKY). Three monitoring sites are located in or near the Calvert City area
(ATKY, BLKY, and TVKY). The final monitoring site is located in Lexington, in
north-central Kentucky (LEKY).
• Six of the eight Kentucky monitoring sites sampled for VOCs, with ASKY-M and
BAKY as the exceptions. Five of the eight sites sampled for PMio metals, with
ASKY, ATKY, and TVKY as the exceptions. GLKY also sampled for PAHs and
carbonyl compounds.
• The number of pollutants failing screens for the Kentucky sites varies from five
(ASKY-M and BAKY) to 13 (GLKY, ATKY, BLKY, and LEKY).
• Of the pollutants of interest, formaldehyde has the highest annual average
concentrations for GLKY. Manganese has the highest annual average concentrations
for ASKY-M, while arsenic has the highest annual average concentrations for BAKY.
Benzene has the highest annual average concentrations for ASKY, while
25-8
-------
1,2-dichloroethane has the highest annual average concentrations for BLKY and
TVKY. Benzene has the highest annual average concentration for ATKY in 2015,
while 1,2-dichloroethane has the highest annual average concentration for the site in
2016. Carbon tetrachloride has the highest annual average concentration for LEKY in
2015; VOC sampling at LEKY was discontinued at the end of July 2016.
ASKY-M has the highest annual average concentrations of arsenic for both 2015 and
2016 among NMP sites sampling PMio metals. BAKY has the third (2015) and fifth
(2016) highest annual concentrations of arsenic.
The Calvert City sites account for the six highest annual average concentrations of
1,2-dichloroethane, four of the five highest annual average concentrations of carbon
tetrachloride, and the highest annual average of 1,3-butadiene, with the highest annual
average for each of these pollutants calculated for TVKY.
Sampling for the site-specific pollutants of interest has occurred at ASKY, ASKY-M,
GLKY, ATKY, BLKY, TVKY, and LEKY for at least 5 consecutive years; thus, a
trends analysis was conducted for each of the site-specific pollutants of interest. Most
notably, concentrations of benzene, 1,3-butadiene, and 1,2-dichloroethane have
decreased at GLKY, while concentrations of formaldehyde have increased.
Concentrations of 1,3-butadiene at BLKY also exhibit a decreasing trend.
Formaldehyde has the highest cancer risk approximations among the pollutants of
interest for GLKY, while arsenic has the highest cancer risk approximations for
ASKY-M and BAKY. Benzene has the highest cancer risk approximations for ASKY
and LEKY (2015 only). For ATKY, BLKY, and TVKY, 1,2-dichloroethane has the
highest cancer risk approximations. The cancer risk approximations for TVKY for
1,2-dichloroethane are among the highest cancer risk approximations calculated for
the site-specific pollutants of interest across the program for both years. None of the
pollutants of interest for which noncancer hazard approximations could be calculated
were greater than an HQ of 1.0.
Benzene is the highest emitted pollutant with a cancer toxicity factor in all six
Kentucky counties with NMP sites. Nickel has the highest cancer toxicity-weighted
emissions for Boyd County; formaldehyde has the highest cancer toxicity-weighted
emissions for Carter, Livingston, and Fayette Counties; naphthalene has the highest
cancer-toxicity weighted emissions for Henderson County; and benzene has the
highest cancer toxicity-weighted emissions for Marshall County.
Toluene is the highest emitted pollutant with a noncancer toxicity factor in Boyd,
Carter, Livingston, and Fayette Counties; carbonyl sulfide is the highest emitted
pollutant with a noncancer toxicity factor in Henderson County; and methanol is the
highest emitted pollutant with a noncancer toxicity factor in Marshall County.
Acrolein has the highest noncancer toxicity-weighted emissions in five of the
Kentucky counties, but ranks second to chlorine in Marshall County.
25-9
-------
Massachusetts.
• The Massachusetts monitoring site (BOMA) is a NATTS site located in Boston.
• Metals (PMio) and PAHs were sampled for at BOMA.
• Nine pollutants failed screens for BOMA, four of which were identified as pollutants
of interest. Arsenic and naphthalene together accounted for nearly 90 percent of the
site's failed screens.
• Of the pollutants of interest, naphthalene has the highest annual average concentration
each year.
• The maximum concentration of nickel measured across the program was measured at
BOMA.
• Sampling for the site-specific pollutants of interest has occurred at BOMA for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Naphthalene concentrations have a decreasing trend at BOMA.
The highest concentrations of nickel and cadmium measured since the onset of
sampling were measured at BOMA in 2016.
• Arsenic has the highest cancer risk approximations for BOMA in 2015 and 2016.
None of the pollutants of interest for BOMA have noncancer hazard approximations
greater than an HQ of 1.0.
• Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in Suffolk
County and has the highest cancer toxicity-weighted emissions. Toluene is the highest
emitted pollutant with a noncancer toxicity factor in Suffolk County, while acrolein
has the highest noncancer toxicity-weighted emissions.
Michigan.
• The Michigan monitoring site (DEMI) is a NATTS site located in Dearborn,
southwest of Detroit.
• VOCs, carbonyl compounds, and PAHs were sampled for at DEMI.
• Sixteen pollutants failed screens for DEMI, 10 of which were identified as pollutants
of interest.
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
DEMI in 2015 and 2016.
• DEMI has the highest (2015) and third highest (2016) annual average concentrations
of naphthalene among NMP sites sampling PAHs. DEMI is one of only two sites with
an annual average concentration of naphthalene greater than 100 ng/m3.
• Sampling for the site-specific pollutants of interest has occurred at DEMI for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
25-10
-------
pollutants of interest. Benzene concentrations exhibit a steady decreasing trend
although concentrations have leveled out in recent years. Concentrations of
acetaldehyde have a slow, steady increasing trend over the last several years of
sampling.
• Formaldehyde has the highest cancer risk approximations for DEMI. None of the
pollutants of interest for DEMI have noncancer hazard approximations greater than
an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Wayne
County, while coke oven emissions (PM) have the highest cancer toxicity-weighted
emissions. Hydrochloric acid is the highest emitted pollutant with a noncancer
toxicity factor in Wayne County, while acrolein has the highest noncancer toxicity-
weighted emissions.
Missouri.
• The NATTS site in Missouri (S4MO) is located in St. Louis.
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at S4MO.
• Twenty-two pollutants failed at least one screen for S4MO, 12 of which contributed
to 95 percent of failed screens. S4MO has the highest number of pollutants failing
screens.
• Of the pollutants of interest for S4MO, formaldehyde and acetaldehyde have the
highest annual average concentrations and are the only pollutants with annual average
concentrations greater than 1 |ig/m3,
• S4MO has the third highest annual average concentration of />dichlorobenzene
(2015) among NMP sites sampling these pollutants.
• Sampling for the site-specific pollutants of interest has occurred at S4MO for at least
5 consecutive years; thus, a trends analysis was conducted for each of the site-specific
pollutants of interest. Concentrations of acetaldehyde, benzene, and ethylbenzene
have decreased significantly over the course of sampling.
• Formaldehyde has the highest cancer risk approximations for S4MO. None of the
pollutants of interest for S4MO have a noncancer hazard approximation greater
than an HQ of 1.0.
• Formaldehyde is the highest emitted pollutant with a cancer toxicity factor in
St. Louis (city and county) and has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions in St. Louis (city and
county).
25-11
-------
New Jersey.
• Three UATMP sites in New Jersey are located in the New York-Newark-Jersey City
CBS A and are located in the towns of Chester (CHNJ), Elizabeth (ELNJ), North
Brunswick (NBNJ). At the end of 2015, the NBNJ site relocated to East Brunswick
(NRNJ). Another UATMP site (CSNJ) is located in the Philadelphia-Camden-
Wilmington CBS A.
• VOCs and carbonyl compounds were sampled for at the New Jersey sites.
• The number of pollutants failing at least one screen for the New Jersey sites ranged
from nine (CHNJ) to 15 (CSNJ). The New Jersey sites have six pollutants of interest
in common: acetaldehyde, formaldehyde, benzene, carbon tetrachloride,
1,3-butadiene, and 1,2-dichloroethane.
• Of the site-specific pollutants of interest, formaldehyde and acetaldehyde have the
highest annual average concentrations for all but one of the New Jersey sites; the
exception is CHNJ for 2015, when annual average concentrations could not be
calculated for the carbonyl compounds.
• Sampling for the site-specific pollutants of interest has occurred at CHNJ, ELNJ, and
NBNJ for at least 5 consecutive years; specifically, ELNJ is the longest running NMP
site still participating in the NMP. As such, a trends analysis was conducted for the
site-specific pollutants of interest for these sites. Concentrations of benzene have
decreased significantly at ELNJ since the onset of sampling. This is also true of
ethylbenzene, although concentrations have leveled out in the last few years.
Concentrations of benzene also have decreasing trends at CHNJ and, to a lesser
extent, NBNJ.
• With the exception of CHNJ in 2015, formaldehyde has the highest cancer risk
approximations for the New Jersey sites. Carbon tetrachloride has the highest cancer
risk approximation for CHNJ in 2015, when annual average concentrations could not
be calculated for carbonyl compounds. None of the pollutants of interest for the New
Jersey sites have noncancer hazard approximations greater than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor for Camden and
Morris Counties, while formaldehyde is the highest emitted pollutant with a cancer
toxicity factor in Union and Middlesex Counties. Formaldehyde has the highest
toxicity-weighted emissions for each New Jersey county.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in Camden,
Union, Middlesex, and Morris Counties. Acrolein has the highest noncancer toxicity-
weighted emissions for each New Jersey county.
New York.
• The New York monitoring sites are located in Bronx (BXNY) and Rochester
(ROCH). Both are NATTS sites.
• PAHs were sampled for at both BXNY and ROCH.
25-12
-------
• Six pollutants failed screens for BXNY, four of which were identified as pollutants of
interest. Naphthalene accounted for nearly 75 percent of failed screens for BXNY.
Eight pollutants failed screens for ROCH, four of which were identified as pollutants
of interest. Naphthalene accounted for nearly 40 percent of failed screens for ROCH.
• Naphthalene has the highest annual average concentrations for BXNY and ROCH,
with the annual average concentrations for BXNY nearly twice the annual average
calculated for ROCH.
• BXNY has the second (2015) and fourth (2016) highest annual average
concentrations of naphthalene among NMP sites sampling PAHs and is one of only
two NMP sites with an annual average concentration greater than 100 ng/m3. ROCH
has the third (2105) and fourth (2016) highest annual average concentrations of
acenaphthene and fluorene among NMP sites sampling PAHs.
• Sampling for the site-specific pollutants of interest has occurred at BXNY and ROCH
for greater than 5 consecutive years; thus, a trends analysis was conducted for each of
the site-specific pollutants of interest. For both sites, the one-year average
concentration of naphthalene is at a minimum for 2016 (based on years when annual
averages could be calculated.)
• Naphthalene has the highest cancer risk approximations among the pollutants of
interest for both ROCH and BXNY. Naphthalene is the only pollutant of interest for
either site with a noncancer toxicity factor; the noncancer hazard approximations for
naphthalene for both sites are considerably less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor for Bronx
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
p-Dichlorobenzene is the highest emitted pollutant with a cancer toxicity factor for
Monroe County and has the highest cancer toxicity-weighted emissions.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor for Bronx
County, while chlorobenzene is the highest emitted pollutant with a noncancer
toxicity factor for Monroe County. Acrolein has the highest noncancer toxicity-
weighted emissions for both counties.
Oklahoma.
• There are seven UATMP sites in Oklahoma: three are located in Tulsa (TOOK,
TMOK, and TROK), three are located in or near Oklahoma City (OCOK, NROK, and
YUOK), and one is located south of Oklahoma City in Bradley (BROK).
• VOCs, carbonyl compounds, and metals (TSP) were sampled for at all three Tulsa
sites, OCOK, and YUOK. VOCs, SNMOCs, and carbonyl compounds were sampled
for at BROK and NROK. In addition, canister samples collected at BROK and
NROK were also analyzed for methane. The Oklahoma sites are the only NMP sites
sampling TSP metals and methane.
25-13
-------
• Seventeen individual pollutants failed screens for each of the three Tulsa sites; 16
pollutants failed screens for OCOK; 15 failed screens for YUOK; 11 failed screens
forBROK; and 10 failed screens for NROK.
• Formaldehyde and acetaldehyde have the highest annual average concentrations for
each of the Oklahoma sites, where annual averages could be calculated.
• BROK's annual average concentration of acetaldehyde for 2015 is the highest annual
average of this pollutant across the program; this site's annual average concentration
of formaldehyde for 2015 ranks fourth highest. BROK's annual averages for 2016 are
significantly less.
• The three Tulsa sites have some of the highest annual average concentrations of
hexachloro-1,3-butadiene (2016 only) among NMP sites sampling VOCs.
• Sampling for the site-specific pollutants of interest has occurred at TOOK, TMOK,
and OCOK for at least 5 consecutive years; thus, a trends analysis was conducted for
the site-specific pollutants of interest. After several years of increasing,
concentrations of several pollutants, including acetaldehyde, benzene, ethylbenzene,
and manganese, decreased at TOOK after 2012 then have remained fairly static in
recent years. Other pollutants exhibit this trend as well but the difference is less
significant. Benzene and acetaldehyde concentrations have also been decreasing at
TMOK but have leveled out in recent years. In addition, the detection rates of
1,2-dichloroethane have been increasing at TOOK, TMOK, and OCOK over the last
five years of sampling.
• Formaldehyde has the highest cancer risk approximations for each of the Oklahoma
monitoring sites, where they could be calculated. None of the pollutants of interest for
the Oklahoma sites have a noncancer hazard approximation greater than an HQ of
1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Oklahoma
and Tulsa Counties while formaldehyde is the highest emitted pollutant with a cancer
toxicity factor in Canadian and Grady Counties. Formaldehyde has the highest cancer
toxicity-weighted emissions for Oklahoma, Canadian, and Grady Counties, while
hexavalent chromium has the highest cancer toxicity-weighted emissions for Tulsa
County.
• Toluene is the highest emitted pollutant with a noncancer toxicity factor in Oklahoma
and Tulsa Counties, while formaldehyde is the highest emitted pollutant with a
noncancer toxicity factor in Canadian and Grady Counties. Acrolein has the highest
noncancer toxicity-weighted emissions for all four counties.
Rhode Island.
• The Rhode Island monitoring site (PRRI) is located in Providence and is a NATTS
site.
• PAHs were sampled for at PRRI.
25-14
-------
• Four pollutants failed screens for PRRI, two of which were identified as pollutants of
interest (naphthalene and fluorene). Ninety-four percent of failed screens for PRRI
are attributable to naphthalene.
• Naphthalene's annual average concentrations for both years are more than ten times
the annual average concentrations of fluorene.
• Sampling for the site-specific pollutants of interest has occurred at PRRI for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of naphthalene have a decreasing trend at PRRI
that has leveled out in recent years. Concentrations of fluorene also decreased for
several years before increasing in 2016.
• Naphthalene had the highest cancer risk approximations for PRRI. Naphthalene is the
only pollutant of interest for PRRI with a noncancer toxicity factor; the noncancer
hazard approximations for naphthalene are considerably less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Providence
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions for Providence
County.
Utah.
• The NATTS site in Utah (BTUT) is located in Bountiful, north of Salt Lake City.
• VOCs, carbonyl compounds, SNMOCs, PAHs, and metals (PMio) were sampled for
at BTUT. This site is one of only two NMP sites sampling both VOCs and SNMOCs
across both years of sampling.
• Twenty-two pollutants failed screens for BTUT, 11 of which contributed to
95 percent of this site's failed screens. BTUT has the second highest number of
individual pollutants failing screens program-wide.
• Of the site-specific pollutants of interest, formaldehyde has the highest annual
average concentrations for BTUT, although annual averages could not be calculated
for the VOCs for 2015 due to low completeness.
• BTUT has the highest annual average concentrations of hexachloro-1,3-butadiene
(2016) and formaldehyde (2015) among NMP sites sampling these pollutants; BTUT
also has the second highest annual average of acetaldehyde (2015) among NMP sites.
• Sampling for the site-specific pollutants of interest has occurred at BTUT for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. The most notable trend is for benzene. Concentrations of
benzene have a decreasing trend at BTUT. Concentrations of 1,3-butadiene and
naphthalene have also decreased in recent years.
25-15
-------
• Formaldehyde has the highest cancer risk approximations for BTUT; formaldehyde's
cancer risk approximation for 2015 is the highest cancer risk approximation
calculated across the program. None of the pollutants of interest have noncancer
hazard approximations greater than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Davis
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor, while
acrolein has the highest noncancer toxicity-weighted emissions for Davis County.
Vermont.
• The NATTS site in Vermont (UNVT) is located in Underhill, just outside Burlington.
• PAHs were sampled for at UNVT.
• Benzo(a)pyrene and naphthalene were the only pollutants to fail screens for UNVT.
One concentration of each failed a screen over the two-year period.
• Both annual average concentrations of naphthalene for UNVT were less than
10 ng/m3, while both annual average concentrations for benzo(a)pyrene were less
than 0.05 ng/m3.
• UNVT has the lowest annual average concentration of naphthalene and
benzo(a)pyrene among NMP sites sampling these pollutants.
• Sampling for the site-specific pollutants of interest has occurred at UNVNT for at
least 5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Naphthalene concentrations exhibit a decreasing trend at
UNVT.
• UNVT's cancer risk approximations for naphthalene and benzo(a)pyrene are less than
1.0 in-a-million. Naphthalene is the only pollutant of interest for UNVT with a
noncancer toxicity factor; the noncancer hazard approximations for naphthalene are
significantly less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Chittenden
County, while POM, Group 3 has the highest cancer toxicity-weighted emissions.
Toluene is the highest emitted pollutant with a noncancer toxicity factor in Chittenden
County, while acrolein has the highest noncancer toxicity-weighted emissions.
Virginia.
• The NATTS site in Virginia (RIVA) is located in East Highland Park, near
Richmond.
• PAHs and hexavalent chromium were sampled for at RIVA. Hexavalent chromium
sampling was discontinued at RIVA at the end of June 2016.
25-16
-------
• Naphthalene and hexavalent chromium failed screens for RIVA, with concentrations
of naphthalene accounting for 99 percent of failed screens, and thus, is the only
pollutant of interest for this site.
• The annual average concentrations for naphthalene were similar to each other in
magnitude, both around 70 ng/m3.
• Sampling for PAHs has occurred at RIVA for at least 5 consecutive years; thus, a
trends analysis was conducted for naphthalene. Concentrations of naphthalene exhibit
a decreasing trend at RIVA, although those concentrations have leveled off over
recent years.
• The cancer risk approximations for naphthalene at RIVA are both less than 3 in-a-
million, consistent with previous years. The noncancer hazard approximations are
significantly less than an HQ of 1.0.
• Benzene is the highest emitted pollutant with a cancer toxicity factor in Henrico
County, while formaldehyde has the highest cancer toxicity-weighted emissions.
Ethylene glycol is the highest emitted pollutant with a noncancer toxicity factor in
Henrico County, while acrolein has the highest noncancer toxicity-weighted
emissions.
Washington.
• The NATTS site in Washington (SEWA) is located in Seattle.
• VOCs, carbonyl compounds, PAHs, and metals (PMio) were sampled for at SEWA.
• Fifteen pollutants failed screens for SEWA, eight of which were identified as
pollutants of interest for this site.
• None of the site-specific pollutants of interest for SEWA have annual average
concentrations greater than 1 |ig/m3, Acetaldehyde has the highest annual average
concentration for 2015 while carbon tetrachloride has the highest annual average for
2016, although the averages for these pollutants are similar to each in magnitude.
• SEWA's annual average concentrations of acetaldehyde and formaldehyde are among
the lowest compared to other NMP sites sampling these pollutants.
• Sampling for the site-specific pollutants of interest has occurred at SEWA for at least
5 consecutive years; thus, a trends analysis was conducted for the site-specific
pollutants of interest. Concentrations of benzene have an overall decreasing trend at
SEWA. Concentrations of naphthalene and 1,3-butadiene exhibit a decreasing trend
over recent years.
• Formaldehyde has the highest cancer risk approximations for SEWA, both of which
are less than 10 in-a-million. Only one NMP site has a lower cancer risk
approximation for formaldehyde among sites for which formaldehyde is a pollutant of
interest. All of the noncancer hazard approximations for the pollutants of interest for
SEWA are considerably less than an HQ of 1.0.
25-17
-------
• Benzene is the highest emitted pollutant with a cancer toxicity factor in King County
while formaldehyde has the highest cancer toxicity-weighted emissions. Toluene is
the highest emitted pollutant with a noncancer toxicity factor in King County, while
acrolein has the highest noncancer toxicity-weighted emissions.
25.1.3 Composite Site-level Results Summary
• Twenty-five pollutants were identified as site-specific pollutants of interest, based on
the risk-based screening process. Benzene, acetaldehyde, and formaldehyde were the
most common pollutants of interest among the monitoring sites. Benzene was
identified as a pollutant of interest for all 34 sites that sampled this pollutant (with
Method TO-15 or SNMOC). Acetaldehyde and formaldehyde were identified as
pollutants of interest for all 33 sites that sampled carbonyl compounds. Naphthalene
was identified as a pollutant of interest for 18 of the 19 sites that sampled PAHs (with
GLKY as the exception). Arsenic was identified as a pollutant of interest for all 19
sites that sampled metals.
• Several pollutants were identified as site-specific pollutants of interest for only one or
two sites. For instance, dichloromethane is a pollutant of interest for only GPCO;
trichloroethylene is a pollutant of interest for only SPIL; and bromomethane is a
pollutant of interest for only CSNJ.
• Table 25-1 summarizes which pollutants of interest were identified for each site, how
many pollutants of interest were identified for each site, and how many sites for
which each pollutant was identified as a pollutant of interest.
• EPA dropped the requirement to sample hexavalent chromium under the NATTS
program beginning in July 2013. RIVA is the only NATTS site to continue sampling
this pollutant beyond 2014, although sampling was discontinued at the end of June
2016. One concentration of hexavalent chromium measured at RIVA failed a screen
(which was measured on July 5, 2015).
• Formaldehyde frequently had the highest site-specific annual average concentration
among the site-specific pollutants of interest; formaldehyde had the highest annual
average concentration for 28 sites. Naphthalene had the next highest at 10 followed
by benzene with five.
• Seven sites have cancer risk approximations greater than 50 in-a-million, eight for
formaldehyde (BTUT and ELNJ's annual averages for both years; BROK's annual
average for 2015; and AZFL, SKFL, and CSNJ's annual averages for 2016) and two
for 1,2-dichloroethane (TVKY's annual averages for both years). Formaldehyde
tended to have the highest cancer risk approximation on a site-specific basis. This is
true for 31 NMP sites. The highest cancer risk approximation for formaldehyde was
calculated for BTUT (109.51 in-a-million, 2015). The second highest annual average-
based cancer risk approximation was calculated for 1,2-dichloroethane (97.62 in-a-
million), based on TVKY's 2015 annual average concentration. Benzene and 1,3-
butadiene are the only other pollutants for which a cancer risk approximation greater
than 10 in-a-million was calculated (two and one, respectively).
25-18
-------
Table 25-1. Summary of Site-Specific Pollutants of Interest
State
Site
# of Pollutants of Interest
Acenaphthene
Acetaldehyde
Arsenic
Benzene
Benzo(a)pyrene
Bromomethane
1,3-Butadiene
Cadmium
Carbon Tetrachloride
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluor anthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Lead
Manganese
Naphthalene
Nickel
p-Dichlorobenzene
Propionaldehyde
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
AZ
PXSS
10
X
X
X
X
X
X
X
X
X
X
AZ
SPAZ
6
X
X
X
X
X
X
CA
CELA
1
X
CA
RUCA
1
X
CA
SJJCA
3
X
X
X
CO
BMCO
3
X
X
X
CO
BRCO
3
X
X
X
CO
GPCO
12
X
X
X
X
X
X
X
X
X
X
X
X
CO
GSCO
4
X
X
X
X
CO
PACO
4
X
X
X
X
CO
RFCO
4
X
X
X
X
CO
RICO
5
X
X
X
X
X
DC
JFADC
1
X
FL
AZFL
2
X
X
FL
ORFL
2
X
X
FL
PAFL
2
X
X
FL
SKFL
3
X
X
X
FL
SYFL
2
X
X
IL
NBIL
12
X
X
X
X
X
X
X
X
X
X
X
X
IL
ROIL
6
X
X
X
X
X
X
IL
SPIL
8
X
X
X
X
X
X
X
X
IN
INDEM
2
X
X
BOLD ITALICS = EPA-designated NATTS Sites
-------
Table 25-1. Summary of Site-Specific Pollutants of Interest (Continued)
to
Lfl
to
o
State
Site
# of Pollutants of Interest
Acenaphthene
Acetaldehyde
Arsenic
Benzene
Benzo(a)pyrene
Bromomethane
1,3-Butadiene
Cadmium
Carbon Tetrachloride
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluor anthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Lead
Manganese
Naphthalene
Nickel
p-Dichlorobenzene
Propionaldehyde
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
IN
WPIN
2
X
X
KY
ASKY
5
X
X
X
X
X
KY
ASKY-M
5
X
X
X
X
X
KY
ATKY
7
X
X
X
X
X
X
X
KY
BAKY
1
X
KY
BLKY
7
X
X
X
X
X
X
X
KY
GLKY
7
X
X
X
X
X
X
X
KY
LEKY
6
X
X
X
X
X
X
KY
TVKY
6
X
X
X
X
X
X
MA
BOMA
4
X
X
X
X
MI
DEMI
10
X
X
X
X
X
X
X
X
X
X
MO
S4MO
12
X
X
X
X
X
X
X
X
X
X
X
X
NJ
CHNJ
6
X
X
X
X
X
X
NJ
CSNJ
9
X
X
X
X
X
X
X
X
X
NJ
ELNJ
7
X
X
X
X
X
X
X
NJ
NBNJ
6
X
X
X
X
X
X
NJ
NRNJ
7
X
X
X
X
X
X
X
NY
BXNY
4
X
X
X
X
NY
ROCH
4
X
X
X
X
OK
BROK
7
X
X
X
X
X
X
X
OK
NROK
8
X
X
X
X
X
X
X
X
OK
OCOK
8
X
X
X
X
X
X
X
X
BOLD ITALICS = EPA-designated NATTS Sites
-------
Table 25-1. Summary of Site-Specific Pollutants of Interest (Continued)
to
Lfl
to
State
Site
# of Pollutants of Interest
Acenaphthene
Acetaldehyde
Arsenic
Benzene
Benzo(a)pyrene
Bromomethane
1,3-Butadiene
Cadmium
Carbon Tetrachloride
1,2-Dichloroethane
Dichloromethane
Ethylbenzene
Fluor anthene
Fluorene
Formaldehyde
Hexachloro-l,3-butadiene
Lead
Manganese
Naphthalene
Nickel
p-Dichlorobenzene
Propionaldehyde
1,1,2-T richloroethane
T richlo roethy lene
Vinyl chloride
OK
TMOK
10
X
X
X
X
X
X
X
X
X
X
OK
TOOK
11
X
X
X
X
X
X
X
X
X
X
X
OK
TOOK
10
X
X
X
X
X
X
X
X
X
X
OK
YUOK
8
X
X
X
X
X
X
X
X
RI
PRRI
2
X
X
UT
BTUT
11
X
X
X
X
X
X
X
X
X
X
X
VA
RIVA
1
X
VT
UNVT
2
X
X
WA
SEWA
8
X
X
X
X
X
X
X
X
Total
297
6
33
19
34
2
1
32
2
29
29
1
12
3
6
33
9
1
2
18
7
10
2
2
1
3
BOLD ITALICS = EPA-designated NATTS Sites
-------
Carbon tetrachloride often had "higher" cancer risk approximations compared to
other pollutants of interest among the monitoring sites, ranging between 3 in-a-
million and 6 in-a-million, but tended to have relatively low emissions and toxicity-
weighted emissions, according to the NEI. This pollutant appears only once in the
emissions-based tables for counties with NMP sites (Marshall County, Kentucky,
where the two of the three Calvert City sites are located).
None of the noncancer hazard approximations based on annual average
concentrations of the site-specific pollutants of interest were greater than an HQ of
1.0. The noncancer hazard approximation calculated for BTUT's annual average
concentration of formaldehyde for 2015 (with an HQ of 0.86) is the highest of all
annual average-based noncancer hazard approximations. Formaldehyde tended to
have the highest noncancer hazard approximations on a site-specific basis, followed
by naphthalene, arsenic, and 1,3-butadiene.
Of those pollutants with cancer UREs, formaldehyde, benzene, acetaldehyde, and
ethylbenzene often had the highest county-level emissions for participating counties.
Benzene, formaldehyde, and 1,3-butadiene typically had the highest toxicity-
weighted emissions (of those with a cancer URE).
Of those pollutants with a noncancer RfC, toluene, xylenes, methanol, and benzene
were often the highest emitted pollutants, although they rarely had the highest
toxicity-weighted emissions. Acrolein tended to have the highest toxicity-weighted
emissions of pollutants with noncancer RfCs, although the quantity of acrolein
emissions were generally low when compared to other pollutants. Acrolein appears
five times among the 10 highest emitted pollutants for counties with NMP sites
(Garfield and Mesa Counties in Colorado, Chittenden County, Vermont, and
Canadian and Grady Counties in Oklahoma). However, due to the high toxicity of
this pollutant, even low emissions translated into high noncancer toxicity-weighted
emissions; the toxicity-weighted value was often several orders of magnitude higher
than other pollutants. Acrolein is a national noncancer risk driver according to
NATA. Besides acrolein, formaldehyde and 1,3-butadiene tended to have the highest
toxicity-weighted emissions among the pollutants with noncancer RfCs.
Although production of carbon tetrachloride has declined sharply over the last
30 years due to its role as an ozone depleting substance, it has a relatively long
atmospheric lifetime and thus, is present at similar levels at nearly any given location.
NMP sites are located in a variety of locations across the country with different
purposes behind the monitoring at each site. In most cases, the concentrations of
carbon tetrachloride measured across the program confirm the ubiquitous nature of
this pollutant. However, carbon tetrachloride concentrations measured at the Calvert
City, Kentucky sites were often higher than levels of this pollutant collected
elsewhere. Vinyl chloride is an industrial-marker and is infrequently measured at
levels above the MDL (this pollutant has a 31 percent detection rate across the
program, though only 10 percent of these measurements were greater than the MDL).
The Calvert City, Kentucky sites together account for more than 30 percent of the
measured detections of vinyl chloride for 2015 and 2016 (and account for highest 146
concentrations of vinyl chloride measured across the program). Individually, these
sites have the highest number of measured detections of vinyl chloride among NMP
25-22
-------
sites sampling VOCs. The Calvert City sites also account for the 174 highest
concentrations of 1,2-dichloroethane measured across the program. These ambient air
measurements agree with corresponding emissions data in the NEI. These three
pollutants appear among the highest emitted pollutants in Marshall County, Kentucky
(among those with a cancer URE) but are not among the highest emitted pollutants
for any other county with an NMP site. From a quantitative standpoint, the emissions
of carbon tetrachloride, 1,2-dichloroethane, and vinyl chloride in Marshall County are
higher than their emissions in any other county with an NMP site.
• For every NMP site for which 1,2-dichloroethane is a pollutant of interest and where
a trends analysis could be conducted for this pollutant (22 sites), a dramatic increase
in the number of measured detections is shown over the last five years of sampling,
particularly for 2012, which was mostly sustained during the years that follow. This
pollutant was detected in less than 10 percent of samples at most sites participating in
the NMP prior to 2010 (and still participating now); the rate increased significantly
since 2010, slowly at first, then significantly in 2012. The detection rate of this
pollutant is between 80 percent and 100 percent for NMP sites sampling this pollutant
in 2015 and/or 2016.
25.1.4 Data Quality Results Summary
Completeness, precision, and accuracy were assessed for the 2015 and 2016 monitoring
efforts. The quality assessments presented in this report show that the 2015-2016 monitoring data
are of a known and high quality, based on the attainment of the established MQOs.
To the largest extent, ambient air concentration datasets met the MQO for completeness;
204 of the 215 site- and method-specific datasets met the 85 percent completeness MQO while
11 datasets (seven from 2015 and four from 2016) did not. Seventy-nine datasets achieved
100 percent completeness.
Method (i.e., sampling and analytical) precision and analytical precision were determined
for the 2015-2016 NMP monitoring efforts using CV calculations based on duplicate, collocated,
and replicate samples. Method precision for most analytical methods utilized during the 2015-
2016 NMP was within the MQO of 15 percent CV (with the exception of Method
TO-13A/PAHs). Analytical precision for each method was determined to be less than 15 percent
CV. The precision calculations presented in this report are based on analytical results greater
than or equal to the sample- and pollutant-specific MDL.
Analytical method accuracy is ensured by using proven methods, as demonstrated by
third-party analysis of proficiency test audit samples and following strict quality control and
quality assurance guidelines. Most of the pollutants for which audit samples were analyzed met
the MQO for accuracy. Of the 143 results analyzed for the 2015 and 2016 audit samples, only
25-23
-------
three exceeded the MQO of ± 25 percent recovery (two VOCs and one PAH), and none failed
multiple audits.
25.2 Conclusions
Conclusions extrapolated from the data analyses of the data generated from the 2015 and
2016 NMP monitoring efforts are presented below.
• Of the 65 pollutants for which the risk screening process was performed,
concentrations of 39 pollutants failed screens. Of these, about one third of
concentrations are greater than their respective risk screening values, particularly for
many of the NATTS MQO Core Analytes. For several of the pollutants, all or nearly
all of the measurements fail screens. Examples of frequently detected pollutants that
typically fail all or nearly all of their screens include benzene, carbon tetrachloride,
formaldehyde, acetaldehyde, and 1,2-dichloroethane. Some less frequently detected
pollutants still fail relatively large numbers of screens. For example, even though
hexachloro-1,3-butadiene was detected relatively infrequently (343 measured
detections), most (298) of those measured detections failed screens. The MDLs for
this pollutant are relatively high (0.34 |ig/m3 for 2015 and 0.42 |ig/m3 for 2016) while
the toxicity factor is relatively low (0.045 |ig/m3). Thus, all or nearly all of the
measured detections fail screens.
• Although the number of concentrations failing screens varies from year to year, the
percentage of failed screens compared to the number of measured detections has been
fairly consistent. Between the 2011 and 2014, the percentage has hovered around
36 percent. The percentage for the combined 2015 and 2016 monitoring effort is
slightly lower at 32 percent. Risk screening values are often updated from year-to-
year, although there were no changes for the 2015-2016 report.
• Among those pollutants for which annual average concentrations could be calculated
and that have available cancer UREs, one cancer risk approximation is greater than
100 in-a-million (BTUT, formaldehyde for 2015). In total, 61 cancer risk
approximations were greater than 10 in-a-million (52 for formaldehyde, six for
1,2-dichloroethane, two for benzene, and one for 1,3-butadiene); and nearly
83 percent were greater than 1.0 in-a-million.
• Among those pollutants for which annual average concentrations could be calculated
and have available noncancer RfCs, none of the noncancer hazard approximations
were greater than an HQ of 1.0.
• When comparing the highest emitted pollutants for a specific county to the pollutants
with the highest toxicity-weighted emissions, the pollutants tended to be more similar
for the pollutants with cancer UREs than for pollutants with noncancer RfCs. This
indicates that pollutants with cancer UREs that are emitted in higher quantities are
often more toxic than pollutants emitted in lower quantities; conversely, the highest
emitted pollutants with noncancer RfCs are not necessarily the most toxic. For
example, toluene is the noncancer pollutant that was emitted in the highest quantities
for many NMP counties (and did not rank less than third for any county with an NMP
site) but was not one of the pollutants with highest toxicity-weighted emissions for
25-24
-------
any of these counties. Conversely, while acrolein had the highest noncancer toxicity-
weighted emissions for all but one county with an NMP site (where it ranked second
rather than first), it was among the highest emitted pollutants for only two counties
with NMP sites (and ranked no higher than eighth).
The number of states and sites participating in the NMP varies from year-to-year. For
example, the number of sites included in the 2014 NMP decreased considerably from
2013, from 66 for 2013 to 51 for 2014. This is predominantly due to the removal of
hexavalent chromium from the NATTS list of required pollutants for which to
sample. Fifty-one NATTS and UATMP sites participated in the NMP for 2015; 51
sites also participated in the NMP for 2016, although the sites participating each year
is slightly different.
Many of the data analyses utilized in this report require data from year-round (or
nearly year-round) sampling. Of the 215 site-method-year combinations, 192
combinations covered an entire calendar year. The 23 exceptions include the
establishment of two new sites in Oklahoma (BROK and NROK), the discontinuation
of sampling at several sites including ROIL, PAFL, and ORFL; the discontinuation of
select methods at sites including LEKY, RIVA, and RFCO; and the relocation of
instrumentation from NBNJ to NRNJ and from BMCO to GSCO and back again.
Of the 53 monitoring sites participating in the 2015 and 2016 NMP, none sampled for
all six available pollutant groups under the NMP through the national contract
laboratory. Two sites (BTUT and NBIL) sampled for five pollutant groups and
another seven sites (BROK, GLKY, NROK, PXSS, GPCO, S4MO, and SEW A)
sampled four pollutant groups. The wide range of pollutant groups sampled for
among the sites, which is often the result of different purposes behind the monitoring
at the sites, makes it difficult to draw definitive conclusions regarding air toxics in
ambient air in a global manner.
The data analyses contained in the 2015 and 2016 NMP report reflect the inclusion of
data from a number of source-oriented monitoring sites. Source-oriented sites include
several of the Kentucky sites and the Camden, New Jersey site. Many of these sites
are the drivers for certain pollutant(s) in the report. This can easily be seen in the
graphical comparisons of the site-specific averages to the program-level average
concentrations contained in Sections 5 through 23. For many of these pollutants,
particularly the VOCs, the highest concentrations were considerably greater than the
majority of measurements, such that the scale in the figures needed to be greatly
reduced.
This report strives to represent data derived from the best laboratory practices and
utilize the best data analysis techniques available. Examples of this include the
improvement of MDLs and the incorporation of updated values for various toxicity
factors. This can lead to adjusting the focus of the report to concentrate on the air
quality issues of highest concern. Thus, the NMP report is dynamic in nature and
scope; yet this approach may prevent the direct comparison of the current report to
past reports. Relatively few major changes were instituted between the 2014 and
2015-2016 NMP reports. The major difference between the 2015-2016 report and
25-25
-------
other reports in recent years is the inclusion of two years of data, and the exclusion of
meteorological data.
25.3 Recommendations
Based on the data summaries and conclusions from the 2015-2016 NMP, a number of
recommendations for future ambient air monitoring efforts are presented below.
• Participate in the National Monitoring Programs year-round. Many of the analyses
presented in the 2015-2016 NMP report require a full year of data to be most useful
and representative of conditions experienced at each specified location. Therefore,
state and local agencies should be encouraged to implement year-long ambient air
monitoring programs in addition to participating in future monitoring efforts.
• Monitor for additional pollutant groups based on the results of data analyses in the
annual report. The risk-based analysis where county-level emissions are weighted
based on toxicity identifies those pollutants whose emissions may result in adverse
health effects in a specific area. If sampling for a pollutant or pollutant group
identified as particularly hazardous for a given area is not being performed, the
responsible agency should consider sampling for those compounds.
• Monitor for additional pollutant groups based on emerging environmental
monitoring needs. With the advent of fracking and the expansion of U.S. oil and gas
production, measurements of several groups of oil and gas-related HAPs emissions
should be added to the NMP and offered to participants.
• Continue to identify and implement improvements to the sampling and analytical
methods. Further research is encouraged to identify method improvements that would
allow for the characterization of an even wider range of components in air pollution
and enhance the ability of the methods to quantify all cancer and noncancer pollutants
to at least their levels of concern (risk screening concentrations). An update to the
Compendium methods is underway at EPA and is an example of potential method
optimization. At the time of publication, a revision to the TO-15 Method was
underway at EPA, with a review of Method TO-11A to follow.
• Perform case studies based on findings from the annual report. Often, the annual
report identifies an interesting tendency or trend, or highlights an event at a particular
site(s). For example, dichloromethane concentrations have been highest at BTUT and
GPCO for multiple years and trichloroethylene concentrations have been highest at
SPIL for multiple years. Further examination of the data in conjunction with
meteorological phenomena and potential emissions events or incidents, or further site
characterization may help state and local agencies pinpoint issues affecting air quality
in their area.
25-26
-------
26.0 References
ACE, 2018. U.S. Army Corps of Engineers. Huntington District, Grayson Lake website.
http://www.lrh.usace.armv.mil/Missions/CivilWorks/Recreation/Kentucky/GravsonLake.
aspx Date Last Accessed: 1/24/2018.
ASTM, 2012. ASTM, International. ASTM D7614-12 Standard Test Method for Determination
of Total Suspended Particulate (TSP) Hexavalent Chromium in Ambient Air Analyzed
by Ion Chromatography (IC) and Spectrophotometric Measurements.
http://www.astm.org/Standards/D7614.htm Date Last Accessed: 7/27/2017.
ASTM, 2013. ASTM, International. ASTM D6209-13 Standard Test Method for Determination
of Gaseous and Particulate Polycyclic Aromatic Hydrocarbons in Ambient Air
(Collection on Sorbent-Backed Filters with Gas Chromatographic/Mass Spectrometric
Analysis). http://www.astm.org/Standards/D6209.htm Date Last Accessed: 7/27/2017.
AZ DOT, 2017. Arizona Department of Transportation. Transportation Data Management
System website. 2010 and 2015 AADTs.
http://adot.ms2soft.com/tcds/tsearch. asp?loc=Adot&mod=
Date Last Accessed: 4/20/2017.
CADOT, 2015. State of California Department of Transportation, Division of Traffic
Operations. 2015 Traffic Volumes on the California State Highway System. Sacramento,
CA. http://traffic-counts.dot.ca.gov/ Date Last Accessed: 4/20/2017.
Calvert City, 2018. Calvert City - Industry website, http://www.calvertcity.com/
Date Last Accessed: 1/24/2018.
Carbondale, 2017. Carbondale, Colorado. The Town website, History.
http s: //www, carb ondal e. com/ab out/hi story/ Date Last Accessed: 12/12/2017.
Census Bureau, 2015. U.S. Census Bureau. July 2015. List 1. Core Based Statistical Areas
(CBSAs), Metropolitan Divisions, and Combined Statistical Areas (CSAs).
https://www.census.gov/geographies/reference-files/time-series/demo/metro-
micro/delineation-files.html Date Last Accessed: 7/27/2017.
Census Bureau, 2017. U.S. Census Bureau. Metropolitan and Micropolitan Glossary website.
https://www.census.gov/programs-survevs/metro-micro/about/glossary.html
Date Last Accessed: 7/27/2017.
CO DOT, 2015. Colorado Department of Transportation. Online Transportation Information
System (OTIS), Traffic Data Explorer. 2015 data, http://dtdapps.coloradodot.info/otis/
Date Last Accessed: 4/20/2017.
DC DOT, 2015. District Department of Transportation. November 2015. 2014 Traffic Volumes
Map. Washington, D.C. http://ddot.dc.gov/page/traffic-volume-maps
Date Last Accessed: 4/20/2017.
26-1
-------
DC WSA, 2017. District of Columbia Water and Sewer Authority. First Street Tunnel Project
website, https://www.dcwater.com/proiects/first-street-tunnel-proiect
Date Last Accessed: 12/26/2017.
EPA, 1986. U.S. EPA. 1986. 40 CFR Part 136, Appendix B. Definition and Procedure for the
Determination of the Method Detection Limit, Revision 1.11. http://www.ecfr.gov/cgi-
bin/text-idx?c=ecfr&tpl=/ecfrbrowse/Title40/40cfrl36 main 02.tpl
Date Last Accessed: 7/27/2017.
EPA, 1998. U.S. EPA. September 1998. Technical Assistance Document for Sampling and
Analysis of Ozone Precursors. EPA/600-R-98/161. Research Triangle Park, NC.
http://www.epa.gov/ttnamtil/pamsguidance.html Date Last Accessed: 7/27/2017.
EPA, 1999a. U.S. EPA. January 1999. Compendium Method TO-15: Determination of Volatile
Organic Compounds (VOCs) in Air Collected in Specially-Prepared Canisters and
Analyzed by Gas Chromatography/Mass Spectrometry (GC/MS). EPA/625/R-96/010b.
Cincinnati, OH. http://www.epa.gov/ttn/amtic/airtox.html
Date Last Accessed: 7/27/2017.
EPA, 1999b. U.S. EPA. January 1999. Compendium Method TO-11A: Determination of
Formaldehyde in Ambient Air Using Adsorbent Cartridge Followed by High
Performance Liquid Chromatography (HPLC). EPA/625/R-96/010b. Cincinnati, OH.
http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 7/27/2017.
EPA, 1999c. U.S. EPA. January 1999. Compendium Method TO-13A: Determination of
Polycyclic Aromatic Hydrocarbons (PAHs) in Ambient Air Using Gas
Chromatography/Mass Spectrometry (GC/MS). EPA/625/R-96/010b. Cincinnati, OH.
http://www.epa.gov/ttn/amtic/airtox.html Date Last Accessed: 7/27/2017.
EPA, 1999d. U.S. EPA. June 1999. Compendium Method 10-3.5: Determination of Metals in
Ambient Particulate Matter Using Inductively Coupled Plasma/Mass Spectrometry
(ICP/MS). EPA/625/R-96/01 Oa. Cincinnati, OH.
http://www.epa.gov/ttn/amtic/files/ambient/inorganic/mthd-3-5.pdf
Date Last Accessed: 7/27/2017.
EPA, 2007. U.S. EPA. May 2007. 1990-2002 NEI HAP Trends: Success of CAA Air Toxic
Programs in Reducing HAP Emissions and Risk. Paper presented by Anne Pope at the
16th Annual International Emission Inventory Conference. Raleigh, NC.
http://www.epa.g0v/ttn/chief/conference/eil6/session6/a.pope.pdf
Date Last Accessed: 11/7/2017.
EPA, 2009a. U.S. EPA. April 1, 2009. Technical Assistance Document for the National Air
Toxics Trends Stations Program, Revision 2. Research Triangle Park, NC.
https://www3.epa.gov/ttnamtil/files/ambient/airtox/nattsTADRevision2 508Compliant.p
dfDate Last Accessed: 7/12/2017.
EPA, 2009b. U.S. EPA. September 10, 2009. Schools Air Toxics Monitoring Activity (2009)
Uses of Health Effects Information in Evaluating Sample Results.
https://www3.epa.gov/air/sat/techinfo.html Date Last Accessed: 11/7/2017.
26-2
-------
EPA, 2010a. U.S. EPA. October 2010. A Preliminary Risk-based Screening Approach for Air
Toxics Monitoring Data Sets, version 2. EPA-904-B-06-001. Atlanta, GA.
nepis.epa.gov/Exe/ZyPURL.cgi?Dockev=P1009A7C.TXT Date Last Accessed: 8/4/2017.
EPA, 2010b. U.S. EPA. December 2010. Data Quality Evaluation Guidelines for Ambient Air
Acrolein Measurements.
http://www.epa.gov/ttn/amtic/files/ambient/airtox/20101217acroleindataqualitveval.pdf
Date Last Accessed: 11/7/2017.
EPA, 2011. U.S. EPA. June 2011. Technical Report for School: Assessing Outdoor Air Near
Schools: Riggins School, North Birmingham Elementary School, and Lewis Elementary
School (Birmingham, Alabama), http://www3.epa.gov/air/sat/NorthBirmi.html
Date Last Accessed: 7/31/2017.
EPA, 2012. U.S. EPA. June 2012. Ambient Air Monitoring Reference and Equivalent Methods:
Designation of Three New Equivalent Methods. Federal Register Volume 77, Number
106, page 32632. http://www.gpo.gov/fdsvs/pkg/FR-2012-06-01/html/2012-1335Q.htm
Date Last Accessed: 7/27/2017.
EPA, 2014. U.S. EPA. September 2014. UATMP, PAMS, andNMOC Support, Contract No.
EP-D-14-030. Date Last Accessed: 7/12/2017.
EPA, 2015a. U.S EPA. December 17, 2015. NATA: Glossary of Terms website.
https://www.epa.gov/national-air-toxics-assessment/nata-glossary-terms
Date Last Accessed: 8/4/2017.
EPA, 2015b. U.S. EPA. December 2015. Technical Support Document EPA's 2011 National-
scale Air Toxics Assessment. Appendix H. https://www.epa.gov/national-air-toxics-
assessment/2011-nata-technical-support-document Date Last Accessed: 7/17/2017.
EPA, 2015c. U.S. EPA. December 2015. Technical Support Document EPA's 2011 National-
scale Air Toxics Assessment. Pg 20-22. https://www.epa.gov/national-air-toxics-
assessment/2011-nata-technical-support-document Date Last Accessed: 7/17/2017.
EPA, 2016. U.S. EPA. September 30, 2016. 2014 National Emissions Inventory, Version 1.
https://www.epa.gov/air-emissions-inventories/2014-national-emissions-inventory-nei-
data Date Last Accessed: 7/27/2017.
EPA, 2017a. U.S. EPA. Air Toxics - National Air Toxics Trends Stations website.
http://www3.epa.gov/ttnamti 1/natts.html Date Last Accessed: 7/12/2017.
EPA, 2017b. U.S. EPA. Air Quality System (AQS) website, http://www.epa.gov/aqs
Date Last Accessed: 7/12/2017.
EPA, 2017c. U.S. EPA. October 2017. National Air Toxics Trends Stations (NATTS) Network
Assessment, 2003-2014. Revised Final. Prepared by Eastern Research Group, Inc.
EPA, 2017d. U.S. EPA. What are Hazardous Air Pollutants? website.
https://www.epa.gov/haps/what-are-hazardous-air-pollutants
Date Last Accessed: 8/4/2017.
26-3
-------
EPA, 2017e. Email from Ted Palma, U.S. EPA, OAQPS. March 30, 2017. Dose Response
Library spreadsheet, January 2017 update.
EPA, 2017f. U.S. EPA. Phaseout of Class I Ozone-Depleting Substances website.
https://www.epa.gov/ods-phaseout/phaseout-class-i-ozone-depleting-substances
Date Last Accessed: 11/14/2017.
EPA, 2017g. EPA Air Toxics - Quality Assurance document listing.
https://www3.epa.gov/ttn/amtic/airtoxqa.html Date Last Accessed: 11/1/2017.
ERG, 2015. Eastern Research Group, Inc. 2015. Support for the EPA National Monitoring
Programs (UATMP, NATTS, C SAT AM, PAMS, andNMOC Support), Quality
Assurance Project Plan, Category 1. Contract No. EP-D-14-030. Morrisville, NC.
Date Last Accessed: 7/27/2017.
ERG, 2016. Eastern Research Group, Inc. 2016. Support for the EPA National Monitoring
Programs (UATMP, NATTS, C SAT AM, PAMS, andNMOC Support), Quality
Assurance Project Plan, Category 1. Contract No. EP-D-14-030. Morrisville, NC.
Date Last Accessed: 7/27/2017.
FAC, 2007. Report of the Federal Advisory Committee on Detection and Quantitation
Approaches and Uses In Clean Water Act Programs, Appendix D: DQ FAC Single
Laboratory Procedure v2.4. December 2007. Submitted to the U.S. EPA.
https://www.epa.gov/cwa-methods/procedures-detection-and-quantitation
Date Last Accessed: 7/27/2017.
FL DOT, 2016. Florida Department of Transportation. Florida Traffic Online Mapping
Application (2016). http://flto.dot.state.fl.us/website/FloridaTrafficOnline/viewer.html
Date Last Accessed: 4/20/2017.
GCA, 2016. Garfield County Administration. Garfield County Profile, 2016. https://garfield-
county.com/economic-development/garfield-countv-profile.aspx
Date Last Accessed: 12/12/2017.
GCPH, 2015. Garfield County Public Health Department. June 2015. Garfield County 2014 Air
Quality Monitoring Report. Page 4-4. Rifle, CO. Prepared by Air Resource Specialists,
Inc., Ft. Collins, CO. http://www.garfield-countv.com/air-quality/
Date Last Accessed: 12/12/2017.
GCPH, 2017. Email from Morgan Hill, Garfield County Public Health. December 19, 2017.
GCRBD, 2014. Garfield County Road and Bridge Department. Garfield County Road Traffic
Counts. 2014 Data, http://www.garfield-countv.com/road-bridge/county-roads.aspx
Date Last Accessed: 4/20/2017.
Girardi, 2015. Girardi, Steven. 2015. Company buys Raytheon property, plans mixed-use
development. Tampa Bay Times. July 27, 2015. http://www.tbo.com/pinellas-
countv/companv-buvs-ravtheon-propertv-plans-mixed-use-development-20150727/
Date Last Accessed: 12/28/2017.
26-4
-------
IL DOT, 2017. Illinois Department of Transportation. Getting Around Illinois website. Average
Annual Daily Traffic map viewer. 2013, 2014, and 2015 data.
http://www.gettingaroundillinois.com/ Date Last Accessed: 4/20/2017.
IN DOT, 2016. Indiana Department of Transportation. Average Daily Traffic and Commercial
Vehicles Interactive map. 2016 Data, http://www.in.gov/indot/2469.htm
Date Last Accessed: 4/20/2017.
KY, 2018. Kentucky State Parks. Grayson Lake History website.
http://parks.kv.gov/parks/recreationparks/gravson-lake/history.aspx
Date Last Accessed: 1/24/2018.
KYTC, 2016. Kentucky Transportation Cabinet. Interactive Statewide Traffic Counts mapping
application. 2012-2016 Data. http://transportation.ky.gov/Planning/Pages/count-
maps.aspx Date Last Accessed: 5/2/2017.
LADCO, 2003. Lake Michigan Air Directors Consortium. December 2003. Phase II: Air Toxics
Monitoring Data: Analyses and Network Design Recommendations. Final technical
report prepared by Battelle Memorial Institute and Sonoma Technology, Inc.
http://www.ladco.org/reports/toxics/battelle 2/ Date Last Accessed: 7/12/2017.
LBWP, 2013. Lawrence Brook Watershed Partnership. June 2013. Lawrence Brook Fish Ladder
Partial Feasibility Assessment. Report prepared by Princeton Hydro, LLC. Ringoes, New
Jersey.
http://www.harborestuarv.org/grants/reports/LawrenceBrookPartialFeasibilitvStudvFINA
Lreport.pdf Date Last Accessed: 7/2/2018.
MA DOT, 2010. Massachusetts Department of Transportation, Highway Division.
Transportation Data Management System. 2010 Traffic Count.
http://www.massdot.state.ma.us/highwav/TrafficVolumeCounts.aspx
Date Last Accessed: 5/2/2017.
MI DOT, 2015. Michigan Department of Transportation. Annual Average Daily Traffic (AADT)
Map. 2015 Data (as confirmed with TMIS report).
http://www.michigan.gOv/mdot/0.4616.7-151-l 1151-22141—.00.html
Date Last Accessed: 5/2/2017.
MO DOT, 2015. Missouri Department of Transportation, Transportation Planning. 2015
St. Louis District Traffic Volume and Commercial Vehicle Count Map. Jefferson City,
MO. http://www.modot.org/safetv/trafficvolumemaps.htm Date Last Accessed: 5/2/2017.
NHMU, 2018. Natural History Museum of Utah. Stefan Brems. Salt Lake Valley Weather
Patterns.
https://nhmu.utah.edu/sites/default/files/attachments/SaltLakeVallevWeatherPatterns.pdf
Date Last Accessed: 4/16/2018.
NJ DOT, 2017. New Jersey Department of Transportation. Roadway Information and Traffic
Monitoring System Program Interactive Traffic Count Reports. 2010, 2012, and 2014
Data, http://www.state.ni.us/transportation/refdata/roadway/traffic counts/
Date Last Accessed: 5/2/2017.
26-5
-------
NJ DOTr, 2008. New Jersey Department of Treasury. January 2008. New Jersey Traffic and
Revenue Study, New Jersey Turnpike & Route 440 Asset Appraisal. Final Report
prepared by Steer Davies Gleave in association with CRA International and EDR Group.
London, England, http://www.nislom.org/nitrstudvpart2.pdf
Date Last Accessed: 5/3/2017.
NOAA, 2017. National Climatic Data Center. Quality Controlled Local Climatological Data
(QCLCD). 2015 and 2016 Data, https://www.ncdc.noaa.gov/data-access/land-based-
station-data/land-based-datasets/qualitv-controlled-local-climatological-data-qclcd
Date Last Accessed: 5/11/2018.
NYS DOT, 2015. New York State Department of Transportation. Traffic Data Viewer mapping
application. 2015 Data.
https://www.dot.nv.gov/divisions/engineering/applications/traffic-data-viewer
Date Last Accessed: 5/2/2017.
OK DEQ, 2017. Email from Randy Ward, Oklahoma Department of Environmental Quality. July
5, 2017.
OK DOT, 2015. Oklahoma Department of Transportation, Strategic Asset and Performance
Management Division, Traffic Analysis Branch. 2015 Annual Average Daily Traffic
Oklahoma Highway System maps. Oklahoma City, OK.
http://www.okladot.state.ok.us/maps/aadt/index.htm Date Last Accessed: 5/3/2017.
Peltier, 2016. Peltier Technical Services, Inc. Peltier Tech Charts for Excel 3.0. Downloaded
March 8, 2016. http://peltiertech.com/Utilitv30/ Date Last Accessed: 5/29/2018.
Podmolik, 2015. Podmolik, Mary Ellen. 2015. Canadian Pacific seeks new use for Schiller Park
intermodal site. Chicago Tribune. January 20, 2015.
http://www.chicagotribune.com/business/breaking/ct-railwav-terminal-development-
0121-biz-20150120-story.html Date Last Accessed: 6/4/2018.
RI DOT, 2016. Rhode Island Department of Transportation. State Highway Map of Rhode
Island, Traffic Flow Map 2016. 2015 Data.
http://www.dot.ri.gov/about/maproom/index.php Date Last Accessed: 5/2/2017.
SAE, 2011. SAE, International. J1151 201109. Methane Measurement Using Gas
Chromatography. September 2011. http://standards.sae.Org/i 1151 201109/
Date Last Accessed: 7/27/2017.
Seattle, 2018. Seattle Parks and Recreation. Jefferson Park website.
http://www.seattle.gov/parks/find/parks/iefferson-park Date Last Accessed: 5/29/2018.
UT DOT, 2014. Utah Department of Transportation. AADT Traffic Map Updated Using 2014
Data (via Google Earth kmz file).
http://www.udot.utah.gov/main/f?p=100:pg:0:::: V.T:.528 Date Last Accessed: 5/3/2017.
UVM, 2015. The University of Vermont. Proctor Maple Research Center Mission website.
http://www.uvm.edu/~pmrc/?Page=mission.htm Date Last Accessed: 3/26/2018.
26-6
-------
VA DOT, 2016. Virginia Department of Transportation, Traffic Engineering Division. Average
Daily Traffic Volumes with Vehicle Classification Data on Interstate, Arterial, and
Primary Routes 2016. http://www.virginiadot.org/info/ct-TrafficCounts.asp
Date Last Accessed: 5/3/2017.
Vtrans, 2015. Vermot Agency of Transportation, Highway Division, Traffic Research Unit. June
2015. 2014 (Route Log) AADTs Major Collectors.
http://vtrans.vermont.gov/operations/technical-services/traffic
Date Last Accessed: 5/3/2017.
WS DOT, 2015. Washington State Department of Transportation. Traffic Data GeoPortal. 2015
Data, http://www.wsdot.wa.gov/mapsdata/tools/trafficplanningtrends.htm
Date Last Accessed: 5/3/2017.
WUSTL, 2013. Email from Jay Turner, Washington University in St. Louis. August 29, 2013.
WUSTL, 2016. Washington University in St. Louis, Air Quality Laboratory. Roxana Air Quality
Study website, http://raqs.seas.wustl.edu/ Date Last Accessed: 6/4/2018.
26-7
------- |